Académique Documents
Professionnel Documents
Culture Documents
2 Tutorial Guide
January 2005
Copyright
c 2005 by Fluent Inc.
All rights reserved. No part of this document may be reproduced or otherwise used in
any form without express written permission from Fluent Inc.
Airpak, FIDAP, FLUENT, FloWizard, GAMBIT, Icemax, Icepak, Icepro, MixSim, and
POLYFLOW are registered trademarks of Fluent Inc. All other products or name
brands are trademarks of their respective holders.
Fluent Inc.
Centerra Resource Park
10 Cavendish Court
Lebanon, NH 03766
Volume 1
1 Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
2 Modeling Periodic Flow and Heat Transfer
3 Modeling External Compressible Flow
4 Modeling Unsteady Compressible Flow
5 Modeling Radiation and Natural Convection
6 Using a Non-Conformal Mesh
7 Using a Single Rotating Reference Frame
8 Using Multiple Rotating Reference Frames
9 Using the Mixing Plane Model
10 Using Sliding Meshes
11 Using Dynamic Meshes
Volume 2
12 Modeling Species Transport and Gaseous Combustion
13 Using the Non-Premixed Combustion Model
14 Modeling Surface Chemistry
15 Modeling Evaporating Liquid Spray
16 Using the VOF Model
17 Modeling Cavitation
18 Using the Mixture and Eulerian Multiphase Models
19 Using the Eulerian Multiphase Model for Granular Flow
20 Modeling Solidification
21 Using the Eulerian Granular Multiphase Model with Heat Transfer
22 Postprocessing
23 Turbo Postprocessing
24 Parallel Processing
Using This Manual
c Fluent Inc. January 11, 2005 i
Using This Manual
You may then want to try a tutorial that demonstrates features that you are going to
use in your application. For example, if you are planning to solve a problem using the
non-premixed combustion model, you should look at Tutorial 13.
You may want to refer to other tutorials for instructions on using specific features, such
as custom field functions, grid scaling, and so on, even if the problem solved in the
tutorial is not of particular interest to you. To learn about postprocessing, you can look
at Tutorial 22, which is devoted entirely to postprocessing (although the other tutorials
all contain some postprocessing as well). For turbomachinery-specific postprocessing, see
Tutorial 23.
• The text interface type style is also used when illustrating exactly what appears on
the screen or exactly what you must type in the text window or in a panel.
• Instructions for performing each step in a tutorial will appear in standard type.
Additional information about a step in a tutorial appears in italicized type.
ii
c Fluent Inc. January 11, 2005
Using This Manual
• A mini flow chart is used to indicate the menu selections that lead you to a specific
command or panel. For example,
Define −→Boundary Conditions...
indicates that the Boundary Conditions... menu item can be selected from the Define
pull-down menu.
The words surrounded by boxes invoke menus (or submenus) and the arrows point
from a specific menu toward the item you should select from that menu.
c Fluent Inc. January 11, 2005 iii
Using This Manual
iv
c Fluent Inc. January 11, 2005
Contents
c Fluent Inc. January 11, 2005 i
CONTENTS
ii
c Fluent Inc. January 11, 2005
CONTENTS
c Fluent Inc. January 11, 2005 iii
CONTENTS
iv
c Fluent Inc. January 11, 2005
CONTENTS
c Fluent Inc. January 11, 2005 v
CONTENTS
vi
c Fluent Inc. January 11, 2005
CONTENTS
c Fluent Inc. January 11, 2005 vii
CONTENTS
viii
c Fluent Inc. January 11, 2005
CONTENTS
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-44
Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13-45
Coal Analysis for Elemental Composition . . . . . . . . . . . . . . . . . . 13-45
Discrete Phase Material Properties . . . . . . . . . . . . . . . . . . . . . 13-46
c Fluent Inc. January 11, 2005 ix
CONTENTS
x
c Fluent Inc. January 11, 2005
CONTENTS
c Fluent Inc. January 11, 2005 xi
CONTENTS
21 Using the Eulerian Granular Multiphase Model with Heat Transfer 21-1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-1
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-1
Problem Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-1
Setup and Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-2
Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-2
Step 1: Grid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21-3
xii
c Fluent Inc. January 11, 2005
CONTENTS
22 Postprocessing 22-1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1
Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-1
Problem Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-2
Setup and Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-2
Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-2
Step 1: Grid Display . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-3
Step 2: Adding Lights . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-5
Step 3: Creating Isosurfaces . . . . . . . . . . . . . . . . . . . . . . . . . 22-9
Step 4: Contours . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-10
Step 5: Velocity Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-13
Step 6: Animation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-18
Step 7: Pathlines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-22
Step 8: Overlaying Velocity Vectors on the Pathline Display . . . . . . . 22-28
Step 9: Exploded Views . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-31
Step 10: Animating the Display of Results in Successive
Streamwise Planes . . . . . . . . . . . . . . . . . . . . . . . . . . 22-36
Step 11: XY Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-38
Step 12: Annotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-42
Step 13: Saving Hardcopy Files . . . . . . . . . . . . . . . . . . . . . . . 22-44
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22-44
c Fluent Inc. January 11, 2005 xiii
CONTENTS
xiv
c Fluent Inc. January 11, 2005
CONTENTS
c Fluent Inc. January 11, 2005 xv
CONTENTS
xvi
c Fluent Inc. January 11, 2005
Tutorial 1. Introduction to Using FLUENT: Fluid Flow and
Heat Transfer in a Mixing Elbow
Introduction
This tutorial illustrates the setup and solution of the two-dimensional turbulent fluid flow
and heat transfer in a mixing junction. The mixing elbow configuration is encountered
in piping systems in power plants and process industries. It is often important to predict
the flow field and temperature field in the neighborhood of the mixing region in order to
properly design the location of inlet pipes.
In this tutorial you will learn how to:
• Set material properties and boundary conditions for a turbulent forced convection
problem
• Adapt the grid based on the temperature gradient to further improve the prediction
of temperature
Prerequisites
This tutorial assumes that you have little experience with FLUENT, but that you are
generally familiar with the interface.
c Fluent Inc. January 11, 2005 1-1
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Problem Description
The problem to be considered is shown schematically in Figure 1.1. A cold fluid at 26◦ C
enters through the large pipe and mixes with a warmer fluid at 40◦ C in the elbow. The
pipe dimensions are in inches, and the fluid properties and boundary conditions are given
in SI units. The Reynolds number at the main inlet is 2.03 × 105 , so that a turbulent
model will be necessary.
1-2
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
/cdrom/fluent6.2/help/tutfiles/
cdrom:\fluent6.2\help\tutfiles\
where cdrom must be replaced by the name of your CD-ROM drive (e.g., E).
2. Unzip introduction.zip.
elbow.msh can be found in the /introduction folder created after unzipping the
file.
c Fluent Inc. January 11, 2005 1-3
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Step 1: Grid
1. Read the grid file elbow.msh.
File −→ Read −→Case...
(a) Select the file elbow.msh by clicking on it under Files and then clicking on OK.
Note: As this grid is read by FLUENT, messages will appear in the console window
reporting the progress of the conversion. After reading the grid file, FLUENT
will report that 918 triangular fluid cells have been read, along with a number
of boundary faces with different zone identifiers.
1-4
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Grid Check
Domain Extents:
x-coordinate: min (m) = 0.000000e+00, max (m) = 6.400001e+01
y-coordinate: min (m) = -4.538534e+00, max (m) = 6.400000e+01
Volume statistics:
minimum volume (m3): 2.782193e-01
maximum volume (m3): 3.926232e+00
total volume (m3): 1.682930e+03
Face area statistics:
minimum face area (m2): 8.015718e-01
maximum face area (m2): 4.118252e+00
Checking number of nodes per cell.
Checking number of faces per cell.
Checking thread pointers.
Checking number of cells per face.
Checking face cells.
Checking bridge faces.
Checking right-handed cells.
Checking face handedness.
Checking element type consistency.
Checking boundary types:
Checking face pairs.
Checking periodic boundaries.
Checking node count.
Checking nosolve cell count.
Checking nosolve face count.
Checking face children.
Checking cell children.
Checking storage.
Done.
Note: The minimum and maximum values may vary slightly when running on
different platforms. The grid check lists the minimum and maximum x and
y values from the grid, in the default SI units of meters, and reports on a
number of other grid features that are checked. Any errors in the grid would
be reported at this time. In particular, you should always make sure that the
minimum volume is not negative, since FLUENT cannot begin a calculation if
this is the case. To scale the grid to the correct units of inches, the Scale Grid
panel will be used.
c Fluent Inc. January 11, 2005 1-5
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
To ensure the best possible grid quality for the calculation, it is good practice to
smooth a triangular or tetrahedral grid after you read it into FLUENT.
(a) Click the Smooth button and then click Swap repeatedly until FLUENT reports
that zero faces were swapped.
If FLUENT cannot improve the grid by swapping, no faces will be swapped.
(b) Close the panel.
1-6
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
(d) The grid is now sized correctly, and the working units for length have been
set to inches. Close the panel.
Note: Because the default SI units will be used for everything but the length, there
will be no need to change any other units in this problem. The choice of inches
for the unit of length has been made by the actions you have just taken. If you
want to change the working units for length to something other than inches,
say, mm, you would have to visit the Set Units panel in the Define pull-down
menu.
(a) Make sure that all of the surfaces are selected and click Display.
c Fluent Inc. January 11, 2005 1-7
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Grid
FLUENT 6.2 (2d, segregated, lam)
Extra: You can use the right mouse button to check which zone number corresponds to
each boundary. If you click the right mouse button on one of the boundaries in the
graphics window, its zone number, name, and type will be printed in the FLUENT
console window. This feature is especially useful when you have several zones of
the same type and you want to distinguish between them quickly.
1-8
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Step 2: Models
1. Keep the default solver settings.
Define −→ Models −→Solver...
c Fluent Inc. January 11, 2005 1-9
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
1-10
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Step 3: Materials
1. Create a new material called water.
Define −→Materials...
(a) Type the name water in the Name text-entry box.
(b) Enter the values shown in the table below under Properties:
Property Value
density 1000 kg/m3
Cp 4216 J/kg-K
thermal conductivity 0.677 W/m-K
viscosity 8 ×10−4 kg/m-s
c Fluent Inc. January 11, 2005 1-11
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
1-12
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
c Fluent Inc. January 11, 2005 1-13
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
1-14
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
3. Repeat this operation for velocity-inlet-6, using the values in the following table:
velocity specification method components
y velocity 1.0 m/s
x velocity 0 m/s
temperature 313 K
turbulence specification method intensity & hydraulic diameter
turbulence intensity 5%
hydraulic diameter 8 in
c Fluent Inc. January 11, 2005 1-15
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
4. Set the boundary conditions for pressure-outlet-7, as shown in the panel below.
These values will be used in the event that flow enters the domain through this
boundary.
1-16
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
c Fluent Inc. January 11, 2005 1-17
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Step 5: Solution
1. Initialize the flow field using the boundary conditions set at velocity-inlet-5.
Solve −→ Initialize −→Initialize...
(a) Choose velocity-inlet-5 from the Compute From list.
(b) Add a Y Velocity value of 0.2 m/sec throughout the domain.
Note: While an initial X Velocity is an appropriate guess for the horizontal
section, the addition of a Y Velocity will give rise to a better initial guess
throughout the entire elbow.
(c) Click Init and Close the panel.
1-18
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
c Fluent Inc. January 11, 2005 1-19
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
1-20
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Keep the Write Binary Files (default) option on so that a binary file will be written.
c Fluent Inc. January 11, 2005 1-21
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Residuals
continuity
x-velocity
1e+03
y-velocity
energy
k 1e+02
epsilon
1e+01
1e+00
1e-01
1e-02
1e-03
1e-04
1e-05
1e-06
1e-07
0 10 20 30 40 50 60
Iterations
Scaled Residuals
FLUENT 6.2 (2d, segregated, ske)
1-22
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Monitors
monitor-1
310.0000
308.0000
306.0000
304.0000
302.0000
Mass 300.0000
Weighted
298.0000
Average
(k) 296.0000
294.0000
292.0000
290.0000
0 10 20 30 40 50 60
Iteration
c Fluent Inc. January 11, 2005 1-23
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
1-24
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
(a) Select Velocity... and then Velocity Magnitude from the drop-down lists under
Contours of.
(b) Select Filled under Options.
(c) Click Display.
Note: Right-clicking on a point in the domain will cause the value of the corre-
sponding contour to be displayed in the console window.
c Fluent Inc. January 11, 2005 1-25
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
1.24e+00
1.18e+00
1.12e+00
1.05e+00
9.93e-01
9.31e-01
8.69e-01
8.07e-01
7.45e-01
6.82e-01
6.20e-01
5.58e-01
4.96e-01
4.34e-01
3.72e-01
3.10e-01
2.48e-01
1.86e-01
1.24e-01
6.20e-02
0.00e+00
1-26
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
(a) Select Temperature... and Static Temperature in the drop-down lists under
Contours of.
(b) Click Display.
c Fluent Inc. January 11, 2005 1-27
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
3.13e+02
3.12e+02
3.11e+02
3.10e+02
3.09e+02
3.08e+02
3.07e+02
3.06e+02
3.05e+02
3.04e+02
3.03e+02
3.02e+02
3.01e+02
3.00e+02
2.99e+02
2.98e+02
2.97e+02
2.96e+02
2.95e+02
2.94e+02
2.93e+02
1-28
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
c Fluent Inc. January 11, 2005 1-29
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
1.40e+00
1.33e+00
1.27e+00
1.20e+00
1.13e+00
1.06e+00
9.96e-01
9.28e-01
8.61e-01
7.94e-01
7.26e-01
6.59e-01
5.91e-01
5.24e-01
4.56e-01
3.89e-01
3.22e-01
2.54e-01
1.87e-01
1.19e-01
5.19e-02
1.40e+00
1.33e+00
1.27e+00
1.20e+00
1.13e+00
1.06e+00
9.96e-01
9.28e-01
8.61e-01
7.94e-01
7.26e-01
6.59e-01
5.91e-01
5.24e-01
4.56e-01
3.89e-01
3.22e-01
2.54e-01
1.87e-01
1.19e-01
5.19e-02
1-30
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
(a) Select Temperature... and Static Temperature in the drop-down lists under the
Y Axis Function.
(b) Select pressure-outlet-7 from the Surfaces list.
(c) Click Plot.
c Fluent Inc. January 11, 2005 1-31
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
pressure-outlet-7
3.10e+02
3.08e+02
3.06e+02
3.04e+02
Static
Temperature 3.02e+02
(k)
3.00e+02
2.98e+02
2.96e+02
48 50 52 54 56 58 60 62 64
Position (in)
Static Temperature
FLUENT 6.2 (2d, segregated, ske)
1-32
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
5. Make an XY plot of the static pressure on the outer wall of the large pipe, wall-8
(Figure 1.10).
(a) Choose Pressure... and Static Pressure from the Y Axis Function drop-down
lists.
(b) Deselect pressure-outlet-7 and select wall-8 from the Surfaces list.
(c) Change the Plot Direction for X to 0, and the Plot Direction for Y to 1.
With a Plot Direction vector of (0,1), FLUENT will plot static pressure at the
cells of wall-8 as a function of y.
(d) Click Plot.
c Fluent Inc. January 11, 2005 1-33
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
wall-8
1.00e+02
0.00e+00
-1.00e+02
-2.00e+02
Static
Pressure -3.00e+02
(pascal)
-4.00e+02
-5.00e+02
-6.00e+02
10 20 30 40 50 60 70
Position (in)
Static Pressure
FLUENT 6.2 (2d, segregated, ske)
Figure 1.10: Pressure Distribution along the Outside Wall of the Bend
1-34
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
6. Define a custom field function for the dynamic head formula (ρ|V |2 /2).
Define −→ Custom Field Functions...
(a) In the Field Functions drop-down list, select Density and click the Select button.
(b) Click the multiplication button, X.
(c) In the Field Functions drop-down list, select Velocity and Velocity Magnitude
and click Select.
(d) Click y^x to raise the last entry to a power, and click 2 for the power.
(e) Click the divide button, /, and then click 2.
(f) Enter the name dynam-head in the New Function Name text entry box.
(g) Click Define, and then Close the panel.
c Fluent Inc. January 11, 2005 1-35
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
(a) Select Custom Field Functions... in the drop-down list under Contours of.
The function you created, dynam-head, will be shown in the lower drop-down
list.
(b) Click Display, and then Close the panel.
Note: You may need to un-zoom your view after the last vector display, if you have
not already done so.
8. Write the case and data files to save the settings for the custom field function.
File −→ Write −→Case & Data...
1-36
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
7.70e+02
7.31e+02
6.93e+02
6.54e+02
6.16e+02
5.77e+02
5.39e+02
5.00e+02
4.62e+02
4.23e+02
3.85e+02
3.46e+02
3.08e+02
2.69e+02
2.31e+02
1.92e+02
1.54e+02
1.15e+02
7.70e+01
3.85e+01
0.00e+00
Contours of dynam-head
FLUENT 6.2 (2d, segregated, ske)
c Fluent Inc. January 11, 2005 1-37
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
(a) Under Discretization, select Second Order for Pressure, Second Order Upwind for
Momentum, Turbulence Kinetic Energy, Turbulence Dissipation Rate, and Energy.
(b) Keep the default Under-Relaxation Factors settings.
Note: You will have to scroll down both the Discretization and Under-Relaxation
Factors lists to see the Energy options.
1-38
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
To save the convergence history for this set of iterations as a separate output file,
you can change the File Name in the Define Surface Monitor to monitor-2.out.
The solution converges in approximately 50 additional iterations (Figure 1.12). The
convergence history is shown in Figure 1.13.
Residuals
continuity
x-velocity
1e+03
y-velocity
energy
k 1e+02
epsilon
1e+01
1e+00
1e-01
1e-02
1e-03
1e-04
1e-05
1e-06
1e-07
0 20 40 60 80 100 120
Iterations
Scaled Residuals
FLUENT 6.2 (2d, segregated, ske)
Note: Whenever you change the solution control parameters, it is natural to see
the residuals jump.
c Fluent Inc. January 11, 2005 1-39
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Monitors
monitor-1
304.4000
304.3000
304.2000
304.1000
Mass 304.0000
Weighted
Average 303.9000
(k)
303.8000
303.7000
303.6000
50 60 70 80 90 100 110
Iteration
1-40
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
4. Write the case and data files for the second-order solution (elbow2.cas and elbow2.dat).
File −→ Write −→ Case & Data...
(a) Enter the name elbow2 in the Case/Data File box.
(b) Click OK.
The files elbow2.cas and elbow2.dat will be created in your directory.
The thermal spreading after the elbow has been reduced from the earlier prediction
(Figure 1.6).
c Fluent Inc. January 11, 2005 1-41
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
3.13e+02
3.12e+02
3.11e+02
3.10e+02
3.09e+02
3.08e+02
3.07e+02
3.06e+02
3.05e+02
3.04e+02
3.03e+02
3.02e+02
3.01e+02
3.00e+02
2.99e+02
2.98e+02
2.97e+02
2.96e+02
2.95e+02
2.94e+02
2.93e+02
1-42
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
c Fluent Inc. January 11, 2005 1-43
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
2. Plot the temperature gradients that will be used for adaption (Figure 1.16).
(a) Select Adaption... and Adaption Function in the Contours of drop-down lists.
(b) Click Display to see the gradients of temperature, displayed on a cell-by-cell
basis.
3.13e+02
3.12e+02
3.11e+02
3.10e+02
3.09e+02
3.08e+02
3.07e+02
3.06e+02
3.05e+02
3.04e+02
3.03e+02
3.02e+02
3.01e+02
3.00e+02
2.99e+02
2.98e+02
2.97e+02
2.96e+02
2.95e+02
2.94e+02
2.93e+02
Figure 1.15: Temperature Contours for the Second-Order Solution: Cell Values
1-44
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
1.25e-01
1.19e-01
1.13e-01
1.06e-01
1.00e-01
9.39e-02
8.76e-02
8.14e-02
7.51e-02
6.88e-02
6.26e-02
5.63e-02
5.01e-02
4.38e-02
3.75e-02
3.13e-02
2.50e-02
1.88e-02
1.25e-02
6.26e-03
1.42e-14
Note: The quantity Adaption Function defaults to the gradient of the variable
whose Max and Min were most recently computed in the Contours panel.
In this example, the static temperature is the most recent variable to have
its Max and Min computed, since this occurs when the Display button is
pushed. Note that for other applications, gradients of another variable
might be more appropriate for performing the adaption.
3. Plot temperature gradients over a limited range in order to mark cells for adaption
(Figure 1.17).
(a) Under Options, deselect Auto Range so that you can change the minimum
temperature gradient value to be plotted.
The Min temperature gradient is 0 K/m, as shown in the Contours panel.
(b) Enter a new Min value of 0.02.
(c) Click Display.
The colored cells in the figure are in the “high gradient” range, so they will be
the ones targeted for adaption.
c Fluent Inc. January 11, 2005 1-45
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
1.25e-01
1.20e-01
1.15e-01
1.09e-01
1.04e-01
9.89e-02
9.36e-02
8.84e-02
8.31e-02
7.78e-02
7.26e-02
6.73e-02
6.21e-02
5.68e-02
5.15e-02
4.63e-02
4.10e-02
3.58e-02
3.05e-02
2.53e-02
2.00e-02
1-46
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
c Fluent Inc. January 11, 2005 1-47
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Note: There are two different ways to adapt. You can click on Adapt in the
Manage Adaption Registers panel as was just done, or Close this panel and
do the adaption in the Gradient Adaption panel. If you use the Adapt
button in the Gradient Adaption panel, FLUENT will recreate an adaption
register. Therefore, once you have the Manage Adaption Registers panel
open, it saves time to use the Adapt button there.
(i) Close the Manage Adaption Registers and Gradient Adaption panels.
1-48
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
c Fluent Inc. January 11, 2005 1-49
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Grid
FLUENT 6.2 (2d, segregated, ske)
7. Write the final case and data files (elbow3.cas and elbow3.dat) using the prefix
elbow3.
File −→ Write −→ Case & Data...
1-50
c Fluent Inc. January 11, 2005
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
Residuals
continuity
x-velocity
1e+03
y-velocity
energy
k 1e+02
epsilon
1e+01
1e+00
1e-01
1e-02
1e-03
1e-04
1e-05
1e-06
1e-07
0 20 40 60 80 100 120 140 160
Iterations
Scaled Residuals
FLUENT 6.2 (2d, segregated, ske)
Monitors
monitor-1
304.2000
304.1500
304.1000
304.0500
Mass 304.0000
Weighted
Average 303.9500
(k)
303.9000
303.8500
303.8000
100 105 110 115 120 125 130 135 140 145 150
Iteration
c Fluent Inc. January 11, 2005 1-51
Introduction to Using FLUENT: Fluid Flow and Heat Transfer in a Mixing Elbow
8. Examine the filled temperature distribution (using node values) on the revised grid
(Figure 1.22).
Display −→ Contours...
3.13e+02
3.12e+02
3.11e+02
3.10e+02
3.09e+02
3.08e+02
3.07e+02
3.06e+02
3.05e+02
3.04e+02
3.03e+02
3.02e+02
3.01e+02
3.00e+02
2.99e+02
2.98e+02
2.97e+02
2.96e+02
2.95e+02
2.94e+02
2.93e+02
Summary
Comparison of the filled temperature contours for the first solution (using the original
grid and first-order discretization) and the last solution (using an adapted grid and
second-order discretization) clearly indicate that the latter is much less diffusive. While
first-order discretization is the default scheme in FLUENT, it is good practice to use
your first-order solution as a starting guess for a calculation that uses a higher-order
discretization scheme and, optionally, an adapted grid.
Note that in this problem, the flow field is decoupled from temperature since all properties
are constant. For such cases, it is more efficient to compute the flow-field solution first
(i.e., without solving the energy equation) and then solve for energy (i.e., without solving
the flow equations). You will use the Solution Controls panel to turn solution of the
equations on and off during this procedure.
1-52
c Fluent Inc. January 11, 2005
Tutorial 2. Modeling Periodic Flow and Heat Transfer
Introduction
Many industrial applications, such as steam generation in a boiler or air cooling in the
coil of an air conditioner, can be modeled as two-dimensional periodic heat flow. This
tutorial illustrates how to set up and solve a periodic heat transfer problem, given a
pregenerated mesh.
The system that is modeled is a bank of tubes containing a flowing fluid at one temper-
ature that is immersed in a second fluid in cross-flow at a different temperature. Both
fluids are water, and the flow is classified as laminar and steady, with a Reynolds number
of approximately 100. The mass flow rate of the cross-flow is known, and the model is
used to predict the flow and temperature fields that result from convective heat transfer.
Due to symmetry of the tube bank and the periodicity of the flow inherent in the tube
bank geometry, only a portion of the geometry will be modeled in FLUENT, with sym-
metry applied to the outer boundaries. The resulting mesh consists of a periodic module
with symmetry. In the tutorial, the inflow boundary will be redefined as a periodic zone,
and the outflow boundary defined as its shadow.
In this tutorial you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
c Fluent Inc. January 11, 2005 2-1
Modeling Periodic Flow and Heat Transfer
Problem Description
This problem considers a 2D section of a tube bank. A schematic of the problem is
shown in Figure 2.1. The bank consists of uniformly spaced tubes with a diameter of 1
cm, that are staggered in the direction of cross-fluid flow. Their centers are separated by
a distance of 2 cm in the x direction, and 1 cm in the y direction. The bank has a depth
of 1 m.
4 cm
Τ
∞ = 300 K
⋅
m = 0.05 kg/s
{ Τ wall = 400 K
0.5 cm
1 cm
3
ρ = 998.2 kg/m
µ = 0.001003 kg/m-s
c p = 4182 J/kg-K
k = 0.6 W/m-K
Figure 2.1: Schematic of the Problem
Because of the symmetry of the tube bank geometry, only a portion of the domain needs
to be modeled. The computational domain is shown in outline in Figure 2.1. A mass
flow rate of 0.05 kg/s is applied to the inflow boundary of the periodic module. The
temperature of the tube wall (Twall ) is 400 K and the bulk temperature of the cross-flow
water (T∞ ) is 300 K. The properties of water that are used in the model are shown in
Figure 2.1.
2-2
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
Step 1: Grid
1. Read the mesh file, tubebank.msh.
File −→ Read −→Case...
2. Check the grid.
Grid −→Check
FLUENT will perform various checks on the mesh and report the progress in the
console window. Make sure that the minimum volume reported is a positive num-
ber.
3. Scale the grid.
Grid −→Scale...
(a) Under Units Conversion, select cm (centimeters) from the drop-down list to
complete the phrase Grid Was Created In cm.
(b) Click Scale to scale the grid and close the panel.
c Fluent Inc. January 11, 2005 2-3
Modeling Periodic Flow and Heat Transfer
Grid
FLUENT 6.2 (2d, segregated, lam)
2-4
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
Quadrilateral cells are used in the regions surrounding the tube walls, and triangular
cells are used for the rest of the domain, resulting in a hybrid mesh (See Figure 2.2).
The quadrilateral cells provide better resolution of the viscous gradients near the
tube walls. The remainder of the computational domain is conveniently filled with
triangular cells.
Extra: Right-click on one of the boundaries in the graphics window to check which
zone number corresponds to the boundary. The zone number, name, and type
will be printed in the FLUENT console window. This feature is especially useful
when there are several zones of the same type and you want to distinguish
between them.
grid/modify-zones/make-periodic
Periodic zone [()] 9
Shadow zone [()] 12
Rotational periodic? (if no, translational) [yes] no
Create periodic zones? [yes] yes
Auto detect translation vector? [yes] yes
zone 12 deleted
c Fluent Inc. January 11, 2005 2-5
Modeling Periodic Flow and Heat Transfer
Step 2: Models
1. Keep the default solver settings.
Define −→ Models −→Solver...
2-6
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
c Fluent Inc. January 11, 2005 2-7
Modeling Periodic Flow and Heat Transfer
Step 3: Materials
Add liquid water to the list of fluid materials by copying it from the materials database.
(b) Select water-liquid (h2o<l>) from the Fluent Fluid Materials list.
This will display the default settings for water-liquid. You will have to scroll
down the Fluent Fluid Materials list to see the entries.
(c) Click Copy and Close the panel.
The Materials panel will now display the copied information of water.
2-8
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
c Fluent Inc. January 11, 2005 2-9
Modeling Periodic Flow and Heat Transfer
2-10
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
c Fluent Inc. January 11, 2005 2-11
Modeling Periodic Flow and Heat Transfer
2-12
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
Step 5: Solution
1. Set the solution parameters.
Solve −→ Controls −→Solution...
c Fluent Inc. January 11, 2005 2-13
Modeling Periodic Flow and Heat Transfer
2-14
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
(a) Under Initial Values, check that the value for Temperature is set to 300 K.
(b) Click Init and Close the panel.
c Fluent Inc. January 11, 2005 2-15
Modeling Periodic Flow and Heat Transfer
Step 6: Postprocessing
1. Display filled contours of static pressure (Figure 2.3).
Display −→Contours...
2. Change the view to mirror the display across the symmetry planes (Figure 2.4).
Display −→Views...
2-16
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
8.18e-02
7.55e-02
6.92e-02
6.28e-02
5.65e-02
5.02e-02
4.38e-02
3.75e-02
3.12e-02
2.48e-02
1.85e-02
1.22e-02
5.82e-03
-5.13e-04
-6.85e-03
-1.32e-02
-1.95e-02
-2.59e-02
-3.22e-02
-3.85e-02
-4.49e-02
(a) Under Mirror Planes, select all of the symmetry zones by clicking the shaded
icon at the right side.
Note: There are four symmetry zones in the Mirror Planes list because the
top and bottom symmetry planes in the domain are each comprised of
two symmetry zones, one on each side of the tube. It is also possible to
generate the same display shown in Figure 2.4 by selecting just one of the
symmetry zones on the top symmetry plane, and one on the bottom.
(b) Click Apply and Close the panel.
c Fluent Inc. January 11, 2005 2-17
Modeling Periodic Flow and Heat Transfer
Use the left button of your mouse to translate the view so that it is centered
in the window.
8.18e-02
7.55e-02
6.92e-02
6.28e-02
5.65e-02
5.02e-02
4.38e-02
3.75e-02
3.12e-02
2.48e-02
1.85e-02
1.22e-02
5.82e-03
-5.13e-04
-6.85e-03
-1.32e-02
-1.95e-02
-2.59e-02
-3.22e-02
-3.85e-02
-4.49e-02
Note: The pressure contours displayed in Figure 2.4 do not include the linear
pressure gradient computed by the solver. Thus the contours are periodic at
the inflow and outflow boundaries.
2-18
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
(a) In the Contours of drop-down list, select Temperature... and Static Temperature
and click Display (Figure 2.5).
The contours reveal the temperature increase in the fluid due to heat transfer from
the tubes. The hotter fluid is confined to the near-wall and wake regions, while a
narrow stream of cooler fluid is convected through the tube bank.
c Fluent Inc. January 11, 2005 2-19
Modeling Periodic Flow and Heat Transfer
4.00e+02
3.94e+02
3.88e+02
3.82e+02
3.75e+02
3.69e+02
3.63e+02
3.57e+02
3.51e+02
3.45e+02
3.39e+02
3.32e+02
3.26e+02
3.20e+02
3.14e+02
3.08e+02
3.02e+02
2.96e+02
2.89e+02
2.83e+02
2.77e+02
2-20
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
(a) In the Color By drop-down list, select Velocity... and Velocity Magnitude.
(b) Set the Scale to 2 and click Display.
This will enlarge the displayed vectors that are displayed, making it easier to
view the flow patterns.
(c) Zoom in on the upper right portion of the left tube using your middle mouse
button, to get the display shown in Figure 2.6.
This zoomed-in view of the velocity vector plot clearly shows the recirculating
flow behind the tube and the boundary layer development along the tube surface.
5. Plot the temperature profiles at three cross sections of the tube bank.
(a) Create an isosurface on the periodic tube bank at x = 0.01 m (through the
first tube).
Create a surface of constant x coordinate for each the cross sections, x = 0.01,
0.02, and 0.03 m. These isosurfaces correspond to the vertical cross sections
c Fluent Inc. January 11, 2005 2-21
Modeling Periodic Flow and Heat Transfer
1.31e-02
1.25e-02
1.18e-02
1.12e-02
1.05e-02
9.85e-03
9.19e-03
8.53e-03
7.88e-03
7.22e-03
6.56e-03
5.91e-03
5.25e-03
4.60e-03
3.94e-03
3.28e-03
2.63e-03
1.97e-03
1.31e-03
6.58e-04
1.93e-06
through the first tube, halfway between the two tubes, and through the second
tube.
Surface −→Iso-Surface...
i. In the Surface of Constant drop-down lists, select Grid... and X-Coordinate.
ii. Enter x=0.01m under New Surface Name.
iii. Enter 0.01 for Iso-Values and click Create.
iv. Follow the same procedure to create surfaces at x = 0.02 m (halfway
between the two tubes) and x = 0.03 m (through the middle of the second
tube).
2-22
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
i. Change the Plot Direction for X to 0, and the Plot Direction for Y to 1.
With a Plot Direction vector of (0,1), FLUENT will plot the selected vari-
c Fluent Inc. January 11, 2005 2-23
Modeling Periodic Flow and Heat Transfer
2-24
c Fluent Inc. January 11, 2005
Modeling Periodic Flow and Heat Transfer
x=0.01m
x=0.02m
x=0.03m
4.00e+02
3.80e+02
3.60e+02
3.40e+02
Static
Temperature 3.20e+02
(k)
3.00e+02
2.80e+02
2.60e+02
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007 0.008 0.009 0.01
Position (m)
Static Temperature
FLUENT 6.2 (2d, segregated, lam)
c Fluent Inc. January 11, 2005 2-25
Modeling Periodic Flow and Heat Transfer
Summary
In this tutorial, periodic flow and heat transfer in a staggered tube bank were modeled
in FLUENT. The model was set up assuming a known mass flow through the tube bank
and constant wall temperatures. Due to the periodic nature of the flow and symmetry of
the geometry, only a small piece of the full geometry was modeled. In addition, the tube
bank configuration lent itself to the use of a hybrid mesh with quadrilateral cells around
the tubes and triangles elsewhere.
The Periodicity Conditions panel makes it easy to run this type of model over a variety
of operating conditions. For example, different flow rates (and hence different Reynolds
numbers) can be studied, or a different inlet bulk temperature can be imposed. The
resulting solution can then be examined to extract the pressure drop per tube row and
overall Nusselt number for a range of Reynolds numbers.
2-26
c Fluent Inc. January 11, 2005
Tutorial 3. Modeling External Compressible Flow
Introduction
The purpose of this tutorial is to compute the turbulent flow past a transonic airfoil at
a non-zero angle of attack. You will use the Spalart-Allmaras turbulence model.
In this tutorial you will learn how to:
• Model compressible flow (using the ideal gas law for density)
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
c Fluent Inc. January 11, 2005 3-1
Modeling External Compressible Flow
Problem Description
The problem considers the flow around an airfoil at an angle of attack α = 4◦ and a free
stream Mach number of 0.8 (M∞ = 0.8). This flow is transonic, and has a fairly strong
shock near the mid-chord (x/c = 0.45) on the upper (suction) side. The chord length is
1 m. The geometry of the airfoil is shown in Figure 3.1.
α = 4°
M∞= 0.8
1m
2. Unzip external_compressible.zip.
airfoil.msh can be found in the /external compressible folder created after
unzipping the file.
3-2
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
Step 1: Grid
1. Read the grid file airfoil.msh.
File −→ Read −→Case...
As FLUENT reads the grid file, it will report its progress in the console window.
(a) Display the grid with the default settings (Figure 3.2).
(b) Use the middle mouse button to zoom in on the image so you can see the mesh
near the airfoil (Figure 3.3).
Quadrilateral cells were used for this simple geometry because they can be
stretched easily to account for different flow gradient magnitudes in different
directions. In the present case, the gradients normal to the airfoil wall are
much greater than those tangent to the airfoil, except near the leading and
trailing edges and in the vicinity of the shock expected on the upper surface.
Consequently, the cells nearest the surface have very high aspect ratios. For
c Fluent Inc. January 11, 2005 3-3
Modeling External Compressible Flow
Grid
FLUENT 6.2 (2d, segregated, lam)
Grid
FLUENT 6.2 (2d, segregated, lam)
3-4
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
geometries that are more difficult to mesh, it may be easier to create a hybrid
mesh comprised of quadrilateral and triangular cells.
A parabola was chosen to represent the far-field boundary because it has no
discontinuities in slope, enabling the construction of a smooth mesh in the
interior of the domain.
Extra: You can use the right mouse button to check which zone number cor-
responds to each boundary. If you click the right mouse button on one
of the boundaries in the graphics window, its zone number, name, and
type will be printed in the FLUENT console window. This feature is espe-
cially useful when you have several zones of the same type and you want
to distinguish between them quickly.
c Fluent Inc. January 11, 2005 3-5
Modeling External Compressible Flow
Step 2: Models
1. Select the Coupled, Implicit solver.
Define −→ Models −→Solver...
The coupled solver is recommended when dealing with applications involving high-
speed aerodynamics with shocks. The implicit solver will generally converge much
faster than the explicit solver, but will use more memory. For this 2D case, memory
is not an issue.
3-6
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
(a) Select the Spalart-Allmaras model and retain the default options and constants.
The Spalart-Allmaras model is a relatively simple one-equation model that solves a mod-
eled transport equation for the kinematic eddy (turbulent) 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 specifically for aerospace applications involving wall-bounded flows and has been
shown to give good results for boundary layers subjected to adverse pressure gradients.
c Fluent Inc. January 11, 2005 3-7
Modeling External Compressible Flow
Step 3: Materials
The default Fluid Material is air, which is the working fluid in this problem. The default
settings need to be modified to account for compressibility and variations of the thermo-
physical properties with temperature.
Define −→Materials...
3-8
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
(a) Click OK to accept the default Three Coefficient Method and parameters.
The Sutherland law for viscosity is well suited for high-speed compressible flows.
3. Click Change/Create in the Materials panel to save these settings, and then close
the panel.
Note: While Density and Viscosity have been made temperature-dependent, Cp and Ther-
mal Conductivity have been left constant. For high-speed compressible flows, thermal
dependency of the physical properties is generally recommended. For simplicity,
Thermal Conductivity and Cp are assumed to be constant in this tutorial.
c Fluent Inc. January 11, 2005 3-9
Modeling External Compressible Flow
When using the coupled solver for flows with Mach numbers greater than 0.1, an operating
pressure of zero is recommended.
See Section 8.13 of the User’s Guide for more information on how to set the operating
pressure.
3-10
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
For external flows, you should choose a viscosity ratio between 1 and 10.
Note: The X- and Y-Component of Flow Direction are set as above because of the 4◦ angle
of attack: cos 4◦ ≈ 0.997564 and sin 4◦ ≈ 0.069756.
c Fluent Inc. January 11, 2005 3-11
Modeling External Compressible Flow
Step 6: Solution
1. Set the solution controls.
Solve −→ Controls −→Solution...
(a) Set the Under-Relaxation Factor for Modified Turbulent Viscosity to 0.9.
Larger (i.e., closer to 1) under-relaxation factors will generally result in faster
convergence. However, instability can arise that may need to be eliminated by
decreasing the under-relaxation factors.
(b) Under Solver Parameters, set the Courant Number to 5.
(c) Under Discretization, select Second Order Upwind for Modified Turbulent Vis-
cosity.
The second-order scheme will resolve the boundary layer and shock more ac-
curately than the first-order scheme.
3-12
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
c Fluent Inc. January 11, 2005 3-13
Modeling External Compressible Flow
This will be sufficient to see where the shock wave is, and the fluctuations of the
solution will have diminished significantly.
5. Increase the Courant number.
Solve −→ Controls −→Solution...
Under Solver Parameters, set the Courant Number to 20.
The solution will generally converge faster for larger Courant numbers, unless the
integration scheme becomes unstable. Since you have performed some initial iter-
ations, and the solution is stable, you can try increasing the Courant number to
speed up convergence. If the residuals increase without bound, or you get a floating
point exception, you will need to decrease the Courant number, read in the previous
data file, and try again.
6. Turn on monitors for lift, drag, and moment coefficients.
Solve −→ Monitors −→Force...
3-14
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
7. Set the reference values that are used to compute the lift, drag, and moment coef-
ficients.
The reference values are used to non-dimensionalize the forces and moments acting
on the airfoil. The dimensionless forces and moments are the lift, drag, and moment
coefficients.
Report −→Reference Values...
(a) In the Compute From drop-down list, select pressure-far-field-1.
FLUENT will update the Reference Values based on the boundary conditions at
the far-field boundary.
c Fluent Inc. January 11, 2005 3-15
Modeling External Compressible Flow
8. Define a monitor for tracking the skin friction coefficient value just upstream of the
shock wave.
(a) Display filled contours of pressure overlaid with the grid.
Display −→Contours...
i. Turn on Filled.
ii. Select Draw Grid.
This will open the Grid Display panel.
iii. Close the Grid Display panel, since there are no changes to be made here.
iv. Click Display in the Contours panel.
v. Zoom in on the airfoil (Figure 3.4).
1.55e+05
1.49e+05
1.44e+05
1.39e+05
1.33e+05
1.28e+05
1.23e+05
1.18e+05
1.12e+05
1.07e+05
1.02e+05
9.65e+04
9.12e+04
8.59e+04
8.06e+04
7.53e+04
7.00e+04
6.48e+04
5.95e+04
5.42e+04
4.89e+04
The shock wave is clearly visible on the upper surface of the airfoil, where
the pressure first jumps to a higher value.
vi. Zoom in on the shock wave, until individual cells adjacent to the upper
surface (wall-top boundary) are visible (Figure 3.5).
3-16
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
1.55e+05
1.49e+05
1.44e+05
1.39e+05
1.33e+05
1.28e+05
1.23e+05
1.18e+05
1.12e+05
1.07e+05
1.02e+05
9.65e+04
9.12e+04
8.59e+04
8.06e+04
7.53e+04
7.00e+04
6.48e+04
5.95e+04
5.42e+04
4.89e+04
The zoomed-in region contains cells just upstream of the shock wave that are
adjacent to the upper surface of the airfoil. In the following step, you will
create a point surface inside a wall-adjacent cell, to be used for the skin friction
coefficient monitor.
(b) Create a point surface just upstream of the shock wave.
Surface −→Point...
i. Under Coordinates, enter 0.53 for x0, and 0.051 for y0.
ii. Click on Create to create the point surface (point-4).
c Fluent Inc. January 11, 2005 3-17
Modeling External Compressible Flow
1.55e+05
1.49e+05
1.44e+05
1.39e+05
1.33e+05
1.28e+05
1.23e+05
1.18e+05
1.12e+05
1.07e+05
1.02e+05
9.65e+04
9.12e+04
8.59e+04
8.06e+04
7.53e+04
7.00e+04
6.48e+04
5.95e+04
5.42e+04
4.89e+04
Note: Here, you have entered the exact coordinates of the point surface so
that your convergence history will match the plots and description in this
tutorial. In general, however, you will not know the exact coordinates in
advance, so you will need to select the desired location in the graphics
window as follows:
i. Click Select Point With Mouse.
ii. Move the mouse to a point located anywhere inside one of the cells
adjacent to the upper surface (wall-top boundary), in the vicinity of
the shock wave. (See Figure 3.6.)
iii. Click the right mouse button.
iv. Click Create to create the point surface.
3-18
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
iii. Select Wall Fluxes... and Skin Friction Coefficient under Report of.
iv. Select point-4 in the Surfaces list.
v. In the Report Type drop-down list, select Vertex Average.
c Fluent Inc. January 11, 2005 3-19
Modeling External Compressible Flow
Monitors
monitor-1
0.0025
0.0023
0.0020
0.0018
0.0015
Average 0.0013
of
0.0010
Surface
Vertex 0.0008
Values
0.0005
0.0003
0.0000
80 100 120 140 160 180 200 220 240
Iteration
Figure 3.7: Skin Friction Convergence History for the Initial Calculation
Note: After about 100 iterations, the residual criteria are satisfied and FLUENT
stops iterating. Since the skin friction monitor indicates that the skin friction
coefficient at point-4 has not converged (Figure 3.7), you will need to decrease
the convergence criterion for the modified turbulent viscosity and continue it-
erating.
3-20
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
11. Reduce the convergence criterion for the modified turbulent viscosity equation.
Solve −→ Monitors −→Residual...
(a) Set the Convergence Criterion for nut to 1e-7 and click OK.
nut stands for νt . This is the residual for the modified turbulent viscosity that
the Spalart-Allmaras model solves for.
c Fluent Inc. January 11, 2005 3-21
Modeling External Compressible Flow
Monitors
monitor-1
0.0025
0.0023
0.0020
0.0018
0.0015
Average 0.0013
of
0.0010
Surface
Vertex 0.0008
Values
0.0005
0.0003
0.0000
100 200 300 400 500 600 700 800
Iteration
0.0800
0.0750
0.0700
Cd 0.0650
0.0600
0.0550
0.0500
100 200 300 400 500 600 700 800
Iterations
3-22
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
0.6000
0.5750
0.5500
0.5250
0.5000
0.4750
Cl
0.4500
0.4250
0.4000
0.3750
0.3500
0.3250
100 200 300 400 500 600 700 800
Iterations
0.0700
0.0600
0.0500
0.0400
0.0300
Cm
0.0200
0.0100
0.0000
-0.0100
-0.0200
100 200 300 400 500 600 700 800
Iterations
c Fluent Inc. January 11, 2005 3-23
Modeling External Compressible Flow
Step 7: Postprocessing
1. Plot the y + distribution on the airfoil.
Plot −→XY Plot...
3-24
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
wall-bottom
wall-top
9.00e+01
8.00e+01
7.00e+01
6.00e+01
5.00e+01
Wall
Yplus 4.00e+01
3.00e+01
2.00e+01
1.00e+01
0.00e+00
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
Wall Yplus
FLUENT 6.2 (2d, coupled imp, S-A)
4. Plot the x component of wall shear stress on the airfoil surface (Figure 3.15).
Plot −→XY Plot...
(a) Under Y Axis Function, choose Wall Fluxes... and X-Wall Shear Stress from the
drop-down lists.
c Fluent Inc. January 11, 2005 3-25
Modeling External Compressible Flow
1.44e+00
1.37e+00
1.30e+00
1.22e+00
1.15e+00
1.08e+00
1.01e+00
9.39e-01
8.68e-01
7.96e-01
7.25e-01
6.53e-01
5.82e-01
5.10e-01
4.39e-01
3.68e-01
2.96e-01
2.25e-01
1.53e-01
8.17e-02
1.02e-02
wall-bottom
wall-top
1.25e+00
1.00e+00
7.50e-01
5.00e-01
2.50e-01
Pressure 0.00e+00
Coefficient
-2.50e-01
-5.00e-01
-7.50e-01
-1.00e+00
-1.25e+00
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
Pressure Coefficient
FLUENT 6.2 (2d, coupled imp, S-A)
3-26
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
wall-bottom
wall-top
2.25e+02
2.00e+02
1.75e+02
1.50e+02
1.25e+02
X-Wall 1.00e+02
Shear
7.50e+01
Stress
(pascal) 5.00e+01
2.50e+01
0.00e+00
-2.50e+01
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
The large, adverse pressure gradient induced by the shock causes the boundary layer
to separate. The point of separation is where the wall shear stress vanishes. Flow
reversal is indicated here by negative values of the x component of the wall shear
stress.
c Fluent Inc. January 11, 2005 3-27
Modeling External Compressible Flow
4.46e+02
4.20e+02
3.94e+02
3.68e+02
3.42e+02
3.16e+02
2.90e+02
2.64e+02
2.38e+02
2.12e+02
1.86e+02
1.60e+02
1.34e+02
1.08e+02
8.17e+01
5.56e+01
2.96e+01
3.57e+00
-2.25e+01
-4.85e+01
-7.45e+01
4.48e+02
4.25e+02
4.03e+02
3.81e+02
3.58e+02
3.36e+02
3.14e+02
2.91e+02
2.69e+02
2.47e+02
2.24e+02
2.02e+02
1.80e+02
1.57e+02
1.35e+02
1.13e+02
9.05e+01
6.82e+01
4.59e+01
2.36e+01
1.27e+00
Figure 3.17: Plot of Velocity Vectors Near Upper Wall, Behind Shock
3-28
c Fluent Inc. January 11, 2005
Modeling External Compressible Flow
Summary
This tutorial demonstrated how to set up and solve an external aerodynamics problem us-
ing the Spalart-Allmaras turbulence model. It showed how to monitor convergence using
residual, force, and surface monitors, and demonstrated the use of several postprocessing
tools to examine the flow phenomena associated with a shock wave.
c Fluent Inc. January 11, 2005 3-29
Modeling External Compressible Flow
3-30
c Fluent Inc. January 11, 2005
Tutorial 4. Modeling Unsteady Compressible Flow
Introduction
In this tutorial, FLUENT’s coupled implicit solver is used to predict the time-dependent
flow through a two-dimensional nozzle. As an initial condition for the transient problem,
a steady-state solution is generated to provide the initial values for the mass flow rate at
the nozzle exit.
In this tutorial you will learn how to:
• Use dynamic mesh adaption for both steady-state and transient flows
• Create an animation of the unsteady flow using FLUENT’s unsteady solution ani-
mation feature
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The geometry to be considered in this tutorial is shown in Figure 4.1. Flow through a
simple nozzle is simulated as a 2D planar model. The nozzle has an inlet height of 0.2 m,
and the nozzle contours have a sinusoidal shape that produces a 10% reduction in flow
area. Due to symmetry, only half of the nozzle is modeled.
c Fluent Inc. January 11, 2005 4-1
Modeling Unsteady Compressible Flow
plane of symmetry
0.2 m p (t )
exit
2. Unzip unsteady_compressible.zip.
nozzle.msh and pexit.c can be found in the /unsteady compressible folder cre-
ated after unzipping the file.
4-2
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
Step 1: Grid
1. Read in the mesh file nozzle.msh.
File −→ Read −→Case...
To make the view more realistic, you will need to mirror it across the centerline.
c Fluent Inc. January 11, 2005 4-3
Modeling Unsteady Compressible Flow
Grid
FLUENT 6.2 (2d, segregated, lam)
4-4
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
Step 2: Units
1. For convenience, define new units for pressure.
The pressure for this problem is specified in atm, which is not the default unit. You
will need to redefine the pressure unit as atm.
Define −→Units...
c Fluent Inc. January 11, 2005 4-5
Modeling Unsteady Compressible Flow
Step 3: Models
1. Select the coupled implicit solver.
The coupled implicit solver is the solver of choice for compressible, transonic flows
without significant regions of low-speed flow. In cases with significant low-speed
flow regions, the segregated solver is preferred. Also, for transient cases with trav-
eling shocks, the coupled explicit solver with explicit time stepping may be the most
efficient.
Define −→ Models −→Solver...
Note: Initially, solve for the steady flow through the nozzle. Later, after obtaining
the steady flow as a starting point, this panel to enable an unsteady calculation.
4-6
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
c Fluent Inc. January 11, 2005 4-7
Modeling Unsteady Compressible Flow
Step 4: Materials
1. Set the properties for air, the default fluid material.
Define −→Materials...
4-8
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
Here, the operating pressure is set to zero and boundary condition inputs for pressure
will be defined in terms of absolute pressures. Boundary condition inputs should
always be relative to the value used for operating pressure.
c Fluent Inc. January 11, 2005 4-9
Modeling Unsteady Compressible Flow
4-10
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
c Fluent Inc. January 11, 2005 4-11
Modeling Unsteady Compressible Flow
4-12
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
(a) Under Discretization, select Second Order Upwind for Modified Turbulent Vis-
cosity.
Second-order discretization provides optimum accuracy.
We will now activate dynamic adaption. Our purpose here is to have the solver
periodically refine the mesh in the vicinity of the shocks as the iterations progress.
We identify the shocks by looking for large gradients of the pressure.
c Fluent Inc. January 11, 2005 4-13
Modeling Unsteady Compressible Flow
4-14
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
(f) Turn on the Dynamic option under Dynamic and set the Interval to 100.
For steady-state flows, it is sufficient to only seldomly adapt the mesh—in
this case an interval of 100 iterations is chosen. For time-dependent flows, a
considerably smaller interval must be used.
(g) Click Apply to store the information.
(h) Click on Controls... to modify the adaption controls.
c Fluent Inc. January 11, 2005 4-15
Modeling Unsteady Compressible Flow
4-16
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
c Fluent Inc. January 11, 2005 4-17
Modeling Unsteady Compressible Flow
The mass flow rate history is shown in Figure 4.3. It shows that the solution is
converged after around 1800 iterations.
-13.5000
-14.0000
-14.5000
-15.0000
Mass
Flow -15.5000
Rate
(kg/s)
-16.0000
-16.5000
-17.0000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Iteration
4-18
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
Although the mass flow rate history indicates that the solution is con-
! verged, you should also check the mass fluxes through the domain to ensure
that mass is being conserved.
The net mass imbalance should be a small fraction (say, 0.5%) of the total
!
flux through the system. If a significant imbalance occurs, you should
decrease your residual tolerances by at least an order of magnitude and
continue iterating.
c Fluent Inc. January 11, 2005 4-19
Modeling Unsteady Compressible Flow
4-20
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
3.45e+02
3.27e+02
3.10e+02
2.93e+02
2.76e+02
2.59e+02
2.41e+02
2.24e+02
2.07e+02
1.90e+02
1.73e+02
1.55e+02
1.38e+02
1.21e+02
1.04e+02
8.64e+01
6.92e+01
5.20e+01
3.48e+01
1.76e+01
3.72e-01
c Fluent Inc. January 11, 2005 4-21
Modeling Unsteady Compressible Flow
11. Display the steady flow contours of static pressure (Figure 4.5).
Display −→Contours...
4-22
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
7.84e-01
7.66e-01
7.47e-01
7.28e-01
7.09e-01
6.90e-01
6.71e-01
6.53e-01
6.34e-01
6.15e-01
5.96e-01
5.77e-01
5.59e-01
5.40e-01
5.21e-01
5.02e-01
4.83e-01
4.65e-01
4.46e-01
4.27e-01
4.08e-01
c Fluent Inc. January 11, 2005 4-23
Modeling Unsteady Compressible Flow
4-24
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
where
ω = circular frequency of unsteady pressure (rad/s)
pexit = mean exit pressure (atm)
In this case, ω = 2200 rad/s, and pexit = 0.7369 atm.
A user-defined function (pexit.c) has been written to define the equation (Equa-
tion 4.1) required for the pressure profile.
Note: To input the value of Equation 4.1 in the correct units, the function pexit.c
has been multiplied by a factor of 101325 to convert from the chosen pressure
unit (atm) to the SI unit required by FLUENT (Pa). This will not affect the
displayed results.
See the separate UDF Manual for details about user-defined functions..
(a) Read in the user-defined function.
Define −→ User-Defined −→ Functions −→Interpreted...
c Fluent Inc. January 11, 2005 4-25
Modeling Unsteady Compressible Flow
i. Select udf unsteady pressure (the user-defined function) in the Gauge Pres-
sure drop-down list.
4-26
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
c Fluent Inc. January 11, 2005 4-27
Modeling Unsteady Compressible Flow
2. Modify the plotting of the mass flow rate at the nozzle exit.
Because each time step requires 30 iterations, a smoother plot will be generated by
plotting at every time step.
Solve −→ Monitors −→Surface...
(a) For monitor-1, select Time Step in the drop-down list under Every.
(b) Click Define... to modify the surface monitor parameters.
i. In the Define Surface Monitor panel, change the File Name to noz uns.out.
ii. In the X Axis drop-down list, select Time Step.
iii. Click OK.
(c) Click OK in the Surface Monitors panel.
4-28
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
Calculation of 600 time steps will require significant CPU resources (about
!
three hours on a single CPU with a 2.6 GHz clock speed). Instead of
calculating, you can read the data file saved after the iterations have been
completed:
noz uns.dat
(The data file is available in the same directory where you found the mesh
and UDF files.)
By requesting 600 time steps, you are asking FLUENT to compute six pressure
cycles. The mass flow rate history is shown in Figure 4.6.
Monitors
monitor-1
-4.0000
-6.0000
-8.0000
-10.0000
Mass -12.0000
Flow
Rate -14.0000
(kg/s)
-16.0000
-18.0000
-20.0000
0 100 200 300 400 500 600
Time Step
c Fluent Inc. January 11, 2005 4-29
Modeling Unsteady Compressible Flow
(a) Set the Autosave Case File Frequency and Autosave Data File Frequency to 10.
(b) In the Filename field, enter noz anim.
(c) Click OK.
If you have constraints on the disc space, you can restrict the number of files
saved by FLUENT using the Maximum Number of Files at Any Instance field.
After saving the specified number of files, FLUENT will overwrite the earliest
existing file. The default value of zero will save all the files.
When FLUENT saves a file, it will append the time step value to the file
name prefix (noz anim). The standard extensions (.cas and .dat) will also
be appended. This will yield file names of the form noz anim0640.cas and
noz anim0640.dat, where 0640 is the time step number.
Optionally, you can add the extension .gz to the end of the file name (e.g.,
noz anim.gz), which will instruct FLUENT to save the case and data files in
compressed format, yielding file names of the form noz anim0640.cas.gz.
4-30
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
2. Create animation sequences for the nozzle pressure and Mach number contour plots.
Solve −→ Animate −→Define...
c Fluent Inc. January 11, 2005 4-31
Modeling Unsteady Compressible Flow
4-32
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
i. In the Contours panel, keep the default selections of Pressure... and Static
Pressure.
ii. Make sure that Filled is selected under Options, and deselect Auto Range.
iii. Enter 0.25 under Min and 1.25 under Max.
This will set a fixed range for the contour plot and subsequent animation.
iv. In the Surfaces list, select all of the surfaces.
v. Click Display.
Figure 4.7 shows the contours of static pressure in the nozzle after 600 time
steps.
(e) Click OK in the Animation Sequence panel.
c Fluent Inc. January 11, 2005 4-33
Modeling Unsteady Compressible Flow
1.25e+00
1.20e+00
1.15e+00
1.10e+00
1.05e+00
1.00e+00
9.50e-01
9.00e-01
8.50e-01
8.00e-01
7.50e-01
7.00e-01
6.50e-01
6.00e-01
5.50e-01
5.00e-01
4.50e-01
4.00e-01
3.50e-01
3.00e-01
2.50e-01
4-34
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
1.30e+00
1.23e+00
1.17e+00
1.11e+00
1.04e+00
9.75e-01
9.10e-01
8.45e-01
7.80e-01
7.15e-01
6.50e-01
5.85e-01
5.20e-01
4.55e-01
3.90e-01
3.25e-01
2.60e-01
1.95e-01
1.30e-01
6.50e-02
0.00e+00
c Fluent Inc. January 11, 2005 4-35
Modeling Unsteady Compressible Flow
When the calculation finishes, you will have ten pairs of case and data files and
there will be 100 pairs of contour plots stored in memory. In the next few steps,
you will play back the animation sequences and examine the results at several time
steps after reading in pairs of newly-saved case and data files.
4-36
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
c Fluent Inc. January 11, 2005 4-37
Modeling Unsteady Compressible Flow
8. Repeat steps 6 and 7, selecting the appropriate active window and sequence name
for the Mach number contours.
Examples of Mach number contours at t = 0.01799 s and t = 0.0191 s are shown
in Figures 4.11 and 4.12.
4-38
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
1.25e+00
1.20e+00
1.15e+00
1.10e+00
1.05e+00
1.00e+00
9.50e-01
9.00e-01
8.50e-01
8.00e-01
7.50e-01
7.00e-01
6.50e-01
6.00e-01
5.50e-01
5.00e-01
4.50e-01
4.00e-01
3.50e-01
3.00e-01
2.50e-01
1.25e+00
1.20e+00
1.15e+00
1.10e+00
1.05e+00
1.00e+00
9.50e-01
9.00e-01
8.50e-01
8.00e-01
7.50e-01
7.00e-01
6.50e-01
6.00e-01
5.50e-01
5.00e-01
4.50e-01
4.00e-01
3.50e-01
3.00e-01
2.50e-01
c Fluent Inc. January 11, 2005 4-39
Modeling Unsteady Compressible Flow
1.30e+00
1.23e+00
1.17e+00
1.11e+00
1.04e+00
9.75e-01
9.10e-01
8.45e-01
7.80e-01
7.15e-01
6.50e-01
5.85e-01
5.20e-01
4.55e-01
3.90e-01
3.25e-01
2.60e-01
1.95e-01
1.30e-01
6.50e-02
0.00e+00
1.30e+00
1.23e+00
1.17e+00
1.11e+00
1.04e+00
9.75e-01
9.10e-01
8.45e-01
7.80e-01
7.15e-01
6.50e-01
5.85e-01
5.20e-01
4.55e-01
3.90e-01
3.25e-01
2.60e-01
1.95e-01
1.30e-01
6.50e-02
0.00e+00
4-40
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
Extra: FLUENT gives you the option of exporting an animation as an MPEG file
or as a series of files in any of the hardcopy formats available in the Graphics
Hardcopy panel (including TIFF and PostScript).
To save an MPEG file, select MPEG from the Write/Record Format drop-down
list in the Playback panel and then click the Write button. The MPEG file will
be saved in your working directory. You can view the MPEG movie using an
MPEG player (e.g., Windows Media Player or another MPEG movie player).
To save a series of TIFF, PostScript, or other hardcopy files, select Hardcopy
Frames in the Write/Record Format drop-down list in the Playback panel. Click
on the Hardcopy Options... button to open the Graphics Hardcopy panel and set
the appropriate parameters for saving the hardcopy files. Click Apply in the
Graphics Hardcopy panel to save your modified settings. In the Playback panel,
click the Write button. FLUENT will replay the animation, saving each frame
to a separate file in your working directory.
If you want to view the solution animation in a later FLUENT session, you
can select Animation Frames as the Write/Record Format and click Write.
Since the solution animation was stored in memory, it will be lost if you
! exit FLUENT without saving it to one of the formats described above. Note
that only the animation-frame format can be read back into the Playback
panel for display in a later FLUENT session.
c Fluent Inc. January 11, 2005 4-41
Modeling Unsteady Compressible Flow
10. Repeat step 9 using case and data files saved for other time steps of interest.
4-42
c Fluent Inc. January 11, 2005
Modeling Unsteady Compressible Flow
2.42e+02
2.30e+02
2.18e+02
2.06e+02
1.94e+02
1.82e+02
1.70e+02
1.58e+02
1.46e+02
1.34e+02
1.22e+02
1.10e+02
9.75e+01
8.54e+01
7.34e+01
6.13e+01
4.92e+01
3.71e+01
2.50e+01
1.30e+01
8.97e-01
Summary
In this tutorial, you modeled the transient flow of air through a nozzle. You learned how
to generate a steady-state solution as an initial condition for the unsteady case, and how
to set solution parameters for implicit time-stepping.
You also learned how to manage the file saving and graphical postprocessing for time-
dependent flows, using file autosaving to automatically save solution information as the
transient calculation proceeds.
Finally, you learned how to use FLUENT’s solution animation tool to create animations
of transient data, and how to view the animations using the playback feature.
c Fluent Inc. January 11, 2005 4-43
Modeling Unsteady Compressible Flow
4-44
c Fluent Inc. January 11, 2005
Tutorial 5. Modeling Radiation and Natural Convection
Introduction
In this tutorial, combined radiation and natural convection are solved in a two-dimensional
square box on a mesh consisting of quadrilateral elements.
In this tutorial you will learn how to:
• Use the radiation models in FLUENT (Rosseland, P-1, DTRM, discrete ordinates
(DO), and surface-to-surface (S2S)) and understand their ranges of application
• Set the boundary conditions for a heat transfer problem involving natural convec-
tion and radiation
• Display velocity vectors and contours of stream function and temperature for flow
visualization
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The problem to be considered is shown schematically in Figure 5.1. A square box of side
L has a hot right wall at T = 2000 K, a cold left wall at T = 1000 K, and adiabatic
top and bottom walls. Gravity points downwards. A buoyant flow develops because of
thermally-induced density gradients. The medium contained in the box is assumed to be
absorbing and emitting, so that the radiant exchange between the walls is attenuated by
absorption and augmented by emission in the medium. All walls are black. The objective
is to compute the flow and temperature patterns in the box, as well as the wall heat flux,
c Fluent Inc. January 11, 2005 5-1
Modeling Radiation and Natural Convection
using the radiation models available in FLUENT, and to compare their performance for
different values of the optical thickness aL.
The working fluid has a Prandtl number of approximately 0.71, and the Rayleigh number
based on L is 5 × 105 . This means the flow is inherently laminar. The Boussinesq
assumption is used to model buoyancy. The Planck number k/(4σLT03 ) is 0.02, and
measures the relative importance of conduction to radiation; here T0 = (Th + Tc )/2.
Three values for the optical thickness are considered: aL = 0, aL = 0.2, and aL = 5.
Note that the values of physical properties and operating conditions (e.g., gravitational
acceleration) have been adjusted to yield the desired Prandtl, Rayleigh, and Planck
numbers.
Adiabatic
ρ = 1000 kg/m3
4
cp= 1.1030x10 J/kgK
k = 15.309 W/mK
-3
µ = 10 kg/ms
Tc= 1000 K
T = 2000 K
-5
h
g ➢ β = 10 1/K
➢
-5
g = -6.96 x 10 m/s2
a = 0, 0.2, 5 1/m
y L=1m
5
x Ra = 5 x 10
Pr = 0.71
Pl = 0.02
L τ = 0.2, 5
2. Unzip radiation_natural_convection.zip.
rad.msh can be found in the /radiation natural convection folder created after
unzipping the file.
5-2
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
Step 1: Grid
1. Read the mesh file rad.msh.
File −→ Read −→Case...
As the mesh is read in, messages will appear in the console window reporting the
progress of the reading. The mesh size will be reported as 2500 cells.
Note: All the walls are currently contained in a single wall zone, wall-4. You will
need to separate them out into four different walls so that you can specify
different boundary conditions for each wall.
c Fluent Inc. January 11, 2005 5-3
Modeling Radiation and Natural Convection
Grid
FLUENT 6.2 (2d, segregated, lam)
(a) Select the Angle separation method (the default) under Options.
(b) Select wall-4 in the Zones list.
(c) Specify 89◦ as the significant Angle.
(d) Click on the Separate button.
Faces with normal vectors that differ by more than 89◦ will be placed in separate
zones. Since the four wall zones are perpendicular (angle = 90◦ ), wall-4 will be
separated into four zones.
5-4
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
c Fluent Inc. January 11, 2005 5-5
Modeling Radiation and Natural Convection
Step 2: Models
As discussed earlier, in this tutorial you will enable each radiation model in turn, obtain
a solution, and postprocess the results. You will start with the Rosseland model, then use
the P-1 model, the discrete transfer radiation model (DTRM), and the discrete ordinates
(DO) model. At the end of the tutorial, you will use the surface-to-surface (S2S) model.
5-6
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
When you click OK in the Radiation Model panel, FLUENT will present an Infor-
mation dialog box telling you that new material properties have been added for the
radiation model. You will be setting properties later, so you can simply click OK in
the dialog box to acknowledge this information.
Note: FLUENT will automatically enable the energy calculation when you enable
a radiation model, so you need not visit the Energy panel.
3. Add the effect of gravity on the model.
Define −→Operating Conditions...
c Fluent Inc. January 11, 2005 5-7
Modeling Radiation and Natural Convection
5-8
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
Step 3: Materials
The default fluid material is air, which is the working fluid in this problem. However,
since you are working with a fictitious fluid whose properties have been adjusted to give
the desired values of the dimensionless parameters, you must change the default properties
for air. You will use an optical thickness aL of 0.2 for this calculation. (Since L = 1, the
absorption coefficient a will be set to 0.2.) Later in the tutorial, results for an optically
thick medium with aL = 5 and non-participating medium with aL = 0 are computed to
show how the different radiation models behave for different optical thicknesses.
Define −→Materials...
1. Select boussinesq in the drop-down list next to Density, and then set the Density to
1000 kg/m3 .
For details about the Boussinesq model, see the User’s Guide.
c Fluent Inc. January 11, 2005 5-9
Modeling Radiation and Natural Convection
6. Keep the default settings for the Scattering Coefficient and the Scattering Phase
Function, since there is no scattering in this problem.
7. Set the Thermal Expansion Coefficient (used by the Boussinesq model) to 1e-5 K−1 .
5-10
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
c Fluent Inc. January 11, 2005 5-11
Modeling Radiation and Natural Convection
5-12
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
(a) Retain the default selected Equations (all of them) and Under-Relaxation Fac-
tors.
(b) Under Discretization, select PRESTO! for Pressure, and Second Order Upwind
for Momentum and Energy.
c Fluent Inc. January 11, 2005 5-13
Modeling Radiation and Natural Convection
5-14
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
c Fluent Inc. January 11, 2005 5-15
Modeling Radiation and Natural Convection
5-16
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
2.10e-04
2.00e-04
1.89e-04
1.79e-04
1.68e-04
1.58e-04
1.47e-04
1.37e-04
1.26e-04
1.16e-04
1.05e-04
9.45e-05
8.40e-05
7.35e-05
6.30e-05
5.25e-05
4.20e-05
3.15e-05
2.10e-05
1.05e-05
3.26e-09
c Fluent Inc. January 11, 2005 5-17
Modeling Radiation and Natural Convection
The recirculatory patterns observed are due to the natural convection in the box.
At a low optical thickness (0.2), radiation should not have a large influence on the
flow. The flow pattern is expected to be similar to that obtained with no radiation
(Figure 5.5). However, the Rosseland model predicts a flow pattern that is very
symmetric (Figure 5.4), and quite different from the pure natural convection case.
This discrepancy occurs because the Rosseland model is not appropriate for small
optical thickness.
6.98e-02
6.63e-02
6.28e-02
5.93e-02
5.58e-02
5.23e-02
4.88e-02
4.54e-02
4.19e-02
3.84e-02
3.49e-02
3.14e-02
2.79e-02
2.44e-02
2.09e-02
1.74e-02
1.40e-02
1.05e-02
6.98e-03
3.49e-03
0.00e+00
Extra: If you want to compute the results without radiation yourself, turn off all
the radiation models in the Radiation Model panel, set the under-relaxation
factor for energy to 0.8, and calculate until convergence. (Remember to reset
the under-relaxation factor to 1 before continuing with the tutorial).
5-18
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
1.97e-02
1.87e-02
1.77e-02
1.67e-02
1.58e-02
1.48e-02
1.38e-02
1.28e-02
1.18e-02
1.08e-02
9.85e-03
8.86e-03
7.88e-03
6.89e-03
5.91e-03
4.92e-03
3.94e-03
2.95e-03
1.97e-03
9.85e-04
0.00e+00
c Fluent Inc. January 11, 2005 5-19
Modeling Radiation and Natural Convection
2.00e+03
1.95e+03
1.90e+03
1.85e+03
1.80e+03
1.75e+03
1.70e+03
1.65e+03
1.60e+03
1.55e+03
1.50e+03
1.45e+03
1.40e+03
1.35e+03
1.30e+03
1.25e+03
1.20e+03
1.15e+03
1.10e+03
1.05e+03
1.00e+03
The Rosseland model predicts a temperature field (Figure 5.6) very different from
that obtained without radiation (Figure 5.7). For the low optical thickness in this
problem, the temperature field predicted by the Rosseland model is not physical.
4. Plot the y velocity along the horizontal centerline of the box.
(a) Create an isosurface at y = 0.5, the horizontal line through the center of the
box.
Surface −→Iso-Surface...
5-20
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
2.00e+03
1.95e+03
1.90e+03
1.85e+03
1.80e+03
1.75e+03
1.70e+03
1.65e+03
1.60e+03
1.55e+03
1.50e+03
1.45e+03
1.40e+03
1.35e+03
1.30e+03
1.25e+03
1.20e+03
1.15e+03
1.10e+03
1.05e+03
1.00e+03
c Fluent Inc. January 11, 2005 5-21
Modeling Radiation and Natural Convection
5-22
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
y=0.5
2.50e-04
2.00e-04
1.50e-04
1.00e-04
5.00e-05
Y 0.00e+00
Velocity
-5.00e-05
(m/s)
-1.00e-04
-1.50e-04
-2.00e-04
-2.50e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
the profile predicted by the Rosseland model exhibits thicker wall layers.
As discussed before, the expected profile for aL = 0.2 is similar to the case
with no radiation.
(c) Save the plot data to a file.
i. Select the Write to File option, and click the Write... button.
ii. In the resulting Select File dialog box, specify rad ross.xy in the XY File
text entry box and click OK.
c Fluent Inc. January 11, 2005 5-23
Modeling Radiation and Natural Convection
6. Save the case and data files (rad ross.cas and rad ross.dat).
File −→ Write −→Case & Data...
Thus far in this tutorial, you have learned how to set up a natural convection problem
using the Rosseland model to compute radiation. You have also learned to postprocess the
results. You will now turn on the P-1 model and compare the results so computed with
those of the Rosseland model.
5-24
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
c Fluent Inc. January 11, 2005 5-25
Modeling Radiation and Natural Convection
2.86e-04
2.72e-04
2.58e-04
2.43e-04
2.29e-04
2.15e-04
2.00e-04
1.86e-04
1.72e-04
1.58e-04
1.43e-04
1.29e-04
1.15e-04
1.00e-04
8.60e-05
7.17e-05
5.74e-05
4.31e-05
2.88e-05
1.45e-05
1.98e-07
(b) Plot the y velocity along the horizontal centerline (Figure 5.10), and save the
plot data to a file called rad p1.xy.
Plot −→XY Plot...
You will need to reselect Y Velocity under Y Axis Function. Also, remember
! to turn off the Write to File option so that you can access the Plot button
to generate the plot.
(c) Compute the total wall heat transfer rate.
Report −→Fluxes ...
The total heat transfer rate reported on the right wall is 8.47 × 105 W. The
heat imbalance at the lateral walls is negligibly small. You will see later that
the Rosseland and P-1 wall heat transfer rates are substantially different from
those obtained by the DTRM and the DO model.
5-26
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
y=0.5
2.50e-04
2.00e-04
1.50e-04
1.00e-04
5.00e-05
0.00e+00
Y
Velocity -5.00e-05
(m/s) -1.00e-04
-1.50e-04
-2.00e-04
-2.50e-04
-3.00e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
Notice how different the velocity vectors and y-velocity profile are from those obtained
using the Rosseland model. The P-1 velocity profiles show a clear momentum boundary
layer along the hot and cold walls. These profiles are much closer to those obtained from
the non-radiating case (Figures 5.11 and 5.12). Though the P-1 model is not appropriate
for this optically thin limit, it yields the correct velocity profiles since the radiation source
in the energy equation, which is proportional to the absorption coefficient, is small. The
Rosseland model uses an effective conductivity to account for radiation, and yields the
wrong temperature field, which in turn results in an erroneous velocity field.
c Fluent Inc. January 11, 2005 5-27
Modeling Radiation and Natural Convection
2.16e-04
2.05e-04
1.94e-04
1.83e-04
1.72e-04
1.62e-04
1.51e-04
1.40e-04
1.29e-04
1.19e-04
1.08e-04
9.70e-05
8.62e-05
7.55e-05
6.47e-05
5.39e-05
4.31e-05
3.24e-05
2.16e-05
1.08e-05
3.92e-08
y=0.5
2.50e-04
2.00e-04
1.50e-04
1.00e-04
5.00e-05
Y 0.00e+00
Velocity
-5.00e-05
(m/s)
-1.00e-04
-1.50e-04
-2.00e-04
-2.50e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
5-28
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
(c) Accept the default settings for Clustering and Angular Discretization by clicking
OK.
When you click OK, FLUENT will open a Select File dialog box so you can
specify a name for the ray file used by the DTRM. The number of Cells Per
Volume Cluster and Faces Per Surface Cluster control the total number of radiat-
ing surfaces and absorbing cells. For a small 2D problem, the default number
of 1 is acceptable. For a large problem, however, you will want to increase
these numbers to reduce the ray tracing expense. The Theta Divisions and Phi
Divisions control the number of rays being created from each surface cluster.
For most practical problems, the default settings will suffice.
c Fluent Inc. January 11, 2005 5-29
Modeling Radiation and Natural Convection
See Section 12.3.3 of the User’s Guide for a more detailed description of the
ray tracing procedure.
(d) In the Ray File text entry box in the Select File dialog box, enter rad dtrm.ray
for the name of the ray file. Then click OK.
FLUENT will print an informational message describing the progress of the ray
tracing procedure.
2. Retain the current under-relaxation factors for pressure, momentum, and energy
(0.3, 0.7, and 1.0).
Solve −→ Controls −→Solution...
You will need to reselect Y Velocity under Y Axis Function. Also, remember
! to turn off the Write to File option so that you can access the Plot button
to generate the plot.
5-30
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
2.88e-04
2.73e-04
2.59e-04
2.44e-04
2.30e-04
2.16e-04
2.01e-04
1.87e-04
1.73e-04
1.58e-04
1.44e-04
1.30e-04
1.15e-04
1.01e-04
8.64e-05
7.21e-05
5.77e-05
4.33e-05
2.90e-05
1.46e-05
2.42e-07
y=0.5
2.50e-04
2.00e-04
1.50e-04
1.00e-04
5.00e-05
0.00e+00
Y
Velocity -5.00e-05
(m/s) -1.00e-04
-1.50e-04
-2.00e-04
-2.50e-04
-3.00e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
c Fluent Inc. January 11, 2005 5-31
Modeling Radiation and Natural Convection
5-32
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
c Fluent Inc. January 11, 2005 5-33
Modeling Radiation and Natural Convection
2. Retain the current under-relaxation factors for pressure, momentum, and energy
(0.3, 0.7, and 1.0), as well as the default under-relaxation of 1 for the discrete
ordinates transport equation.
Solve −→ Controls −→Solution...
2.89e-04
2.75e-04
2.60e-04
2.46e-04
2.31e-04
2.17e-04
2.03e-04
1.88e-04
1.74e-04
1.59e-04
1.45e-04
1.30e-04
1.16e-04
1.01e-04
8.69e-05
7.25e-05
5.80e-05
4.36e-05
2.91e-05
1.47e-05
2.23e-07
5-34
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
(b) Plot the y velocity along the horizontal centerline (Figure 5.16), and save the
plot data to a file called rad do.xy.
Plot −→XY Plot...
You will need to reselect Y Velocity under Y Axis Function. Also, remember
! to turn off the Write to File option so that you can access the Plot button
to generate the plot.
y=0.5
3.00e-04
2.00e-04
1.00e-04
Y 0.00e+00
Velocity
(m/s)
-1.00e-04
-2.00e-04
-3.00e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
c Fluent Inc. January 11, 2005 5-35
Modeling Radiation and Natural Convection
2. Click on Plot.
Extra: You can click Curves... in the File XY Plot panel to open the Curves panel,
where you can define different styles for different plot curves. In Figure 5.17,
different symbols have been selected for each curve.
3. Resize and move the legend box so that you can read the information inside it.
(a) To resize the box, press any mouse button on a corner and drag the mouse to
the desired position.
(b) To move the legend box, press any mouse button anywhere else on the box
and drag it to the desired location.
5-36
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
Y Velocity
Y Velocity
3.00e-04
Y Velocity (rad_dtrm.xy)
Y Velocity (rad_p1.xy)
Y Velocity (rad_ross.xy)
2.00e-04
1.00e-04
Y 0.00e+00
Velocity
-1.00e-04
-2.00e-04
-3.00e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
Notice in Figure 5.17 that the velocity profiles for the P-1 model, DTRM, and DO model
are nearly identical even though the reported wall heat transfer rates are different. This
is because in an optically thin problem, the velocity field is essentially independent of the
radiation field, and all three models give a flow solution very close to the non-radiating
case. The Rosseland model gives substantially erroneous solutions for an optically thin
case.
c Fluent Inc. January 11, 2005 5-37
Modeling Radiation and Natural Convection
For the DTRM calculation, you may need to click the Iterate button re-
! peatedly until the radiation field is updated. Since the number of Flow
Iterations Per Radiation Iteration in the Radiation Model panel is 10, it is
possible that the radiation field will not be updated for as many as 9 it-
erations, although FLUENT will report that the solution is converged. If
this happens, keep clicking the Iterate button until the radiation field is
updated and the solution proceeds for multiple iterations.
(d) Save the new case and data files using a different file name (e.g., rad ros5.cas
and
rad ros5.dat).
File −→ Write −→Case & Data...
(e) Compute the total wall heat transfer rate.
Report −→Fluxes...
(f) Plot the y velocity along the horizontal centerline, and save the plot data to
a file (e.g., rad ros5.xy).
Plot −→XY Plot...
5-38
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
2. Compare the computed heat transfer rates for the four models.
The wall heat transfer rates predicted by the four radiation models range from 3.50×
105 to 3.97 × 105 W.
Y Velocity
Y Velocity
Y Velocity (rad_do5.xy)
5.00e-04
Y Velocity (rad_p15.xy)
Y Velocity (rad_ross5.xy)
4.00e-04
3.00e-04
2.00e-04
1.00e-04
Y 1.36e-20
Velocity
-1.00e-04
-2.00e-04
-3.00e-04
-4.00e-04
-5.00e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
The XY plots of y velocity are nearly identical for the P-1 model, DO model, and
DTRM. The Rosseland model gives somewhat different velocities, but is still within
10% of the other results. The Rosseland and P-1 models are suitable for the opti-
cally thick limit; the DTRM and DO models are valid across the range of optical
thicknesses. Consequently, they yield similar answers at aL = 5. For many ap-
plications with large optical thicknesses, the Rosseland and P-1 models provide a
simple low-cost alternative.
c Fluent Inc. January 11, 2005 5-39
Modeling Radiation and Natural Convection
1. Turn on the surface-to-surface (S2S) radiation model and define the view factor
and cluster parameters.
Define −→ Models −→Radiation...
(a) Select Surface to Surface under Model.
The panel will expand to show additional inputs for the S2S model.
5-40
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
c Fluent Inc. January 11, 2005 5-41
Modeling Radiation and Natural Convection
The size of the viewfactor file can be very large if not compressed. It is
i highly recommended to compress the view factor file by providing .gz or
.Z extension after the name (i.e. rad s2s.gz or rad s2s.Z).
2. Retain the current under-relaxation factors for pressure, momentum, and energy
(0.3, 0.7, and 1.0).
Solve −→ Controls −→Solution...
5-42
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
2.48e-04
2.36e-04
2.23e-04
2.11e-04
1.99e-04
1.86e-04
1.74e-04
1.61e-04
1.49e-04
1.37e-04
1.24e-04
1.12e-04
9.94e-05
8.70e-05
7.46e-05
6.22e-05
4.98e-05
3.74e-05
2.50e-05
1.26e-05
2.29e-07
(b) Plot the y velocity along the horizontal centerline (Figure 5.20), and save the
plot data to a file called rad s2s.xy.
Plot −→XY Plot...
You will have to reselect Y Velocity under Y Axis Function. Also, remember
! to turn off the Write to File option to access the Plot button to generate
the plot.
c Fluent Inc. January 11, 2005 5-43
Modeling Radiation and Natural Convection
y=0.5
2.50e-04
2.00e-04
1.50e-04
1.00e-04
5.00e-05
Y 0.00e+00
Velocity
-5.00e-05
(m/s)
-1.00e-04
-1.50e-04
-2.00e-04
-2.50e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
5-44
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
For the DTRM calculation, you may need to click the Iterate button re-
! peatedly until the radiation field is updated. Since the number of Flow
Iterations Per Radiation Iteration in the Radiation Model panel is 10, it is
possible that the radiation field will not be updated for as many as 9 it-
erations, although FLUENT will report that the solution is converged. If
this happens, keep clicking the Iterate button until the radiation field is
updated and the solution proceeds for multiple iterations.
(d) Save the new case and data files using a different file name (e.g., rad dtr0.cas
and rad dtr0.dat).
File −→ Write −→Case & Data...
(e) Compute the total wall heat transfer rate.
Report −→Fluxes...
(f) Plot the y velocity along the horizontal centerline, and save the plot data to
a file (e.g., rad dtr0.xy)
Plot −→XY Plot...
2. Compare the computed heat transfer rates for the three models.
For the S2S model, the total heat transfer rate on the right wall was 6.77 × 105 W.
This is about 5% higher than that predicted by the DTRM and 1.5% higher than DO.
Although the S2S, DO, and DTRM values are comparable to each other, this problem
involves enclosure radiative transfer without participating media. Therefore, the
S2S model provides the most accurate solution.
c Fluent Inc. January 11, 2005 5-45
Modeling Radiation and Natural Convection
Y Velocity
Y Velocity
3.00e-04
Y Velocity (rad_dtrm.xy)
Y Velocity (rad_dtr0.xy)
2.00e-04
1.00e-04
Y 0.00e+00
Velocity
-1.00e-04
-2.00e-04
-3.00e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
In Figure 5.21, the velocity profiles for the DTRM, DO, and S2S models are almost
identical even though the wall heat transfer rates are different.
5-46
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
Step 14: S2S Model Definition, Solution and Postprocessing with Partial
Enclosure
As mentioned earlier, when the S2S model is used, you also have the option to define
a ”partial enclosure”; i.e., you can disable the view factor calculation for walls with
negligible emission/absorption, or walls that have uniform temperature. Even though the
view factor will not be computed for these walls, they will still emit radiation at a fixed
temperature called the ”partial enclosure temperature”. The main advantage of this is to
speed up the view factor and the radiosity calculation.
For this problem, specify the left wall boundary as the non-participating wall in S2S
radiation. Consequently, you need to specify the partial enclosure temperature for the wall
boundary that is not participating in S2S radiation. Note that if multiple wall boundaries
are not participating in S2S radiation and each has a different temperature, then the
partial enclosure option may not yield accurate results. This is because the same partial
enclosure temperature is specified for each of the non-participating walls.
1. Read in the case and data files for the S2S model (rad s2s.cas and rad s2s.dat).
2. In the Radiation Model panel, retain Surface to Surface (S2S) as the radiation model.
Define −→ Models −→Radiation...
c Fluent Inc. January 11, 2005 5-47
Modeling Radiation and Natural Convection
You should compute the view factors only when you have specified the
! boundaries that will participate in the radiation model using the Boundary
Conditions panel. If you first compute the view factors and then make a
change to the boundary conditions, FLUENT will use the view factor file
stored earlier for calculating a solution, in which case, the changes that
you made to the model will not be used for the calculation. Therefore,
you should recompute the view factors and save the case file whenever you
modify the number of objects that will participate in radiation.
5-48
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
7. Retain the current under-relaxation factors for pressure, momentum, and energy
(0.3, 0.7, and 1.0).
Solve −→ Controls −→Solution...
You will have to reselect Y Velocity under Y Axis Function. Also, remember
! to turn off the Write to File option to access the Plot button to generate
the plot.
(c) Compute the total wall heat transfer rate.
Report −→Fluxes ...
c Fluent Inc. January 11, 2005 5-49
Modeling Radiation and Natural Convection
2.47e-04
2.35e-04
2.23e-04
2.10e-04
1.98e-04
1.85e-04
1.73e-04
1.61e-04
1.48e-04
1.36e-04
1.24e-04
1.11e-04
9.90e-05
8.67e-05
7.43e-05
6.20e-05
4.96e-05
3.73e-05
2.49e-05
1.26e-05
2.27e-07
Figure 5.22: Velocity Vectors for the S2S Model with Partial Enclosure
y=0.5
2.50e-04
2.00e-04
1.50e-04
1.00e-04
5.00e-05
Y 0.00e+00
Velocity
-5.00e-05
(m/s)
-1.00e-04
-1.50e-04
-2.00e-04
-2.50e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (m)
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
Figure 5.23: XY Plot of Centerline y Velocity for the S2S Model with Partial Enclosure
5-50
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
The total heat transfer rate on the right wall is 6.77 × 105 W. Note that the
total heat transfer rate on the left wall is reported as zero because the report
utility in the current version of FLUENT does not account for the radiation
heat transfer rate by this wall, as it should.
c Fluent Inc. January 11, 2005 5-51
Modeling Radiation and Natural Convection
Step 15: Comparison of S2S Models with and without Partial Enclosure
1. Compare the computed heat transfer rates for the two S2S models.
Y Velocity
Y Velocity
Y Velocity (rad_s2sp.xy)
2.50e-04
2.00e-04
1.50e-04
1.00e-04
5.00e-05
Y 0.00e+00
Velocity
-5.00e-05
-1.00e-04
-1.50e-04
-2.00e-04
-2.50e-04
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position
Y Velocity
FLUENT 6.2 (2d, segregated, lam)
In Figure 5.24, the velocity profiles for the S2S model without partial enclosure and the
S2S model with partial enclosure are almost identical.
5-52
c Fluent Inc. January 11, 2005
Modeling Radiation and Natural Convection
Summary
In this tutorial, you studied combined natural convection and radiation in a square box
and compared the performance of four radiation models in FLUENT for optically thin
and optically thick cases, and the performance of three radiation models for a non-
participating medium.
• For the optically thin case, the Rosseland and P-1 models are not appropriate; the
DTRM and the DO model are applicable, and yield similar results.
• In the optically thick limit, all four models are appropriate and yield similar results.
In this limit, the less computationally-expensive Rosseland and P-1 models may be
adequate for many engineering applications.
• The S2S radiation model is appropriate for modeling the enclosure radiative transfer
without participating media, where the methods for participating radiation may not
always be efficient.
See Section 12.3 of the User’s Guide for more information about the applicability of the
different radiation models.
c Fluent Inc. January 11, 2005 5-53
Modeling Radiation and Natural Convection
5-54
c Fluent Inc. January 11, 2005
Tutorial 6. Using a Non-Conformal Mesh
Introduction
Film cooling is a process that is used to protect turbine vanes in a gas turbine engine
from exposure to hot combustion gases. This tutorial illustrates how to set up and solve
a film cooling problem using a non-conformal mesh. The system that is modeled consists
of three parts: a duct, a hole array, and a plenum. The duct is modeled with a hexahedral
mesh, and the plenum and hole regions are modeled using a tetrahedral mesh. These two
meshes are merged together to form a “hybrid” mesh, with a non-conformal interface
boundary between them.
Due to symmetry of the hole array, only a portion of the geometry is modeled in FLUENT,
with symmetry applied to the outer boundaries. The duct contains a high-velocity fluid in
streamwise flow (Figure 6.1). An array of holes intersects the duct at an inclined angle,
and a cooler fluid is injected into the holes from a plenum. The coolant that moves
through the holes acts to cool the surface of the duct, downstream of the injection. Both
fluids are air, and the flow is classified as turbulent. The velocity and temperature of the
streamwise and cross-flow fluids are known, and FLUENT is used to predict the flow and
temperature fields that result from convective heat transfer.
In this tutorial you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
c Fluent Inc. January 11, 2005 6-1
Using a Non-Conformal Mesh
Problem Description
This problem considers a model of a 3D section of a film cooling test rig. A schematic is
shown in Figures 6.1 and 6.2. The problem consists of a duct, 24.5 in long, with cross-
sectional dimensions of 0.75 in × 5 in. An array of uniformly spaced holes is located at
the bottom of the duct. Each hole has a diameter of 0.5 inches, is inclined at 35 degrees,
and is spaced 1.5 inches apart laterally. Cooler injected air enters the system through
the plenum, with cross-sectional dimensions of 3.3 in × 1.25 in.
Because of the symmetry of the geometry, only a portion of the domain needs to be
modeled. The computational domain is shown in outline in Figure 6.2. The bulk tem-
perature of the streamwise air (T∞ ) is 273 K, and the velocity of the air stream is 20
m/s. The bottom wall of the duct that intersects the hole array is assumed to be a
completely insulated (adiabatic) wall. The secondary (injected) air enters the plenum at
a uniform velocity of 0.4559 m/s. The temperature of the injected air (Tinject ) is 136.6 K.
The properties of air that are used in the model are shown in Figure 6.2.
9.5 in 0.5 in 14.5 in
v = 20 m/s
Τ
∞ = 273 K Duct
y 5 in
x
1.25 in Hole 35°
1.25 in Plenum
3.3 in
v = 0.4559 m/s
T = 136.6 K
inject
6-2
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
z
Τ
∞ = 273 K
0.25 in 0.75 in
µ = 0.000017894 kg/m-s
c p = 1006.43 J/kg-K
2. Unzip non_conformal_mesh.zip.
film hex.msh and film tet.msh can be found in the /non conformal mesh folder
created after unzipping the file.
c Fluent Inc. January 11, 2005 6-3
Using a Non-Conformal Mesh
2. Provide the mesh file names film tet.msh and film hex.msh as prompted. Pro-
vide scaling of 1 and translations and rotations of zero for each mesh file. Save the
new merged mesh file as filmcool.msh.
The mesh files must be read into tmerge in this order for the tutorial to run
! as written. Otherwise, zone names and numbers will be assigned differently
when the files are merged together. In general, however, you can specify
files to be read into tmerge in any order.
6-4
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
Step 2: Grid
1. Start the 3D version of FLUENT.
(a) In the Unit Conversion drop-down list, select in to complete the phrase Grid
Was Created In in (inches).
(b) Click Scale to scale the grid.
(c) Click Change Length Units to set inches as the working units for length.
The final Domain Extents should appear as in the panel above.
(d) Close the panel.
c Fluent Inc. January 11, 2005 6-5
Using a Non-Conformal Mesh
Y
Z X
Grid
FLUENT 6.2 (3d, segregated, ske)
6-6
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
(c) Zoom in using your middle mouse button to get the view displayed in Fig-
ure 6.4.
Y
Z X
Grid
FLUENT 6.2 (3d, segregated, ske)
In Figure 6.4 you can see the quadrilateral faces of the hexahedral cells that are
used to model the duct region, and the triangular faces of the tetrahedral cells that
are used to model the plenum and hole regions, resulting in a hybrid mesh.
Extra: You can use the right mouse button to check which zone number corre-
sponds to each boundary. If you click the right mouse button on one of the
boundaries in the graphics window, its zone number, name, and type will be
printed in the FLUENT console window. This feature is especially useful when
you have several zones of the same type and you want to distinguish between
them quickly.
c Fluent Inc. January 11, 2005 6-7
Using a Non-Conformal Mesh
Step 3: Models
1. Keep the default solver settings.
Define −→ Models −→Solver...
6-8
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
c Fluent Inc. January 11, 2005 6-9
Using a Non-Conformal Mesh
Step 4: Materials
Define −→Materials...
1. Retain air as the Fluent Fluid Materials , and use the incompressible-ideal-gas law to
compute density. Retain the default values for all other properties.
6-10
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
c Fluent Inc. January 11, 2005 6-11
Using a Non-Conformal Mesh
1. Set the boundary conditions for the streamwise flow inlet (velocity-inlet-1).
2. Set the boundary conditions for the injected stream inlet (velocity-inlet-14).
(a) Change the Zone Name from velocity-inlet-14 to velocity-inlet-plenum.
(b) Set the Velocity Magnitude to 0.4559 m/s.
(c) Set the Temperature to 136.6 K.
(d) In the Turbulence Specification Method drop-down list, select Intensity and Vis-
cosity Ratio.
6-12
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
(e) Set the Turbulence Intensity to 1% and keep the Turbulent Viscosity Ratio default
of 10.
In the absence of any identifiable length scale for turbulence, the Intensity and Vis-
cosity Ratio method should be used.
See Chapter 11 of the User’s Guide for more information on how to set the boundary
conditions for turbulence.
c Fluent Inc. January 11, 2005 6-13
Using a Non-Conformal Mesh
6-14
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
c Fluent Inc. January 11, 2005 6-15
Using a Non-Conformal Mesh
5. Set the conditions for the fluid in the plenum and hole (fluid-17).
(a) Change the Zone Name from fluid-17 to fluid-plenum.
(b) Keep the default selection of air as the Material Name.
6-16
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
6. Keep the default boundary conditions for the plenum and hole wall (wall-15).
c Fluent Inc. January 11, 2005 6-17
Using a Non-Conformal Mesh
X Y
Z
Grid
FLUENT 6.2 (3d, segregated, ske)
6-18
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
(b) Click Yes when asked OK to change wall-1’s type from wall to interface?.
(c) Change the Zone Name to interface-duct.
(d) Repeat this procedure to convert wall-12 to an interface boundary zone named
interface-hole.
c Fluent Inc. January 11, 2005 6-19
Using a Non-Conformal Mesh
When one interface zone is smaller than the other, it is recommended that
! you choose the smaller zone as Interface Zone 1.
2. Select interface-duct in the Interface Zone 2 list.
4. Click Create.
Note: In the process of creating the grid interface, FLUENT creates two new wall
boundary zones: wall-11 and wall-18. You will not be able to display these
walls.
wall-11 is the non-overlapping region of the interface-hole zone that results from
the intersection of the interface-hole and interface-duct boundary zones, and is
listed under Boundary Zone 1 in the Grid Interfaces panel. wall-11 is empty,
since interface-hole is completely contained within the interface-duct boundary.
wall-18 is the non-overlapping region of the interface-duct zone that results from
the intersection of the two interface zones, and is listed under Boundary Zone
2 in the Grid Interfaces panel.
6-20
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
In general, you will need to set boundary conditions for wall-18 (since it is
!
not empty). In this case, default settings are used.
Step 8: Solution
1. Set the solution parameters.
Solve −→ Controls −→Solution...
(a) Under Discretization, select Second Order Upwind for Momentum and Turbulence
Kinetic Energy.
(b) Scroll down the list and select Second Order Upwind for Turbulence Dissipation
Rate and Energy.
c Fluent Inc. January 11, 2005 6-21
Using a Non-Conformal Mesh
6-22
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
c Fluent Inc. January 11, 2005 6-23
Using a Non-Conformal Mesh
Step 9: Postprocessing
1. Display filled contours of static pressure (Figure 6.6).
Display −→Contours...
6-24
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
3.36e+02
3.12e+02
2.88e+02
2.64e+02
2.39e+02
2.15e+02
1.91e+02
1.67e+02
1.42e+02
1.18e+02
9.37e+01
6.95e+01
4.52e+01
2.09e+01
-3.33e+00
-2.76e+01
-5.19e+01
-7.61e+01
-1.00e+02 Y
-1.25e+02 Z X
-1.49e+02
c Fluent Inc. January 11, 2005 6-25
Using a Non-Conformal Mesh
6-26
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
2.73e+02
2.59e+02
2.46e+02
2.32e+02
2.18e+02
2.05e+02
1.91e+02
1.78e+02
1.64e+02
1.50e+02
1.37e+02
1.23e+02
1.09e+02
9.56e+01
8.19e+01
6.83e+01
5.46e+01
4.10e+01
2.73e+01 Y
1.37e+01 Z X
0.00e+00
2.73e+02
2.59e+02
2.46e+02
2.32e+02
2.18e+02
2.05e+02
1.91e+02
1.78e+02
1.64e+02
1.50e+02
1.37e+02
1.23e+02
1.09e+02
9.56e+01
8.19e+01
6.83e+01
5.46e+01
4.10e+01
2.73e+01 Y
1.37e+01 Z X
0.00e+00
c Fluent Inc. January 11, 2005 6-27
Using a Non-Conformal Mesh
(a) Select Velocity... and Velocity Magnitude in the Color by drop-down lists.
(b) Change the Scale to 2.
This will enlarge the vectors that are displayed, making it easier to view the
flow patterns.
(c) In the Surfaces list, select interface-duct and interface-hole.
(d) Scroll down the Surfaces list and select symmetry-1, symmetry-tet, and wall-15.
(e) In the Vectors panel, click Display.
(f) Zoom in on the view to get the display shown in Figure 6.9.
The flow pattern in the vicinity of the coolant hole shows the level of penetration
of the coolant jet into the main flow. Notice that the velocity field varies smoothly
across the non-conformal interface.
6-28
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
2.13e+01
2.02e+01
1.92e+01
1.81e+01
1.71e+01
1.60e+01
1.49e+01
1.39e+01
1.28e+01
1.17e+01
1.07e+01
9.60e+00
8.53e+00
7.47e+00
6.40e+00
5.33e+00
4.27e+00
3.20e+00
2.14e+00 Y
1.07e+00 Z X
9.16e-03
4. Plot the temperature profile along a horizontal cross-section of the duct, 0.1 inches
above the bottom.
(a) Create an isosurface on the duct surface at y = 0.1 in.
Surface −→Iso-Surface...
c Fluent Inc. January 11, 2005 6-29
Using a Non-Conformal Mesh
6-30
c Fluent Inc. January 11, 2005
Using a Non-Conformal Mesh
y=0.1in
2.80e+02
2.60e+02
2.40e+02
2.20e+02
Static 2.00e+02
Temperature
(k) 1.80e+02
1.60e+02
1.40e+02
1.20e+02
-10 -7.5 -5 -2.5 0 2.5 5 7.5 10 12.5 15 17.5
Y
Z X Position (in)
Static Temperature
FLUENT 6.2 (3d, segregated, ske)
In this plot you can see how the temperature of the fluid changes as the cool
air from the injection hole mixes with the primary flow. As expected, the tem-
perature is coolest just downstream of the hole. Note that you could also make
a similar plot on the lower wall itself, to examine the wall surface temperature.
Summary
This tutorial demonstrates how FLUENT’s non-conformal grid interface capability can
be used to handle hybrid meshes for complex geometries, such as the film cooling hole
configuration examined here. One of the principal advantages of this approach is that
it allows you to merge existing component meshes together to create a larger, more
complex mesh system, without requiring that the different components have the same
node locations on their shared boundaries. Thus, you can perform parametric studies
by merging the desired meshes, creating the non-conformal interface(s), and solving the
model. For example, in the present case, you can
c Fluent Inc. January 11, 2005 6-31
Using a Non-Conformal Mesh
6-32
c Fluent Inc. January 11, 2005
Tutorial 7. Using a Single Rotating Reference Frame
Introduction
This tutorial considers the flow within a 2D, axisymmetric, co-rotating disk cavity system.
Understanding the behavior of such flows is important in the design of secondary air
passages for turbine disk cooling.
In this tutorial you will learn how to:
• Use the standard k- and RNG k- turbulence models with the enhanced near-wall
treatment
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The problem to be considered is shown schematically in Figure 7.1. This case is similar
to a disk cavity configuration that was extensively studied by Pincombe [1].
Air enters the cavity between two co-rotating disks. The disks are 88.6 cm in diameter
and the air enters at 1.146 m/s through a circular bore 8.86 cm in diameter. The disks,
which are 6.2 cm apart, are spinning at 71.08 rpm, and the air enters with no swirl. As
the flow is diverted radially, the rotation of the disk has a significant effect on the viscous
flow developing along the surface of the disk.
c Fluent Inc. January 11, 2005 7-1
Using a Single Rotating Reference Frame
Outflow
6.2 cm
71.08 rpm
Inflow 4.43 cm
As noted by Pincombe [1], there are two nondimensional parameters that characterize
this type of disk cavity flow: the volume flow rate coefficient, Cw , and the rotational
Reynolds number, Reφ . These parameters are defined as follows:
Q
Cw = (7.1)
ν rout
2
Ωrout
Reφ = (7.2)
ν
where Q is the volumetric flow rate, Ω is the rotational speed, ν is the kinematic viscosity,
and rout is the outer radius of the disks. Here, you will consider a case for which Cw =
1092 and Reφ = 105 .
7-2
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
2. Unzip single_rotating.zip.
disk.msh can be found in the /single rotating folder created after unzipping the
file.
Step 1: Grid
1. Read the grid file (disk.msh).
File −→ Read −→Case...
As FLUENT reads the grid file, it will report its progress in the console window.
c Fluent Inc. January 11, 2005 7-3
Using a Single Rotating Reference Frame
Grid
FLUENT 6.2 (2d, segregated, lam)
Extra: You can use the right mouse button to check which zone number corresponds to
each boundary. If you click the right mouse button on one of the boundaries in the
graphics window, its zone number, name, and type will be printed in the FLUENT
console window. This feature is especially useful when you have several zones of
the same type and you want to distinguish between them quickly.
7-4
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
Step 2: Units
1. For convenience, define new units for angular velocity and length.
In the problem description, angular velocity and length are specified in rpm and cm,
respectively. These are not the default units for these quantities.
Define −→Units...
c Fluent Inc. January 11, 2005 7-5
Using a Single Rotating Reference Frame
Step 3: Models
1. Specify the solver formulation to be used for the model calculation, and enable the
modeling of axisymmetric swirl.
Define −→ Models −→Solver...
7-6
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
2. Turn on the standard k- turbulence model with the enhanced near-wall treatment.
c Fluent Inc. January 11, 2005 7-7
Using a Single Rotating Reference Frame
Step 4: Materials
1. Accept the default properties for air.
Define −→Materials...
For the present analysis, you will model air as an incompressible fluid with a density
of 1.225 kg/m3 and a dynamic viscosity of 1.7894×10−5 kg/m-s. Since these are
the default values, no change is required in the Materials panel.
Extra: You can modify the fluid properties for air at any time or copy another
material from the database.
See Chapter 8 of the User’s Guide for details.
7-8
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
1. Define the rotating reference frame for the fluid zone (fluid-7).
(a) Select Moving Reference Frame in the Motion Type drop-down list.
(b) Scroll down below Motion Type and set the Speed (under Rotational Velocity)
to 71.08 rpm.
c Fluent Inc. January 11, 2005 7-9
Using a Single Rotating Reference Frame
Note: FLUENT will use the backflow conditions only if the fluid is flowing into
the computational domain through the outlet. Since backflow might occur at
some point during the solution procedure, you should set reasonable backflow
conditions to prevent convergence from being adversely affected.
7-10
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
Note: For a rotating reference frame, FLUENT assumes by default that all walls
rotate at the speed of the moving reference frame, and hence are moving with
respect to the stationary (absolute) reference frame. To specify a non-rotating
wall, you must specify a rotational speed of 0 in the absolute frame.
c Fluent Inc. January 11, 2005 7-11
Using a Single Rotating Reference Frame
(a) Under Discretization, select PRESTO! from the drop-down list to the right of
Pressure.
The PRESTO! scheme is well suited for steep pressure gradients involved in
rotating flows. It provides improved pressure interpolation in situations where
large body forces or strong pressure variations are present as in swirling flows.
(b) Select Second Order Upwind from the adjacent drop-down list for Momentum,
Swirl Velocity, Turbulence Kinetic Energy, and Turbulence Dissipation Rate.
7-12
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
c Fluent Inc. January 11, 2005 7-13
Using a Single Rotating Reference Frame
i. Select Mass Flow Rate from the Report Type drop-down list.
ii. Select pressure-outlet-3 in the Surfaces list.
iii. Click OK to define the monitor.
(d) Click OK in the Surface Monitors panel to enable the monitor.
7-14
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
4. Initialize the flow field using the boundary conditions set at velocity-inlet-2.
Solve −→ Initialize −→Initialize...
Throughout the calculation, FLUENT will report reversed flow at the exit. This is
reasonable for the current case.
The solution should be sufficiently converged after about 230 iterations. The mass
flow rate history is shown in Figure 7.3.
c Fluent Inc. January 11, 2005 7-15
Using a Single Rotating Reference Frame
0.0000
-0.0500
-0.1000
Mass -0.1500
Flow
Rate
-0.2000
-0.2500
-0.3000
0 25 50 75 100 125 150 175 200 225 250
Iteration
Although the mass flow rate history indicates that the solution is con-
! verged, you should also check the net mass fluxes through the domain to
ensure that mass is being conserved.
7-16
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
The net mass imbalance should be a small fraction (say, 0.5%) of the total
!
flux through the system. If a significant imbalance occurs, you should
decrease your residual tolerances by at least an order of magnitude and
continue iterating.
c Fluent Inc. January 11, 2005 7-17
Using a Single Rotating Reference Frame
7-18
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
3.27e+00
3.10e+00
2.94e+00
2.78e+00
2.62e+00
2.45e+00
2.29e+00
2.13e+00
1.97e+00
1.80e+00
1.64e+00
1.48e+00
1.32e+00
1.15e+00
9.91e-01
8.29e-01
6.66e-01
5.04e-01
3.42e-01
1.79e-01
1.67e-02
Figure 7.4: Magnified View of Velocity Vectors within the Disk Cavity
c Fluent Inc. January 11, 2005 7-19
Using a Single Rotating Reference Frame
(a) Select Pressure... and Static Pressure in the Contours of drop-down list.
(b) Turn on the Filled option.
(c) Click Display.
The pressure contours are displayed in Figure 7.5. Notice the high pressure that
occurs on the right disk near the hub due to the stagnation of the flow entering from
the bore.
7-20
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
8.01e-01
7.34e-01
6.66e-01
5.98e-01
5.31e-01
4.63e-01
3.95e-01
3.28e-01
2.60e-01
1.92e-01
1.25e-01
5.70e-02
-1.07e-02
-7.84e-02
-1.46e-01
-2.14e-01
-2.81e-01
-3.49e-01
-4.17e-01
-4.84e-01
-5.52e-01
(a) Select Grid... and Y-Coordinate in the Surface of Constant drop-down lists.
(b) Click Compute to update the minimum and maximum values.
(c) Enter 37 in the Iso-Values field.
This is the radial position along which you will plot the radial velocity profile.
c Fluent Inc. January 11, 2005 7-21
Using a Single Rotating Reference Frame
7-22
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
y=37
3.50e-01
3.00e-01
2.50e-01
2.00e-01
Radial 1.50e-01
Velocity
(m/s) 1.00e-01
5.00e-02
0.00e+00
-5.00e-02
0 1 2 3 4 5 6 7
Position (cm)
Radial Velocity
FLUENT 6.2 (axi, swirl, segregated, ske)
c Fluent Inc. January 11, 2005 7-23
Using a Single Rotating Reference Frame
1. Turn on the RNG k- turbulence model with the enhanced near-wall treatment.
Define −→ Models −→Viscous...
7-24
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
3. Save the case and data files (disk rng.cas and disk rng.dat).
File −→ Write −→Case & Data...
c Fluent Inc. January 11, 2005 7-25
Using a Single Rotating Reference Frame
7-26
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
(e) Click the Curves... button to define a different curve symbol for the RNG k-
data.
This will open the Curves - Solution XY Plot panel.
y=37cm
y=37cm
4.00e-01
3.50e-01
3.00e-01
2.50e-01
2.00e-01
Radial 1.50e-01
Velocity
1.00e-01
(m/s)
5.00e-02
0.00e+00
-5.00e-02
-1.00e-01
0 1 2 3 4 5 6 7
Position (cm)
Figure 7.7: Radial Velocity Distribution: RNG and Standard k- Solutions
The plot should be similar to the one shown in Figure 7.7. The peak velocity
predicted by the RNG solution is higher than that predicted by the k- solution.
This is due to the less diffusive character of the RNG k- model.
c Fluent Inc. January 11, 2005 7-27
Using a Single Rotating Reference Frame
(g) Adjust the range of the x axis to magnify the region of the peaks.
i. In the Solution XY Plot panel, click the Axes... button to specify the x-axis
range.
This will open the Axes - Solution XY panel.
7-28
c Fluent Inc. January 11, 2005
Using a Single Rotating Reference Frame
y=37cm
y=37cm
4.00e-01
3.50e-01
3.00e-01
2.50e-01
Radial 2.00e-01
Velocity
(m/s) 1.50e-01
1.00e-01
5.00e-02
0.00e+00
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Position (cm)
Radial Velocity
FLUENT 6.2 (axi, swirl, segregated, rngke)
Figure 7.8: Radial Velocity Distribution: RNG and Standard k- Solutions (x = 0 cm to
x = 1 cm)
Summary
This tutorial has demonstrated how to set up an axisymmetric disk cavity problem in
FLUENT. The ability to calculate a swirl velocity permits the use of a 2D mesh, so
the calculation is simpler and more economical to run. This is especially important
for problems where the enhanced wall treatment is used, and the near-wall flow field is
resolved through the viscous sublayer and buffer zones (that is, the first grid point away
from the wall is placed at a y + on the order of 1).
See Section 11.10 of the User’s Guide for more information about grid considerations for
turbulence modeling.
c Fluent Inc. January 11, 2005 7-29
Using a Single Rotating Reference Frame
Further Improvements
The case modeled in this tutorial lends itself to parametric study due to its relatively
small size. Here are some things you may wish to try:
• Use adaption to see if resolving the high velocity and pressure-gradient region of
the flow has a significant effect on the solution.
• Introduce a non-zero swirl at the inlet or use a velocity profile for fully-developed
pipe flow. This is probably more realistic than the constant axial velocity used
here, since the flow at the inlet is typically being supplied by a pipe.
• Model compressible flow (using the ideal gas law for density) rather than assuming
incompressible flow.
References
1. Pincombe, J.R., “Velocity Measurements in the Mk II - Rotating Cavity Rig with a
Radial Outflow”, Thermo-Fluid Mechanics Research Centre, University of Sussex,
Brighton, UK, 1981.
7-30
c Fluent Inc. January 11, 2005
Tutorial 8. Using Multiple Rotating Reference Frames
Introduction
Many engineering problems involve rotating flow domains. One example is the centrifugal
blower unit that is typically used in automotive climate control systems. For problems
where all the moving parts (fan blades, hub and shaft surfaces, etc.) are rotating at a
prescribed angular velocity, and the stationary walls (e.g., shrouds, duct walls) are sur-
faces of revolution with respect to the axis of rotation, the entire domain can be referred
to as a single rotating frame of reference. However, when each of several parts is rotating
about a different axis of rotation, or about the same axis at different speeds, or when the
stationary walls are not surfaces of revolution (such as the volute around a centrifugal
blower wheel), a single rotating coordinate system is not sufficient to “immobilize” the
computational domain so as to predict a steady-state flow field.
In FLUENT, the flow features associated with multiple rotating parts can be analyzed
using the multiple reference frame (MRF) capability. This model is powerful in that
multiple rotating reference frames can be included in a single domain. The resulting flow
field is representative of a snapshot of the transient flow field in which the rotating parts
are moving. However, in many cases the interface can be chosen in such a way that the
flow field at this location is independent of the orientation of the moving parts. In other
words, if an interface can be drawn on which there is little or no angular dependence,
the model can be a reliable tool for simulating time-averaged flow fields. It is therefore
very useful in complicated situations where one or more rotating parts are present.
This tutorial illustrates the procedure for setting up and solving a problem using the
MRF capability. As an example, the flow field on a 2D section of a centrifugal blower
will be calculated. The example will be limited to a single rotating reference frame.
The following FLUENT features will be demonstrated in this tutorial:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
c Fluent Inc. January 11, 2005 8-1
Using Multiple Rotating Reference Frames
In general, to solve problems using the MRF feature, you should be familiar with the
concept of creating multiple fluid zones in your grid generator.
Problem Description
This problem considers a 2D section of a generic centrifugal blower. A schematic of the
problem is shown in Figure 8.1. The blower consists of 32 blades, each with a chord length
of 13.5 mm. The blades are located approximately 56.5 mm (measured from the leading
edge) from the center of rotation. The radius of the outer wall varies logarithmically
from 80 mm to 146.5 mm. The total pressure at the inlet is defined to be 200 Pa and the
flow discharges to ambient conditions (static pressure = 0 Pa). The blades are rotating
with an angular velocity of 261 rad/s. The flow is assumed to be turbulent.
Pressure-inlet-5
261 rad/s
35 mm
56.5 mm
blower blades
(13.5 mm chord length)
Pressure-Outlet-9
145 mm
8-2
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
2. Unzip multiple_rotating.zip.
blower.msh can be found in the /multiple rotating folder created after unzipping
the file.
Step 1: Grid
1. Read in the mesh file (blower.msh).
File −→ Read −→Case...
c Fluent Inc. January 11, 2005 8-3
Using Multiple Rotating Reference Frames
The fluid zones are separated by wall boundaries. These boundaries were used in
the grid generator to separate the fluid zones, and will be converted to interior zones
when the boundary conditions are set later in this tutorial. Each of these wall zones
also has an associated “shadow wall” which was created by FLUENT when it read
the grid. Shadow walls are created whenever a wall has fluid zones on both sides.
8-4
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
Grid
FLUENT 6.2 (2d, segregated, lam)
c Fluent Inc. January 11, 2005 8-5
Using Multiple Rotating Reference Frames
Step 2: Models
1. Keep the default solver settings.
Define −→ Models −→Solver...
8-6
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
c Fluent Inc. January 11, 2005 8-7
Using Multiple Rotating Reference Frames
Step 3: Materials
You will use the default material, air, with its predefined properties, for all fluid zones.
No action is required in the panel.
Define −→Materials...
Extra: If needed, you could modify the fluid properties for air or copy another material
from the database.
8-8
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
(a) Select wall-2 in the Zone list and then select interior in the Type list.
FLUENT will prompt for confirmation before changing the zone type.
(b) Click Yes and FLUENT will fuse wall-2 and wall-2-shadow together to form
interior-2.
c Fluent Inc. January 11, 2005 8-9
Using Multiple Rotating Reference Frames
2. Identify the rotating fluid zone (i.e., the zone containing the blades) by displaying
the mesh for each zone.
Display −→Grid...
It is unclear when you read the grid which fluid zone corresponds to which interior
zone. While the interior zones can be selected individually in the Grid Display panel,
the fluid zones cannot. Commands in the text interface, however, can be used to
make this association.
(a) Deselect all surfaces by clicking on the unshaded icon to the right of Surfaces.
(b) Click the Outline button at the bottom of the panel to select only the outline
surfaces of the domain.
(c) Click Display.
Only the domain boundaries and interior walls will be displayed.
(d) In the console window, type the commands shown in boxes in the dialog below.
Hint: You may need to press the <Enter> key to get the > prompt.
> display
/display> zone-grid
()
zone id/name(1) [()] 13
zone id/name(2) [()] <Enter>
The resulting display (Figure 8.3) shows that zone fluid-13 corresponds to the ro-
tating region.
8-10
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
c Fluent Inc. January 11, 2005 8-11
Using Multiple Rotating Reference Frames
Note: Since the other fluid zones are stationary, you do not need to set any bound-
ary conditions for them. If one of the remaining fluid zones was also rotating,
you would need to set the appropriate rotational speed for it.
4. Set the following conditions (see Figure 8.1) for the flow inlet (pressure-inlet-5).
8-12
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
Note: All pressures that you specify in FLUENT are gauge pressures, relative to
the operating pressure specified in the Operating Conditions panel. By default,
the operating pressure is 101325 Pa.
See Section 8.13 of the User’s Guide for details.
5. Set the backflow turbulence parameters for the flow outlet (pressure-outlet-9) to the
same values used for pressure-inlet-5.
Note: The backflow values are used only if reversed flow occurs at the outlet, but it
is a good idea to use reasonable values, even if you do not expect any backflow
to occur.
6. Define the velocity of the wall zone representing the blades (wall-7) relative to the
moving fluid zone.
With fluid-13 set to a rotating reference frame, wall-7 becomes a moving wall.
(a) In the Momentum section of the Wall panel, enable the Moving Wall option.
The panel will expand to show the wall motion parameters.
(b) Under Motion, select Relative to Adjacent Cell Zone and Rotational.
(c) Set the (relative) Speed to 0 rad/s.
The Rotation-Axis Origin should be located at x = 0 m and y = 0 m. With
these settings, the blades will move at the same speed as the surrounding fluid.
c Fluent Inc. January 11, 2005 8-13
Using Multiple Rotating Reference Frames
8-14
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
Step 5: Solution
1. Choose the second-order discretization scheme for the governing equations.
Solve −→ Controls −→Solution...
(a) In the drop-down lists next to Momentum, Turbulence Kinetic Energy, and
Turbulence Dissipation Rate, select Second Order Upwind.
The second-order scheme will provide a more accurate solution.
(b) Keep the default parameters for all other solution controls.
c Fluent Inc. January 11, 2005 8-15
Using Multiple Rotating Reference Frames
8-16
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
Note: In this tutorial, you chose an Absolute reference frame for initializing the
solution. In certain cases, Relative to Cell Zone may help the solution converge
faster.
See Chapter 26.15 of the User’s Guide for guidelines.
During the calculation, FLUENT will report that there is reversed flow occurring at
the exit. This is due to the sudden expansion, which results in a recirculating flow
near the exit.
The solution will converge in around 160 iterations (when all residuals have dropped
below 0.001).
c Fluent Inc. January 11, 2005 8-17
Using Multiple Rotating Reference Frames
Step 6: Postprocessing
1. Display filled contours of total pressure (Figure 8.4).
Display −→Contours...
(a) Select Pressure... and Total Pressure in the Contours of drop-down lists.
(b) Select Filled under Options.
(c) Click Display.
Total pressure contours show the expected pressure jump across the blower blades.
8-18
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
1.13e+03
1.03e+03
9.25e+02
8.23e+02
7.22e+02
6.20e+02
5.18e+02
4.16e+02
3.15e+02
2.13e+02
1.11e+02
9.50e+00
-9.22e+01
-1.94e+02
-2.96e+02
-3.97e+02
-4.99e+02
-6.01e+02
-7.03e+02
-8.04e+02
-9.06e+02
c Fluent Inc. January 11, 2005 8-19
Using Multiple Rotating Reference Frames
8-20
c Fluent Inc. January 11, 2005
Using Multiple Rotating Reference Frames
6.83e+01
6.49e+01
6.15e+01
5.81e+01
5.47e+01
5.13e+01
4.78e+01
4.44e+01
4.10e+01
3.76e+01
3.42e+01
3.08e+01
2.74e+01
2.39e+01
2.05e+01
1.71e+01
1.37e+01
1.03e+01
6.87e+00
3.45e+00
3.93e-02
c Fluent Inc. January 11, 2005 8-21
Using Multiple Rotating Reference Frames
Summary
This tutorial illustrates the procedure for setting up and solving problems with multiple
reference frames using FLUENT. Although this tutorial considers only one rotating fluid
zone, extension to multiple rotating fluid zones is straightforward as long as you delineate
each fluid zone.
Note that this tutorial was solved using the default absolute velocity formulation. For
some problems involving rotating reference frames, you may wish to use the relative
velocity formulation. See the User’s Guide for details.
8-22
c Fluent Inc. January 11, 2005
Tutorial 9. Using the Mixing Plane Model
Introduction
This tutorial considers the flow in an axial fan with a rotor in front and stators (vanes)
in the rear. This configuration is typical of a single-stage axial flow turbomachine. By
considering the rotor and stator together in a single calculation, you can determine the
interaction between these components.
In this tutorial you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The problem to be considered is shown schematically in Figure 9.1. The rotor and stator
consist of 9 and 12 blades, respectively. A steady-state solution for this configuration
using only one rotor blade and one stator blade is desired. Since the periodic angles for
the rotor and stator are different, a mixing plane must be used at the interface.
The mixing plane is defined at the rotor outlet/stator inlet. The grid is set up with
periodic boundaries on either side of the rotor and stator blades. A pressure inlet is used
at the upstream boundary and a pressure outlet at the downstream boundary. Ambient
air is drawn into the fan (at 0 Pa gauge total pressure) and is exhausted back out to the
ambient environment (0 Pa static pressure). The hub and blade of the rotor are assumed
to be rotating at 1800 rpm.
c Fluent Inc. January 11, 2005 9-1
Using the Mixing Plane Model
stator
rotor
outlet
inlet
z x ω = 1800 rpm
2. Unzip mixing_plane.zip.
fanstage.msh can be found in the /mixing plane folder created after unzipping
the file.
9-2
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
Step 1: Grid
1. Read the grid file (fanstage.msh).
File −→ Read −→Case...
As FLUENT reads the grid file, it will report its progress in the console window.
c Fluent Inc. January 11, 2005 9-3
Using the Mixing Plane Model
Z X
Grid
FLUENT 6.2 (3d, segregated, ske)
Extra: You can use the right mouse button to check which zone number corresponds to
each boundary. If you click the right mouse button on one of the boundaries in the
graphics window, its zone number, name, and type will be printed in the FLUENT
console window. This feature is especially useful when you have several zones of
the same type and you want to distinguish between them quickly.
9-4
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
Step 2: Units
1. For convenience, define new units for angular velocity.
The angular velocity for this problem is known in rpm, which is not the default unit
for angular velocity. You will need to redefine the angular velocity units as rpm.
Define −→Units...
c Fluent Inc. January 11, 2005 9-5
Using the Mixing Plane Model
Step 3: Models
1. Keep the default solver settings.
Define −→ Models −→Solver...
9-6
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
2. Turn on the standard k- turbulence model with standard wall functions.
Define −→ Models −→Viscous...
c Fluent Inc. January 11, 2005 9-7
Using the Mixing Plane Model
3. Click Create.
FLUENT will name the mixing plane by combining the names of the zones selected
as the Upstream Zone and Downstream Zone. This new name will be displayed in
the Mixing Plane list.
The essential idea behind the mixing plane concept is that each fluid zone (stator
and rotor) is solved as a steady-state problem. At some prescribed iteration inter-
val, the flow data at the mixing plane interface are averaged in the circumferential
direction on both the rotor outlet and the stator inlet boundaries. FLUENT uses
these circumferential averages to define “profiles” of flow properties. These profiles
are then used to update boundary conditions along the two zones of the mixing plane
interface.
In this example, profiles of averaged total pressure (p0 ), static pressure (ps ), direc-
tion cosines of the local flow angles in the radial, tangential, and axial directions
(αr , αt , αz ), total temperature (T0 ), turbulence kinetic energy (k), and turbulence
dissipation rate () are computed at the rotor exit and used to update boundary
conditions at the stator inlet. Likewise, the same profiles, except for that of total
9-8
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
pressure are computed at the stator inlet and used as a boundary condition on the
rotor exit. You can view the profiles computed at the rotor exit and stator inlet in
the Boundary Profiles panel.
Define −→Profiles...
You will also see that these profiles appear in the boundary conditions panels for
the rotor exit and stator inlet.
See Section 10.4 of the User’s Guide for more information on mixing planes.
c Fluent Inc. January 11, 2005 9-9
Using the Mixing Plane Model
Step 5: Materials
1. Accept the default properties for air.
Define −→Materials...
For the present analysis, you will model air as an incompressible fluid with a density
of 1.225 kg/m3 and a dynamic viscosity of 1.7894× 10−5 kg/m-s. Since these are
the default values, no change is required in the materials panel.
9-10
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
c Fluent Inc. January 11, 2005 9-11
Using the Mixing Plane Model
3. Specify rotational periodicity for the periodic boundary of the rotor (periodic-11).
9-12
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
4. Specify rotational periodicity for the periodic boundary of the stator (periodic-22).
5. Set the following conditions for the pressure inlet of the rotor (pressure-inlet-rotor).
To model ambient conditions, you use P0 = 0 gauge. The turbulence level is as-
sumed to be low (1% ) and the hydraulic diameter is used as the length scale.
c Fluent Inc. January 11, 2005 9-13
Using the Mixing Plane Model
6. Examine the conditions for the pressure inlet of the stator (pressure-inlet-stator).
The profiles computed at the rotor outlet are used to update the boundary conditions
at the stator inlet. These profiles were set for you automatically when the mixing
plane was created. Therefore, you do not need to set any parameters in this panel.
7. Examine the conditions for the pressure outlet of the rotor (pressure-outlet-rotor).
The Backflow Direction Specification Method was set to Direction Vector when you
created the mixing plane, and the Coordinate System to Cylindrical (like for the stator
inlet ). The values for the direction cosines are taken from the profiles at the stator.
9-14
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
8. Set the conditions for the pressure outlet of the stator (pressure-outlet-stator).
c Fluent Inc. January 11, 2005 9-15
Using the Mixing Plane Model
9. Set the conditions for the inlet hub of the rotor (rotor-inlet-hub).
9-16
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
10. Set the conditions for the shroud of the rotor inlet (rotor-inlet-shroud).
c Fluent Inc. January 11, 2005 9-17
Using the Mixing Plane Model
11. Set the following conditions for the rotor shroud (rotor-shroud).
9-18
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
c Fluent Inc. January 11, 2005 9-19
Using the Mixing Plane Model
9-20
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
Step 7: Solution
1. Set the solution parameters.
Solve −→ Controls −→Solution...
c Fluent Inc. January 11, 2005 9-21
Using the Mixing Plane Model
9-22
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
i. Select Mass Flow Rate from the Report Type drop-down list.
ii. Select pressure-outlet-stator in the Surfaces list.
iii. Click on OK to define the monitor.
(d) Click on OK in the Surface Monitors panel to enable the monitor.
c Fluent Inc. January 11, 2005 9-23
Using the Mixing Plane Model
9-24
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
Calculating until the mass flow rate converges will require significant CPU
! resources. Instead of calculating the solution, you can read the data file
(fanstage.dat) with the pre-calculated solution, and proceed to the post-
processing section of the tutorial (Step 8). This data file can be found in
the directory where you found the mesh file.
The solution will converge after about 640 iterations. However, the residual history
plot is only one indication of solution convergence. Notice that the mass flow rate
has not reached a constant value. To remedy this, you will reduce the convergence
criterion for the continuity equation and iterate until the mass flow rate reaches a
constant value.
7. Save the case and data file (fanstage.cas & fanstage.dat).
File −→ Write −→Case & Data...
8. Reduce the convergence criterion for the continuity equation.
Solve −→ Monitors −→Residual...
(a) Set the Convergence Criterion for continuity to 1e-05 and click OK.
Note: In this case, the reason for continuing the calculation is to obtain bet-
ter global mass conservation; thus, only the convergence tolerance for the
continuity equation is adjusted. In general, the convergence behavior of
the continuity equation is a good indicator of the overall convergence of
the solution.
c Fluent Inc. January 11, 2005 9-25
Using the Mixing Plane Model
Monitors
monitor-1
-0.0020
-0.0040
-0.0060
-0.0080
-0.0100
Mass -0.0120
Flow
-0.0140
Rate
(kg/s) -0.0160
-0.0180
-0.0200
-0.0220
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Y
Z X
Iteration
10. Save the case and data file (fanstage1.cas & fanstage1.dat).
File −→ Write −→Data...
Although the mass flow rate history indicates that the solution is con-
! verged, you should also check the mass fluxes through the domain to ensure
that mass is being conserved.
9-26
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
The net mass imbalance should be a small fraction (say, 0.5%) of the total
!
flux through the system. If a significant imbalance occurs, you should
decrease your residual tolerances by at least an order of magnitude and
continue iterating.
Note: The fluxes for the portions of the rotor and stator that have been modeled are
different. However, the flux for the whole rotor and the whole stator are very
nearly equal: approximately 0.23274 kg/s (0.02586 × 9 rotor blades), versus
approximately 0.23328 kg/s (0.01944 × 12 stator blades).
c Fluent Inc. January 11, 2005 9-27
Using the Mixing Plane Model
Step 8: Postprocessing
1. Create two surfaces for postprocessing, one at y = 0.12 m and one at z = −0.1 m.
Surface −→Iso-Surface...
The surface y = 0.12 m is a midspan slice through the grid. This view is good
for looking at the blade-to-blade flow field. The surface z = −0.1 m is an axial
plane downstream of the stator. This will be used to plot circumferentially-averaged
profiles.
9-28
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
c Fluent Inc. January 11, 2005 9-29
Using the Mixing Plane Model
2.77e+01
2.63e+01
2.49e+01
2.35e+01
2.22e+01
2.08e+01
1.94e+01
1.80e+01
1.66e+01
1.53e+01
1.39e+01
1.25e+01
1.11e+01
9.73e+00
8.35e+00
6.97e+00
5.58e+00
4.20e+00
2.82e+00
1.44e+00 X
Y
5.63e-02
Z
Plotting the velocity field in this manner gives a good indication of the midspan
flow over the stator. For the rotor, it is instructive to similarly plot the relative
velocity field.
> plot
/plot> circum-avg-radial
averages of> total-pressure
on surface [] 17
number of bands [5] 15
9-30
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
(b) Enter the name of the output file as circum-plot.xy when prompted.
c Fluent Inc. January 11, 2005 9-31
Using the Mixing Plane Model
3.75e+01
3.50e+01
3.25e+01
3.00e+01
2.75e+01
Total
Pressure 2.50e+01
2.25e+01
2.00e+01
1.75e+01
1.50e+01
0.10 0.11 0.11 0.12 0.12 0.12 0.13 0.14 0.14
Y
Z X Radius
Circumferential Averages
FLUENT 6.2 (3d, segregated, ske)
Figure 9.5: Plot of Circumferential Average of the Total Pressure on the Plane z = −0.1.
9-32
c Fluent Inc. January 11, 2005
Using the Mixing Plane Model
5.04e+02
4.66e+02
4.29e+02
3.91e+02
3.53e+02
3.15e+02
2.77e+02
2.39e+02
2.02e+02
1.64e+02
1.26e+02
8.82e+01
5.04e+01
1.26e+01
-2.52e+01
-6.30e+01
-1.01e+02
-1.39e+02
-1.76e+02 Y
-2.14e+02
Z X
-2.52e+02
Figure 9.6: Contours of Total Pressure for the Rotor Blade and Hub
Summary
This tutorial has demonstrated the use of the mixing plane model for a typical axial flow
turbomachine configuration. The mixing plane model is useful for predicting steady-
state flow in a turbomachine stage, where local interaction effects (such as wake and
shock wave interaction) are secondary. If local effects are important, then an unsteady,
sliding mesh calculation is required.
c Fluent Inc. January 11, 2005 9-33
Using the Mixing Plane Model
9-34
c Fluent Inc. January 11, 2005
Tutorial 10. Using Sliding Meshes
Introduction
In this tutorial, the sliding mesh capability of FLUENT is used to predict the time-
dependent flow through a two-dimensional rotor-stator blade row. The time-varying
rotor-stator interaction is modeled by allowing the mesh associated with the moving
rotor to translate (slide) relative to the stationary mesh associated with the stator blade.
In this tutorial you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The rotor-stator geometry considered in this tutorial is shown in Figure 10.1. The geom-
etry consists of a planar slice through the rotor and stator blades, extracted by unrolling
a plane of constant radius (R = 0.686 m) in an axial flow turbomachine. The speed
of rotation, 410 RPM, yields a linear velocity of the rotor, RΩ, equal to 29.4 m/s, as
indicated in the figure. The fluid, assumed to be air, enters the stator row at the specified
total pressure and temperature and exits the rotor at the specified exit static pressure.
The inlet Mach number is 0.07 and the flow will be treated as compressible.
c Fluent Inc. January 11, 2005 10-1
Using Sliding Meshes
Cx
= 0.1524 m
P01 = 101325 Pa
blade pitch = 0.1959 m
T01 = 300 K
M = 0.07
P2 = 97576 Pa
10-2
c Fluent Inc. January 11, 2005
Using Sliding Meshes
2. Unzip sliding_mesh.zip.
rotor.msh and stator.msh can be found in the /sliding mesh folder created after
unzipping the file.
Note: The geometries of the rotor and stator flow domains have been meshed separately.
This is the usual procedure when the sliding mesh capability is used: separate mesh
files are created for the sliding and stationary mesh regions. This ensures that the
sliding interface between the two regions is defined by two separate boundary zones
which share no common nodes. The two separate mesh files must be merged prior
to reading them into FLUENT, as detailed in Step 1, below.
c Fluent Inc. January 11, 2005 10-3
Using Sliding Meshes
2. Provide the mesh file names, rotor.msh and stator.msh, as prompted. Provide
scaling factors of 1 and translations and rotations of zero for each mesh file. Save
the new merged mesh file as slide.msh.
The mesh files must be read into tmerge in this order for the tutorial to run
! as written. Otherwise, zone names and numbers will be assigned differently
when the files are merged together. In general, however, you can specify
files to be read into tmerge in any order.
10-4
c Fluent Inc. January 11, 2005
Using Sliding Meshes
Step 2: Grid
1. Start the 2D version of FLUENT.
(a) Select in in the Units Conversion drop-down list to complete the phrase Grid
Was Created In in (inches).
(b) Click on Scale to scale the grid.
The final domain extents should appear as in the panel above.
c Fluent Inc. January 11, 2005 10-5
Using Sliding Meshes
Note: You can use the mouse probe button (right button, by default) to find out the
boundary zone labels. As annotated in Figure 10.3, the upstream boundary is a
pressure inlet, the downstream boundary is a pressure outlet, and the lateral top
and bottom boundaries are periodic. The stator blade and stator-side fluid are
identified as wall-7 and fluid-9. The rotor blade and rotor-side fluid are wall-16
and fluid-18. (Your mouse probe will report the fluid regions as interior-8 and
interior-17 zones, for the stator and rotor sides, respectively. These are the
face zones associated with the fluid regions.)
To determine which fluid zone is the stator-side fluid, you can create the fluid-
9 and fluid-18 display surfaces using the Zone Surface panel. Then, display the
zones (one at a time) using the Grid Display panel.
If you wish to annotate your own graphics display, you can use the Annotate
panel.
Display −→Annotate...
10-6
c Fluent Inc. January 11, 2005
Using Sliding Meshes
Grid
FLUENT 6.2 (2d, segregated, lam)
periodic-8
periodic-15
fluid-18
pressure-
pressure-outlet-14
inlet-3 wall-7
wall-16
Grid
FLUENT 6.2 (2d, segregated, lam)
c Fluent Inc. January 11, 2005 10-7
Using Sliding Meshes
Step 3: Models
1. Select the coupled explicit solver.
Define −→ Models −→Solver...
Note: Initially, you will solve for the steady flow through the blade passage. Later,
after obtaining the steady flow as the starting point for the transient calcula-
tion, you will revisit this panel to turn on time-dependent flow.
10-8
c Fluent Inc. January 11, 2005
Using Sliding Meshes
Note: The Reynolds number of the flow is about 105 , and the flow will be treated
as fully turbulent.
c Fluent Inc. January 11, 2005 10-9
Using Sliding Meshes
Step 4: Materials
1. Select air (the default material) as the fluid material, and use the ideal-gas law to
compute density. Retain the default values for all other properties.
Define −→Materials...
10-10
c Fluent Inc. January 11, 2005
Using Sliding Meshes
Here, the operating pressure is set to zero and boundary condition inputs for pressure
will be defined in terms of absolute pressures. Boundary condition inputs should
always be relative to the value used for operating pressure.
c Fluent Inc. January 11, 2005 10-11
Using Sliding Meshes
10-12
c Fluent Inc. January 11, 2005
Using Sliding Meshes
c Fluent Inc. January 11, 2005 10-13
Using Sliding Meshes
3. Keep the default Momentum boundary conditions for the stator blades (wall-7) and
the rotor blades (wall-16).
The velocity of a “non-moving” wall is assumed to match that of the adjacent mesh
region, yielding a no-slip condition in the reference frame of the mesh. Thus, FLU-
ENT will assume that the stator blade is stationary in the non-moving reference
frame of the stator mesh. Similarly, FLUENT will assume that the rotor blade is
moving at the grid speed in the sliding rotor region. Therefore, you will not modify
the wall velocity of the rotor (wall-16), even though the rotor is moving. The default
setting of a non-moving wall is correct and implies zero velocity in the moving ref-
erence frame of the sliding region. (The motion of the mesh region will be defined
as a boundary condition for the fluid zone, below.)
10-14
c Fluent Inc. January 11, 2005
Using Sliding Meshes
c Fluent Inc. January 11, 2005 10-15
Using Sliding Meshes
10-16
c Fluent Inc. January 11, 2005
Using Sliding Meshes
c Fluent Inc. January 11, 2005 10-17
Using Sliding Meshes
10-18
c Fluent Inc. January 11, 2005
Using Sliding Meshes
c Fluent Inc. January 11, 2005 10-19
Using Sliding Meshes
10-20
c Fluent Inc. January 11, 2005
Using Sliding Meshes
c Fluent Inc. January 11, 2005 10-21
Using Sliding Meshes
4. Enable the monitoring of the lift force on the rotor blade (wall-16).
Note: Monitoring forces on the rotor blade provides a good measure of convergence
during the initial steady-state flow prediction. Here, you will request dynamic
plotting of lift as the solution proceeds. In addition, you will write the lift
information to a file, cl-hist.ss. You could choose to monitor any other
variable (e.g., mass flow rate), including a custom field function.
Solve −→ Monitors −→Force...
10-22
c Fluent Inc. January 11, 2005
Using Sliding Meshes
c Fluent Inc. January 11, 2005 10-23
Using Sliding Meshes
0.0000
-2.0000
-4.0000
-6.0000
Cl -8.0000
-10.0000
-12.0000
-14.0000
-16.0000
0 50 100 150 200 250 300 350 400 450 500
Iterations
8. Save the case and data files (slide ss.cas and slide ss.dat).
File −→ Write −→Case & Data...
Note: If you choose a file name that already exists in the current directory, FLU-
ENT will prompt you for confirmation to overwrite the file.
10-24
c Fluent Inc. January 11, 2005
Using Sliding Meshes
c Fluent Inc. January 11, 2005 10-25
Using Sliding Meshes
10. Enable the display of three periodic repeats of the solution domain (Figure 10.5).
Display −→Views...
10-26
c Fluent Inc. January 11, 2005
Using Sliding Meshes
1.42e+02
1.35e+02
1.28e+02
1.20e+02
1.13e+02
1.06e+02
9.93e+01
9.23e+01
8.53e+01
7.82e+01
7.12e+01
6.41e+01
5.71e+01
5.00e+01
4.30e+01
3.60e+01
2.89e+01
2.19e+01
1.48e+01
7.79e+00
7.49e-01
c Fluent Inc. January 11, 2005 10-27
Using Sliding Meshes
11. Display the steady flow contours of static pressure (Figure 10.6).
Display −→Contours...
The steady flow prediction shows the expected pressure distribution through the pas-
sage, with low pressure on the suction surfaces and stagnation around the impinge-
ment point on the rotor.
10-28
c Fluent Inc. January 11, 2005
Using Sliding Meshes
1.02e+05
1.01e+05
1.01e+05
9.99e+04
9.92e+04
9.86e+04
9.79e+04
9.72e+04
9.65e+04
9.59e+04
9.52e+04
9.45e+04
9.38e+04
9.31e+04
9.25e+04
9.18e+04
9.11e+04
9.04e+04
8.97e+04
8.91e+04
8.84e+04
c Fluent Inc. January 11, 2005 10-29
Using Sliding Meshes
10-30
c Fluent Inc. January 11, 2005
Using Sliding Meshes
where CFL is the Courant number. Explicit time-stepping might be the optimum
solution strategy if you are modeling a traveling shock wave, where the Courant
condition is ideal for determination of the time step value.
(a) Scroll down and select Moving Mesh in the Motion Type drop-down list.
(b) Change the Translational Velocity in the Y direction to -29.445.
3. Save the case and data files (slide un.cas and slide un.dat).
File −→ Write −→Case & Data...
The mesh changes during the preview so be sure to save the case before mesh pre-
view.
c Fluent Inc. January 11, 2005 10-31
Using Sliding Meshes
5. Read the case and data files back into FLUENT (slide un.cas and slide un.dat).
Note: As the mesh preview option advances the time step, thereby updating the
mesh, it is essential to reread the case file to ensure that the mesh is at the
proper position before the calculation is started. If the case file is not read
again and the solution is initialized after the mesh preview, the solution time
is initialized to zero but the mesh does not go back to its original position.
10-32
c Fluent Inc. January 11, 2005
Using Sliding Meshes
The Courant number controls the time step used by FLUENT during the inner iter-
ations performed during each time step. A Courant number of 1 is a conservative
setting that should ensure the stability of the inner iterations.
c Fluent Inc. January 11, 2005 10-33
Using Sliding Meshes
If you do not clear the old force-monitoring data, FLUENT will plot them
! with the new data, corrupting the new lift coefficient plot.
(e) Click on the Axes... button.
This will open the Axes - Force Monitor Plot panel.
10-34
c Fluent Inc. January 11, 2005
Using Sliding Meshes
c Fluent Inc. January 11, 2005 10-35
Using Sliding Meshes
Note: Monitoring the lift force is an ideal way to determine when the transient flow
prediction becomes time-periodic (independent of the initial condition). In the
time-periodic solution, the lift force variation will repeat identically from one
passing period to the next.
Using a time step of 0.0001 second, 67 time steps will be performed as the
rotor performs one pass.
10-36
c Fluent Inc. January 11, 2005
Using Sliding Meshes
The maximum number of iterations per time step should be set large
! enough so that the inner iterations converge before the solution moves
to the next time value. The value of 20, selected here, is quite small: it
is likely that the initial time steps will not fully converge within 20 inner
iterations. While this would be unsuitable for prediction of most time-
dependent flows, the current simulation does not require high accuracy
during the initial time steps. The rotor-stator flow prediction will be con-
tinued in time until a time-periodic flow is obtained. Low accuracy during
the initial passing periods is acceptable as long as convergence is achieved
during each time step of the final passing periods.
c Fluent Inc. January 11, 2005 10-37
Using Sliding Meshes
Calculation of 1000 time steps will require significant CPU resources. In-
! stead of calculating, you can read the case and data files saved after 0.1
seconds:
After reading the files, skip to Step 11: Postprocessing at t=0.1 Second.
(The case and data files are available in the same directory where you found
the mesh files.)
By requesting 1000 time steps, you are asking FLUENT to compute until time is
equal to 0.1 second. This will include roughly 15 passing periods (15 × 6.7e-3 sec
= 0.10 sec). Experience shows that the flow becomes time-periodic after about 12
passing periods. The lift history display allows you to confirm this. Your lift force
history should be similar to that shown in Figure 10.7. Notice the periodicity of the
lift coefficient after approximately 0.05 seconds.
1.00e+01
5.00e+00
0.00e+00
-5.00e+00
Cl
-1.00e+01
-1.50e+01
-2.00e+01
-2.50e+01
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
Time
10-38
c Fluent Inc. January 11, 2005
Using Sliding Meshes
c Fluent Inc. January 11, 2005 10-39
Using Sliding Meshes
-7.4000
-7.6000
-7.8000
-8.0000
-8.2000
Cl
-8.4000
-8.6000
-8.8000
-9.0000
-9.2000
0.020 0.030 0.040 0.050 0.060 0.070 0.080 0.090 0.100 0.110
Time
7. Save the case and data files at t = 0.1 second (slide01.cas and slide01.dat).
File −→ Write −→Case & Data...
When the sliding mesh model is used, you must save a case file whenever a
! data file is saved. This is because the case file contains the grid information,
which is changing with time.
10-40
c Fluent Inc. January 11, 2005
Using Sliding Meshes
For sliding mesh cases, it is important that you read the associated case
! file whenever you read a data file, because the case file contains the grid
information, which is changing with time.
9.27e+01
8.82e+01
8.36e+01
7.91e+01
7.45e+01
6.99e+01
6.54e+01
6.08e+01
5.63e+01
5.17e+01
4.72e+01
4.26e+01
3.80e+01
3.35e+01
2.89e+01
2.44e+01
1.98e+01
1.52e+01
1.07e+01
6.13e+00
1.57e+00
Note: The velocity vectors in Figure 10.9 are displayed with respect to the absolute
reference frame (default). If you want to display the vectors with respect to the
moving reference frame, you can select Relative Velocity in the Vectors of drop-
down list and select the rotor fluid zone as the Reference Zone in the Reference
Values panel.
c Fluent Inc. January 11, 2005 10-41
Using Sliding Meshes
1.01e+05
1.01e+05
1.01e+05
1.00e+05
9.98e+04
9.94e+04
9.90e+04
9.86e+04
9.82e+04
9.78e+04
9.75e+04
9.71e+04
9.67e+04
9.63e+04
9.59e+04
9.55e+04
9.51e+04
9.47e+04
9.43e+04
9.39e+04
9.35e+04
Figure 10.10: Contours of Static Pressure at Time = 0.1 Second: Unsteady Flow
Note: Slight discontinuities in the pressure contours along the sliding interface are
expected. This is because the contour plotting uses one-sided interpolation on
either side of the sliding plane. This is purely a display issue.
10-42
c Fluent Inc. January 11, 2005
Using Sliding Meshes
4. Plot the instantaneous pressure coefficient on the rotor blade at t = 0.1 second
(Figure 10.11).
Plot −→XY Plot...
c Fluent Inc. January 11, 2005 10-43
Using Sliding Meshes
(a) Select Pressure... and Pressure Coefficient in the Y Axis Function drop-down
lists.
(b) In the Surfaces list, select wall-16.
(c) Click Plot.
wall-16
-2.00e+00
-4.00e+00
-6.00e+00
-8.00e+00
-1.00e+01
Pressure -1.20e+01
Coefficient
-1.40e+01
-1.60e+01
-1.80e+01
-2.00e+01
-2.20e+01
0.175 0.2 0.225 0.25 0.275 0.3 0.325 0.35 0.375 0.4
Position (m)
Figure 10.11: Pressure Coefficient on the Moving Rotor at Time = 0.1 Second
10-44
c Fluent Inc. January 11, 2005
Using Sliding Meshes
(a) Set the Autosave Case File Frequency and Autosave Data File Frequency to 5.
(b) In the Filename field, enter slide.gz.
(c) Click OK.
FLUENT will append the time step value to the file name prefix (slid). The
standard extensions (.cas and .dat will also be appended. This will yield file
names of the form slid1005.cas and slid1005.dat, where 1005 is the time
step number.
The extension “.gz”, supplied in the panel above, instructs FLUENT to save
the case and data files in compressed format, yielding file names of the form
slid1005.cas.gz and slid1005.dat.gz. If you leave off the “.gz” exten-
sion, FLUENT will save the files as slid1005.cas and slid1005.dat, etc.
When the sliding mesh model is used, you must save a case file whenever a
! data file is saved. This is because the case file contains the grid information,
which is changing with time.
Extra: If you want to generate a solution animation by plotting, for example, pres-
sure contours every 5 time steps, you can use the Solution Animation panel to
set up the animation before you begin the calculation. Tutorial 4 demonstrates
how to do this.
c Fluent Inc. January 11, 2005 10-45
Using Sliding Meshes
10-46
c Fluent Inc. January 11, 2005
Using Sliding Meshes
-7.90e+00
-8.00e+00
-8.10e+00
Cl -8.20e+00
-8.30e+00
-8.40e+00
-8.50e+00
0.1 0.101 0.102 0.103 0.104 0.105 0.106 0.107
Time
4. Examine the results at different time steps within a single passing period.
(a) Read in the case and data files of interest.
File −→ Read −→Case & Data...
(b) Display contours of static pressure.
The display of pressure contours every five time steps will show the time-
varying pressure distribution and the motion of the rotor. Examples of two
pressure contour plots at t = 0.1030 and t = 1.070 seconds are shown in
Figures 10.13 and 10.14, respectively.
Display −→Contours...
Extra: If you generated a solution animation during the calculation as mentioned
earlier in this tutorial, you could play it back inside FLUENT to see the pressure
contour animation over time. See Tutorial 4 for details.
c Fluent Inc. January 11, 2005 10-47
Using Sliding Meshes
1.01e+05
1.01e+05
1.01e+05
1.00e+05
1.00e+05
9.97e+04
9.93e+04
9.90e+04
9.86e+04
9.82e+04
9.79e+04
9.75e+04
9.72e+04
9.68e+04
9.65e+04
9.61e+04
9.58e+04
9.54e+04
9.51e+04
9.47e+04
9.43e+04
1.01e+05
1.01e+05
1.01e+05
1.00e+05
9.98e+04
9.94e+04
9.91e+04
9.87e+04
9.83e+04
9.79e+04
9.75e+04
9.71e+04
9.68e+04
9.64e+04
9.60e+04
9.56e+04
9.52e+04
9.48e+04
9.45e+04
9.41e+04
9.37e+04
10-48
c Fluent Inc. January 11, 2005
Using Sliding Meshes
Summary
In this tutorial, you have modeled the time-periodic flow involving rotor-stator interac-
tion. You have learned how to merge the separate rotor and stator meshes using tmerge,
and to create the grid-interface zones along the sliding interface. Similar procedures can
be used to tie together meshes for non-sliding mesh analysis.
You have used FLUENT’s time-dependent flow prediction capability, and you have learned
how to set solution parameters for implicit time-stepping. These time-dependent flow
prediction procedures can also be applied to other, non-sliding mesh, analysis. The
procedures in this tutorial, however, are applicable to time-periodic calculations, in which
the initial condition and initial phase of the transient calculation are treated without
concern for time accuracy.
You have also learned how to manage the file saving and graphical postprocessing for
time-dependent flows, using file autosaving and command monitors to automatically save
solution information as the transient calculation proceeds.
c Fluent Inc. January 11, 2005 10-49
Using Sliding Meshes
10-50
c Fluent Inc. January 11, 2005
Tutorial 11. Using Dynamic Meshes
Introduction
This tutorial provides information for performing basic dynamic mesh calculations. In
addition to combining the basic mesh-motion schemes, this tutorial will introduce rigid-
body motion of a cell zone. This is useful for a multitude of realistic cases with moving
meshes.
In this tutorial you will learn how to:
• Use the dynamic mesh capability of FLUENT to solve a simple flow-driven rigid-
body motion problem.
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1. Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The problem to be considered is shown schematically in Figure 11.1. A 2D axisymmetric
valve geometry is used, consisting of a pressurized cavity on the left, driving the motion
of a poppet that toggles the flow to the circumferential pressure outlets. A spring force
is also acting on the poppet. In this case the transient closure of the valve is studied.
c Fluent Inc. January 11, 2005 11-1
Using Dynamic Meshes
pressure outlets
mass
flow
inlet
moving
poppet
2. Unzip dynamic_mesh.zip.
valve.msh and valve.c can be found in the /dynamic mesh folder created after
unzipping the file.
A user-defined function will be used to define the rigid-body motion of the poppet
in the valve geometry. This function has already been written (valve.c). You will
only need to compile it within FLUENT.
11-2
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
Step 1: Grid
1. Read the grid file valve.msh.
File −→ Read −→Case...
(a) Under Unit Conversion, select in from the drop-down list to complete the phrase
Grid Was Created In in (inches).
(b) Click Scale to scale the grid.
(c) Click Change Length Units to set inches as the working units for length, and
then close the panel.
c Fluent Inc. January 11, 2005 11-3
Using Dynamic Meshes
Grid
FLUENT 6.2 (2d, segregated, lam)
11-4
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
Step 2: Units
1. For convenience, define new units for pressure and mass flow.
In the problem description, pressure, length, and mass flow are specified in psi, in,
and gpm, respectively. While the units for length were switched while scaling the
grid in the previous step, psi and gpm are not the default units for pressure and
mass flow.
Define −→Units...
c Fluent Inc. January 11, 2005 11-5
Using Dynamic Meshes
Step 3: Models
1. Enable an axisymmetric time-dependent calculation.
Define −→ Models −→Solver...
Dynamic mesh simulations currently work only with first-order time ad-
! vancement.
(d) Click OK.
11-6
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
(a) Select k-epsilon as the Model, and retain the default setting of Standard under
k-epsilon Model.
(b) Click OK.
c Fluent Inc. January 11, 2005 11-7
Using Dynamic Meshes
Step 4: Materials
You will create a new material called oil.
Define −→Materials...
4. Click Change/Create.
11-8
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
c Fluent Inc. January 11, 2005 11-9
Using Dynamic Meshes
1. Set the conditions for the mass flow inlet (inlet) as shown in the following figure.
2. Click OK.
11-10
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
3. Set the conditions for the exit boundary (outlet) as shown in the following figure.
5. Select oil as the Material Name from the drop-down list, and keep the default Motion
Type as Stationary.
This replaces air with oil as a working media.
6. Repeat this procedure for the other two fluid zones, inlet zone and outlet zone.
c Fluent Inc. January 11, 2005 11-11
Using Dynamic Meshes
11-12
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
Note that in general, you will need to set boundary conditions for these new wall
zones, when they are not empty. In this case, default settings are used.
c Fluent Inc. January 11, 2005 11-13
Using Dynamic Meshes
11-14
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
c Fluent Inc. January 11, 2005 11-15
Using Dynamic Meshes
i. Under Options, be sure that the Must Improve Skewness option is selected.
ii. Specify other parameters as shown the following table:
Parameter Value
Minimum Length Scale 2.0e-5
Maximum Length Scale 0.007
Maximum Cell Skewness 0.7
Size Remesh Interval 1
If a cell exceeds these limits, the cell is marked for remeshing. Therefore,
you will always need to specify problem-specific values under Remeshing
Parameters.
(e) Click OK.
4. Specify the motion of the poppet, the adjacent walls, and the fluid region left of
the poppet.
The poppet motion and the motion of the deforming wall side-wall-3 are specified by
means of the UDF valve.
Define −→ Dynamic Mesh −→Zones...
11-16
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
c Fluent Inc. January 11, 2005 11-17
Using Dynamic Meshes
11-18
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
(c) Specify the motion of the deforming wall corresponding to the radial boundary
of the deforming fluid zone next to the poppet (int int).
i. In the Zone Names drop-down list, select int int.
ii. Under Type, keep the previous selection of Deforming.
iii. Click the Geometry Definition tab.
iv. In the Definition drop-down list, select plane.
The panel will expand again to show the inputs for a planar geometry.
v. Under Point on Plane, enter 0, 0.22625.
vi. Under Plane Normal, keep the previous setting of 0, 1.
vii. Click the Meshing Options tab.
viii. Under Methods, be sure that Smoothing and Remeshing are selected, and
keep the previous settings for Maximum Skewness to value of 1.
ix. Specify a value of 0.0045 for Minimum Length Scale and a value of 0.0055
for Maximum Length Scale.
x. Click Create.
In many MDM problems, you may want to preview the mesh motion before proceed-
ing any further. In this problem, the mesh motion is driven by the pressure exerted
by the fluid on the poppet and acting against the inertia of the poppet and the force
of a preloaded spring attached to it. Hence, for this problem, mesh motion in the
absence of a flow field solution is meaningless, and you will not use this feature
here.
c Fluent Inc. January 11, 2005 11-19
Using Dynamic Meshes
Step 9: Solution
1. Set the solution parameters.
Solve −→ Controls −→Solution...
(a) Keep all default discretization schemes and values for under-relaxation factors.
This problem has been found to converge satisfactorily with these default set-
tings. Alternatively, you may want to try the PISO discretization scheme for
Pressure-Velocity Coupling in conjunction with higher under-relaxation factors
in order to reduce the overall CPU time needed for this simulation.
(b) Click OK.
3. Request that case and data files are automatically saved every 10 time steps.
File −→ Write −→Autosave...
11-20
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
(a) Set the Autosave Case File Frequency and Autosave Data File Frequency to 10.
To retain all files, keep the Overwrite Existing Files inactive.
(b) In the Filename field, enter valve.
When FLUENT saves a file, it will append the time step value to the file name
prefix (valve). The standard extensions (.cas and .dat) will also be ap-
pended.
(c) Click OK.
c Fluent Inc. January 11, 2005 11-21
Using Dynamic Meshes
5. Create animation sequences for the static pressure contour plots and velocity vectors
plots in the valve.
You will use FLUENT’s solution animation feature to save contour plots of temper-
ature every 5 time steps. After the calculation is complete, you will use the solution
animation playback feature to view the animated temperature plots over time.
Solve −→ Animate −→Define...
11-22
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
c Fluent Inc. January 11, 2005 11-23
Using Dynamic Meshes
8.00e+01
7.67e+01
7.35e+01
7.02e+01
6.69e+01
6.37e+01
6.04e+01
5.71e+01
5.39e+01
5.06e+01
4.74e+01
4.41e+01
4.08e+01
3.76e+01
3.43e+01
3.10e+01
2.78e+01
2.45e+01
2.12e+01
1.80e+01
1.47e+01
11-24
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
c Fluent Inc. January 11, 2005 11-25
Using Dynamic Meshes
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
3.10e+00
11-26
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
7. Save the initial case and data files (valve.cas and valve.dat).
File −→ Write −→Case & Data...
Extra: If you decide to read in the case file that is provided for this tutorial on the
documentation CD, you will need to compile the UDF associated with this tutorial
in your working directory. This is necessary because FLUENT will expect to find
the correct UDF libraries in your working directory when reading the case file.
The UDF (valve.c) that is provided can be edited and customized by changing the
parameters as required for your case. In this tutorial, the values necessary for this
case were preset in the source code. These values may be modified to best suit your
model.
c Fluent Inc. January 11, 2005 11-27
Using Dynamic Meshes
8.17e+02
7.60e+02
7.04e+02
6.47e+02
5.90e+02
5.33e+02
4.77e+02
4.20e+02
3.63e+02
3.07e+02
2.50e+02
1.93e+02
1.36e+02
7.97e+01
2.29e+01
-3.38e+01
-9.05e+01
-1.47e+02
-2.04e+02
-2.61e+02
-3.17e+02
11-28
c Fluent Inc. January 11, 2005
Using Dynamic Meshes
7.77e+01
7.38e+01
6.99e+01
6.60e+01
6.21e+01
5.83e+01
5.44e+01
5.05e+01
4.66e+01
4.28e+01
3.89e+01
3.50e+01
3.11e+01
2.73e+01
2.34e+01
1.95e+01
1.56e+01
1.18e+01
7.88e+00
4.00e+00
1.23e-01
c Fluent Inc. January 11, 2005 11-29
Using Dynamic Meshes
(b) Set the slider bar above Replay Speed about halfway in between Slow and Fast.
(c) Keep the default settings in the rest of the panel and click the play button
(the second from the right in the group of buttons under Playback).
See Tutorial 4 and Section 26.20 of the User’s Guide for additional information
on animating the solution.
Summary
In this tutorial you learned how to use the dynamic mesh feature of FLUENT to simulate
the rigid-body motion of a valve poppet in a flow field, driven by the flow-generated
forces, and spring and inertial forces, by means of a user defined function (UDF).
11-30
c Fluent Inc. January 11, 2005
Tutorial 12. Modeling Species Transport and Gaseous
Combustion
Introduction
This tutorial examines chemical species mixing and combustion of a gaseous fuel. A cylin-
drical combustor burning methane (CH4 ) in air is studied using the finite-rate chemistry
model in FLUENT.
In this tutorial you will learn how to:
• Enable physical models, select material properties, and define boundary conditions
for a turbulent flow with chemical species mixing and reaction
• Initiate and solve the combustion simulation using the segregated solver
• Compare the results computed with constant and variable specific heat
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
You may find it helpful to read Chapter 13 of the User’s Guide to learn more about
chemical reaction modeling. Otherwise, no previous experience with chemical reaction
or combustion modeling is assumed.
Problem Description
The cylindrical combustor considered in this tutorial is shown in Figure 12.1. The flame
considered is a turbulent diffusion flame. A small nozzle in the center of the combustor
introduces methane at 80 m/s. Ambient air enters the combustor coaxially at 0.5 m/s.
c Fluent Inc. January 11, 2005 12-1
Modeling Species Transport and Gaseous Combustion
The overall equivalence ratio is approximately 0.76 (about 28% excess air). The high-
speed methane jet initially expands with little interference from the outer wall, and
entrains and mixes with the low-speed air. The Reynolds number based on the methane
jet diameter is approximately 5.7 × 103 .
0.225 m
0.005m
1.8 m
12-2
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
Background
In this tutorial, you will use the generalized finite-rate chemistry model to analyze the
methane-air combustion system. The combustion will be modeled using a global one-
step reaction mechanism, assuming complete conversion of the fuel to CO2 and H2 O. The
reaction equation is
2. Unzip species_transport.zip.
gascomb.msh can be found in the /species transport folder created after unzip-
ping the file.
c Fluent Inc. January 11, 2005 12-3
Modeling Species Transport and Gaseous Combustion
Step 1: Grid
1. Read the grid file gascomb.msh.
File −→ Read −→Case...
After reading the grid file, FLUENT will report that 1615 quadrilateral fluid cells
have been read, along with a number of boundary faces with different zone identi-
fiers.
Extra: You can use the right mouse button to check which zone number corresponds
to each boundary. If you click the right mouse button on one of the boundaries
in the graphics window, its zone number, name, and type will be printed in the
FLUENTconsole window. This feature is especially useful when you have several
zones of the same type and you want to distinguish between them quickly.
12-4
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
Grid
FLUENT 6.2 (2d, segregated, lam)
c Fluent Inc. January 11, 2005 12-5
Modeling Species Transport and Gaseous Combustion
Step 2: Models
1. Define the domain as axisymmetric, and keep the default (segregated) solver.
Define −→ Models −→Solver...
12-6
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
The panel will expand to provide further options. Click OK to accept the default
Standard model and parameters.
c Fluent Inc. January 11, 2005 12-7
Modeling Species Transport and Gaseous Combustion
12-8
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
In this axisymmetric model, the centerline should be treated using the axis
boundary condition instead of symmetry. You will change the symmetry zone
to an axis boundary in Step 4: Boundary Conditions.
The console window will also list the properties that are required for the models
you have enabled. You will see an Information dialog box, reminding you to
confirm the property values that have been extracted from the database.
c Fluent Inc. January 11, 2005 12-9
Modeling Species Transport and Gaseous Combustion
Step 3: Materials
Define −→Materials...
The Materials panel shows the mixture material, methane-air, that was enabled in the
Species Model panel. The properties for this mixture material have been copied from the
FLUENT database and can be modified by you.
Here, you will modify the default setting for the mixture by enabling the gas law. By
default, the mixture material uses constant properties: you will retain this constant prop-
erty assumption for now, allowing only the mixture density to vary with temperature and
composition. The influence of variable property inputs on the combustion prediction will
be examined in a later part of this tutorial.
12-10
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
You can add or remove species from the mixture material using this panel. Here,
the species that make up the methane-air mixture are predefined and require no
modification.
4. In the Materials panel, click the Edit... button to the right of the Reaction drop-
down list.
This will open the Reactions panel.
c Fluent Inc. January 11, 2005 12-11
Modeling Species Transport and Gaseous Combustion
The eddy-dissipation reaction model ignores chemical kinetics (the Arrhenius rate)
and uses only the Mixing Rate parameters in the Reactions panel. The Arrhenius
Rate section of the panel is therefore inactive. (The Rate Exponent and Arrhenius
Rate entries are included in the database and are employed when the alternate finite-
rate/eddy-dissipation model is used.) See the User’s Guide for details.
5. Accept the default settings for the Mixing Rate constants by clicking the OK button.
6. In the Materials panel, select constant from the drop-down list next to Cp and enter
1000 for the specific heat value.
7. Use the scroll bar to review the remaining properties. Click on the Change/Create
button to accept the material property settings and then Close the panel.
12-12
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
As noted above, the initial calculation will be performed assuming that all properties except
density are constant. Using constant transport properties (viscosity, thermal conductivity,
and mass diffusion coefficients) is acceptable here because the flow is fully turbulent. The
molecular transport properties will play a minor role compared to turbulent transport. The
assumption of constant specific heat, in contrast, has a strong effect on the combustion
solution, and you will change this property definition in Step 6: Solution Using Non-
Constant Heat Capacity.
c Fluent Inc. January 11, 2005 12-13
Modeling Species Transport and Gaseous Combustion
(a) Select symmetry-5 in the Zone list and then select axis in the Type list.
You will be prompted to accept the change of boundary type:
(c) In the resulting Axis panel, click OK to accept the default axis zone name.
12-14
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
(a) Rename the boundary as air-inlet in the Zone Name text entry box.
(b) Set the boundary conditions at the air inlet as shown in the panel.
c Fluent Inc. January 11, 2005 12-15
Modeling Species Transport and Gaseous Combustion
(a) Rename this zone fuel-inlet and assign inlet conditions as shown in the
panel.
12-16
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
Note: The Backflow values in this panel are utilized only when backflow occurs
at the pressure outlet. Reasonable values should always be assigned, since
backflow may occur during intermediate iterations and could affect the solution
stability.
c Fluent Inc. January 11, 2005 12-17
Modeling Species Transport and Gaseous Combustion
(a) Rename this boundary outer-wall in the Zone Name text entry box.
(b) Set the thermal condition to Temperature and keep the default temperature of
300 K.
(c) Retain the default settings in the Momentum and Species sections of the panel.
12-18
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
6. Set the boundary conditions for wall-2, which represents the small fuel inlet nozzle.
(a) Rename this boundary nozzle in the Zone Name text entry box.
(b) Accept the default thermal condition of Heat Flux with a value of zero (adia-
batic wall).
(c) Retain the default settings in the Momentum and Species sections of the panel.
c Fluent Inc. January 11, 2005 12-19
Modeling Species Transport and Gaseous Combustion
12-20
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
(a) Use the slider bar next to the Under-Relaxation Factors list to locate each
species and set the under-relaxation factor for all the species to 0.9.
c Fluent Inc. January 11, 2005 12-21
Modeling Species Transport and Gaseous Combustion
12-22
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
7. Review the current state of the solution by viewing contours of temperature (Fig-
ure 12.3).
Display −→Contours...
c Fluent Inc. January 11, 2005 12-23
Modeling Species Transport and Gaseous Combustion
3.06e+03
2.92e+03
2.78e+03
2.65e+03
2.51e+03
2.37e+03
2.23e+03
2.09e+03
1.96e+03
1.82e+03
1.68e+03
1.54e+03
1.40e+03
1.27e+03
1.13e+03
9.90e+02
8.52e+02
7.14e+02
5.76e+02
4.38e+02
3.00e+02
12-24
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
(a) In the drop-down list next to Cp, select mixing-law as the specific heat method.
(b) Click on the Change/Create button to render the mixture specific heat based
on a local mass-fraction-weighted average of all the species.
c Fluent Inc. January 11, 2005 12-25
Modeling Species Transport and Gaseous Combustion
12-26
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
4. Save the new case and data files (gascomb2.cas and gascomb2.dat).
File −→ Write −→Case & Data...
c Fluent Inc. January 11, 2005 12-27
Modeling Species Transport and Gaseous Combustion
Step 7: Postprocessing
Review the solution by examining graphical displays of the results and performing surface
integrations at the combustor exit.
2.30e+03
2.20e+03
2.10e+03
2.00e+03
1.90e+03
1.80e+03
1.70e+03
1.60e+03
1.50e+03
1.40e+03
1.30e+03
1.20e+03
1.10e+03
9.99e+02
8.99e+02
8.00e+02
7.00e+02
6.00e+02
5.00e+02
4.00e+02
3.00e+02
12-28
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
The contours are shown in Figure 12.5. The mixture specific heat is largest where
the CH4 is concentrated, near the fuel inlet, and where the temperature and com-
bustion product concentrations are large. The increase in heat capacity, relative to
the constant value used before, substantially lowers the peak flame temperature.
2.77e+03
2.68e+03
2.59e+03
2.50e+03
2.41e+03
2.33e+03
2.24e+03
2.15e+03
2.06e+03
1.98e+03
1.89e+03
1.80e+03
1.71e+03
1.62e+03
1.54e+03
1.45e+03
1.36e+03
1.27e+03
1.19e+03
1.10e+03
1.01e+03
c Fluent Inc. January 11, 2005 12-29
Modeling Species Transport and Gaseous Combustion
12-30
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
(c) In the Vectors panel, reset the Scale to 0.01 and click Display.
The velocity vectors are shown in Figure 12.6.
9.26e+01
8.80e+01
8.34e+01
7.88e+01
7.42e+01
6.96e+01
6.50e+01
6.03e+01
5.57e+01
5.11e+01
4.65e+01
4.19e+01
3.73e+01
3.26e+01
2.80e+01
2.34e+01
1.88e+01
1.42e+01
9.57e+00
4.95e+00
3.38e-01
c Fluent Inc. January 11, 2005 12-31
Modeling Species Transport and Gaseous Combustion
(a) Select Velocity... and Stream Function in the Contours of drop-down list.
(b) Click Display.
The stream function contours are shown in Figure 12.7. The entrainment of air
into the high-velocity methane jet is clearly visible in the streamline display.
12-32
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
1.55e-02
1.48e-02
1.40e-02
1.32e-02
1.24e-02
1.17e-02
1.09e-02
1.01e-02
9.32e-03
8.55e-03
7.77e-03
6.99e-03
6.21e-03
5.44e-03
4.66e-03
3.88e-03
3.11e-03
2.33e-03
1.55e-03
7.77e-04
0.00e+00
1.00e+00
9.50e-01
9.00e-01
8.50e-01
8.00e-01
7.50e-01
7.00e-01
6.50e-01
6.00e-01
5.50e-01
5.00e-01
4.50e-01
4.00e-01
3.50e-01
3.00e-01
2.50e-01
2.00e-01
1.50e-01
1.00e-01
5.00e-02
0.00e+00
c Fluent Inc. January 11, 2005 12-33
Modeling Species Transport and Gaseous Combustion
2.30e-01
2.19e-01
2.07e-01
1.96e-01
1.84e-01
1.73e-01
1.61e-01
1.50e-01
1.38e-01
1.27e-01
1.15e-01
1.04e-01
9.20e-02
8.05e-02
6.90e-02
5.75e-02
4.60e-02
3.45e-02
2.30e-02
1.15e-02
0.00e+00
1.47e-01
1.39e-01
1.32e-01
1.25e-01
1.17e-01
1.10e-01
1.03e-01
9.53e-02
8.80e-02
8.06e-02
7.33e-02
6.60e-02
5.86e-02
5.13e-02
4.40e-02
3.66e-02
2.93e-02
2.20e-02
1.47e-02
7.33e-03
0.00e+00
12-34
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
1.20e-01
1.14e-01
1.08e-01
1.02e-01
9.60e-02
9.00e-02
8.40e-02
7.80e-02
7.20e-02
6.60e-02
6.00e-02
5.40e-02
4.80e-02
4.20e-02
3.60e-02
3.00e-02
2.40e-02
1.80e-02
1.20e-02
6.00e-03
0.00e+00
c Fluent Inc. January 11, 2005 12-35
Modeling Species Transport and Gaseous Combustion
T ρ~v · dA
R
~
T = R (12.2)
ρ~v · dA~
(c) Select pressure-outlet-9 as the surface over which to perform the integration.
(d) Click Compute.
The mass-weighted average exit temperature is about 1796 K.
12-36
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
(e) Select Area-Weighted Average as the Report Type and Velocity Magnitude as
the Field Variable.
The area-weighted velocity-magnitude average will be computed as
1Z
v̄ = v dA (12.3)
A
(f) Click Compute.
The area-averaged exit velocity is about 3.14 m/s.
c Fluent Inc. January 11, 2005 12-37
Modeling Species Transport and Gaseous Combustion
12-38
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
Carbon Number is the number of carbon atoms per molecule of fuel and is used
in the prompt NOx prediction. The Fuel Species designation is also used in the
prompt NOx model.
(g) Click OK to accept these changes.
2. Enable the calculation of only the NO species, and set the under-relaxation factor
for this equation.
Solve −→ Controls −→Solution...
(a) In the Equations list, deselect all variables except the NO species.
(b) Increase the NO under-relaxation factor to 1.0.
You will predict NOx formation in a “postprocessing” mode, with the flow field,
temperature, and hydrocarbon combustion species concentrations fixed. Thus, only
the NO equation is computed. Prediction of NO in this mode is justified on the
grounds that the NO concentrations are very low and have negligible impact on the
hydrocarbon combustion prediction.
c Fluent Inc. January 11, 2005 12-39
Modeling Species Transport and Gaseous Combustion
5. Save the new case and data files (gascomb3.cas and gascomb3.dat).
12-40
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
4.11e-03
3.90e-03
3.70e-03
3.49e-03
3.28e-03
3.08e-03
2.87e-03
2.67e-03
2.46e-03
2.26e-03
2.05e-03
1.85e-03
1.64e-03
1.44e-03
1.23e-03
1.03e-03
8.21e-04
6.16e-04
4.11e-04
2.05e-04
0.00e+00
(a) Select Mass-Weighted Average in the Report Type drop-down list and NOx...
and Mass fraction of NO in the Field Variable drop-down list.
(b) Select pressure-outlet-9 as the surface over which to perform the integration.
(c) Click Compute.
The mass-weighted average exit NO mass fraction is about 0.00469.
c Fluent Inc. January 11, 2005 12-41
Modeling Species Transport and Gaseous Combustion
8. Disable the prompt NOx mechanism and solve for thermal NOx only.
Define −→ Models −→ Species −→NOx...
(a) Turn off Prompt NO under Models to disable the prompt NOx mechanism, and
click Apply.
(b) Request 50 iterations.
Solve −→Iterate...
The solution converges in about 10 iterations.
(c) Review the thermal NOx solution by viewing contours of NO mass fraction
(Figure 12.13).
Display −→Contours...
i. Check that NOx... and Mass fraction of NO are selected in the Contours
of drop-down list.
ii. Click Display.
The NO mass fraction contours are shown in Figure 12.13. The concen-
tration of NO is slightly lower without the prompt NOx mechanism.
4.07e-03
3.87e-03
3.67e-03
3.46e-03
3.26e-03
3.06e-03
2.85e-03
2.65e-03
2.44e-03
2.24e-03
2.04e-03
1.83e-03
1.63e-03
1.43e-03
1.22e-03
1.02e-03
8.15e-04
6.11e-04
4.07e-04
2.04e-04
3.19e-20
(d) Compute the average exit NO mass fraction with only thermal NOx formation.
12-42
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
The mass-weighted average exit NO mass fraction, with thermal but no prompt
NOx formation, is about 0.00457.
9. Solve for prompt NOx production only.
Define −→ Models −→ Species −→NOx...
(a) Turn off Thermal NO and turn on Prompt NO under Models, and click Apply.
(b) Request 50 iterations.
The solution converges in about 10 iterations.
Solve −→Iterate...
(c) Review the prompt NOx solution by viewing contours of NO mass fraction
(Figure 12.14).
Display −→Contours...
The NO mass fraction contours are shown in Figure 12.14. The prompt NOx
mechanism is most significant in fuel-rich flames. In this case the flame is
lean and prompt NO production is low.
6.98e-05
6.63e-05
6.28e-05
5.93e-05
5.58e-05
5.23e-05
4.88e-05
4.53e-05
4.19e-05
3.84e-05
3.49e-05
3.14e-05
2.79e-05
2.44e-05
2.09e-05
1.74e-05
1.40e-05
1.05e-05
6.98e-06
3.49e-06
6.04e-27
(d) Compute the average exit NO mass fraction with only prompt NOx formation.
Report −→Surface Integrals...
Hint: Follow the same procedure you used earlier for the calculation with both
thermal and prompt NOx formation.
The mass-weighted average exit NO mass fraction, with only prompt NOx for-
mation, is about 0.000071.
c Fluent Inc. January 11, 2005 12-43
Modeling Species Transport and Gaseous Combustion
Note: The individual thermal and prompt NO mass fractions do not add up
to the levels predicted with the two models combined. This is because
reversible reactions are involved. NO produced in one reaction can be
destroyed in another reaction.
10. Use a custom field function to compute NO parts per million (ppm).
Define −→Custom Field Functions...
NO ppm is computed from the following equation:
1
mixture MW = X mass fraction (12.6)
MW
i
12-44
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
c Fluent Inc. January 11, 2005 12-45
Modeling Species Transport and Gaseous Combustion
7.80e+01
7.41e+01
7.02e+01
6.63e+01
6.24e+01
5.85e+01
5.46e+01
5.07e+01
4.68e+01
4.29e+01
3.90e+01
3.51e+01
3.12e+01
2.73e+01
2.34e+01
1.95e+01
1.56e+01
1.17e+01
7.80e+00
3.90e+00
5.80e-21
Contours of no-ppm
FLUENT 6.2 (axi, segregated, spe, ske)
12-46
c Fluent Inc. January 11, 2005
Modeling Species Transport and Gaseous Combustion
Summary
In this tutorial you used FLUENT to model the transport, mixing, and reaction of chemical
species. The reaction system was defined by using and modifying a mixture-material
entry in the FLUENT database. The procedures used here for simulation of hydrocarbon
combustion can be applied to other reacting flow systems.
This exercise illustrated the important role of the mixture heat capacity in the prediction
of flame temperature. The combustion modeling results are summarized in the following
table. (Note that some of the values in the table were not explicitly calculated during
the tutorial.)
Further Improvements
Further improvements can be expected by including the effects of intermediate species
and radiation, both of which will result in lower predicted combustion temperatures.
The single-step reaction process used in this tutorial cannot account for the moderating
effects of intermediate reaction products, such as CO and H2 . Multiple-step reactions
can be used to address these species. If a multi-step Magnussen model is used, consid-
erably more computational effort is required to solve for the additional species. Where
applicable, the non-premixed combustion model can be used to account for intermediate
species at a reduced computational cost.
See Chapter 15 of the User’s Guide for more details on the non-premixed combustion
model.
Radiation heat transfer tends to make the temperature distribution more uniform, thereby
lowering the peak temperature. In addition, radiation heat transfer to the wall can be
c Fluent Inc. January 11, 2005 12-47
Modeling Species Transport and Gaseous Combustion
very significant (especially here, with the wall temperature set at 300 K). The large
influence of radiation can be anticipated by computing the Boltzmann number for the
flow:
where σ is the Boltzmann constant (5.729×10−8 W/m2 -K4 ) and TAF is the adiabatic
flame temperature. For a quick estimate, assume ρ = 1 kg/m3 , U = 0.5 m/s, and
cp = 1000 J/kg-K (the majority of the inflow is air). Assume TAF = 2000 K. The
resulting Boltzmann number is Bo = 1.09, which shows that radiation is just about as
important as convection for this problem.
See Section 12.3 of the User’s Guide and Tutorial 5 for details on radiation modeling.
12-48
c Fluent Inc. January 11, 2005
Tutorial 13. Using the Non-Premixed Combustion Model
Introduction
A pulverized coal combustion simulation involves modeling a continuous gas phase flow
field and its interaction with a discrete phase of coal particles. The coal particles, traveling
through the gas, will devolatilize and undergo char combustion, creating a source of fuel
for reaction in the gas phase.
The reaction can be modeled using either the species transport model or the non-premixed
combustion model. In this tutorial you will model a simplified coal combustion furnace
using the non-premixed combustion model for the reaction chemistry.
In this tutorial you will learn how to:
• Prepare a Probability Density Function (PDF) table in FLUENT for pulverized coal
The non-premixed combustion model uses a modeling approach that solves transport
equations for one or two conserved scalars, the mixture fractions. Multiple chemical
species, including radicals and intermediate species, may be included in the problem
definition. Their concentrations will be derived from the predicted mixture fraction
distribution.
Property data for the species are accessed through a chemical database and turbulence-
chemistry interaction is modeled using a β-function for the PDF. See Chapter 15 of the
User’s Guide for details on the non-premixed combustion modeling approach.
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1. Some steps in the setup and solution procedure will not
be shown explicitly.
c Fluent Inc. January 11, 2005 13-1
Using the Non-Premixed Combustion Model
Problem Description
The coal combustion system considered in this tutorial is a simple 10 m by 1 m two-
dimensional duct depicted in Figure 13.1. As the model is symmetric, only half of the
domain width is modeled.
The inlet of the 2D duct is split into two streams. A high-speed stream near the center
of the duct enters at 50 m/s and spans 0.125 m. The other stream enters at 15 m/s and
spans 0.375 m. Both streams are air at 1500 K. Coal particles enter the furnace near the
center of the high-speed stream with a mass flow rate of 0.1 kg/s (total flow rate in the
furnace is 0.2 kg/s).
The duct wall has a constant temperature of 1200 K. The Reynolds number based on
the inlet dimension and the average inlet velocity is about 100,000. Thus, the flow is
turbulent.
Details of the coal composition and size distribution are included in Step 3: Non Adiabatic
PDF Table and Step 5: Materials.
T = 1200 K
w
Air: 15 m/s, 1500 K
0.5 m
Coal Injection: 0.1 kg/s
Symmetry
Plane
10 m
13-2
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
2. Unzip non_premix_combustion.zip.
coal.msh can be found in the /non premix combustion folder created after unzip-
ping the file.
The mesh file, coal.msh is a quadrilateral mesh describing the system geometry
shown in Figure 13.1.
Step 1: Grid
1. Read the 2D mesh file, coal.msh.
File −→ Read −→Case...
The FLUENT console window reports that the mesh contains 1357 quadrilateral
cells.
c Fluent Inc. January 11, 2005 13-3
Using the Non-Premixed Combustion Model
velocity-inlet-2 wall-7
velocity-inlet-8 symmetry
Grid
FLUENT 6.2 (2d, segregated, lam)
Figure 13.2: 2D Coal Furnace Mesh Display with Annotated Boundary Types
13-4
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
Step 2: Models
1. Accept the default segregated solver.
The non-premixed combustion model is available only with the segregated solver.
Define −→ Models −→Solver...
c Fluent Inc. January 11, 2005 13-5
Using the Non-Premixed Combustion Model
Coal particles can radiate significantly, and the P-1 radiation model is appropriate
for combustors larger than 1m, so the optical thickness is higher than 1.
13-6
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
c Fluent Inc. January 11, 2005 13-7
Using the Non-Premixed Combustion Model
13-8
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
You can define either a single fuel stream, or a fuel stream plus a secondary
stream. Enabling a secondary stream allows you to keep track of two mixture
fractions. For coal combustion, this will allow you to track volatile matter (the
secondary stream) separately from the char (fuel stream). However, the two
mixture fraction has a substantially greater computational expense than the
single mixture fraction model.
For complex hydrocarbons, like coal, the individual species components are most
often unknown. The empirical model allows you to define the fuel composition
in terms of the ultimate analysis (atomic fractions of C, H, O, N, and S),
along with the lower heating value and heat capacity.
(d) Specify the Empirical Parameters for fuel.
i. Specify the Fuel Lower Calorific Value as 3.53e+07 j/kg.
ii. Specify the value of Fuel Specific Heat as 1000 j/kg-k.
The Fuel Rich Flammability Limit allows you to perform a “partial equilibrium”
calculation, suspending equilibrium calculations when the mixture fraction ex-
ceeds the specified rich limit. This increases the efficiency of the PDF cal-
culation, allowing you to bypass the complex equilibrium calculations in the
fuel-rich region. This is also more physically realistic than the assumption of
full equilibrium. For empirically defined streams, the rich limit is always 1.0
and cannot be altered.
c Fluent Inc. January 11, 2005 13-9
Using the Non-Premixed Combustion Model
13-10
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
(d) Under Temperature, enter the inlet temperatures for Fuel as 400 and Oxid as
1500 and click Apply.
The system pressure and inlet stream temperatures are required for the equi-
librium chemistry calculation. The fuel stream inlet temperature for coal com-
bustion should be the temperature at the onset of devolatilization. The oxidizer
inlet temperature should correspond to the air inlet temperature. In this tuto-
rial, the coal devolatilization temperature will be set to 400 K and the air inlet
temperature is 1500 K. The system pressure is one atmosphere.
4. In the Control tab, retain default species to be excluded from the equilibrium cal-
culation.
(a) In the Table tab, retain the default values for all the Table Parameters and
click Apply.
The maximum number of species determines the number of most preponderant
species to consider after the equilibrium calculation is performed.
The minimum temperature should be a few degrees lower than the lowest bound-
ary condition temperature (e.g., the inlet temperature or wall temperature).
In coal combustion systems, the minimum system temperature should also be
set below the temperature at which the volatiles begin to evolve from the coal.
Here, the vaporization temperature at which devolatilization begins will be set
to 400 K. Thus, the minimum system temperature is set to 298 K (the default).
c Fluent Inc. January 11, 2005 13-11
Using the Non-Premixed Combustion Model
13-12
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
Z
Y
X
Mean Temperature(K)
FLUENT 6.2 (2d, segregated, pdf20, ske)
Figure 13.3: Non-Adiabatic Temperature Look-Up Table on the Adiabatic Enthalpy Slice
The maximum and minimum values for mean temperature and the corresponding
mean mixture fraction are also reported in the console. The maximum mean tem-
perature is reported as 2782 K at a mean mixture fraction of 0.09.
(b) Similarly, plot the mean mole fractions for CO (Figure 13.5).
c Fluent Inc. January 11, 2005 13-13
Using the Non-Premixed Combustion Model
Z
Y
X
Z
Y
X
Mole Fraction of co
FLUENT 6.2 (2d, segregated, pdf20, ske)
13-14
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
1.20e-01
1.00e-01
8.00e-02
Mole 6.00e-02
Fraction
of
co2 4.00e-02
2.00e-02
0.00e+00
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Z
Y Mean Mixture Fraction
X
c Fluent Inc. January 11, 2005 13-15
Using the Non-Premixed Combustion Model
1. Enable the discrete phase coupling to the continuous phase flow prediction.
Define −→ Models −→Discrete Phase...
13-16
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
(a) Under Interaction, turn on the Interaction with Continuous Phase option.
This option enables coupling, in which the discrete phase trajectories (along
with heat and mass transfer to the particles) are allowed to impact the gas
phase equations.
If you leave this option turned off, you can track particles but they will have
no impact on the continuous phase flow.
(b) Set the coupling parameter, the Number of Continuous Phase Iterations per
DPM Iteration, to 20.
Use higher values of this parameter in problems that include a high particle
mass loading or a larger grid size. Less frequent trajectory updates can be
beneficial in such problems, in order to converge the gas phase equations more
completely before repeating the trajectory calculation.
(c) Under Tracking Parameters, set the Max. Number of Steps to 10000.
The limit on the number of trajectory time steps is used to abort trajectories
of particles that are trapped in the domain (e.g., in a recirculation).
(d) Retain the default Step Length Factor of 5.
(e) In the Physical Models tab, turn on Particle Radiation Interaction.
(f) Click OK to close the Discrete Phase Model panel.
c Fluent Inc. January 11, 2005 13-17
Using the Non-Premixed Combustion Model
13-18
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
c Fluent Inc. January 11, 2005 13-19
Using the Non-Premixed Combustion Model
13-20
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
The Injections panel can be used to copy and delete injection definitions. You
list the initial conditions of particle streams defined by an injection in the
console window.
The listing for the injection-0 group will show 10 particle streams, each with a
unique diameter between the specified minimum and maximum value, obtained
from the Rosin-Rammler distribution, and a unique mass flow rate.
c Fluent Inc. January 11, 2005 13-21
Using the Non-Premixed Combustion Model
Step 5: Materials
Define −→Materials...
13-22
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
(b) Retain the default selection (coal-mv) in the Fluent Combusting Particle Mate-
rials list.
c Fluent Inc. January 11, 2005 13-23
Using the Non-Premixed Combustion Model
This is the combusting particle material type that you selected from the list of
database options in the Set Injection Properties panel. Additional combusting
particle materials can be copied from the property database, if required. Click
the Fluent Database... button to view the combusting-particle materials that
are available. Modify the property settings for the selected material, coal-mv.
How FLUENTuses these inputs are defined in the Appendix: Discrete Phase
Material Properties.
(c) Modify the coal-mv material properties as per following constant values:
Properties Values
Density kg/m3 1300
Cp J/kg-K 1000
Thermal Conductivity w/m-k 0.0454
Latent Heat 0
Vaporization Temperature K 400
Volatile Component Fraction (%) 28
Binary Diffusivity w/m-k 5e-4
Particle Emissivity 0.9
Particle Scattering Factor 0.6
Swelling Coefficient 2
Burnout Stoichiometric Ratio 2.67
Combustible Fraction (%) 64
Note: The values for the Vaporization Temperature should be consistent with
the fuel temperature considered in Step 3-3: Non Adiabatic PDF Table. The
Volatile Component Fraction and Combustible Fraction should be consistent with
the volatiles and char ratios in the proximate analysis of the coal shown in the
Appendix: Analysis for Elemental Composition of Coal.
(d) For Devolatilization Model, select single rate in the drop-down list and accept
default values in the Single Rate Devolatilization Model panel.
(e) For the Combustion Model, select kinetics/diffusion limited and accept default
values in the Kinetics/Diffusion-Limited Combustion Model panel.
13-24
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
c Fluent Inc. January 11, 2005 13-25
Using the Non-Premixed Combustion Model
13-26
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
Hint: You can click the mouse probe button (right button, by default) on the required
boundary zone in the graphics display window to select that zone in the Boundary
Conditions panel.
1. Set the following conditions for the zone, velocity-inlet-2 (low-speed inlet boundary).
Properties Values
Velocity Specification Method Magnitude and Direction
Velocity Magnitude(m/s) 15
Temperature(k) 1500
Turbulent Specification Method Intensity and Hydraulic Diameter
Turbulence Intensity(%) 10
Hydraulic Diameter(m) 0.75
c Fluent Inc. January 11, 2005 13-27
Using the Non-Premixed Combustion Model
Turbulence parameters are defined based on intensity and hydraulic diameter. The
relatively large turbulence intensity of 10% may be typical for combustion air flows.
The hydraulic diameter has been set to twice the height of the 2D inlet stream.
For the non-premixed combustion calculation, you have to define the inlet Mean
Mixture Fraction and Mixture Fraction Variance. For coal combustion, all fuel comes
from the discrete phase and thus the gas phase inlets have zero mixture fraction.
Therefore, you can accept the zero default settings.
2. Set the following conditions for the velocity-inlet-8 zone (high-speed inlet boundary).
Properties Values
Velocity Specification Method Magnitude and Direction
Velocity Magnitude(m/s) 50
Temperature(k) 1500
Turbulent Specification Method Intensity and Hydraulic Diameter
Turbulence Intensity(%) 10
Hydraulic Diameter(m) 0.25
13-28
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
3. Set the following conditions for the pressure-outlet-6 zone (exit boundary).
Properties Values
Backflow Total Temperature 2000
Turbulent Specification Method Intensity and Hydraulic Diameter
Turbulence Intensity(%) 10
Hydraulic Diameter(m) 1
The exit gauge pressure of zero defines the system pressure at the exit to be the
operating pressure. The backflow conditions for scalars (temperature, mixture frac-
tion, turbulence parameters) will be used only if flow is entrained into the domain
through the exit. It is a good idea to use reasonable values in case flow reversal
occurs at the exit at some point during the solution process.
c Fluent Inc. January 11, 2005 13-29
Using the Non-Premixed Combustion Model
13-30
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
Step 8: Solution
1. Set the P1 under-relaxation factor to 1.
Solve −→ Controls −→Solution...
Use the Init button to initialize the flow field data.The Apply button does
! not initialize the flow field data. It only allows you to store your initializa-
tion parameters for later use.
Note: Here, with very high pre-heat of the oxidizer stream, you can start the com-
bustion calculation from the inlet-based initialization. In general, you may
need to start your coal combustion calculations by patching a high-temperature
region and performing a discrete phase trajectory calculation. This provides the
initial volatile and char release required to initiate combustion. The Solve/Initialize/
Patch... menu item and the solve/dpm-update text command can be used to
perform this initialization.
c Fluent Inc. January 11, 2005 13-31
Using the Non-Premixed Combustion Model
Note: The default convergence criteria will be met in about 140 iterations.
13-32
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
Step 9: Postprocessing
1. Display the predicted temperature field (Figure 13.7).
Display −→Contours...
Hint: Use the Views panel (Display/Views...) to mirror the display about the sym-
metry plane.
The peak temperature in the system is about 2280 K.
3. Display the char burnout rate (Figure 13.9) by selecting DPM Burnout from the
lower drop-down list.
The char burnout occurs after complete devolatilization. Figure 13.9 shows that
burnout is complete at about three-quarters of the furnace.
c Fluent Inc. January 11, 2005 13-33
Using the Non-Premixed Combustion Model
2.25e+03
2.20e+03
2.15e+03
2.09e+03
2.04e+03
1.99e+03
1.94e+03
1.88e+03
1.83e+03
1.78e+03
1.73e+03
1.67e+03
1.62e+03
1.57e+03
1.52e+03
1.46e+03
1.41e+03
1.36e+03
1.31e+03 Y
1.25e+03 Z X
1.20e+03
2.54e-03
2.41e-03
2.29e-03
2.16e-03
2.03e-03
1.91e-03
1.78e-03
1.65e-03
1.52e-03
1.40e-03
1.27e-03
1.14e-03
1.02e-03
8.89e-04
7.62e-04
6.35e-04
5.08e-04
3.81e-04
2.54e-04 Y
1.27e-04 Z X
0.00e+00
13-34
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
3.89e-04
3.70e-04
3.51e-04
3.31e-04
3.12e-04
2.92e-04
2.73e-04
2.53e-04
2.34e-04
2.14e-04
1.95e-04
1.75e-04
1.56e-04
1.36e-04
1.17e-04
9.74e-05
7.79e-05
5.84e-05
3.89e-05 Y
1.95e-05 Z X
0.00e+00
c Fluent Inc. January 11, 2005 13-35
Using the Non-Premixed Combustion Model
3.71e-02
3.52e-02
3.34e-02
3.15e-02
2.96e-02
2.78e-02
2.59e-02
2.41e-02
2.22e-02
2.04e-02
1.85e-02
1.67e-02
1.48e-02
1.30e-02
1.11e-02
9.26e-03
7.41e-03
5.56e-03
3.71e-03 Y
1.85e-03 Z X
0.00e+00
6. Similarly, display mass fraction distributions for other species — CO2 (Figure 13.12),
H2 O (Figure 13.13), CO (Figure 13.14).
13-36
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
2.33e-01
2.27e-01
2.22e-01
2.16e-01
2.11e-01
2.05e-01
2.00e-01
1.94e-01
1.89e-01
1.83e-01
1.78e-01
1.72e-01
1.67e-01
1.61e-01
1.55e-01
1.50e-01
1.44e-01
1.39e-01
1.33e-01 Y
1.28e-01 Z X
1.22e-01
1.21e-01
1.15e-01
1.09e-01
1.03e-01
9.67e-02
9.07e-02
8.46e-02
7.86e-02
7.25e-02
6.65e-02
6.04e-02
5.44e-02
4.84e-02
4.23e-02
3.63e-02
3.02e-02
2.42e-02
1.81e-02
1.21e-02 Y
6.04e-03 Z X
0.00e+00
c Fluent Inc. January 11, 2005 13-37
Using the Non-Premixed Combustion Model
1.64e-02
1.56e-02
1.48e-02
1.40e-02
1.31e-02
1.23e-02
1.15e-02
1.07e-02
9.85e-03
9.03e-03
8.21e-03
7.39e-03
6.57e-03
5.75e-03
4.93e-03
4.10e-03
3.28e-03
2.46e-03
1.64e-03 Y
8.21e-04 Z X
0.00e+00
4.54e-03
4.31e-03
4.08e-03
3.86e-03
3.63e-03
3.40e-03
3.18e-03
2.95e-03
2.72e-03
2.50e-03
2.27e-03
2.04e-03
1.81e-03
1.59e-03
1.36e-03
1.13e-03
9.07e-04
6.81e-04
4.54e-04 Y
2.27e-04 Z X
0.00e+00
13-38
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
(a) Retain Particle Variables... and Particle Residence Time under Color by.
(b) Select injection-0 in the Release from Injections list.
(c) Turn on Track Single Particle Stream, set the Stream ID to 5 and click Display.
c Fluent Inc. January 11, 2005 13-39
Using the Non-Premixed Combustion Model
4.21e-01
4.00e-01
3.79e-01
3.58e-01
3.37e-01
3.16e-01
2.95e-01
2.74e-01
2.53e-01
2.32e-01
2.11e-01
1.90e-01
1.69e-01
1.47e-01
1.26e-01
1.05e-01
8.43e-02
6.32e-02
4.21e-02 Y
2.11e-02 Z X
0.00e+00
13-40
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
2. Compute the volume sources of heat transferred between the gas and discrete par-
ticle phase.
Report −→Volume Integrals...
c Fluent Inc. January 11, 2005 13-41
Using the Non-Premixed Combustion Model
13-42
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
Fate Number
Elapsed Time (s) Injection,
Min Max Avg Std Dev Min Max
---- ----- -------- --------- --------- ---------- -------- --------
Escaped-Zone6 100 2.511e-01 4.793e-01 3.31e-01 5.248e-02 inj..n-0 1 inj..n-0 7
The report shows that the average residence time of the coal particles is about 0.33
seconds. Volatiles are completely released within the domain and the char conver-
sion is 99.99% .
c Fluent Inc. January 11, 2005 13-43
Using the Non-Premixed Combustion Model
Extra: You can obtain a detailed report of the particle position, velocity, diameter, and
temperature along the trajectories of individual particles. A detailed track reporting
is useful if you are trying to understand unusual or important details in the discrete
model behavior. To generate the report, visit the Particle Tracks panel and do the
following:
1. Select Step By Step under Report Type.
2. Select File under Report to.
3. Enable the Track Single Particle Stream option, and set the Stream ID to the
desired particle stream.
4. Click Write... to bring up the Select File dialog box and enter the name of the
file to be written.
This file can be viewed with a text editor.
Summary
Coal combustion modeling involves the prediction of volatile evolution and char burnout
from pulverized coal along with simulation of the combustion chemistry occurring in the
gas phase. In this tutorial you learned how to use the non-premixed combustion model to
represent the gas phase combustion chemistry. In this approach the fuel composition was
defined empirically and the fuel was assumed to react according to the equilibrium system
data. This equilibrium chemistry model can be applied to other turbulent, diffusion-
reaction systems. You can also model coal combustion using the finite-rate chemistry
model.
You also learned how to set up and solve a problem involving a discrete phase of com-
busting particles. You created discrete phase injections, activated coupling to the gas
phase, and defined the discrete phase material properties. These procedures can be used
to set up other simulations involving reacting or inert particles.
13-44
c Fluent Inc. January 11, 2005
Using the Non-Premixed Combustion Model
Appendix
Coal Analysis for Elemental Composition
From proximate analysis, the fuel considered here consists of 28% volatiles, 64% char,
and 8% ash. Use this information, along with the ultimate analysis, to define the coal
composition. Define the composition of the fuel stream (char) to be 100% C(S). The fuel
stream composition (char and volatiles) is derived as follows.
Begin by converting the proximate data to a dry-ash-free basis:
Proximate Analysis Wt % Wt %
(dry) (DAF)
Volatiles 28 30.4
Char (C(s)) 64 69.6
Ash 8 -
Element Wt % (DAF)
C 89.3
H 5.0
O 3.4
N 1.5
S 0.8
For simplifying modeling, the sulfur content of the coal can be combined into the nitrogen
mass fraction, to yield:
Element Wt % (DAF)
C 89.3
H 5.0
O 3.4
N 2.3
S -
Combine the proximate and ultimate analysis data to yield the following elemental com-
position of the volatile stream:
c Fluent Inc. January 11, 2005 13-45
Using the Non-Premixed Combustion Model
• Density impacts the particle inertia and body forces (when the gravitational accel-
eration is non-zero).
• Latent Heat is the heat required to vaporize the volatiles. This can usually be set
to zero when the non-premixed combustion model is used for coal combustion. If
the volatile composition has been selected in order to preserve the heating value of
the fuel, the latent heat has been effectively included. (You would, however, use a
non-zero latent heat if water content had been included in the volatile definition as
vapor phase H2 O.)
• Volatile Component Fraction determines the mass of each coal particle that is de-
volatilized.
• Binary Diffusivity is the diffusivity of oxidant to the particle surface and is used in
the diffusion-limited char burnout rate.
• Combustible Fraction is the mass fraction of char in the coal particle. It determines
the mass of each coal particle that is consumed by the char burnout submodel.
13-46
c Fluent Inc. January 11, 2005
Tutorial 14. Modeling Surface Chemistry
Introduction
In chemically reacting laminar flows, such as those encountered in chemical vapor de-
position (CVD) applications, accurate modeling of time-dependent hydrodynamics, heat
and mass transfer, and chemical reactions (including wall surface reactions) is important.
Tutorials 12 and 13 deal with reacting flows with applications in gaseous fuel and coal
combustion.
In this tutorial, surface reactions are considered. Modeling the reactions taking place at
gas-solid interfaces is complex and involves several elementary physico-chemical processes
like adsorption of gas-phase species on the surface, chemical reactions occurring on the
surface, and desorption of gases from the surface back to the gas phase.
In this tutorial, you will learn how to:
• Enable physical models and define boundary conditions for a chemically reacting
laminar flow involving wall surface reactions
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
Before beginning, you should read Chapter 13 of the User’s Guide for more information
about species transport, chemically reacting flows, wall surface reaction modeling, and
chemical vapor deposition.. In particular, you should be familiar with the Arrhenius rate
equation, as this equation is used for the surface reactions modeled in this tutorial.
c Fluent Inc. January 11, 2005 14-1
Modeling Surface Chemistry
Inlet
Rotating
Disk
Outlet
Problem Description
A rotating disk CVD reactor for the growth of Gallium Arsenide (GaAs) shown in Fig-
ure 14.1 will be modeled.
The process gases, Trimethyl Gallium (Ga(CH3 )3 ) and Arsine (AsH3 ) enter the reactor
at 293 K through the inlet at the top. These gases flow over the hot, spinning disk
depositing thin layers of gallium and arsenide on it in a uniform, repeatable manner.
The disk rotation generates a radially pumping effect, which forces the gases to flow in a
laminar manner down to the growth surface, outward across the disk, and finally to be
discharged from the reactor.
The semiconductor materials Ga(s) and As(s) are deposited on the heated surface gov-
erned by the following surface reactions.
As mentioned earlier, the inlet gas is a mixture of trimethyl gallium and arsine. In the
inlet mixture the mass fraction of Ga(CH3 )3 is 0.15 and AsH3 is 0.4. The mixture velocity
at the inlet is 0.02189 m/s. The disk rotates at 80 rad/sec, and the top wall (wall-1) is
heated to 473 K, and the sidewalls (wall-2) of the reactor are maintained at 343 K. The
susceptor (wall-4) is heated to a uniform temperature of 1023 K, and the bottom wall
(wall-6) is at 303 K. These CVD reactors are typically known as cold-wall reactors, where
only the wafer surface is heated to higher temperatures, while the remaining reactor walls
are maintained at low temperatures.
In this tutorial, simultaneous deposition of Ga and As is simulated and examined.
The mixture properties and the mass diffusivity are determined based on kinetic the-
14-2
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
ory. Detailed surface reactions with multiple sites and site species, and full multi-
component/thermal diffusion effects are also included in the simulation.
The purpose of this tutorial is to demonstrate surface reaction capabilities in FLUENT.
Convective heat transfer is considered to be the dominant mechanism compared to ra-
diative heat transfer, thus radiation effects are ignored.
2. Unzip surface_chem.zip.
surface.msh can be found in the /surface chem folder created after unzipping the
file.
c Fluent Inc. January 11, 2005 14-3
Modeling Surface Chemistry
Step 1: Grid
1. Read in the mesh file surface.msh.
File −→ Read −→Case...
(a) In the Units Conversion drop-down list, select cm to complete the phrase Grid
Was Created In cm (centimeters).
(b) Click Scale to scale the grid.
The final Domain Extents should appear as in the panel above.
Note: Because the default SI units will be used in this tutorial, there is no need to
change any units.
14-4
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
Extra: You can use the left mouse button to rotate the image and view it from
different angles. You can use the right mouse button to check which zone
number corresponds to each boundary. If you click the right mouse button
on one of the boundaries in the graphics window, its name and type will be
printed in the FLUENT console window. This feature is especially useful when
you have several zones of the same type and you want to distinguish between
them quickly. Use the middle mouse button to zoom the image.
c Fluent Inc. January 11, 2005 14-5
Modeling Surface Chemistry
14-6
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
Step 2: Models
In this problem, the energy equation and the species conservation equations will be solved,
along with the momentum and continuity equations.
c Fluent Inc. January 11, 2005 14-7
Modeling Surface Chemistry
14-8
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
Note: This includes the effect of enthalpy transport due to species diffusion
in the energy equation, which contributes to the energy balance, especially
for the case of Lewis numbers far from unity.
(e) Select Full Multicomponent Diffusion and Thermal Diffusion.
Note: The Full Multicomponent Diffusion activates Stefan-Maxwell’s equa-
tions and computes the diffusive fluxes of all species in the mixture to
all concentration gradients. The Thermal Diffusion effects cause heavy
molecules to diffuse less rapidly, and light molecules to diffuse more rapidly,
towards heated surfaces.
(f) Click OK.
The console window will list the properties that are required for the models
that you have enabled. You will see an Information dialog box, reminding you
to confirm the property values that have been extracted from the database.
c Fluent Inc. January 11, 2005 14-9
Modeling Surface Chemistry
Step 3: Materials
Define −→Materials...
1. Create the gas-phase species (AsH3 , Ga(CH3 )3 , CH3 , H2 ), the site species (Ga s
and As s), and solid species (Ga and As).
(a) Create species AsH3 .
i. In the Materials panel, select fluid under Material Type.
ii. Select nitrogen under Fluid Materials to create the new material.
iii. Set the Mixture to none.
iv. Enter arsine under Name.
v. Enter ash3 under Chemical Formula.
vi. Specify the following for each of the properties:
14-10
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
Parameter Value
Cp kinetic-theory
Thermal Conductivity kinetic-theory
Viscosity kinetic-theory
Molecular Weight 77.95
Standard State Enthalpy 0
Standard State Entropy 130579.1
Reference Temperature 298.15
L-J Characteristic Length 4.145
L-J Energy Parameter 259.8
Degrees of Freedom 0
Ignore the Density parameter as the density will be set to incompressible-
ideal-gas-law for mixture.
vii. Click Change/Create to create the new material.
viii. FLUENT will ask if you would like to overwrite nitrogen. Click No in the
Question panel.
c Fluent Inc. January 11, 2005 14-11
Modeling Surface Chemistry
(b) Create the other species following the same procedure as for AsH3 .
i. The parameter values for each of the species is as per the table given
below:
14-12
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
The species should appear in the same order as shown in the above table.
i
(f) To set the species follow the procedure listed below:
i. To remove an unwanted species from the Selected Species list, select the
species and click Remove under Selected Species.
ii. Select the required species in the Available Materials list.
iii. Click Add under the corresponding species list.
iv. Click OK after all the species are set under the respective categories.
c Fluent Inc. January 11, 2005 14-13
Modeling Surface Chemistry
14-14
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
ii. Change the ID to 2 and set the parameters for the second equation as
shown in the above table.
iii. Click OK to save your data and close the panel.
c Fluent Inc. January 11, 2005 14-15
Modeling Surface Chemistry
14-16
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
6. Set Cp to mixing-law.
c Fluent Inc. January 11, 2005 14-17
Modeling Surface Chemistry
2. Enable Gravity.
14-18
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
c Fluent Inc. January 11, 2005 14-19
Modeling Surface Chemistry
14-20
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
(d) Under the Species section of the panel, enable Reaction and set Mechanisms to
gaas-ald.
c Fluent Inc. January 11, 2005 14-21
Modeling Surface Chemistry
8. Use the TUI commands to turn off diffusion at the inlet. In the console window,
type the commands shown in boxes in the dialog below.
Hint: You may need to enter press the <Enter> key to get the > prompt.
> define/models/species/inlet-diffusion?
Include diffusion at inlets? [yes] no
14-22
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
Step 6: Solution
1. Set the solution parameters.
Solve −→ Controls −→Solution...
(a) Change the Under-Relaxation Factor as follows:
Parameter URF Parameter URF
Pressure 0.1 ash3 1
Density 0.3 ga<ch3>3 1
Body Forces 1 ch3 1
Momentum 0.2 Energy 0.9
Hint: You will need to scroll down the Under-Relaxation Factors list to see the
species and Energy.
(b) Under Discretization, retain the default First Order Upwind for Momentum, all
the species and Energy.
c Fluent Inc. January 11, 2005 14-23
Modeling Surface Chemistry
2. Initialize the flow field using the boundary conditions set at velocity-inlet.
Solve −→ Initialize −→Initialize...
14-24
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
c Fluent Inc. January 11, 2005 14-25
Modeling Surface Chemistry
Residuals
continuity
x-velocity
1e+01
y-velocity
z-velocity
energy 1e+00
ash3
gach3 1e-01
ch3
1e-02
1e-03
1e-04
1e-05
1e-06
1e-07
1e-08
1e-09
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Y
Z X Iterations
Scaled Residuals
FLUENT 6.2 (3d, dp, segregated, spe, lam)
14-26
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
Step 7: Postprocessing
1. Create an iso-surface near wall-4.
Surface −→Iso-Surface...
(a) In the Iso-Surface panel, select Grid and Z-Coordinate under Surface of Constant.
(b) Click Compute.
(c) Enter 0.075438 under Iso-Values.
(d) Enter z=0.07 under New Surface Name.
(e) Click Create.
c Fluent Inc. January 11, 2005 14-27
Modeling Surface Chemistry
1.02e+03
9.86e+02
9.50e+02
9.14e+02
8.77e+02
8.40e+02
8.04e+02
7.68e+02
7.31e+02
6.94e+02
6.58e+02
6.22e+02
5.85e+02
5.48e+02
5.12e+02
4.76e+02
4.39e+02
4.02e+02
3.66e+02
3.30e+02 Z X
2.93e+02
Y
The temperature contours shows the temperature distribution across a plane just
above the rotating disk. You can see that the disk has a temperature of 1023 K.
14-28
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
3.89e-05
3.70e-05
3.50e-05
3.31e-05
3.11e-05
2.92e-05
2.72e-05
2.53e-05
2.34e-05
2.14e-05
1.95e-05
1.75e-05
1.56e-05
1.36e-05
1.17e-05
9.73e-06
7.78e-06
5.84e-06
3.89e-06
1.95e-06 Z X
0.00e+00
Y
Figure 14.5 shows the gradient of surface deposition rate of ga. The maximum
deposition is seen at the center of the disk.
c Fluent Inc. January 11, 2005 14-29
Modeling Surface Chemistry
1.63e-01
1.55e-01
1.46e-01
1.38e-01
1.30e-01
1.22e-01
1.14e-01
1.06e-01
9.77e-02
8.95e-02
8.14e-02
7.32e-02
6.51e-02
5.70e-02
4.88e-02
4.07e-02
3.26e-02
2.44e-02
1.63e-02
8.14e-03 Z X
0.00e+00
Y
14-30
c Fluent Inc. January 11, 2005
Modeling Surface Chemistry
(a) Select Species... and Surface Deposition Rate of ga in the Y Axis Function
drop-down list.
(b) Under Options, deselect Node Values.
The source/sink terms due to the surface reaction are deposited in the cell
adjacent to the wall cells, so it is necessary to plot the cell values and not the
node values.
(c) In the Surfaces list, select line-9.
(d) In the Solution XY Plot, click Plot.
The peak of the surface deposition rate occurs at the center of wall-4 (where the
concentration of the mixture is highest).
c Fluent Inc. January 11, 2005 14-31
Modeling Surface Chemistry
line-9
4.00e-05
3.80e-05
3.60e-05
3.40e-05
3.20e-05
Surface
Deposition 3.00e-05
Rate
of 2.80e-05
ga
(kg/m2-s) 2.60e-05
2.40e-05
2.20e-05
-0.02 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Z Position (m)
Y
X
Extra: You can also perform all the above postprocessing steps to analyze the deposition
of As.
Summary
The main focus of this tutorial is the accurate modeling of macroscopic gas flow, heat
and mass transfer, species diffusion, and chemical reactions (including surface reactions)
in a rotating disk CVD reactor. In this tutorial, you learned how to use the two-step
surface reactions involving site species, and computed simultaneous deposition of gallium
and arsenide from a mixture of precursor gases on a rotating susceptor. Note that the
same approach is valid if you are simulating multi-step reactions with multiple sites/site
species.
14-32
c Fluent Inc. January 11, 2005
Tutorial 15. Modeling Evaporating Liquid Spray
Introduction
In this tutorial, FLUENT’s air-blast atomizer model is used to predict the behavior of an
evaporating methanol spray. The air flow is modeled first without droplets. To predict
the behavior of the spray, several other discrete-phase models, including collision and
breakup, are used
In this tutorial you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The geometry to be considered in this tutorial is shown in Figure 15.1. Methanol is cooled
to −10◦ C before being introduced into an air-blast atomizer. The atomizer contains an
inner air stream surrounded by a swirling annular stream. (The species include the
components of air as well as water vapor, so the model can be expanded to include
combustion, if desired.) To make use of the periodicity of the problem, only a 30-degree
section of the atomizer will be modeled.
c Fluent Inc. January 11, 2005 15-1
Modeling Evaporating Liquid Spray
Y
Z
X
2. Unzip evaporate_liquid.zip.
sector.msh can be found in the /evaporate liquid folder created after unzipping
the file.
15-2
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
Step 1: Grid
1. Read in the mesh file sector.msh.
File −→ Read −→Case...
c Fluent Inc. January 11, 2005 15-3
Modeling Evaporating Liquid Spray
15-4
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
Y
Z
X
Grid
FLUENT 6.2 (3d, segregated, lam)
c Fluent Inc. January 11, 2005 15-5
Modeling Evaporating Liquid Spray
5. Using the text interface, change zones periodic-a and periodic-b from wall zones
to periodic zones.
(a) In the console window, type the commands shown in boxes in the dialog below.
> grid
/grid> modify-zones
/grid/modify-zones> list-zones
id name type material kind
---- ---------------- ----------------- ------------------ ----
1 fluid fluid air cell
2 atomizer-wall wall aluminum face
3 central_air mass-flow-inlet face
4 co-flow-air velocity-inlet face
5 outlet pressure-outlet face
6 swirling_air velocity-inlet face
7 periodic-a wall aluminum face
8 periodic-b wall aluminum face
9 outer-wall wall aluminum face
11 default-interior interior face
/grid/modify-zones> make-periodic
Periodic zone [()] 7
Shadow zone [()] 8
Rotational periodic? (if no, translational) [yes] yes
Create periodic zones? [yes] yes
zone 8 deleted
15-6
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
c Fluent Inc. January 11, 2005 15-7
Modeling Evaporating Liquid Spray
Step 2: Models
1. Keep the default solver settings.
Define −→ Models −→Solver...
15-8
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
The realizable k- model gives a more accurate prediction of the spreading rate of
both planar and round jets than the standard k- model.
c Fluent Inc. January 11, 2005 15-9
Modeling Evaporating Liquid Spray
When you click OK, the console window will list the properties that are
! required for the models you have enabled. You will see an Information
dialog box, reminding you to confirm the property values that have been
extracted from the database.
(c) Click OK in the Information dialog box to continue.
15-10
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
1. Set the following conditions for the inner air stream (central air).
c Fluent Inc. January 11, 2005 15-11
Modeling Evaporating Liquid Spray
2. Set the following conditions for the air stream surrounding the atomizer (co-flow-
air).
15-12
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
c Fluent Inc. January 11, 2005 15-13
Modeling Evaporating Liquid Spray
4. Set the following conditions for the swirling annular stream (swirling air).
15-14
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
5. Set the following conditions for the outer wall of the atomizer (outer-wall).
c Fluent Inc. January 11, 2005 15-15
Modeling Evaporating Liquid Spray
15-16
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
c Fluent Inc. January 11, 2005 15-17
Modeling Evaporating Liquid Spray
15-18
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
7. Create a clip plane to examine the flow field at the midpoint of the atomizer section.
Surface −→Iso-Surface...
(a) Select Grid... and Angular Coordinate in the Surface of Constant lists.
(b) Click on Compute to update the minimum and maximum values.
(c) Enter 15 in the Iso-Values field.
(d) Enter angle=15 for the New Surface Name.
(e) Click on Create to create the isosurface.
c Fluent Inc. January 11, 2005 15-19
Modeling Evaporating Liquid Spray
8. Review the current state of the solution by examining contours of velocity magni-
tude (Figure 15.3).
Display −→Contours...
(a) Select Velocity... and Velocity Magnitude in the Contours Of drop-down list.
(b) Under Options, select Filled and Draw Grid.
This will open the Grid Display panel.
15-20
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
(c) Keep the current grid display settings and close the Grid Display panel.
(d) In the Contours panel, select angle=15 in the Surfaces list.
(e) Click Display.
(f) Use your mouse to obtain the view shown in Figure 15.3.
9.87e+01
9.38e+01
8.88e+01
8.39e+01
7.90e+01
7.40e+01
6.91e+01
6.42e+01
5.92e+01
5.43e+01
4.94e+01
4.44e+01
3.95e+01
3.45e+01
2.96e+01
2.47e+01
1.97e+01
1.48e+01
9.87e+00 Y
4.94e+00 Z
0.00e+00 X
c Fluent Inc. January 11, 2005 15-21
Modeling Evaporating Liquid Spray
9. Display path lines of the air in the swirling annular stream (Figure 15.4).
Display −→Path Lines...
15-22
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
4.90e+01
4.65e+01
4.41e+01
4.17e+01
3.92e+01
3.68e+01
3.43e+01
3.19e+01
2.94e+01
2.70e+01
2.45e+01
2.20e+01
1.96e+01
1.71e+01
1.47e+01
1.22e+01
9.80e+00
7.35e+00
4.90e+00 Y
2.45e+00 Z
0.00e+00 X
c Fluent Inc. January 11, 2005 15-23
Modeling Evaporating Liquid Spray
15-24
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
c Fluent Inc. January 11, 2005 15-25
Modeling Evaporating Liquid Spray
15-26
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
c Fluent Inc. January 11, 2005 15-27
Modeling Evaporating Liquid Spray
15-28
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
c Fluent Inc. January 11, 2005 15-29
Modeling Evaporating Liquid Spray
Step 6: Solution
1. Uncheck the convergence criteria for all the residuals.
Solve −→ Monitors −→Residuals...
2. Reduce the Under-Relaxation Factor for Discrete Phase Sources to 0.1.
Solve −→ Controls −→Solution...
3. Request for 200 iterations.
Solve −→Iterate...
4. Save the case and data files (spray2.cas and spray2.dat).
File −→ Write −→Case...
5. Display the trajectories of the droplets in the spray injection (Figure 15.5).
This will allow you to review the location of the droplets.
Display −→Particle Tracks...
15-30
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
(e) Keep the current display settings and close the panel.
(f) In the Particle Tracks panel, select Particle Variables... and Particle Diameter in
the Color by drop-down list.
This will display the location of the droplets colored by their diameters.
(g) In the Release from Injections list, select injection-0.
(h) Click Display.
(i) Use your mouse to obtain the view shown in Figure 15.5.
c Fluent Inc. January 11, 2005 15-31
Modeling Evaporating Liquid Spray
1.43e-04
1.37e-04
1.31e-04
1.24e-04
1.18e-04
1.12e-04
1.06e-04
9.93e-05
9.31e-05
8.68e-05
8.05e-05
7.42e-05
6.79e-05
6.17e-05
5.54e-05
4.91e-05
4.28e-05
3.65e-05
3.03e-05
Z
2.40e-05 Y
1.77e-05 X
Figure 15.5: Particle Tracks for the Spray Injection After 200 Iterations
The air-blast atomizer model assumes that a cylindrical liquid sheet exits the
atomizer, which then disintegrates into ligaments and droplets. Appropriately,
the model determines that the droplets should be input into the domain in a
ring. The radius of this disk is determined from the inner and outer radii of
the injector.
Note that the maximum diameter of the droplets is about
10−4 m, or 0.1 mm. This is slightly smaller than the film height, which makes
sense. Recall that the inner diameter and outer diameter of the injector are
3.5 mm and 4.5 mm, respectively. The film height is then 12 (4.5 − 3.5) =
0.5 mm. The range in the droplet sizes is due to the fact that the air-blast
atomizer automatically uses a distribution of droplet sizes.
Also note that the droplets are placed a slight distance away from the injector.
Once the droplets are injected into the domain, they can collide/coalesce with
other droplets as determined by the secondary models (breakup and collision).
However, once a droplet has been introduced into the domain, the air-blast
atomizer model no longer affects the droplet.
7. Save the new case and data files (spray3.cas and spray3.dat).
15-32
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
Step 7: Postprocessing
1. Display the particle trajectories again, to see how the droplets have dispersed.
Display −→Particle Tracks...
(a) Click Display in the Particle Tracks panel.
(b) Use your mouse to obtain the view shown in Figure 15.6.
1.76e-04
1.68e-04
1.60e-04
1.52e-04
1.44e-04
1.36e-04
1.28e-04
1.20e-04
1.13e-04
1.05e-04
9.67e-05
8.88e-05
8.09e-05
7.30e-05
6.50e-05
5.71e-05
4.92e-05
4.13e-05
3.34e-05 Y
Z
2.54e-05
1.75e-05 X
Figure 15.6: Particle Tracks for the Spray Injection After 400 Iterations.
c Fluent Inc. January 11, 2005 15-33
Modeling Evaporating Liquid Spray
(a) Select Species... and Mass fraction of ch3oh in the Surface of Constant lists.
(b) Click on Compute to update the minimum and maximum values.
(c) Enter 0.0015 in the Iso-Values field.
(d) Enter methanol-mf=0.0015 for the New Surface Name.
(e) Click on Create to create the isosurface.
15-34
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
c Fluent Inc. January 11, 2005 15-35
Modeling Evaporating Liquid Spray
15-36
c Fluent Inc. January 11, 2005
Modeling Evaporating Liquid Spray
XY
Z
Grid
FLUENT 6.2 (3d, segregated, spe, rke)
Figure 15.7: Full Atomizer Display with Surface of Constant Methanol Mass Fraction
Summary
In this tutorial, you defined a spray injection for an air-blast atomizer and calculated
the solution using FLUENT’s under relaxation factor for discrete phase source. You
viewed the location of methanol droplet particles after they had exited the atomizer and
examined an isosurface of the methanol mass fraction.
c Fluent Inc. January 11, 2005 15-37
Modeling Evaporating Liquid Spray
15-38
c Fluent Inc. January 11, 2005
Tutorial 16. Using the VOF Model
Introduction
This tutorial illustrates the setup and solution of the two-dimensional turbulent fluid
flow in a partially filled spinning bowl.
In this tutorial you will learn how to:
• Set up and solve a transient free-surface problem using the segregated solver
• Examine the fluid flow and the free-surface shape using velocity vectors and volume
fraction contours
Prerequisites
This tutorial requires a basic familiarity with FLUENT. You may also find it helpful to
read about VOF multiphase flow modeling in the FLUENT by reading Section 24.2 of the
User’s Guide for more information. Otherwise, no previous experience with multiphase
modeling is required.
Problem Description
The information relevant to this problem is shown in Figure 16.1. A large bowl, 1 m in
radius, is one-third filled with water and is open to the atmosphere. The bowl spins with
an angular velocity of 3 rad/sec. Based on the rotating water, the Reynolds number is
about 106 , so the flow is modeled as turbulent.
c Fluent Inc. January 11, 2005 16-1
Using the VOF Model
2m
1m
Bowl: Ω = 3 rad/s
Air: ρ = 1.225 kg/m 3
-5
µ = 1.7894 x 10 kg/m-s
Water: ρ = 998.2 kg/m 3
µ = 1 x 10
-3
kg/m-s
2. Unzip vof.zip.
bowl.msh can be found in the /vof folder created after unzipping the file.
The mesh file bowl.msh is a quadrilateral mesh describing the system geometry
shown in Figure 16.1.
16-2
c Fluent Inc. January 11, 2005
Using the VOF Model
Step 1: Grid
1. Read the 2D grid file, bowl.msh.
File −→ Read −→Case...
As shown in Figure 16.2, half of the bowl is modeled, with a symmetry boundary at
the centerline. The bowl is shown lying on its side, with the region to be modeled
extending from the centerline to the outer wall. When you begin to display data
graphically, you will need to rotate the view and mirror it across the centerline to
obtain a more realistic view of the model. This step will be performed later in the
tutorial.
c Fluent Inc. January 11, 2005 16-3
Using the VOF Model
Grid
FLUENT 6.2 (2d, segregated, lam)
16-4
c Fluent Inc. January 11, 2005
Using the VOF Model
Step 2: Models
1. Specify a transient model with axisymmetric swirl.
Define −→ Models −→Solver...
c Fluent Inc. January 11, 2005 16-5
Using the VOF Model
16-6
c Fluent Inc. January 11, 2005
Using the VOF Model
(a) Select k-epsilon as the Model, and retain the default setting of Standard under
k-epsilon Model.
c Fluent Inc. January 11, 2005 16-7
Using the VOF Model
Step 3: Materials
1. Copy water from the FLUENT database materials so that it can be used for the
secondary phase.
Define −→Materials...
(a) Click on the Fluent Database... button to open the Fluent Database Materials
panel.
(b) In the Fluent Fluid Materials list (near the bottom), select water-liquid.
(c) Click Copy and close the Fluent Database Materials and Materials panels.
16-8
c Fluent Inc. January 11, 2005
Using the VOF Model
Step 4: Phases
Here, water is defined as the secondary phase mainly for convenience in setting up the
problem. When you define the initial solution, you will be patching an initial swirl velocity
in the bottom third of the bowl, where the water is. It is more convenient to patch a water
volume fraction of 1 there than to patch an air volume fraction of 1 in the rest of the
domain. Also, the default volume fraction at the pressure inlet is 0, which is the correct
value if water is the secondary phase.
In general, you can specify the primary and secondary phases whichever way you prefer.
It is a good idea, especially in more complicated problems, to consider how your choice
will affect the ease of problem setup.
ii. In the Primary Phase panel, enter air for the Name.
iii. Keep the default selection of air for the Phase Material.
c Fluent Inc. January 11, 2005 16-9
Using the VOF Model
ii. In the Secondary Phase panel, enter water for the Name.
iii. Select water-liquid from the Phase Material drop-down list.
16-10
c Fluent Inc. January 11, 2005
Using the VOF Model
c Fluent Inc. January 11, 2005 16-11
Using the VOF Model
(d) Click OK in the Axis panel to accept the default Zone Name.
16-12
c Fluent Inc. January 11, 2005
Using the VOF Model
ii. Set the Turb. Kinetic Energy to 2.25e-2 and the Turb. Dissipation Rate to
7.92e-3.
Since there is initially no flow passing through the pressure inlet, you need
to specify k and explicitly rather than using one of the other turbulence
specification methods. All of the other methods require you to specify the
turbulence intensity, which is 0 in this case.
The values for k and are computed as follows:
k = (Iwwall )2
0.093/4 k 3/2
=
`
where the turbulence intensity I is 0.05 (close to zero), wwall is 3 m/s,
and ` is 0.07 (obtained by multiplying 0.07 by the maximum radius of the
bowl, which is 1).
See Section 7.2.2 of the User’s Guide for details about the specification of
turbulence boundary conditions at flow inlets and exits.
(b) Check the volume fraction of the secondary phase.
i. In the Boundary Conditions panel, select water from the Phase drop-down
list and click Set....
c Fluent Inc. January 11, 2005 16-13
Using the VOF Model
16-14
c Fluent Inc. January 11, 2005
Using the VOF Model
c Fluent Inc. January 11, 2005 16-15
Using the VOF Model
Step 7: Solution
In simple flows, the under-relaxation factors can usually be increased at the start of the
calculation. This is particularly true when the VOF model is used, where high under-
relaxation on all variables can greatly improve the performance of the solver.
Be sure to use the scroll bar to access the under-relaxation factors that are
! initially out of view.
(b) Under Discretization, choose the PRESTO! scheme in the drop-down list next
to Pressure.
(c) Under Pressure-Velocity Coupling, select PISO.
PISO is recommended for transient flow calculations.
16-16
c Fluent Inc. January 11, 2005
Using the VOF Model
3. Enable the plotting of the axial velocity of water near the outer edge of the bowl
during the calculation.
For transient calculations, it is often useful to monitor the value of a particular
variable to see how it changes over time. Here you will first specify the point at
which you want to track the velocity, and then define the monitoring parameters.
(a) Define a point surface near the outer edge of the bowl.
Surface −→Point...
c Fluent Inc. January 11, 2005 16-17
Using the VOF Model
16-18
c Fluent Inc. January 11, 2005
Using the VOF Model
iv. Click on Define... to specify the surface monitor parameters in the Define
Surface Monitor panel.
c Fluent Inc. January 11, 2005 16-19
Using the VOF Model
5. Patch the initial distribution of water (i.e., water volume fraction of 1.0) and a
swirl velocity of 3 rad/s in the bottom third of the bowl (where the water is).
In order to patch a value in just a portion of the domain, you will need to define
a cell “register” for that region. You will use the same tool that is used to mark a
region of cells for adaption. Also, you will need to define a custom function for the
swirl velocity.
(a) Define a register for the bottom third of the domain.
Adapt −→Region...
16-20
c Fluent Inc. January 11, 2005
Using the VOF Model
i. Select the register (hexahedron-r0) in the Registers list and click Display.
The graphics display will show the bottom third of the bowl in red.
c Fluent Inc. January 11, 2005 16-21
Using the VOF Model
16-22
c Fluent Inc. January 11, 2005
Using the VOF Model
7. Patch the water volume fraction in the bottom third of the bowl.
Solve −→ Initialize −→Patch...
c Fluent Inc. January 11, 2005 16-23
Using the VOF Model
(f) Patch the swirl velocity in the bottom third of the bowl.
16-24
c Fluent Inc. January 11, 2005
Using the VOF Model
c Fluent Inc. January 11, 2005 16-25
Using the VOF Model
(a) Select axis-2 in the Mirror Planes list and click Apply.
(b) Use your middle and left mouse buttons to zoom and translate the view so
that the entire bowl is visible in the graphics display.
(c) Click on the Camera... button to open the Camera Parameters panel.
(d) Using your left mouse button, rotate the dial clockwise until the bowl appears
upright in the graphics window (90◦ ).
(e) Close the Camera Parameters panel.
(f) In the Views panel, click on the Save button under Actions to save the mirrored,
upright view, and then close the panel.
When you do this, view-0 will be added to the list of Views.
The upright view of the bowl in Figure 16.3 correctly shows that w = 3r in the
region of the bowl that is filled with water.
16-26
c Fluent Inc. January 11, 2005
Using the VOF Model
2.35e+00
2.23e+00
2.12e+00
2.00e+00
1.88e+00
1.76e+00
1.65e+00
1.53e+00
1.41e+00
1.29e+00
1.18e+00
1.06e+00
9.41e-01
8.23e-01
7.06e-01
5.88e-01
4.70e-01
3.53e-01
2.35e-01
1.18e-01
0.00e+00
c Fluent Inc. January 11, 2005 16-27
Using the VOF Model
(a) Select Phases... and Volume fraction of water in the Contours of lists.
(b) Select water in the Phase drop-down list.
(c) Set the number of contour Levels to 2 and click Display.
There are only two possible values for the volume fraction at this point: 0 or
1.
16-28
c Fluent Inc. January 11, 2005
Using the VOF Model
1.00e+00
5.00e-01
0.00e+00
Figure 16.4 correctly shows that the bottom third of the bowl contains water.
c Fluent Inc. January 11, 2005 16-29
Using the VOF Model
Figure 16.5 shows the time history for the axial velocity. The velocity is clearly
oscillating, and the oscillations appear to be decaying over time (as the peaks
become smaller). This periodic oscillation has a cycle of 1 second. The switch
from a positive to a negative axial velocity indicates that the water is sloshing
up and down the sides of the bowl in an attempt to reach an equilibrium po-
sition. The fact that the amplitude is decaying suggests that equilibrium will
be reached at some point. The periodic behavior in evidence will therefore be
present only during the initial startup phase of the bowl rotation.
0.3000
0.2000
0.1000
Vertex 0.0000
Average
Axial
Velocity -0.1000
-0.2000
-0.3000
0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 1.2000 1.4000 1.6000 1.8000 2.0000
Flow Time
16-30
c Fluent Inc. January 11, 2005
Using the VOF Model
Step 8: Postprocessing
As indicated by changes in axial velocity in Figure 16.5, the flow field is oscillating peri-
odically. In this step, you will examine the flow field at several different times. (Recall
that you have saved the data files for t = 0.4, 0.6, 0.8, and, 1.0.)
c Fluent Inc. January 11, 2005 16-31
Using the VOF Model
1.00e+00
5.00e-01
0.00e+00
1.00e+00
5.00e-01
0.00e+00
16-32
c Fluent Inc. January 11, 2005
Using the VOF Model
1.00e+00
5.00e-01
0.00e+00
1.00e+00
5.00e-01
0.00e+00
c Fluent Inc. January 11, 2005 16-33
Using the VOF Model
16-34
c Fluent Inc. January 11, 2005
Using the VOF Model
2.77e+01
2.59e+01
2.40e+01
2.22e+01
2.03e+01
1.85e+01
1.66e+01
1.48e+01
1.29e+01
1.11e+01
9.25e+00
7.40e+00
5.55e+00
3.70e+00
1.85e+00
0.00e+00
2.47e+01
2.30e+01
2.14e+01
1.97e+01
1.81e+01
1.64e+01
1.48e+01
1.31e+01
1.15e+01
9.86e+00
8.22e+00
6.57e+00
4.93e+00
3.29e+00
1.64e+00
0.00e+00
c Fluent Inc. January 11, 2005 16-35
Using the VOF Model
4.68e+01
4.37e+01
4.06e+01
3.74e+01
3.43e+01
3.12e+01
2.81e+01
2.50e+01
2.18e+01
1.87e+01
1.56e+01
1.25e+01
9.36e+00
6.24e+00
3.12e+00
0.00e+00
7.02e+00
6.55e+00
6.08e+00
5.61e+00
5.14e+00
4.68e+00
4.21e+00
3.74e+00
3.27e+00
2.81e+00
2.34e+00
1.87e+00
1.40e+00
9.35e-01
4.68e-01
0.00e+00
16-36
c Fluent Inc. January 11, 2005
Using the VOF Model
c Fluent Inc. January 11, 2005 16-37
Using the VOF Model
16-38
c Fluent Inc. January 11, 2005
Using the VOF Model
1.91e+00
1.82e+00
1.72e+00
1.63e+00
1.53e+00
1.44e+00
1.34e+00
1.25e+00
1.15e+00
1.06e+00
9.61e-01
8.66e-01
7.71e-01
6.75e-01
5.80e-01
4.85e-01
3.89e-01
2.94e-01
1.99e-01
1.03e-01
7.98e-03
Figure 16.14: Velocity Vectors for the Air and Water at t = 0.4
1.96e+00
1.87e+00
1.77e+00
1.67e+00
1.57e+00
1.47e+00
1.37e+00
1.28e+00
1.18e+00
1.08e+00
9.82e-01
8.84e-01
7.86e-01
6.88e-01
5.90e-01
4.91e-01
3.93e-01
2.95e-01
1.97e-01
9.88e-02
6.55e-04
Figure 16.15: Velocity Vectors for the Air and Water at t = 0.6
c Fluent Inc. January 11, 2005 16-39
Using the VOF Model
2.12e+00
2.01e+00
1.91e+00
1.80e+00
1.70e+00
1.59e+00
1.48e+00
1.38e+00
1.27e+00
1.17e+00
1.06e+00
9.56e-01
8.51e-01
7.45e-01
6.39e-01
5.34e-01
4.28e-01
3.22e-01
2.17e-01
1.11e-01
5.18e-03
Figure 16.16: Velocity Vectors for the Air and Water at t = 0.8
2.12e+00
2.01e+00
1.91e+00
1.80e+00
1.70e+00
1.59e+00
1.48e+00
1.38e+00
1.27e+00
1.17e+00
1.06e+00
9.55e-01
8.49e-01
7.44e-01
6.38e-01
5.32e-01
4.26e-01
3.20e-01
2.14e-01
1.08e-01
2.22e-03
16-40
c Fluent Inc. January 11, 2005
Using the VOF Model
In Figure 16.16 you can see that the flow is rising up more quickly in the middle of
the bowl, and in Figure 16.17 you can see that the flow is still moving upward, but
more slowly. These patterns correspond to the volume fraction plots at these times.
As the upward motion in the center of the bowl decreases, you can expect the flow
to reverse as the water again seeks to reach a state of equilibrium.
Summary
In this tutorial, you have learned how to use the VOF free surface model to solve a
problem involving a spinning bowl of water. The time-dependent VOF formulation is
used in this problem to track the shape of the free surface and the flow field inside the
spinning bowl.
You observed the changing pattern of the water and air in the bowl by displaying volume
fraction contours, stream function contours, and velocity vectors at t = 0.4, t = 0.6,
t = 0.8, and t = 1 second.
c Fluent Inc. January 11, 2005 16-41
Using the VOF Model
16-42
c Fluent Inc. January 11, 2005
Tutorial 17. Modeling Cavitation
Introduction
This tutorial examines the pressure-driven cavitating flow of water
through a sharp-edged orifice. This is a typical configuration in fuel injectors, and brings
a challenge to the physics and numerics of cavitation models, because of the high pres-
sure differentials involved, and the high ratio of liquid to vapor density. Using FLUENT’s
multiphase modeling capability, you will be able to predict the strong cavitation near the
orifice after flow separation at a sharp edge. In this tutorial you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1. Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The problem considers the cavitation caused by the flow separation after a sharp-edged
orifice. The flow is pressure driven, with an inlet pressure of 5 × 105 Pa, and an outlet
pressure of 9.5 × 104 Pa. The orifice diameter is 4 × 10− 3 m, and geometrical param-
eters of the orifice are D/d = 2.88 and L/r = 8, where D, d, and L are inlet diameter,
orifice diameter, and orifice length respectively. The geometry of the orifice is shown in
Figure 17.1.
c Fluent Inc. January 11, 2005 17-1
Modeling Cavitation
2. Unzip cavitation.zip.
cav.msh can be found in the /cavitation folder created after unzipping the file.
17-2
c Fluent Inc. January 11, 2005
Modeling Cavitation
Step 1: Grid
1. Read the grid file (cav.msh).
File −→ Read −→Case...
As FLUENT reads the grid file, it will report its progress in the console window.
c Fluent Inc. January 11, 2005 17-3
Modeling Cavitation
Grid
FLUENT 6.2 (2d, segregated, lam)
17-4
c Fluent Inc. January 11, 2005
Modeling Cavitation
Step 2: Models
1. Specify a steady-state axisymmetric model.
Define −→ Models −→Solver...
The segregated solver must be used for multiphase calculations.
c Fluent Inc. January 11, 2005 17-5
Modeling Cavitation
17-6
c Fluent Inc. January 11, 2005
Modeling Cavitation
i. Activate Cavitation.
The panel will expand again to show the cavitation inputs.
ii. Enter 3540 for the Vaporization Pressure (pascal).
The vaporization pressure is a property of the working liquid, which de-
pends mainly on the liquid temperature. The default value is the vapor-
ization pressure of water at a temperature of 300 K.
iii. Enter 0.0717 for the Surface Tension Coefficient.
Like the vaporization pressure, the liquid-vapor surface tension is a prop-
erty of the liquid, which depends mainly on temperature. Here too, the
default value is the surface tension for water and vapor at a temperature
of 300 K.
iv. Enter 1.5e-5 for Non-Condensable Gas Mass Fraction .
This is the mass fraction of non condensable gas dissolved in the working
liquid. 1.5e − 5 (15 ppm) is a typical value for air dissolved in water.
4. Turn on the standard k- turbulence model with standard wall functions.
Define −→ Models −→Viscous...
(a) Select k-epsilon as the Model.
c Fluent Inc. January 11, 2005 17-7
Modeling Cavitation
17-8
c Fluent Inc. January 11, 2005
Modeling Cavitation
Step 3: Materials
1. Create a new material to be used for the primary phase. Copy water vapor from
the materials database so that it can be used for the secondary phase, and modify
its density.
Define −→Materials...
(a) In the Name field, type water.
(b) Clear the Chemical Formula field.
(c) In the Density drop-down list, keep the default selection of constant, and enter
a value of 1000.
(d) In the Viscosity drop-down list, keep the default selection of constant, and enter
a value of 0.001.
(e) Click Change/Create, and then click Yes in the dialog box prompting whether
you want to overwrite the definition of air.
(f) Click the Fluent Database... button in the Materials panel.
The Fluent Database Materials panel will open.
c Fluent Inc. January 11, 2005 17-9
Modeling Cavitation
17-10
c Fluent Inc. January 11, 2005
Modeling Cavitation
Step 4: Phases
1. Define the liquid water and water vapor phases that flow through the orifice.
Define −→Phases...
ii. In the Primary Phase panel, enter liquid for the Name.
iii. Select water from the Phase Material drop-down list.
(b) Specify water vapor as the secondary phase.
i. Select phase-2 and click the Set... button.
ii. In the Secondary Phase panel, enter vapor for the Name.
iii. Select water-vapor from the Phase Material drop-down list.
c Fluent Inc. January 11, 2005 17-11
Modeling Cavitation
17-12
c Fluent Inc. January 11, 2005
Modeling Cavitation
c Fluent Inc. January 11, 2005 17-13
Modeling Cavitation
17-14
c Fluent Inc. January 11, 2005
Modeling Cavitation
iii. Select inlet-1 in the From Zone list, and then select inlet-2 in the To Zones
list.
iv. Click Copy and close the panel.
c Fluent Inc. January 11, 2005 17-15
Modeling Cavitation
17-16
c Fluent Inc. January 11, 2005
Modeling Cavitation
Step 7: Solution
1. Set the solution parameters.
Solve −→ Controls −→Solution...
(a) Under Under-Relaxation Factors, set the under-relaxation factor for both Pres-
sure and Momentum to 0.4.
(b) Scroll down and set the under-relaxation factors for Turbulence Kinetic Energy,
Turbulence Dissipation Rate, and Turbulent Viscosity to 0.5.
(c) Keep the under-relaxation factor for Vaporization Mass to 1.
FLUENT’s new cavitation model follows a different numerical approach from
the previous one. In general it is more robust and gives more accurate results.
Typically, for more complex cases, with very high pressure drops or large liquid-
vapor density ratios, the under-relaxation factors may need to be reduced to
between 0.1 and 0.2. For the Vaporization Mass, it is advised to use a value of
0.1. But in this tutorial a value of 1 is used, since the solution diverges for a
value of 0.1.
(d) Under Discretization, select Linear in the Pressure drop-down list.
(e) Select SIMPLEC in the Pressure-Velocity Coupling drop-down list.
c Fluent Inc. January 11, 2005 17-17
Modeling Cavitation
(a) Change all convergence criteria, except for vf-vapor, to 1e-5 for improved
accuracy.
(b) Select Plot under Options, and click OK.
17-18
c Fluent Inc. January 11, 2005
Modeling Cavitation
3. Initialize the solution from either of the pressure inlet zones (inlet-1 or inlet-2).
Solve −→ Initialize −→Initialize...
c Fluent Inc. January 11, 2005 17-19
Modeling Cavitation
Step 8: Postprocessing
1. Plot the pressure in the orifice.
Display −→Contours...
(a) Select Pressure... and Static Pressure in the drop-down lists under Contours of.
(b) Select Filled under Options.
(c) Click Display.
Note the dramatic pressure drop at the flow restriction in Figure 17.3. Low
static pressure is the major factor to cause cavitation, though turbulence also
contributes to cavitation, due to the effect of pressure fluctuation and turbulent
diffusion, as will be shown in the following plots.
To make the view more realistic, you will need to mirror it across the center-
line.
17-20
c Fluent Inc. January 11, 2005
Modeling Cavitation
4.99e+05
4.75e+05
4.50e+05
4.25e+05
4.00e+05
3.75e+05
3.51e+05
3.26e+05
3.01e+05
2.76e+05
2.51e+05
2.27e+05
2.02e+05
1.77e+05
1.52e+05
1.27e+05
1.03e+05
7.79e+04
5.31e+04
2.83e+04
3.54e+03
c Fluent Inc. January 11, 2005 17-21
Modeling Cavitation
(a) Select symm-1 and symm-2 in the Mirror Planes list and click Apply.
4.99e+05
4.75e+05
4.50e+05
4.25e+05
4.00e+05
3.75e+05
3.51e+05
3.26e+05
3.01e+05
2.76e+05
2.51e+05
2.27e+05
2.02e+05
1.77e+05
1.52e+05
1.27e+05
1.03e+05
7.79e+04
5.31e+04
2.83e+04
3.54e+03
17-22
c Fluent Inc. January 11, 2005
Modeling Cavitation
2.91e+01
2.77e+01
2.62e+01
2.48e+01
2.33e+01
2.18e+01
2.04e+01
1.89e+01
1.75e+01
1.60e+01
1.46e+01
1.31e+01
1.17e+01
1.02e+01
8.74e+00
7.29e+00
5.83e+00
4.37e+00
2.92e+00
1.46e+00
6.03e-03
In this example, the grid used is fairly coarse. However, in cavitating flows the
pressure distribution is the dominant factor, and is not very sensitive to grid size.
c Fluent Inc. January 11, 2005 17-23
Modeling Cavitation
1.00e-00
9.50e-01
9.00e-01
8.50e-01
8.00e-01
7.50e-01
7.00e-01
6.50e-01
6.00e-01
5.50e-01
5.00e-01
4.50e-01
4.00e-01
3.50e-01
3.00e-01
2.50e-01
2.00e-01
1.50e-01
1.00e-01
5.00e-02
0.00e+00
Note that the high turbulent kinetic energy region near the neck of the orifice (Fig-
ure 17.5) coincides with the highest volume fraction of vapor in Figure 17.6. This
indicates the correct prediction of a localized high phase change rate. The vapor
then gets convected downstream by the main flow.
Summary
This tutorial demonstrated how to set up and resolve a strongly cavitating pressure
driven flow through an orifice, using FLUENT’s multiphase mixture model with cavitation
effects. You learned how to set the boundary conditions for an internal flow. A steady-
state solution was calculated to simulate the formation of a vapor bubble in the neck
of the flow after the section restriction at the orifice. A more computationally-intensive
unsteady calculation is necessary to accurately simulate the irregular cyclic process of
bubble formation, growth, filling by water jet re-entry, and breakoff.
17-24
c Fluent Inc. January 11, 2005
Tutorial 18. Using the Mixture and Eulerian Multiphase
Models
Introduction
This tutorial examines the flow of water and air in a tee junction. First you will solve
the problem using the less computationally-intensive mixture model, and then you will
turn to the more accurate Eulerian model. Finally, you will compare the results obtained
with the two approaches.
In this tutorial you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1. Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
This problem considers an air-water mixture flowing upwards in a duct and then splitting
in a tee-junction. The ducts are 25 mm in width, the inlet section of the duct is 125 mm
long, and the top and the side ducts are 250 mm long. The geometry and data for the
problem are shown in Figure 18.1.
c Fluent Inc. January 11, 2005 18-1
Using the Mixture and Eulerian Multiphase Models
velocity inlet
water: v = - 0.31 m/s
air: v = - 0.45 m/s
pressure outlet
velocity inlet
water: air:
ρ=1000 kg/m3 ρ=1.2 kg/m3
µ=9e-4 kg/m-s µ=2e-5 kg/m-s
v=1.53 m/s v=1.6 m/s
vol frac=0.02
bubble diam=1
2. Unzip mix_eulerian_multiphase.zip.
tee.msh can be found in the /mix eulerian multiphase folder created after un-
zipping the file.
18-2
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
Step 1: Grid
1. Read the grid file (tee.msh).
File −→ Read −→Case...
As FLUENT reads the grid file, it will report its progress in the console window.
(a) Display the grid using the default settings (Figure 18.2).
Extra: You can use the right mouse button to check which zone number cor-
responds to each boundary. If you click the right mouse button on one
of the boundaries in the graphics window, its zone number, name, and
type will be printed in the FLUENT console window. This feature is espe-
cially useful when you have several zones of the same type and you want
to distinguish between them quickly.
c Fluent Inc. January 11, 2005 18-3
Using the Mixture and Eulerian Multiphase Models
Grid
FLUENT 6.2 (2d, segregated, lam)
18-4
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
Step 2: Models
1. Keep the default settings for the 2D segregated steady-state solver.
Define −→ Models −→Solver...
The segregated solver must be used for multiphase calculations.
c Fluent Inc. January 11, 2005 18-5
Using the Mixture and Eulerian Multiphase Models
18-6
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
3. Turn on the standard k- turbulence model with standard wall functions.
Define −→ Models −→Viscous...
c Fluent Inc. January 11, 2005 18-7
Using the Mixture and Eulerian Multiphase Models
18-8
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
Step 3: Materials
1. Copy liquid water from the materials database so that it can be used for the primary
phase.
Define −→Materials...
(a) Click the Fluent Database... button in the Materials panel.
The Fluent Database Materials panel will open.
c Fluent Inc. January 11, 2005 18-9
Using the Mixture and Eulerian Multiphase Models
Step 4: Phases
1. Define the liquid water and air phases that flow in the tee junction.
Define −→Phases...
18-10
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
c Fluent Inc. January 11, 2005 18-11
Using the Mixture and Eulerian Multiphase Models
18-12
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
ii. In the Turbulence Specification Method drop-down list, select Intensity and
Length Scale.
iii. Set Turbulence Intensity to 10% and Turbulence Length Scale to 0.025 m.
c Fluent Inc. January 11, 2005 18-13
Using the Mixture and Eulerian Multiphase Models
ii. Keep the default Velocity Specification Method and Reference Frame.
iii. Set the Velocity Magnitude to 1.53.
(c) Set the conditions for the secondary phase.
i. In the Boundary Conditions panel, select air from the Phase drop-down list
and click Set....
ii. Keep the default Velocity Specification Method and Reference Frame.
iii. Set the Velocity Magnitude to 1.6.
iv. Set the Volume Fraction to 0.02.
18-14
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
ii. In the Turbulence Specification Method drop-down list, select Intensity and
Length Scale.
iii. Set Turbulence Intensity to 10% and Turbulence Length Scale to 0.025 m.
(b) Set the conditions for the primary phase.
i. In the Boundary Conditions panel, select water from the Phase drop-down
list and click Set....
ii. Keep the default Velocity Specification Method and Reference Frame.
iii. Set the Velocity Magnitude to -0.31.
In this problem, outflow characteristics at the upper velocity inlet are as-
sumed to be known, and therefore imposed as a boundary condition.
c Fluent Inc. January 11, 2005 18-15
Using the Mixture and Eulerian Multiphase Models
ii. Keep the default Velocity Specification Method and Reference Frame.
iii. Set the Velocity Magnitude to -0.45.
iv. Set the Volume Fraction to 0.02.
18-16
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
ii. In the Turbulence Specification Method drop-down list, select Intensity and
Length Scale.
iii. Set the Backflow Turbulence Intensity to 10%.
iv. Set the Backflow Turbulence Length Scale to 0.025.
(b) Set the conditions for the secondary phase.
i. In the Boundary Conditions panel, select air from the Phase drop-down list
and click Set....
ii. Set the Backflow Volume Fraction to 0.02.
c Fluent Inc. January 11, 2005 18-17
Using the Mixture and Eulerian Multiphase Models
18-18
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
c Fluent Inc. January 11, 2005 18-19
Using the Mixture and Eulerian Multiphase Models
(a) Select Pressure... and Static Pressure in the Contours of drop-down lists.
(b) Select Filled under Options.
(c) Click Display.
18-20
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
2.35e+03
2.15e+03
1.96e+03
1.76e+03
1.57e+03
1.37e+03
1.18e+03
9.80e+02
7.85e+02
5.90e+02
3.96e+02
2.01e+02
5.63e+00
-1.89e+02
-3.84e+02
-5.79e+02
-7.74e+02
-9.69e+02
-1.16e+03
-1.36e+03
-1.55e+03
c Fluent Inc. January 11, 2005 18-21
Using the Mixture and Eulerian Multiphase Models
2.24e+00
2.12e+00
2.01e+00
1.90e+00
1.79e+00
1.68e+00
1.57e+00
1.45e+00
1.34e+00
1.23e+00
1.12e+00
1.01e+00
8.95e-01
7.83e-01
6.71e-01
5.59e-01
4.47e-01
3.35e-01
2.24e-01
1.12e-01
0.00e+00
18-22
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
8.80e-01
8.36e-01
7.92e-01
7.48e-01
7.04e-01
6.60e-01
6.16e-01
5.72e-01
5.28e-01
4.84e-01
4.40e-01
3.96e-01
3.52e-01
3.08e-01
2.64e-01
2.20e-01
1.76e-01
1.32e-01
8.80e-02
4.40e-02
4.33e-13
c Fluent Inc. January 11, 2005 18-23
Using the Mixture and Eulerian Multiphase Models
18-24
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
2. Specify the drag law to be used for computing the interphase momentum transfer.
Define −→Phases...
(a) Click the Interaction... button in the Phases panel.
(b) In the Phase Interaction panel, keep the default selection of schiller-naumann in
the Drag Coefficient drop-down list.
Note: For this problem there are no parameters to be set for the individual phases,
other than those that you specified when you set up the phases for the mixture
model calculation. If you use the Eulerian model for a flow involving a granular
secondary phase, there are additional parameters that you need to set. There
are also other options in the Phase Interaction panel that may be relevant for
other applications.
See Section 24.8 of the User’s Guide for complete details on setting up an Eulerian
multiphase calculation.
c Fluent Inc. January 11, 2005 18-25
Using the Mixture and Eulerian Multiphase Models
(a) Under k-epsilon Multiphase Model, keep the default selection of Mixture.
The mixture turbulence model is applicable when phases separate, for stratified
(or nearly stratified) multiphase flows, and when the density ratio between
phases is close to 1. In these cases, using mixture properties and mixture
velocities is sufficient to capture important features of the turbulent flow.
See Chapter 24 of the User’s Guide for more information on turbulence models
for the Eulerian multiphase model.
18-26
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
c Fluent Inc. January 11, 2005 18-27
Using the Mixture and Eulerian Multiphase Models
2.40e+03
2.21e+03
2.02e+03
1.83e+03
1.64e+03
1.45e+03
1.27e+03
1.08e+03
8.87e+02
6.97e+02
5.08e+02
3.18e+02
1.29e+02
-6.04e+01
-2.50e+02
-4.39e+02
-6.29e+02
-8.18e+02
-1.01e+03
-1.20e+03
-1.39e+03
18-28
c Fluent Inc. January 11, 2005
Using the Mixture and Eulerian Multiphase Models
2.37e+00
2.25e+00
2.13e+00
2.01e+00
1.90e+00
1.78e+00
1.66e+00
1.54e+00
1.42e+00
1.30e+00
1.19e+00
1.07e+00
9.48e-01
8.30e-01
7.11e-01
5.93e-01
4.74e-01
3.56e-01
2.37e-01
1.19e-01
0.00e+00
9.89e-01
9.40e-01
8.90e-01
8.41e-01
7.91e-01
7.42e-01
6.93e-01
6.43e-01
5.94e-01
5.44e-01
4.95e-01
4.45e-01
3.96e-01
3.46e-01
2.97e-01
2.47e-01
1.98e-01
1.48e-01
9.89e-02
4.95e-02
0.00e+00
c Fluent Inc. January 11, 2005 18-29
Using the Mixture and Eulerian Multiphase Models
Summary
This tutorial demonstrated how to set up and solve a multiphase problem using the
mixture model and the Eulerian model. You learned how to set boundary conditions for
the mixture and both phases. The solution obtained with the mixture model was used as a
starting point for the calculation with the Eulerian model. After completing calculations
with both models, you compared the results obtained with the two approaches.
18-30
c Fluent Inc. January 11, 2005
Tutorial 19. Using the Eulerian Multiphase Model for
Granular Flow
Introduction
Mixing tanks are used to maintain solid particles or droplets of heavy fluids in suspension.
Mixing may be required to enhance reaction during chemical processing or to prevent
sedimentation. In this tutorial, you will use the Eulerian multiphase model to solve the
particle suspension problem. The Eulerian multiphase model solves momentum equations
for each of the phases, which are allowed to mix in any proportion.
In this tutorial you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1. Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The problem involves the transient startup of an impeller-driven mixing tank. The
primary phase is water, while the secondary phase consists of sand particles with a 111
micron diameter. The sand is initially settled at the bottom of the tank, to a level just
above the impeller. A schematic of the mixing tank and the initial sand position is shown
in Figure 19.1. The domain is modeled as 2D axisymmetric.
c Fluent Inc. January 11, 2005 19-1
Using the Eulerian Multiphase Model for Granular Flow
.4446 m
.016 m
water
.4446 m
impeller
settled
sand .1728 m
.0864 m
bed .116 m
The fixed-values option will be used to simulate the impeller. Experimental data are used
to represent the time-averaged velocity and turbulence values at the impeller location.
This approach avoids the need to model the impeller itself. These experimental data are
provided in a user-defined function.
2. Unzip eulerian_multiphase_granular.zip.
mixtank.msh and fix.c can be found in the /eulerian multiphase granular
folder created after unzipping the file.
19-2
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
Step 1: Grid
1. Read the grid file (mixtank.msh).
File −→ Read −→Case...
As FLUENT reads the grid file, it will report its progress in the console window.
Grid
FLUENT 6.2 (2d, segregated, lam)
Extra: You can use the right mouse button to check which zone number corre-
sponds to each boundary. If you click the right mouse button on one of the
boundaries in the graphics window, its zone number, name, and type will be
printed in the FLUENT console window. This feature is especially useful when
you have several zones of the same type and you want to distinguish between
them quickly.
c Fluent Inc. January 11, 2005 19-3
Using the Eulerian Multiphase Model for Granular Flow
Grid
FLUENT 6.2 (axi, segregated, lam)
19-4
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
Grid
FLUENT 6.2 (axi, segregated, lam)
c Fluent Inc. January 11, 2005 19-5
Using the Eulerian Multiphase Model for Granular Flow
i. With left mouse button drag the indicator of the dial in the counter-
clockwise direction till the upright view is displayed (Figure 19.5).
ii. Click Apply and close the Camera Parameters.
Grid
FLUENT 6.2 (axi, segregated, lam)
19-6
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
Step 2: Models
1. Specify a transient, axisymmetric model.
Define −→ Models −→Solver...
c Fluent Inc. January 11, 2005 19-7
Using the Eulerian Multiphase Model for Granular Flow
19-8
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
c Fluent Inc. January 11, 2005 19-9
Using the Eulerian Multiphase Model for Granular Flow
Step 3: Materials
In this step, you will add liquid water to the list of fluid materials by copying it from the
Fluent Database for materials, and create a new material called sand.
Define −→Materials...
1. Copy liquid water from the materials database so that it can be used for the primary
phase.
(a) Click the Fluent Database... button in the Materials panel.
The Fluent Database Materials panel will open.
19-10
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
c Fluent Inc. January 11, 2005 19-11
Using the Eulerian Multiphase Model for Granular Flow
Step 4: Phases
Define −→Phases...
19-12
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
c Fluent Inc. January 11, 2005 19-13
Using the Eulerian Multiphase Model for Granular Flow
3. Specify the drag law to be used for computing the interphase momentum transfer.
(a) Click the Interaction... button in the Phases panel.
(b) In the Phase Interaction panel, select gidaspow in the Drag Coefficient drop-
down list and click OK.
19-14
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
Variable A1 A2 A3 A4 A5 A6
u velocity -7.1357e-2 54.304 -3.1345e+3 4.5578e+4 -1.966e+5 –
v velocity 3.1131e-2 -10.313 9.5558e+2 -2.0051e+4 1.186e+5 –
kinetic energy 2.2723e-2 6.7989 -424.18 9.4615e+3 -7.725e+4 1.8410e+5
dissipation -6.5819e-2 88.845 -5.3731e+3 1.1643e+5 -9.120e+5 1.9567e+6
See the separate UDF Manual about setting up a UDF using the DEFINE PROFILE macro.
Note that, while this macro is usually used to specify a profile condition on a boundary
face zone, it is used in fix.c to specify the condition in a fluid cell zone. The arguments
of the macro have been changed accordingly.
c Fluent Inc. January 11, 2005 19-15
Using the Eulerian Multiphase Model for Granular Flow
2. Set the conditions for the fluid zone representing the impeller (fix-zone).
Specify the conditions for water and sand separately using the UDF. For the mixture
(i.e., conditions that apply to all phases), the default conditions are acceptable.
Define −→Boundary Conditions...
(a) Set the water boundary conditions on fix-zone.
i. In the Boundary Conditions panel, select water under Phase and click Set....
19-16
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
c Fluent Inc. January 11, 2005 19-17
Using the Eulerian Multiphase Model for Granular Flow
19-18
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
Step 6: Solution
1. Set the solution parameters.
Solve −→ Controls −→Solution...
(a) For the Under-Relaxation Factors, set Pressure to 0.5, Momentum to 0.2, and
Turbulent Viscosity to 0.8.
(b) Under Discretization, keep the default settings.
c Fluent Inc. January 11, 2005 19-19
Using the Eulerian Multiphase Model for Granular Flow
19-20
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
6. Save the initial case and data files (mixtank.cas and mixtank.dat).
File −→ Write −→Case & Data...
The problem statement is now complete. As a precaution, you should review the
impeller velocity fixes and sand bed patch after running the calculation for a single
time step. Since you are using a UDF for the velocity profiles, you need to perform
one time step in order for the profiles to be calculated and available for viewing.
7. Run the calculation for 0.005 seconds.
Solve −→Iterate...
(a) Set the Number of Time Steps to 1 and click Iterate.
8. Check the initial velocities and sand volume fraction.
In order to display the initial fixed velocities in the fluid zone (fix-zone), you will
need to create a surface for this zone.
c Fluent Inc. January 11, 2005 19-21
Using the Eulerian Multiphase Model for Granular Flow
19-22
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
c Fluent Inc. January 11, 2005 19-23
Using the Eulerian Multiphase Model for Granular Flow
8.20e-01
7.79e-01
7.38e-01
6.97e-01
6.56e-01
6.15e-01
5.74e-01
5.33e-01
4.92e-01
4.51e-01
4.10e-01
3.69e-01
3.28e-01
2.87e-01
2.46e-01
2.05e-01
1.64e-01
1.23e-01
8.20e-02
4.10e-02
4.25e-05
8.25e-01
7.84e-01
7.42e-01
7.01e-01
6.60e-01
6.19e-01
5.77e-01
5.36e-01
4.95e-01
4.54e-01
4.12e-01
3.71e-01
3.30e-01
2.89e-01
2.47e-01
2.06e-01
1.65e-01
1.24e-01
8.25e-02
4.12e-02
0.00e+00
19-24
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
c Fluent Inc. January 11, 2005 19-25
Using the Eulerian Multiphase Model for Granular Flow
5.62e-01
5.34e-01
5.06e-01
4.78e-01
4.50e-01
4.22e-01
3.94e-01
3.65e-01
3.37e-01
3.09e-01
2.81e-01
2.53e-01
2.25e-01
1.97e-01
1.69e-01
1.41e-01
1.12e-01
8.43e-02
5.62e-02
2.81e-02
0.00e+00
10. Save the case and data files (mixtank1.cas and mixtank1.dat).
File −→ Write −→Case & Data...
19-26
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
8.20e-01
7.79e-01
7.38e-01
6.97e-01
6.56e-01
6.15e-01
5.74e-01
5.33e-01
4.92e-01
4.51e-01
4.10e-01
3.69e-01
3.28e-01
2.87e-01
2.46e-01
2.05e-01
1.64e-01
1.23e-01
8.20e-02
4.10e-02
4.25e-05
c Fluent Inc. January 11, 2005 19-27
Using the Eulerian Multiphase Model for Granular Flow
8.25e-01
7.84e-01
7.42e-01
7.01e-01
6.60e-01
6.19e-01
5.77e-01
5.36e-01
4.95e-01
4.54e-01
4.12e-01
3.71e-01
3.30e-01
2.89e-01
2.47e-01
2.06e-01
1.65e-01
1.24e-01
8.25e-02
4.12e-02
0.00e+00
5.48e-01
5.21e-01
4.94e-01
4.66e-01
4.39e-01
4.11e-01
3.84e-01
3.57e-01
3.29e-01
3.02e-01
2.74e-01
2.47e-01
2.19e-01
1.92e-01
1.65e-01
1.37e-01
1.10e-01
8.23e-02
5.48e-02
2.74e-02
0.00e+00
19-28
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
13. Save the case and data files (mixtank20.cas and mixtank20.dat).
File −→ Write −→Case & Data...
c Fluent Inc. January 11, 2005 19-29
Using the Eulerian Multiphase Model for Granular Flow
Step 7: Postprocessing
You will now examine the progress of the sand and water in the mixing tank after a total
of 20 seconds.The mixing tank has nearly, but not quite, reached a steady flow solution.
8.91e-01
8.47e-01
8.02e-01
7.58e-01
7.13e-01
6.69e-01
6.24e-01
5.79e-01
5.35e-01
4.90e-01
4.46e-01
4.01e-01
3.57e-01
3.12e-01
2.68e-01
2.23e-01
1.79e-01
1.34e-01
8.96e-02
4.51e-02
5.63e-04
19-30
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
8.86e-01
8.42e-01
7.98e-01
7.53e-01
7.09e-01
6.65e-01
6.20e-01
5.76e-01
5.32e-01
4.87e-01
4.43e-01
3.99e-01
3.55e-01
3.10e-01
2.66e-01
2.22e-01
1.77e-01
1.33e-01
8.86e-02
4.43e-02
0.00e+00
c Fluent Inc. January 11, 2005 19-31
Using the Eulerian Multiphase Model for Granular Flow
2.47e-01
2.34e-01
2.22e-01
2.10e-01
1.97e-01
1.85e-01
1.73e-01
1.60e-01
1.48e-01
1.36e-01
1.23e-01
1.11e-01
9.87e-02
8.63e-02
7.40e-02
6.17e-02
4.93e-02
3.70e-02
2.47e-02
1.23e-02
0.00e+00
4. Display filled contours of static pressure for the mixture (Figure 19.15).
Display −→Contours...
(a) Select mixture in the Phase drop-down list.
(b) Select Pressure... and Static Pressure in the Contours of drop-down lists.
(c) Click Display.
Figure 19.15 shows the pressure distribution after 20 seconds of operation.
The pressure field represents the hydrostatic pressure except for some slight
deviations due to the flow of the impeller near the bottom of the tank.
19-32
c Fluent Inc. January 11, 2005
Using the Eulerian Multiphase Model for Granular Flow
7.87e+01
4.14e+00
-7.04e+01
-1.45e+02
-2.19e+02
-2.94e+02
-3.69e+02
-4.43e+02
-5.18e+02
-5.92e+02
-6.67e+02
-7.41e+02
-8.16e+02
-8.90e+02
-9.65e+02
-1.04e+03
-1.11e+03
-1.19e+03
-1.26e+03
-1.34e+03
-1.41e+03
Summary
This tutorial demonstrated how to set up and solve a granular multiphase problem us-
ing the Eulerian multiphase model. The problem involved the 2D modeling of particle
suspension in a mixing tank, and postprocessing showed the near-steady-state behavior
of the sand in the mixing tank, under the assumptions made.
c Fluent Inc. January 11, 2005 19-33
Using the Eulerian Multiphase Model for Granular Flow
19-34
c Fluent Inc. January 11, 2005
Tutorial 20. Modeling Solidification
Introduction
This tutorial illustrates how to set up and solve a problem involving solidification. In
this tutorial, you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1. Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
This tutorial demonstrates the setup and solution procedure for a fluid flow and heat
transfer problem involving solidification, namely the Czochralski growth process. The
geometry considered is a 2D axisymmetric bowl (shown in Figure 20.1), containing a liq-
uid metal. The bottom and sides of the bowl are heated above the liquidus temperature,
as is the free surface of the liquid. The liquid is solidified by heat loss from the crystal
and the solid is pulled out of the domain at a rate of 0.001 m/s and a temperature of
500 K. There is a steady injection of liquid at the bottom of the bowl with a velocity of
1.01 × 10−3 and a temperature of 1300 K. Material properties are listed in Figure 20.1.
Starting with an existing 2D mesh, the details regarding the setup and solution proce-
dure for the solidification problem are presented. The steady conduction solution for
this problem is computed as an initial condition. Then, the fluid flow is turned on to
investigate the effect of natural and Marangoni convection in an unsteady fashion.
c Fluent Inc. January 11, 2005 20-1
Modeling Solidification
20-2
c Fluent Inc. January 11, 2005
Modeling Solidification
2. Unzip solidification.zip.
solid.msh can be found in the /solidification folder created after unzipping the
file.
Step 1: Grid
1. Read the mesh file solid.msh.
File −→ Read −→Case...
As this mesh is read by FLUENT, messages appear in the console window reporting
the progress of the reading.
c Fluent Inc. January 11, 2005 20-3
Modeling Solidification
Grid
FLUENT 6.2 (2d, segregated, lam)
20-4
c Fluent Inc. January 11, 2005
Modeling Solidification
Step 2: Models
1. Enable the modeling of axisymmetric swirl.
Define −→ Models −→Solver...
c Fluent Inc. January 11, 2005 20-5
Modeling Solidification
(b) Under Parameters, keep the default value for the Mushy Zone Constant.
The default value of 100000 is acceptable for most cases.
(c) Turn on Include Pull Velocities.
The panel will expand to show an additional input.
• Including the pull velocities accounts for the movement of the solidified
material as it is continuously withdrawn from the domain in the continu-
ous casting process.
• It is possible to have FLUENT compute the pull velocities during the cal-
culation, but this approach is computationally expensive, and is recom-
mended only if the pull velocities are strongly dependent on the location
of the liquid-solid interface. In this tutorial, you will patch values for the
pull velocities instead of having FLUENT compute them. See the User’s
Guide for more information.
• When you click OK in the Solidification and Melting panel, FLUENT will present
an Information dialog box telling you that available material properties have
changed for the solidification model. You will be setting properties later, so
you can simply click OK in the dialog box to acknowledge this information.
• FLUENT will automatically enable the energy calculation when you enable the
solidification model, so you need not visit the Energy panel.
20-6
c Fluent Inc. January 11, 2005
Modeling Solidification
c Fluent Inc. January 11, 2005 20-7
Modeling Solidification
Step 3: Materials
In this step, you will create a new material and specify its properties, including the melting
heat, solidus temperature, and liquidus temperature.
Define −→Materials...
20-8
c Fluent Inc. January 11, 2005
Modeling Solidification
c Fluent Inc. January 11, 2005 20-9
Modeling Solidification
20-10
c Fluent Inc. January 11, 2005
Modeling Solidification
c Fluent Inc. January 11, 2005 20-11
Modeling Solidification
20-12
c Fluent Inc. January 11, 2005
Modeling Solidification
c Fluent Inc. January 11, 2005 20-13
Modeling Solidification
20-14
c Fluent Inc. January 11, 2005
Modeling Solidification
c Fluent Inc. January 11, 2005 20-15
Modeling Solidification
20-16
c Fluent Inc. January 11, 2005
Modeling Solidification
c Fluent Inc. January 11, 2005 20-17
Modeling Solidification
(a) In the Field Functions drop-down lists, select Grid... and Radial Coordinate.
(b) Click the Select button.
radial-coordinate will appear in the Definition field. If you make a mistake,
click the DEL button on the calculator pad to delete the last item you added to
the function definition.
(c) Click the X button on the calculator pad.
(d) Click on 1.
(e) Enter omegar as the New Function Name.
(f) Click Define and close the panel.
If you wish to check the function definition, click on Manage... and select
omegar.
20-18
c Fluent Inc. January 11, 2005
Modeling Solidification
c Fluent Inc. January 11, 2005 20-19
Modeling Solidification
6. Save the initial case and data files (solid0.cas and solid0.dat).
File −→ Write −→Case & Data...
20-20
c Fluent Inc. January 11, 2005
Modeling Solidification
c Fluent Inc. January 11, 2005 20-21
Modeling Solidification
1.40e+03
1.36e+03
1.31e+03
1.27e+03
1.22e+03
1.18e+03
1.13e+03
1.09e+03
1.04e+03
9.95e+02
9.50e+02
9.05e+02
8.60e+02
8.15e+02
7.70e+02
7.25e+02
6.80e+02
6.35e+02
5.90e+02
5.45e+02
5.00e+02
The thickness of the mushy zone can be determined from the contours of tempera-
ture. The mushy zone is the region where the temperature is between the liquidus
temperature and solidus temperature.
9. Save the case and data files for the steady conduction solution (solid.cas and
solid.dat).
File −→ Write −→Case & Data...
20-22
c Fluent Inc. January 11, 2005
Modeling Solidification
c Fluent Inc. January 11, 2005 20-23
Modeling Solidification
3. Save the initial case and data files (solid01.cas and solid01.dat).
File −→ Write −→Case & Data...
(a) Under Time, set the Time Step Size to 0.1 seconds.
(b) Set the Number of Time Steps to 2.
(c) Under Iteration, retain the default value of 20 for Max Iterations per Time Step.
(d) Click Iterate.
20-24
c Fluent Inc. January 11, 2005
Modeling Solidification
1.40e+03
1.36e+03
1.31e+03
1.27e+03
1.22e+03
1.18e+03
1.13e+03
1.09e+03
1.04e+03
9.95e+02
9.50e+02
9.05e+02
8.60e+02
8.15e+02
7.70e+02
7.25e+02
6.80e+02
6.35e+02
5.90e+02
5.45e+02
5.00e+02
The temperature contours show the gradient in temperature from the hot
walls on the left to the cooler zone on the right.
(b) Display contours of stream function (Figure 20.5).
i. Under Options, deselect Filled.
ii. Select Velocity... and Stream Function in the Contours of drop-down lists.
iii. Click Display.
c Fluent Inc. January 11, 2005 20-25
Modeling Solidification
2.18e-02
2.07e-02
1.97e-02
1.86e-02
1.75e-02
1.64e-02
1.53e-02
1.42e-02
1.31e-02
1.20e-02
1.09e-02
9.83e-03
8.74e-03
7.64e-03
6.55e-03
5.46e-03
4.37e-03
3.28e-03
2.18e-03
1.09e-03
0.00e+00
20-26
c Fluent Inc. January 11, 2005
Modeling Solidification
1.00e+00
9.50e-01
9.00e-01
8.50e-01
8.00e-01
7.50e-01
7.00e-01
6.50e-01
6.00e-01
5.50e-01
5.00e-01
4.50e-01
4.00e-01
3.50e-01
3.00e-01
2.50e-01
2.00e-01
1.50e-01
1.00e-01
5.00e-02
0.00e+00
The liquid fraction contours show the current position of the melt front.
Note that in Figure 20.6, the mushy zone divides the liquid and solid
regions roughly in half.
c Fluent Inc. January 11, 2005 20-27
Modeling Solidification
1.40e+03
1.36e+03
1.31e+03
1.26e+03
1.22e+03
1.18e+03
1.13e+03
1.08e+03
1.04e+03
9.95e+02
9.50e+02
9.05e+02
8.60e+02
8.15e+02
7.70e+02
7.25e+02
6.80e+02
6.35e+02
5.90e+02
5.45e+02
5.00e+02
As shown in Figure 20.7, the temperature contours are fairly uniform through
the melt front and solid material. The distortion of the temperature field due
to the recirculating liquid is also clearly evident.
In a continuous casting process, it is important to pull out the solidified ma-
terial at the proper time. If the material is pulled out too soon, it will not
have solidified; that is, it will still be in a mushy state. If it is pulled out too
late, it solidifies in the casting pool and cannot be pulled out in the required
shape. The optimal rate of pull can be determined from the contours of liquidus
temperature and solidus temperature.
20-28
c Fluent Inc. January 11, 2005
Modeling Solidification
1.41e-01
1.34e-01
1.27e-01
1.20e-01
1.13e-01
1.05e-01
9.84e-02
9.14e-02
8.44e-02
7.73e-02
7.03e-02
6.33e-02
5.63e-02
4.92e-02
4.22e-02
3.52e-02
2.81e-02
2.11e-02
1.41e-02
7.03e-03
0.00e+00
Note that the flow has developed more fully now, as compared with Figure 20.5
after 0.2 seconds. The main eddy, driven by natural convection and Marangoni
stress, dominates the flow.
To examine the position of the melt front and the extent of the mushy zone,
you will plot the contours of liquid fraction.
c Fluent Inc. January 11, 2005 20-29
Modeling Solidification
1.00e+00
9.50e-01
9.00e-01
8.50e-01
8.00e-01
7.50e-01
7.00e-01
6.50e-01
6.00e-01
5.50e-01
5.00e-01
4.50e-01
4.00e-01
3.50e-01
3.00e-01
2.50e-01
2.00e-01
1.50e-01
1.00e-01
5.00e-02
0.00e+00
The introduction of liquid material at the left of the domain is balanced by the
pulling of the solidified material from the right. After 5 seconds, the equilib-
rium position of the melt front is beginning to be established.
8. Save the case and data files for the solution at 5 seconds (solid5.cas and solid5.dat).
File −→ Write −→Case & Data...
20-30
c Fluent Inc. January 11, 2005
Modeling Solidification
Summary
In this tutorial, you studied the setup and solution for a fluid flow problem involving
solidification for the Czochralski growth process.
The solidification model in FLUENT can be used to model the continuous casting process
where a solid material is continuously pulled out from the casting domain. In this tutorial,
you patched a constant value and a custom field function for the pull velocities instead of
computing them. For cases where the pull velocity is not changing over the domain, this
approach is used as it is computationally less expensive than having FLUENT compute
the pull velocities during the calculation.
See Chapter 25 of the User’s Guide for more information about the solidification/melting
model.
c Fluent Inc. January 11, 2005 20-31
Modeling Solidification
20-32
c Fluent Inc. January 11, 2005
Tutorial 21. Using the Eulerian Granular Multiphase Model
with Heat Transfer
Introduction
This tutorial examines the flow of air and a granular solid phase consisting of glass beads
in a hot gas fluidized bed, under uniform minimum fluidization conditions. The results
obtained for the local wall-to-bed heat transfer coefficient in FLUENT are compared with
the analytical results.
In this tutorial you will learn how to:
• Use a user defined function (UDF) to specify a phase-specific velocity inlet profile
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1. Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
This problem considers a hot gas fluidized bed in which air flows upwards through the
bottom of the domain and through an additional small orifice next to a heated wall. A
uniformly fluidized bed is examined, for the sake of comparison with analytical results
[1]. The geometry and data for the problem are shown in Figure 21.1.
c Fluent Inc. January 11, 2005 21-1
Using the Eulerian Granular Multiphase Model with Heat Transfer
pressure outlet
101325 Pa
0.598
volume
fraction
of solids
2. Unzip eulerian_granular_heat.zip.
fluid-bed.msh and gasvel.c can be found in the /eulerian granular heat folder
created after unzipping the file.
21-2
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
Step 1: Grid
1. Read the grid file (fluid-bed.msh).
File −→ Read −→Case...
As FLUENT reads the grid file, it will report its progress in the console window.
c Fluent Inc. January 11, 2005 21-3
Using the Eulerian Granular Multiphase Model with Heat Transfer
Grid
FLUENT 6.2 (2d, dp, segregated, lam)
21-4
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
Step 2: Models
1. Enable the segregated unsteady solver.
The segregated solver must be used for multiphase calculations.
Define −→ Models −→Solver...
(a) Under Time, select Unsteady and click OK.
c Fluent Inc. January 11, 2005 21-5
Using the Eulerian Granular Multiphase Model with Heat Transfer
21-6
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
c Fluent Inc. January 11, 2005 21-7
Using the Eulerian Granular Multiphase Model with Heat Transfer
Step 3: Materials
1. Compile the user-defined function, gasvel.c, that will be used to define the thermal
conductivity for the gas and solid phase.
Define −→ User-Defined −→ Functions −→Compiled...
21-8
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
2. Modify the properties for air, which will be used for the primary phase.
The properties used for air are modified to match data used by Kuipers et al. [1]
Define −→Materials...
c Fluent Inc. January 11, 2005 21-9
Using the Eulerian Granular Multiphase Model with Heat Transfer
3. Define a new fluid material for the granular phase (the glass beads).
21-10
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
Step 4: Phases
Define −→Phases...
(b) In the Primary Phase panel, enter air for the Name.
(c) Select air from the Phase Material drop-down list and click OK.
c Fluent Inc. January 11, 2005 21-11
Using the Eulerian Granular Multiphase Model with Heat Transfer
(b) In the Secondary Phase panel, enter solids for the Name.
(c) Select solids from the Phase Material drop-down list.
(d) Turn on Granular.
(e) Under Granular Temperature Model retain Phase Property option.
(f) Set the Diameter to 0.0005 m.
(g) In the Granular Viscosity drop-down list, select syamlal-obrien.
(h) In the Granular Bulk Viscosity drop-down list, select lun-et-al.
(i) Click OK.
21-12
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
c Fluent Inc. January 11, 2005 21-13
Using the Eulerian Granular Multiphase Model with Heat Transfer
ii. Keep the default Velocity Specification Method and Reference Frame.
iii. Set the Velocity Magnitude to 0.25.
iv. Set the Temperature to 293.
v. Click OK.
(b) Set the conditions at v uniform for the secondary phase.
i. In the Boundary Conditions panel, select solids from the Phase drop-down
list and click Set....
ii. Keep the default Velocity Specification Method and Reference Frame.
iii. Keep the default value of 0 for the Velocity Magnitude.
iv. Set the Temperature to 293.
v. Set the Granular Temperature to 0.0001.
vi. Keep the default value of 0 for the Volume Fraction.
vii. Click OK.
21-14
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
ii. Keep the default Velocity Specification Method and Reference Frame.
iii. Set the Velocity Magnitude to 0.25.
In order for a comparison with analytical results [1] to be meaningful, in
this simulation you will use a uniform value for the air velocity equal to
the minimum fluidization velocity at both inlets on the bottom of the bed.
iv. Set the Temperature to 293.
v. Click OK.
c Fluent Inc. January 11, 2005 21-15
Using the Eulerian Granular Multiphase Model with Heat Transfer
ii. Keep the default Velocity Specification Method and Reference Frame.
iii. Keep the default value of 0 for the Velocity Magnitude.
iv. Set the Temperature to 293.
v. Set the Granular Temperature to 0.0001.
vi. Keep the default value of 0 for the Volume Fraction.
vii. Click OK.
21-16
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
c Fluent Inc. January 11, 2005 21-17
Using the Eulerian Granular Multiphase Model with Heat Transfer
4. Set the boundary conditions for the heated wall (wall hot).
For the heated wall, you will set thermal conditions for the mixture, and momentum
conditions (zero shear) for both phases.
(a) Set the conditions for the mixture.
i. In the Boundary Conditions panel, select mixture from the Phase drop-down
list and click Set....
ii. Select Temperature under Thermal Conditions, and input 373 for the Tem-
perature.
iii. Click OK.
21-18
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
c Fluent Inc. January 11, 2005 21-19
Using the Eulerian Granular Multiphase Model with Heat Transfer
5. Set the boundary conditions for the adiabatic wall (wall ins).
For the adiabatic wall, you will retain the default thermal conditions for the mixture
(zero heat flux), and set momentum conditions (zero shear) for both phases.
(a) Set the conditions for the primary phase.
i. In the Boundary Conditions panel, select air from the Phase drop-down list
and click Set....
ii. Select Specified Shear under Shear Condition (the panel will expand), and
keep the default values of 0 for the X-Component and Y-Component.
iii. Click OK.
(b) Set the conditions for the secondary phase.
For the secondary phase, you will set the same conditions of zero shear as for
the primary phase.
21-20
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
Step 6: Solution
1. Set the solution parameters.
Solve −→ Controls −→Solution...
c Fluent Inc. January 11, 2005 21-21
Using the Eulerian Granular Multiphase Model with Heat Transfer
21-22
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
c Fluent Inc. January 11, 2005 21-23
Using the Eulerian Granular Multiphase Model with Heat Transfer
4. Define the point surface in the cell next to the wall, on the plane y = 0.24.
Surface −→Point...
(a) Under Coordinates, enter 0.28494 for x0, and 0.24 for y0.
(b) Enter y=0.24 for the New Surface Name.
(c) Click Create.
21-24
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
c Fluent Inc. January 11, 2005 21-25
Using the Eulerian Granular Multiphase Model with Heat Transfer
7. Define an adaption register for the lower half of the fluidized bed.
This register is used to patch the initial volume fraction of solids in the next step.
Adapt −→Region...
21-26
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
(a) Under Input Coordinates, specify the Xmaximum value as 0.3 and the Ymaxi-
mum value as 0.5.
(b) Click Mark.
(c) Click Manage...
The Manage Adaption Registers panel will open.
(d) Under Registers, in the Manage Adaption Registers panel, select hexahedron-r0,
and click Display.
8. Patch the initial volume fraction of solids in the lower half of the fluidized bed.
Solve −→ Initialize −→Patch...
(a) In the Phase drop-down list, select solids.
(b) In the Variable drop-down list, select Volume Fraction.
(c) In the Value field, enter 0.598.
(d) In the Registers to Patch list, select hexahedron-r0.
(e) Click Patch.
c Fluent Inc. January 11, 2005 21-27
Using the Eulerian Granular Multiphase Model with Heat Transfer
At this point, it is good practice to display contours of the variable you just patched,
to ensure that the desired field was obtained.
5.98e-01
5.68e-01
5.38e-01
5.08e-01
4.78e-01
4.49e-01
4.19e-01
3.89e-01
3.59e-01
3.29e-01
2.99e-01
2.69e-01
2.39e-01
2.09e-01
1.79e-01
1.50e-01
1.20e-01
8.97e-02
5.98e-02
2.99e-02
0.00e+00
21-28
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
11. Set a time step size of 0.00025 s and run the calculation for 7000 time steps.
Solve −→Iterate...
The plot of the value of the mixture-averaged heat transfer coefficient in the cell
next to the heated wall versus time is in excellent agreement with results published
for the same case [1].
3000.0000
2500.0000
2000.0000
Vertex 1500.0000
Average
ave_htc
1000.0000
500.0000
0.0000
0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 1.2000 1.4000 1.6000 1.8000
Flow Time
Figure 21.5: Plot of Mixture-Averaged Heat Transfer Coefficient in the Cell Next to the
Heated Wall Versus Time
12. Save the case and data files (fluid-bed.cas and fluid-bed.dat).
File −→ Write −→Case & Data...
Extra: If you decide to read in the case file that is provided for this tutorial on the
documentation CD, you will need to compile the UDF associated with this tutorial
in your working directory. This is necessary because FLUENT will expect to find
the correct UDF libraries in your working directory when reading the case file.
c Fluent Inc. January 11, 2005 21-29
Using the Eulerian Granular Multiphase Model with Heat Transfer
Step 7: Postprocessing
1. Display the pressure field in the fluidized bed (Figure 21.6).
Display −→Contours...
(a) Select Pressure... and Static Pressure in the Contours of drop-down lists.
(b) Select Filled under Options.
(c) Click Display.
Note the build-up of static pressure in the granular phase.
2. Display the volume fraction of solids (Figure 21.7).
Display −→Contours...
(a) Select Phases... and Volume fraction in the Contours of drop-down lists.
(b) Select solids in the Phase drop-down list.
(c) Click Display.
(d) Zoom in to show the contours close to the region where the change in volume
fraction is the greatest.
Note that the region occupied by the granular phase has expanded slightly, as a result
of fluidization.
21-30
c Fluent Inc. January 11, 2005
Using the Eulerian Granular Multiphase Model with Heat Transfer
7.79e+03
7.40e+03
7.01e+03
6.62e+03
6.23e+03
5.84e+03
5.45e+03
5.06e+03
4.67e+03
4.29e+03
3.90e+03
3.51e+03
3.12e+03
2.73e+03
2.34e+03
1.95e+03
1.56e+03
1.17e+03
7.79e+02
3.89e+02
-8.77e-02
6.29e-01
5.98e-01
5.67e-01
5.35e-01
5.04e-01
4.72e-01
4.41e-01
4.09e-01
3.78e-01
3.46e-01
3.15e-01
2.83e-01
2.52e-01
2.20e-01
1.89e-01
1.57e-01
1.26e-01
9.44e-02
6.29e-02
3.15e-02
0.00e+00
c Fluent Inc. January 11, 2005 21-31
Using the Eulerian Granular Multiphase Model with Heat Transfer
Summary
This tutorial demonstrated how to set up and solve a granular multiphase problem with
heat transfer, using the Eulerian model. You learned how to set boundary conditions
for the mixture and both phases. The solution obtained is in excellent agreement with
analytical results from Kuipers et al. [1].
References
1. J. A. M. Kuipers, W. Prins, and W. P. M. Van Swaaij “Numerical Calculation
of Wall-to-Bed Heat Transfer Coefficients in Gas-Fluidized Beds”, Department of
Chemical Engineering, Twente University of Technology, in AIChE Journal, July
1992, Vol. 38, No. 7.
21-32
c Fluent Inc. January 11, 2005
Tutorial 22. Postprocessing
Introduction
This tutorial demonstrates the postprocessing capabilities of FLUENT using a 3D model
of a flat circuit board with a heat generating electronic chip mounted on it. The flow
over the chip is laminar and involves conjugate heat transfer.
The heat transfer involves conduction in the chip and conduction and convection in the
surrounding fluid. The physics of conjugate heat transfer such as this, is common in many
engineering applications, including the design and cooling of electronic components.
In this tutorial, you will read the case and data files (without doing the calculation) and
perform a number of postprocessing exercises. In the process, you will learn how to:
• Create animations
• Display pathlines
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
c Fluent Inc. January 11, 2005 22-1
Postprocessing
Problem Description
The problem considered is shown schematically in Figure 22.1. The configuration consists
of a series of side-by-side electronic chips, or modules, mounted on a circuit board. Air
flow, confined between the circuit board and an upper wall, cools the modules. To take
advantage of the symmetry present in the problem, the model will extend from the middle
of one module to the plane of symmetry between it and the next module.
As shown in the figure, each half-module is assumed to generate 2.0 Watts and to have a
bulk conductivity of 1.0 W/m2 -K. The circuit board conductivity is assumed to be one
order of magnitude lower: 0.1 W/m2 -K. The air flow enters the system at 298 K with
a velocity of 1 m/s. The Reynolds number of the flow, based on the module height, is
about 600. The flow is therefore treated as laminar.
Symmetry
Planes
Top Wall
Externally Cooled
Bottom Wall
Externally Cooled
Electronic Module (one half)
Air Flow k = 1.0 W/m2-K
1.0 m/s Q = 2.0 Watts
298 K
Circuit Board
k = 0.1 W/m2-K
22-2
c Fluent Inc. January 11, 2005
Postprocessing
c Fluent Inc. January 11, 2005 22-3
Postprocessing
Y X
Z
Grid
FLUENT 6.2 (3d, segregated, lam)
Extra: You can click the right mouse button on one of the boundaries displayed
in the graphics window and its zone number, name, and type will be printed
in the console window. This feature is especially useful when you have several
zones of the same type and you want to distinguish between them.
22-4
c Fluent Inc. January 11, 2005
Postprocessing
c Fluent Inc. January 11, 2005 22-5
Postprocessing
Y
X
Z
Grid
FLUENT 6.2 (3d, segregated, lam)
22-6
c Fluent Inc. January 11, 2005
Postprocessing
Extra: You can use the left mouse button to rotate the ball in the Active Lights window
to gain a perspective view on the relative locations of the lights that are currently
active, and see the shading effect on the ball at the center.
You can also change the color of one or more of the lights by typing the name of a
color in the Color field or moving the Red, Green, and Blue sliders.
c Fluent Inc. January 11, 2005 22-7
Postprocessing
Y
X
Z
Grid
FLUENT 6.2 (3d, segregated, lam)
Y
X
Z
Grid
FLUENT 6.2 (3d, segregated, lam)
22-8
c Fluent Inc. January 11, 2005
Postprocessing
(a) In the Surface of Constant drop-down lists, select Grid... and Y-Coordinate.
(b) Click Compute.
The Min and Max fields will display the y extents of the domain.
(c) Enter 0.25 under Iso-Values.
(d) Enter y=0.25in under New Surface Name.
(e) Click Create, and Close the panel.
c Fluent Inc. January 11, 2005 22-9
Postprocessing
Step 4: Contours
1. Display filled contours of temperature on the symmetry plane (Figure 22.6).
Display −→Contours...
Hint: If the display disappears from the screen at any time, or if you are having
difficulty manipulating it with the mouse, you can open the Views panel from
the Display pull-down menu and use the Default button to reset the view. Al-
ternatively, you can revert to a previous graphics display using the keyboard
shortcut <Ctrl>-L.
The peak temperatures in the chip appear where the heat is generated, along with
the higher temperatures in the wake where the flow is recirculating.
22-10
c Fluent Inc. January 11, 2005
Postprocessing
4.09e+02
4.03e+02
3.98e+02
3.92e+02
3.87e+02
3.81e+02
3.76e+02
3.70e+02
3.64e+02
3.59e+02
3.53e+02
3.48e+02
3.42e+02
3.37e+02
3.31e+02
3.26e+02
3.20e+02
3.15e+02
3.09e+02 Y
3.04e+02 X
Z
2.98e+02
c Fluent Inc. January 11, 2005 22-11
Postprocessing
In the contour display (Figure 22.7), the high temperatures in the wake of
the module are clearly visible. You can also display other quantities using the
Contours panel, such as velocity magnitude or pressure.
22-12
c Fluent Inc. January 11, 2005
Postprocessing
1. Display velocity vectors on the symmetry plane through the module centerline
(Figure 22.8).
Display −→Vectors...
c Fluent Inc. January 11, 2005 22-13
Postprocessing
1.41e+00
1.34e+00
1.27e+00
1.20e+00
1.13e+00
1.06e+00
9.89e-01
9.20e-01
8.50e-01
7.81e-01
7.11e-01
6.41e-01
5.72e-01
5.02e-01
4.33e-01
3.63e-01
2.94e-01
2.24e-01
1.54e-01 Y
8.49e-02
1.53e-02 Z X
Note: The vectors in Figure 22.8 are shown without arrowheads. You can modify
the arrow style in the Vectors panel by selecting a different option from the
Style drop-down list.
Extra: If you want to decrease the number of vectors displayed, you can increase
the Skip factor to a non-zero value.
2. Plot velocity vectors in the horizontal plane intersecting the module (Figure 22.9).
After plotting the vectors, you will enhance the view by mirroring the display about
the module centerline and displaying the module surfaces.
Display −→Vectors...
(a) Deselect all surfaces by clicking the unshaded icon to the right of Surfaces.
(b) In the Surfaces list, select y=0.25in.
(c) Set the Scale to 3.8.
(d) Under Options, select Draw Grid to display the module surfaces.
This will open the Grid Display panel.
22-14
c Fluent Inc. January 11, 2005
Postprocessing
c Fluent Inc. January 11, 2005 22-15
Postprocessing
1.41e+00
1.34e+00
1.27e+00
1.20e+00
1.13e+00
1.06e+00
9.89e-01
9.20e-01
8.50e-01
7.81e-01
7.11e-01
6.41e-01
5.72e-01
5.02e-01
4.33e-01
3.63e-01
2.94e-01
2.24e-01
1.54e-01 Y
8.49e-02 X
1.53e-02
Z
3. Mirror the view about the chip symmetry plane (Figure 22.10).
Display −→Views...
22-16
c Fluent Inc. January 11, 2005
Postprocessing
1.41e+00
1.34e+00
1.27e+00
1.20e+00
1.13e+00
1.06e+00
9.89e-01
9.20e-01
8.50e-01
7.81e-01
7.11e-01
6.41e-01
5.72e-01
5.02e-01
4.33e-01
3.63e-01
2.94e-01
2.24e-01
1.54e-01 Y
8.49e-02 X
1.53e-02 Z
c Fluent Inc. January 11, 2005 22-17
Postprocessing
Step 6: Animation
Using FLUENT, you can animate the solution and also a scene. For information on
animating the solution, see Tutorial 11, Steps 9 and 10. In this tutorial you will animate
a scene between two static views of the graphics display.
Display the surface temperature distribution on the module and the circuit board by select-
ing the corresponding boundaries. Create the key frames and view the transition between
the key frames, dynamically, using the animation feature.
1. Display filled contours of surface temperature on the board-top and chip surfaces.
(Figure 22.11).
Display −→Contours...
22-18
c Fluent Inc. January 11, 2005
Postprocessing
4.09e+02
4.03e+02
3.98e+02
3.92e+02
3.87e+02
3.81e+02
3.76e+02
3.70e+02
3.64e+02
3.59e+02
3.53e+02
3.48e+02
3.42e+02
3.37e+02
3.31e+02
3.26e+02
3.20e+02
3.15e+02
3.09e+02 Y
3.04e+02 X
2.98e+02 Z
Figure 22.11: Filled Temperature Contours on the Chip and Board Top Surfaces
Figure 22.11 shows the high temperatures on the downstream portions of the module
and the relatively localized heating of the circuit board around the module.
You will use the current display (Figure 22.11) as the starting view for the anima-
tion (Frame = 1).
c Fluent Inc. January 11, 2005 22-19
Postprocessing
The zoomed view will be the tenth keyframe of the animation, with intermediate
displays (2 through 9) to be filled in during the animation.
(e) Rotate the view and un-zoom the display so that the downstream side of the
module is in the foreground, as shown in Figure 22.12.
(f) Change the Frame number to 20.
(g) Click Add.
This will store the new display as Key-20.
(a) Click on the “play” arrow button ( ) (second from the right in the row of
playback buttons) in the Playback section of the Animate panel.
While effective animation is best conducted on “high-end” graphics workstations,
you can view scene animations on any workstation. If the graphics display speed is
slow, the animation playback will take some time and will appear choppy, with the
redrawing very obvious. On fast graphics workstations, the animation will appear
smooth and continuous and will provide an excellent visualization of the display
from a variety of spatial orientations. On many machines, you can improve the
22-20
c Fluent Inc. January 11, 2005
Postprocessing
4.09e+02
4.03e+02
3.98e+02
3.92e+02
3.87e+02
3.81e+02
3.76e+02
3.70e+02
3.64e+02
3.59e+02
3.53e+02
3.48e+02
3.42e+02
3.37e+02
3.31e+02
3.26e+02
3.20e+02
3.15e+02
3.09e+02 Y
3.04e+02
X
2.98e+02Z
Figure 22.12: Filled Temperature Contours on the Chip and Board Top Surfaces
smoothness of the animation by turning on the Double Buffering option in the Display
Options panel.
Note: You can also make use of FLUENT’s animation tools for transient cases as demon-
strated in Tutorial 4.
Extra: You can change the Playback mode if you want to “auto repeat” or “auto reverse”
the animation. When you are in either of these Playback modes, you can click on
the “stop” button (square) to stop the continuous animation.
c Fluent Inc. January 11, 2005 22-21
Postprocessing
Step 7: Pathlines
Pathlines are the lines traveled by neutrally buoyant particles in equilibrium with the fluid
motion. Pathlines are an excellent tool for visualization of complex three-dimensional
flows. In this example, you will use pathlines to examine the flow around and in the wake
of the module.
22-22
c Fluent Inc. January 11, 2005
Postprocessing
c Fluent Inc. January 11, 2005 22-23
Postprocessing
9.00e+00
8.55e+00
8.10e+00
7.65e+00
7.20e+00
6.75e+00
6.30e+00
5.85e+00
5.40e+00
4.95e+00
4.50e+00
4.05e+00
3.60e+00
3.15e+00
2.70e+00
2.25e+00
1.80e+00
1.35e+00
9.00e-01 Y
4.50e-01 X
0.00e+00 Z
22-24
c Fluent Inc. January 11, 2005
Postprocessing
(a) In the Release From Surfaces list, keep the selection of pathline-rake.
(b) Select sphere from the Style drop-down list.
(c) Click the Style Attributes button.
This displays the Path Style Attributes panel.
c Fluent Inc. January 11, 2005 22-25
Postprocessing
9.00e+00
8.55e+00
8.10e+00
7.65e+00
7.20e+00
6.75e+00
6.30e+00
5.85e+00
5.40e+00
4.95e+00
4.50e+00
4.05e+00
3.60e+00
3.15e+00
2.70e+00
2.25e+00
1.80e+00
1.35e+00
9.00e-01 Y
4.50e-01 X
0.00e+00 Z
5. In the Path Lines panel, select Surface ID in the lower Color By drop-down list and
click Display (Figure 22.15).
This will color the path lines by the surface they are released from.
Note: You can also create solution animations for path lines using the Solve/Animate/
Define menu and the Animation Sequence panel.
22-26
c Fluent Inc. January 11, 2005
Postprocessing
5.00e+00
4.80e+00
4.60e+00
4.40e+00
4.20e+00
4.00e+00
3.80e+00
3.60e+00
3.40e+00
3.20e+00
3.00e+00
2.80e+00
2.60e+00
2.40e+00
2.20e+00
2.00e+00
1.80e+00
1.60e+00
1.40e+00 Y
1.20e+00 X
Z
1.00e+00
c Fluent Inc. January 11, 2005 22-27
Postprocessing
22-28
c Fluent Inc. January 11, 2005
Postprocessing
c Fluent Inc. January 11, 2005 22-29
Postprocessing
9.00e+00
8.55e+00
8.10e+00
7.65e+00
7.20e+00
6.75e+00
6.30e+00
5.85e+00
5.40e+00
4.95e+00
4.50e+00
4.05e+00
3.60e+00
3.15e+00
2.70e+00
2.25e+00
1.80e+00
1.35e+00
9.00e-01 Y
4.50e-01 X
Z
0.00e+00
Note: The final display (Figure 22.16) does not require mirroring about the symmetry
plane because the vectors obscure the mirrored image. You may turn off the mir-
roring option in the Views panel at any stage during this exercise.
22-30
c Fluent Inc. January 11, 2005
Postprocessing
1. Delete the velocity vectors and pathlines from the current display.
Display −→Scene...
(a) In the Names list, select the velocity vectors and pathlines.
(b) Click Delete Geometry.
(c) Click Apply.
The Scene Description panel should then contain only the two grid surfaces
(board-top and chip).
2. Create a plotting surface at x=3 inches (named x=3.0in), just downstream of the
trailing edge of the module.
Surface −→Iso-Surface...
c Fluent Inc. January 11, 2005 22-31
Postprocessing
Hint: If you forget how to create an isosurface, see Step 3: Isosurface Creation.
22-32
c Fluent Inc. January 11, 2005
Postprocessing
(b) Deselect all surfaces by clicking on the unshaded icon to the right of Surfaces.
(c) In the Surfaces list, select x=3.0in.
(d) Click Display, and Close the panel.
The filled temperature contours will be displayed on the x=3.0 in. surface.
(a) Under Options, select Draw Grid and close the panel that opens.
(b) Deselect all surfaces by clicking on the unshaded icon to the right of Surfaces.
(c) In the Surfaces list, select x=3.0in.
(d) Increase the Skip to 2.
(e) Change the Scale to 1.9.
(f) Click Display.
The display will show the vectors superimposed on the contours of temperature
at x=3.0 in.
c Fluent Inc. January 11, 2005 22-33
Postprocessing
5. Create the exploded view (Figure 22.17) by translating the contour display, placing
it above the vectors.
Display −→Scene...
(a) In the Scene Description panel, select contour-6-temperature in the Names list.
(b) Click Transform....
This will open the Transformations panel.
22-34
c Fluent Inc. January 11, 2005
Postprocessing
1.41e+00
1.34e+00
1.27e+00
1.20e+00
1.13e+00
1.06e+00
9.89e-01
9.20e-01
8.50e-01
7.81e-01
7.11e-01
6.41e-01
5.72e-01
5.02e-01
4.33e-01
3.63e-01
2.94e-01
2.24e-01
1.54e-01 Y
8.49e-02 X
1.53e-02 Z
c Fluent Inc. January 11, 2005 22-35
Postprocessing
2. Use the mouse to un-zoom the view in the graphics window so that the entire board
surface is visible.
3. Generate contours of velocity magnitude and sweep them through the domain along
the x axis.
Display −→Sweep Surface...
22-36
c Fluent Inc. January 11, 2005
Postprocessing
(b) Under Animation, set the Initial Value to 0 m and the Final Value to 0.1651 m.
The units for the initial and final values are in meters, regardless of the
! length units being used in the model. Here, the initial and final values are
set to the Min Value and Max Value, to generate an animation through the
entire domain.
(c) Set the number of Frames to 20.
(d) Select Contours under Display Type.
This will open the Contours panel.
Note: You can also make use of FLUENT’s animation tools for transient cases as demon-
strated in Tutorial 4.
c Fluent Inc. January 11, 2005 22-37
Postprocessing
22-38
c Fluent Inc. January 11, 2005
Postprocessing
2. Plot the temperature distribution along the top centerline of the module (Fig-
ure 22.18).
Plot −→XY Plot...
(a) Select Temperature... and Static Temperature in the Y Axis Function drop-down
lists.
(b) In the Surfaces list, select top-center-line.
(c) Keep the default Plot Direction of X.
This will plot temperature vs. the x coordinate along the selected line (top-
center-line).
(d) Click Axes... to modify the axis range.
This will open the Axes - Solution XY Plot panel .
c Fluent Inc. January 11, 2005 22-39
Postprocessing
22-40
c Fluent Inc. January 11, 2005
Postprocessing
top-center-line
4.02e+02
4.00e+02
3.98e+02
3.96e+02
Static
Temperature 3.94e+02
(k)
3.92e+02
3.90e+02
3.88e+02
1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8
Y
X Position (in)
Z
Static Temperature
FLUENT 6.2 (3d, segregated, lam)
c Fluent Inc. January 11, 2005 22-41
Postprocessing
1. In the Annotation Text field, enter the text describing the plot (e.g., Temperature
Along the Top Centerline).
2. Click Add.
A Working dialog box will appear telling you to select the desired location of the text
using the mouse-probe button, which is, by default, the right button.
3. Click the right mouse button in the graphics display window where you want the
text to appear, and you will see the text displayed at the desired location (Fig-
ure 22.19).
Extra: If you want to move the text to a new location on the screen, click Delete
Text in the Annotate panel, and click Add once again, defining a new position
with the mouse.
Note: Depending on the size of the graphics window and the hardcopy file format
you choose, the font size of the annotation text you see on the screen may be
different from the font size in a hardcopy file of that graphics window. The
annotation text font size is absolute, while the rest of the items in the graphics
window are scaled to the proportions of the hardcopy.
22-42
c Fluent Inc. January 11, 2005
Postprocessing
top-center-line
Temperature Along the Top Centerline
4.02e+02
4.00e+02
3.98e+02
3.96e+02
Static
Temperature 3.94e+02
(k)
3.92e+02
3.90e+02
3.88e+02
1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8
Y
X Position (in)
Z
Static Temperature
FLUENT 6.2 (3d, segregated, lam)
c Fluent Inc. January 11, 2005 22-43
Postprocessing
3. Click Save....
This will open the Select File dialog box.
4. In the Select File dialog box, enter a name for the hardcopy file.
Summary
This tutorial has demonstrated the use of many of the extensive postprocessing features
available in FLUENT.
See Chapter 29 and Chapter 30 of the User’s Guide for more information on these and
related features.
22-44
c Fluent Inc. January 11, 2005
Tutorial 23. Turbo Postprocessing
Introduction
This tutorial demonstrates the turbomachinery postprocessing capabilities of FLUENT.
In this example, you will read the case and data files (without doing the calculation) and
perform a number of turbomachinery-specific postprocessing exercises. In the process,
you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The problem considered in this tutorial is a centrifugal compressor shown schematically
in Figure 23.1. The model consists of a single 3D sector of the compressor, to take
advantage of the circumferential periodicity in the problem. The flow of air through
the compressor is simulated and the postprocessing capabilities of FLUENT are used to
display realistic full 360-degree images of the solution obtained.
c Fluent Inc. January 11, 2005 23-1
Turbo Postprocessing
inlet
shroud side
hub side
outlet
23-2
c Fluent Inc. January 11, 2005
Turbo Postprocessing
3. Deselect all surfaces, and click on the Outline button under the Surface Types list.
4. Click Display.
5. Use the left mouse button to rotate the view, and the middle mouse button to
zoom the view to obtain an isometric display of the compressor duct, as shown in
Figure 23.2.
Extra: You can use the right mouse button to check which zone number corresponds
to each boundary. If you click the right mouse button on one of the boundaries
displayed in the graphics window, its zone number, name, and type will be printed
in the console window. This feature is especially useful when you have several zones
of the same type and you want to distinguish between them quickly.
c Fluent Inc. January 11, 2005 23-3
Turbo Postprocessing
Y
X
Z
Grid
FLUENT 6.2 (3d, coupled imp, rke)
23-4
c Fluent Inc. January 11, 2005
Turbo Postprocessing
c Fluent Inc. January 11, 2005 23-5
Turbo Postprocessing
FLUENT will inform you that the turbomachinery postprocessing functions have been ac-
tivated, and the Turbo menu will appear in FLUENT menu bar at the top of the console
window.
In the Turbo Topology panel, you can define any number of turbo topologies. This is
especially useful when you have a model consisting of multiple blade rows and you need to
define more than one blade row simultaneously. Each topology can be assigned a specific
name and accessed using the drop-down list in the Turbo Topology panel.
See Section 29.9.1 of the User’s Guide for more information on defining turbo topologies.
Note: You can display the selected surfaces by clicking on Display at the bottom of the
panel. This is useful as a graphical check to ensure that all relevant surfaces have
been selected.
23-6
c Fluent Inc. January 11, 2005
Turbo Postprocessing
(a) In the Surface of Constant drop-down lists, select Grid... and Meridional Coor-
dinate.
(b) Enter 0.2 under Iso-Values.
(c) Enter meridional-0.2 under New Surface Name.
(d) Click Create.
Note: The iso-values you enter for these turbo-specific surfaces are expressed
as a percentage of the entire domain (i.e., you just defined a surface of
meridional coordinate equal to 20% of the path along the duct).
(e) Repeat the steps above to define surfaces of meridional coordinates equal to
0.4, 0.6, and 0.8.
c Fluent Inc. January 11, 2005 23-7
Turbo Postprocessing
(a) In the Surface of Constant drop-down lists, select Grid... and Spanwise Coordi-
nate
(b) Enter 0.25 under Iso-Values.
(c) Enter spanwise-0.25 under New Surface Name.
(d) Click Create.
(e) Repeat the steps above to define surfaces of spanwise coordinates equal to 0.5
and 0.75.
23-8
c Fluent Inc. January 11, 2005
Turbo Postprocessing
Step 5: Contours
1. Plot filled contours of pressure on the meridional isosurfaces (Figure 23.3).
Display −→Contours...
c Fluent Inc. January 11, 2005 23-9
Turbo Postprocessing
1.84e+00
1.78e+00
1.73e+00
1.67e+00
1.62e+00
1.56e+00
1.50e+00
1.45e+00
1.39e+00
1.34e+00
1.28e+00
1.22e+00
1.17e+00
1.11e+00
1.06e+00
1.00e+00
9.44e-01
8.88e-01
8.32e-01
Y
7.76e-01
X
7.20e-01 Z
1.04e+00
9.85e-01
9.35e-01
8.85e-01
8.35e-01
7.84e-01
7.34e-01
6.84e-01
6.34e-01
5.83e-01
5.33e-01
4.83e-01
4.33e-01
3.82e-01
3.32e-01
2.82e-01
2.32e-01
1.81e-01
1.31e-01
Y
8.07e-02
X
3.05e-02 Z
23-10
c Fluent Inc. January 11, 2005
Turbo Postprocessing
In Figure 23.4, you can observe locations at which the flow becomes slightly super-
sonic, about halfway through the duct.
3. Plot filled contours of Mach number on the spanwise isosurfaces (Figure 23.5).
(a) In the Surfaces list, deselect all surfaces, and then select spanwise-0.25, spanwise-
0.5, and spanwise-0.75.
(b) Click Display.
1.04e+00
9.85e-01
9.35e-01
8.85e-01
8.35e-01
7.84e-01
7.34e-01
6.84e-01
6.34e-01
5.83e-01
5.33e-01
4.83e-01
4.33e-01
3.82e-01
3.32e-01
2.82e-01
2.32e-01
1.81e-01
1.31e-01
Y
8.07e-02
X
3.05e-02 Z
The display in Figure 23.5 allows you to further study the variation of the Mach
number inside the duct. You may want to explore using different combinations of
surfaces to display the same or additional variables.
4. Display a 360-degree image of the Mach number contours on the 0.5 spanwise
isosurface (Figure 23.6).
(a) Redisplay the contours, just on the 0.5 spanwise isosurface.
i. In the Surfaces list, deselect spanwise-0.25 and spanwise-0.75.
ii. Click Display.
c Fluent Inc. January 11, 2005 23-11
Turbo Postprocessing
23-12
c Fluent Inc. January 11, 2005
Turbo Postprocessing
1.04e+00
9.85e-01
9.35e-01
8.85e-01
8.35e-01
7.84e-01
7.34e-01
6.84e-01
6.34e-01
5.83e-01
5.33e-01
4.83e-01
4.33e-01
3.82e-01
3.32e-01
2.82e-01
2.32e-01
1.81e-01
1.31e-01
Y
8.07e-02 X
3.05e-02 Z
Figure 23.6: Filled Contours of Mach Number on the 0.5 Spanwise Iso Surface
Note: This step demonstrates a typical view-manipulation task. See Tutorial 22 for
further examples of postprocessing features.
c Fluent Inc. January 11, 2005 23-13
Turbo Postprocessing
23-14
c Fluent Inc. January 11, 2005
Turbo Postprocessing
2. Click Compute.
(a) In the Contours Of drop-down lists, select Pressure... and Static Pressure.
(b) Click Display.
c Fluent Inc. January 11, 2005 23-15
Turbo Postprocessing
1.80e+00
1.76e+00
1.72e+00
1.67e+00
1.63e+00
1.58e+00
1.54e+00
1.50e+00
1.45e+00
1.41e+00
1.36e+00
1.32e+00
1.28e+00
1.23e+00
1.19e+00
1.14e+00
1.10e+00
1.06e+00
1.01e+00 Y
9.68e-01
Z X
9.24e-01
23-16
c Fluent Inc. January 11, 2005
Turbo Postprocessing
Step 8: 2D Contours
In postprocessing a turbomachinery solution, it is often desirable to display contours on
constant spanwise coordinates, and then project these contours onto a plane. This permits
easier evaluation of the contours, especially for surfaces that are highly three-dimensional.
FLUENT allows you to display contours in this manner using the Turbo 2D Contours panel.
(a) In the Contours Of drop-down lists, select Velocity... and Mach Number.
(b) Under Normalised Spanwise Coordinates, enter 0.5.
Note: For highly curved edges, if a surface is created very close to the curved
edge the resulting surface may have some void spaces in it.
(c) Click Display.
(d) Use the mouse to obtain the view shown in Figure 23.8.
c Fluent Inc. January 11, 2005 23-17
Turbo Postprocessing
9.12e-01
8.69e-01
8.26e-01
7.83e-01
7.40e-01
6.96e-01
6.53e-01
6.10e-01
5.67e-01
5.24e-01
4.81e-01
4.37e-01
3.94e-01
3.51e-01
3.08e-01
2.65e-01
2.22e-01
1.79e-01
1.35e-01 X
9.22e-02 Z
Y
4.91e-02
23-18
c Fluent Inc. January 11, 2005
Turbo Postprocessing
(a) In the Y Axis Function drop-down lists, select Temperature... and Static Tem-
perature.
(b) In the X Axis Function drop-down list, select Meridional Distance.
(c) Under Fractional Distance, enter 0.9.
(d) Click Plot.
c Fluent Inc. January 11, 2005 23-19
Turbo Postprocessing
3.50e+02
3.40e+02
3.30e+02
3.20e+02
Static
Temperature 3.10e+02
(k)
3.00e+02
2.90e+02
2.80e+02
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Y
Z X
Meridional Distance
Averaged XY - temperature
FLUENT 6.2 (3d, coupled imp, rke)
Figure 23.9: Averaged XY Plot of Static Temperature on Spanwise Surface of 0.9 Isovalue
23-20
c Fluent Inc. January 11, 2005
Turbo Postprocessing
Summary
This tutorial has demonstrated the use of some of the turbomachinery-specific postpro-
cessing features of FLUENT. These features can be accessed once you have defined the
topology of the problem. More extensive general-purpose postprocessing features are
demonstrated in Tutorial 22.Also see Chapter 29 and Chapter 30 of the User’s Guide for
additional information.
c Fluent Inc. January 11, 2005 23-21
Turbo Postprocessing
23-22
c Fluent Inc. January 11, 2005
Tutorial 24. Parallel Processing
Introduction
This tutorial illustrates the setup and solution of a simple 2D problem using FLUENT’s
parallel processing capabilities. In order to be run in parallel, the mesh must be divided
into smaller, evenly sized partitions. Each FLUENT process, called a compute node,
will solve on a single partition, and information will be passed back and forth across all
partition interfaces. FLUENT’s solver allows parallel processing on a dedicated parallel
machine, or a network of heterogeneous workstations running UNIX, or a network of
workstations running Windows.
The tutorial assumes that both FLUENT and network communication software have been
correctly installed (see the separate installation instructions and related information for
details). The case chosen is the mixing elbow problem you solved in Tutorial 1.
In this tutorial you will learn how to:
Prerequisites
This tutorial assumes that you are familiar with the menu structure in FLUENT and that
you have completed Tutorial 1 . Some steps in the setup and solution procedure will not
be shown explicitly.
Problem Description
The problem to be considered is shown schematically in Figure 24.1. A cold fluid at 26◦ C
enters through the large pipe and mixes with a warmer fluid at 40◦ C in the elbow. The
pipe dimensions are in inches, and the fluid properties and boundary conditions are given
in SI units. The Reynolds number at the main inlet is 2.03 × 105 , therefore, a turbulence
model will be necessary.
c Fluent Inc. January 11, 2005 24-1
Parallel Processing
3
Density: ρ= 1000 kg/m
-4
Viscosity: µ = 8 x 10 Pa-s
Conductivity: k = 0.677 W/m-K 32 ″
Specific Heat: C p = 4216 J/kg-K
39.9
3°
39.9
16 ″
3°
24-2
c Fluent Inc. January 11, 2005
Parallel Processing
2. Unzip parallel_process.zip.
elbow3.cas can be found in the /parallel process folder created after unzipping
the file.
You can partition the grid before or after you set up the problem (define models,
boundary conditions, etc.). It is best to partition after the problem is set up, since
partitioning has some model dependencies (e.g., sliding-mesh and shell-conduction
encapsulation). Since you have already followed the procedure for setting up the
mixing elbow in Tutorial 1, elbow3.cas is provided to save you the effort of re-
defining the models and boundary conditions.
c Fluent Inc. January 11, 2005 24-3
Parallel Processing
fluent 2d -t2
24-4
c Fluent Inc. January 11, 2005
Parallel Processing
c Fluent Inc. January 11, 2005 24-5
Parallel Processing
fluent 2d -t2
24-6
c Fluent Inc. January 11, 2005
Parallel Processing
c Fluent Inc. January 11, 2005 24-7
Parallel Processing
(a) Specify the machine on which you want to spawn the process.
i. Under Host Entry, specify the machine name in the Hostname field.
ii. Enter your user ID in the Username field.
iii. Click Add.
The machine will be added to the Available Hosts list.
Note: It is possible to create a list of available machines and add them to the
hosts database, rather than adding machines manually. See Chapter 30 of
the User’s Guide for details.
(b) Select the newly added host in the Available Hosts list.
Note: If you do not have access to another machine to use for this tutorial,
you can spawn the second node on your own machine by selecting it from
the Available Hosts list, although you will incur a performance penalty on
a single processor machine.
(c) Under Spawn Count, keep the default value of 1.
This will give you the desired total number of 2 computational nodes.
24-8
c Fluent Inc. January 11, 2005
Parallel Processing
c Fluent Inc. January 11, 2005 24-9
Parallel Processing
For information about all defined compute nodes, you will select node 0, since
this is the node from which all other nodes are spawned.
(b) Click Print.
--------------------------------------------------------------------
ID Comm. Hostname O.S. PID Mach ID HW ID Name
--------------------------------------------------------------------
n1 net dori hpux 11681 1 7 Fluent Node
host net bilbo hpux 12697 0 3 Fluent Host
n0* net bilbo hpux 12698 0 -1 Fluent Node
24-10
c Fluent Inc. January 11, 2005
Parallel Processing
1. At the DOS command prompt, enter the following to start the 2D network parallel
version with one process:
2. Spawn an additional compute node, following the procedure described in Step 1C,
substep 3, for a network of UNIX machines.
c Fluent Inc. January 11, 2005 24-11
Parallel Processing
If the Case File option is turned on (the default setting), and there exists a valid
partition section in the case file (i.e., one where the number of partitions in the case
file divides evenly into the number of compute nodes), then that partition informa-
tion will be used rather than repartitioning the mesh. You need to turn off the Case
File option only if you want to change other parameters in the Auto Partition Grid
panel.
(a) Keep all defaults in the Auto Partition Grid panel and click OK.
When you keep the Case File option turned on, FLUENT will automatically
select a partitioning method for you. This is the preferred initial approach for
most problems. In the next step you will inspect the partitions created and be
able to change them, if you so choose.
24-12
c Fluent Inc. January 11, 2005
Parallel Processing
Grid
FLUENT 6.2 (2d, segregated, ske)
c Fluent Inc. January 11, 2005 24-13
Parallel Processing
24-14
c Fluent Inc. January 11, 2005
Parallel Processing
----------------------------------------------------------------------
Collective Partition Statistics: Minimum Maximum Total
----------------------------------------------------------------------
Cell count 619 620 1239
Mean cell count deviation -0.1% 0.1%
Partition boundary cell count 10 14 24
Partition boundary cell count ratio 1.6% 2.3% 1.9%
c Fluent Inc. January 11, 2005 24-15
Parallel Processing
i. In the Contours Of drop-down lists, select Cell Info... and Active Cell
Partition.
ii. Under Options, select Filled.
iii. Set the number of Levels to 2, the number of compute nodes.
iv. Click Display.
As shown in Figure 24.3, the cell partitions are acceptable for this problem.
The position of the interface reveals that the criteria mentioned above will be
matched. If you were unsatisfied with the partitions, you could use the Partition
Grid panel to repartition the grid. Recall that, if you wish to use the modified
partitions for a calculation, you will need to make the Stored Cell Partition the
Active Cell Partition by either clicking on the Use Stored Partitions button in the
Partition Grid panel or saving the case file and reading it back into FLUENT.
24-16
c Fluent Inc. January 11, 2005
Parallel Processing
1.00e+00
5.00e-01
0.00e+00
See Section 32.5.3 of the User’s Guide for details about the procedure and
options for manually partitioning a grid.
c Fluent Inc. January 11, 2005 24-17
Parallel Processing
Step 3: Solution
1. Initialize the flow field using the boundary conditions set at velocity-inlet-5.
Solve −→ Initialize −→Initialize...
24-18
c Fluent Inc. January 11, 2005
Parallel Processing
The most accurate way to evaluate parallel performance is by running the same paral-
lel problem on 1 CPU and on n CPUs, and comparing the Total wall-clock time
(elapsed time for the iterations) in both cases. Ideally you would want to have the Total
wall-clock time with n CPUs be 1/n times the Total wall-clock time with 1 CPU.
In practice, this improvement will be reduced by the performance of the communication
subsystem of your hardware, and the overhead of the parallel process itself. As a rough
estimate of parallel performance, you can compare the Total wall-clock time with
the CPU time. In this case, the CPU time was approximately 1.98 times the Total
wall-clock time. For a parallel process run on two compute nodes, this reveals very
good parallel performance, even though the advantage over a serial calculation is small,
as expected for this simple 2D problem.
c Fluent Inc. January 11, 2005 24-19
Parallel Processing
Step 5: Postprocessing
See Tutorial 1 for complete postprocessing exercises for this example. Here, two plots are
generated so that you can confirm that the results you obtained with the parallel solver
are the same as those you obtained with the serial solver.
(a) Select Temperature... and Static Temperature in the Y Axis Function drop-down
lists.
(b) Select pressure-outlet-7 in the Surfaces list.
(c) Click on Plot.
24-20
c Fluent Inc. January 11, 2005
Parallel Processing
pressure-outlet-7
3.12e+02
3.10e+02
3.08e+02
3.05e+02
Static 3.02e+02
Temperature
(k) 3.00e+02
2.98e+02
2.95e+02
2.92e+02
48 50 52 54 56 58 60 62 64
Position (in)
Static Temperature
FLUENT 6.2 (2d, segregated, ske)
c Fluent Inc. January 11, 2005 24-21
Parallel Processing
2. Display filled contours of the custom field function dynam-head (Figure 24.5).
Display −→ Contours...
(a) Select Custom Field Functions... in the drop-down list under Contours Of.
The function you created in Tutorial 1, dynam-head, will be shown in the lower
drop-down list.
(b) Change the number of Levels back to 20.
(c) Click on Display, and then Close the panel.
24-22
c Fluent Inc. January 11, 2005
Parallel Processing
8.06e+02
7.65e+02
7.25e+02
6.85e+02
6.44e+02
6.04e+02
5.64e+02
5.24e+02
4.83e+02
4.43e+02
4.03e+02
3.62e+02
3.22e+02
2.82e+02
2.42e+02
2.01e+02
1.61e+02
1.21e+02
8.06e+01
4.03e+01
0.00e+00
Contours of dynam-head
FLUENT 6.2 (2d, segregated, ske)
Summary
In this tutorial you learned how to solve a simple 2D problem using FLUENT’s parallel
solver. Here the automatic grid partitioning performed by FLUENT when you read the
mesh into the parallel version was found to be acceptable. You also learned how to check
the performance of the parallel solver to determine if optimizations are required.
See Section 32.6 of the User’s Guide for additional details about using the parallel solver.
c Fluent Inc. January 11, 2005 24-23
Parallel Processing
24-24
c Fluent Inc. January 11, 2005