Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Star CD User

Download as pdf or txt
Download as pdf or txt
You are on page 1of 406

CCM USER GUIDE

STAR-CD VERSION 4.02

CONFIDENTIAL — FOR AUTHORISED USERS ONLY

© 2006 CD-adapco
TABLE OF CONTENTS

OVERVIEW

1 COMPUTATIONAL ANALYSIS PRINCIPLES


Introduction ............................................................................................................... 1-1
The Basic Modelling Process .................................................................................... 1-1
Spatial description and volume discretisation ........................................................... 1-2
Solution domain definition .............................................................................. 1-3
Mesh definition ................................................................................................ 1-4
Mesh distortion ................................................................................................ 1-5
Mesh distribution and density ......................................................................... 1-6
Mesh distribution near walls ........................................................................... 1-7
Moving mesh features ..................................................................................... 1-8
Problem characterisation and material property definition ....................................... 1-8
Nature of the flow ............................................................................................ 1-9
Physical properties ........................................................................................... 1-9
Force fields and energy sources ...................................................................... 1-9
Initial conditions ............................................................................................ 1-10
Boundary description .............................................................................................. 1-10
Boundary location ......................................................................................... 1-11
Boundary conditions ...................................................................................... 1-11
Numerical solution control ..................................................................................... 1-13
Selection of solution procedure ..................................................................... 1-13
Transient flow calculations with PISO .......................................................... 1-13
Steady-state flow calculations with PISO ..................................................... 1-15
Steady-state flow calculations with SIMPLE ................................................ 1-16
Transient flow calculations with SIMPLE .................................................... 1-17
Effect of round-off errors .............................................................................. 1-18
Choice of the linear equation solver .............................................................. 1-19
Monitoring the calculations .................................................................................... 1-19
Model evaluation .................................................................................................... 1-20
2 BASIC STAR-CD FEATURES
Introduction ............................................................................................................... 2-1
Running a STAR-CD Analysis ................................................................................. 2-2
Using the script-based procedure .................................................................... 2-3
Using STAR-Launch ....................................................................................... 2-8
pro-STAR Initialisation .......................................................................................... 2-12
Input/output window ..................................................................................... 2-13
Main window ................................................................................................. 2-15

Version 4.02 i
The menu bar .................................................................................................2-16
General Housekeeping and Session Control ...........................................................2-18
Basic set-up ....................................................................................................2-18
Screen display control ....................................................................................2-18
Error messages ...............................................................................................2-19
Error recovery ................................................................................................2-20
Session termination ........................................................................................2-21
Set Manipulation .....................................................................................................2-21
Table Manipulation .................................................................................................2-24
Basic functionality .........................................................................................2-24
The table editor ..............................................................................................2-26
Useful points ..................................................................................................2-31
Plotting Functions ....................................................................................................2-31
Basic set-up ....................................................................................................2-31
Advanced screen control ................................................................................2-32
Screen capture ................................................................................................2-33
The Users Tool ........................................................................................................2-35
Getting On-line Help ...............................................................................................2-35
The STAR GUIde Environment ..............................................................................2-38
Panel navigation system .................................................................................2-40
STAR GUIde usage .......................................................................................2-41
General Guidelines ..................................................................................................2-41
3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Introduction ...............................................................................................................3-1
The Cell Table ...........................................................................................................3-1
Cell indexing ....................................................................................................3-3
Multi-Domain Property Setting .................................................................................3-5
Setting up models .............................................................................................3-6
Compressible Flow ....................................................................................................3-9
Setting up compressible flow models ..............................................................3-9
Useful points on compressible flow ...............................................................3-10
Non-Newtonian Flow ..............................................................................................3-11
Setting up non-Newtonian models .................................................................3-11
Useful points on non-Newtonian flow ...........................................................3-11
Turbulence Modelling .............................................................................................3-12
Wall functions ................................................................................................3-13
Two-layer models ..........................................................................................3-13
Low Re models ..............................................................................................3-14
Hybrid wall boundary condition ....................................................................3-14

ii Version 4.02
Reynolds Stress models ................................................................................. 3-15
DES models ................................................................................................... 3-15
LES models ................................................................................................... 3-15
Changing the turbulence model in use .......................................................... 3-16
Heat Transfer In Solid-Fluid Systems ..................................................................... 3-16
Setting up solid-fluid heat transfer models .................................................... 3-17
Heat transfer in baffles .................................................................................. 3-18
Useful points on solid-fluid heat transfer ...................................................... 3-19
Buoyancy-driven Flows and Natural Convection ................................................... 3-20
Setting up buoyancy-driven models .............................................................. 3-20
Useful points on buoyancy-driven flow ........................................................ 3-20
Fluid Injection ......................................................................................................... 3-21
Setting up fluid injection models ................................................................... 3-22
4 BOUNDARY AND INITIAL CONDITIONS
Introduction ............................................................................................................... 4-1
Boundary Location .................................................................................................... 4-1
Command-driven facilities .............................................................................. 4-2
Boundary set selection facilities ...................................................................... 4-3
Boundary listing .............................................................................................. 4-3
Boundary Region Definition ..................................................................................... 4-5
Inlet Boundaries ........................................................................................................ 4-9
Introduction ..................................................................................................... 4-9
Useful points .................................................................................................. 4-10
Outlet Boundaries ................................................................................................... 4-11
Introduction ................................................................................................... 4-11
Useful points .................................................................................................. 4-12
Pressure Boundaries ................................................................................................ 4-12
Introduction ................................................................................................... 4-12
Useful points .................................................................................................. 4-13
Stagnation Boundaries ............................................................................................ 4-14
Introduction ................................................................................................... 4-14
Useful points .................................................................................................. 4-15
Non-reflective Pressure and Stagnation Boundaries ............................................... 4-16
Introduction ................................................................................................... 4-16
Useful points .................................................................................................. 4-18
Wall Boundaries ...................................................................................................... 4-19
Introduction ................................................................................................... 4-19
Thermal radiation properties ......................................................................... 4-20
Solar radiation properties .............................................................................. 4-20

Version 4.02 iii


Other radiation modelling considerations ......................................................4-21
Useful points ..................................................................................................4-22
Baffle Boundaries ....................................................................................................4-23
Introduction ....................................................................................................4-23
Setting up models ...........................................................................................4-24
Thermal radiation properties ..........................................................................4-25
Solar radiation properties ...............................................................................4-26
Other radiation modelling considerations ......................................................4-26
Useful points ..................................................................................................4-27
Symmetry Plane Boundaries ...................................................................................4-27
Cyclic Boundaries ...................................................................................................4-27
Introduction ....................................................................................................4-27
Setting up models ...........................................................................................4-28
Useful points ..................................................................................................4-30
Cyclic set manipulation ..................................................................................4-31
Free-stream Transmissive Boundaries ....................................................................4-32
Introduction ....................................................................................................4-32
Useful points ..................................................................................................4-33
Transient-wave Transmissive Boundaries ...............................................................4-34
Introduction ....................................................................................................4-34
Useful points ..................................................................................................4-35
Riemann Boundaries ...............................................................................................4-36
Introduction ....................................................................................................4-36
Useful points ..................................................................................................4-37
Attachment Boundaries ...........................................................................................4-38
Useful points ..................................................................................................4-39
Radiation Boundaries ..............................................................................................4-39
Useful points ..................................................................................................4-40
Phase-Escape (Degassing) Boundaries ...................................................................4-40
Monitoring Regions .................................................................................................4-40
Boundary Visualisation ...........................................................................................4-41
Solution Domain Initialisation ................................................................................4-42
Steady-state problems ....................................................................................4-42
Transient problems .........................................................................................4-42
5 CONTROL FUNCTIONS
Introduction ...............................................................................................................5-1
Analysis Controls for Steady-State Problems ...........................................................5-1
Analysis Controls for Transient Problems ................................................................5-4
Default (single-transient) solution mode .........................................................5-4

iv Version 4.02
Load-step based solution mode ....................................................................... 5-6
Load step characteristics .................................................................................. 5-6
Load step definition ......................................................................................... 5-8
Solution procedure outline .............................................................................. 5-9
Other transient functions ............................................................................... 5-14
Solution Control with Mesh Changes ..................................................................... 5-15
Mesh-changing procedure ............................................................................. 5-15
Solution-Adapted Mesh Changes ........................................................................... 5-17
6 POROUS MEDIA FLOW
Setting Up Porous Media Models ............................................................................. 6-1
Useful Points ............................................................................................................. 6-4
7 THERMAL AND SOLAR RADIATION
Radiation Modelling for Surface Exchanges ............................................................ 7-1
Radiation Modelling for Participating Media ........................................................... 7-3
Capabilities and Limitations of the DTRM Method ................................................. 7-5
Capabilities and Limitations of the DORM Method ................................................. 7-7
Radiation Sub-domains ............................................................................................. 7-8
8 CHEMICAL REACTION AND COMBUSTION
Introduction ............................................................................................................... 8-1
Local Source Models ................................................................................................ 8-2
Presumed Probability Density Function (PPDF) Models ......................................... 8-3
Single-fuel PPDF ............................................................................................. 8-3
Multiple-fuel PPDF ......................................................................................... 8-9
Regress Variable Models ........................................................................................ 8-10
Complex Chemistry Models ................................................................................... 8-11
Setting Up Chemical Reaction Schemes ................................................................. 8-14
Useful general points for local source and regress variable schemes ........... 8-16
Chemical Reaction Conventions ................................................................... 8-18
Useful points for PPDF schemes ................................................................... 8-18
Useful points for complex chemistry models ................................................ 8-21
Useful points for ignition models .................................................................. 8-21
Setting Up Advanced I.C. Engine Models .............................................................. 8-22
Coherent Flame model (CFM) ...................................................................... 8-24
Extended Coherent Flame model (ECFM) .................................................... 8-26
Extended Coherent Flame model 3Z (ECFM-3Z) — spark ignition ............ 8-28
Extended Coherent Flame model 3Z (ECFM-3Z) — compression ignition . 8-29
Useful points for ECFM models .................................................................... 8-30
Level Set model ............................................................................................. 8-31
Write Data sub-panel ..................................................................................... 8-32

Version 4.02 v
The Arc and Kernel Tracking ignition model (AKTIM) ...............................8-33
Useful points for the AKTIM model .............................................................8-35
The Double-Delay autoignition model ..........................................................8-37
NOx Modelling ........................................................................................................8-39
Soot Modelling ........................................................................................................8-39
Coal Combustion Modelling ...................................................................................8-41
Stage 1 ............................................................................................................8-41
Stage 2 ............................................................................................................8-42
Useful notes ...................................................................................................8-44
Switches and constants for coal modelling ....................................................8-45
Special settings for the Mixed-is-Burnt and Eddy Break-Up models ............8-46
9 LAGRANGIAN MULTI-PHASE FLOW
Setting Up Lagrangian Multi-Phase Models .............................................................9-1
Data Post-Processing .................................................................................................9-4
Static displays ..................................................................................................9-5
Trajectory displays ...........................................................................................9-8
Engine Combustion Data Files ..................................................................................9-9
Useful Points ...........................................................................................................9-10
10 EULERIAN MULTI-PHASE FLOW
Introduction .............................................................................................................10-1
Setting up multi-phase models ................................................................................10-1
Useful points on Eulerian multi-phase flow ..................................................10-4
11 FREE SURFACE AND CAVITATION
Free Surface Flows ..................................................................................................11-1
Setting up free surface cases ..........................................................................11-1
Cavitating Flows ......................................................................................................11-5
Setting up cavitation cases .............................................................................11-5
12 ROTATING AND MOVING MESHES
Rotating Reference Frames .....................................................................................12-1
Models for a single rotating reference frame .................................................12-1
Useful points on single rotating frame problems ...........................................12-1
Models for multiple rotating reference frames (implicit treatment) ..............12-2
Useful points on multiple implicit rotating frame problems ..........................12-4
Models for multiple rotating reference frames (explicit treatment) ...............12-5
Useful points on multiple explicit rotating frame problems ..........................12-8
Moving Meshes .......................................................................................................12-9
Basic concepts ................................................................................................12-9
Setting up models .........................................................................................12-10
Useful points ................................................................................................12-13

vi Version 4.02
Automatic Event Generation for Moving Piston Problems ......................... 12-13
Cell-layer Removal/Addition ................................................................................ 12-14
Basic concepts ............................................................................................. 12-14
Setting up models ........................................................................................ 12-15
Useful points ................................................................................................ 12-18
Sliding Meshes ...................................................................................................... 12-18
Regular sliding interfaces ............................................................................ 12-18
Cell Attachment and Change of Fluid Type ......................................................... 12-22
Basic concepts ............................................................................................. 12-22
Setting up models ........................................................................................ 12-23
Useful points ................................................................................................ 12-27
Mesh Region Exclusion ........................................................................................ 12-28
Basic concepts ............................................................................................. 12-28
Moving Mesh Pre- and Post-processing ............................................................... 12-28
Introduction ................................................................................................. 12-28
Action commands ........................................................................................ 12-29
Status setting commands ............................................................................. 12-30
13 OTHER PROBLEM TYPES
Multi-component Mixing ........................................................................................ 13-1
Setting up multi-component models .............................................................. 13-1
Useful points on multi-component mixing .................................................... 13-3
Aeroacoustic Analysis ............................................................................................ 13-3
Setting up aeroacoustic models ..................................................................... 13-3
Useful points on aeroacoustic analyses ......................................................... 13-4
Liquid Films ............................................................................................................ 13-5
Setting up liquid film models ........................................................................ 13-5
Film stripping ................................................................................................ 13-7
14 USER PROGRAMMING
Introduction ............................................................................................................. 14-1
Subroutine Usage .................................................................................................... 14-1
Useful points .................................................................................................. 14-4
Description of UFILE Routines .............................................................................. 14-5
Boundary condition subroutines .................................................................... 14-5
Material property subroutines ........................................................................ 14-6
Turbulence modelling subroutines ................................................................ 14-9
Source subroutines ....................................................................................... 14-10
Radiation modelling subroutines ................................................................. 14-11
Free surface / cavitation subroutines ........................................................... 14-11
Lagrangian multi-phase subroutines ............................................................ 14-12

Version 4.02 vii


Liquid film subroutines ................................................................................14-14
Eulerian multi-phase subroutines .................................................................14-14
Chemical reaction subroutines .....................................................................14-15
Rotating reference frame subroutines ..........................................................14-16
Moving mesh subroutines ............................................................................14-16
Miscellaneous flow characterisation subroutines ........................................14-17
Solution control subroutines ........................................................................14-18
Sample Listing .......................................................................................................14-19
New Coding Practices ...........................................................................................14-20
User Coding in parallel runs ..................................................................................14-22
15 PROGRAM OUTPUT
Introduction .............................................................................................................15-1
Permanent Output ....................................................................................................15-1
Input-data summary .......................................................................................15-1
Run-time output .............................................................................................15-3
Printout of Field Values ..........................................................................................15-3
Optional Output .......................................................................................................15-3
Example Output .......................................................................................................15-4
16 pro-STAR CUSTOMISATION
Set-up Files ..............................................................................................................16-1
Panels .......................................................................................................................16-2
Panel creation .................................................................................................16-2
Panel definition files ......................................................................................16-5
Panel manipulation .........................................................................................16-6
Macros .....................................................................................................................16-6
Function Keys ..........................................................................................................16-9
17 OTHER STAR-CD FEATURES AND CONTROLS
Introduction .............................................................................................................17-1
File Handling ...........................................................................................................17-1
Naming conventions ......................................................................................17-1
Commonly used files .....................................................................................17-1
File relationships ............................................................................................17-7
File manipulation ...........................................................................................17-9
Special pro-STAR Features ...................................................................................17-12
pro-STAR environment variables ................................................................17-12
Resizing pro-STAR ......................................................................................17-13
Special pro-STAR executables ....................................................................17-14
Use of temporary files by pro-STAR ...........................................................17-14
The StarWatch Utility ...........................................................................................17-15

viii Version 4.02


Running StarWatch ..................................................................................... 17-15
Choosing the monitored values ................................................................... 17-17
Controlling STAR ....................................................................................... 17-17
Manipulating the StarWatch display ........................................................... 17-20
Monitoring another job ................................................................................ 17-21
Hard Copy Production .......................................................................................... 17-21
Neutral plot file production and use ............................................................ 17-21
Scene file production and use ...................................................................... 17-23

APPENDICES
A pro-STAR CONVENTIONS
Command Input Conventions .................................................................................. A-1
Help Text / Prompt Conventions ............................................................................. A-3
Control and Function Key Conventions .................................................................. A-4
File Name Conventions ............................................................................................ A-4
B FILE TYPES AND THEIR USAGE
C PROGRAM UNITS
D pro-STAR X-RESOURCES
E USER INTERFACE TO MESSAGE PASSING ROUTINES
F STAR RUN OPTIONS
Usage .........................................................................................................................F-1
Options ......................................................................................................................F-1
Parallel Options .........................................................................................................F-3
Resource Allocation ..................................................................................................F-6
Default Options .........................................................................................................F-7
Cluster Computing ....................................................................................................F-8
Batch Runs Using STAR-NET .................................................................................F-8
Running under IBM Loadleveler using STAR-NET .......................................F-8
Running under LSF using STAR-NET ...........................................................F-9
Running under OpenPBS using STAR-NET ................................................F-10
Running under PBSPro using STAR-NET ....................................................F-11
Running under SGE using STAR-NET .........................................................F-11
Running under Torque using STAR-NET .....................................................F-12
G BIBLIOGRAPHY

INDEX

INDEX OF COMMANDS

Version 4.02 ix
OVERVIEW
Purpose
The Methodology volume presents the mathematical modelling practices embodied
in the STAR-CD system and the numerical solution procedures employed. In this
volume, the focus is on the structure of the system itself and how to use it. This
presentation assumes that the reader is familiar with the background information
provided in the Methodology volume.

Contents
Chapter 1 introduces some of the fundamental principles of computational
continuum mechanics, including an outline of the basic steps involved in setting up
and using a successful computer model. The important factors to consider at each
step are mostly explained independently of the computer system used to perform the
analysis. However, reference is also made to the particular capabilities of the
STAR-CD system, where appropriate.
Chapter 2 outlines the basic features of STAR-CD, including GUI facilities,
session control and plotting utilities. Chapters 3 to 5 provide the reader with detailed
instructions on how to use some of the basic code facilities, e.g. boundary condition
specification, material property definition, etc., and an overview of the GUI panels
appropriate to each of them. The description covers all facilities (other than mesh
generation) that might be employed for modelling most common continuum
mechanics problems. Mesh generation itself is covered in a separate volume, the
Meshing User Guide.
Chapters 2 to 5 should be read at least once to gain an understanding of the
general housekeeping principles of pro-STAR and to help with any problems
arising from routine operations. It is recommended that users refer to the
appropriate chapter repeatedly when setting up a model for general guidance and an
overview of the relevant GUI panels.
Chapters 6 to 13 describe additional STAR-CD capabilities relevant to models
of a more specialised nature, i.e. rotating systems, combustion processes,
buoyancy-driven flows, etc. Users interested in a particular topic should consult the
appropriate section for a summary of commands or options specially designed for
that purpose, plus hints and tips on performing a successful simulation.
Chapter 14 outlines the user programmability features available and provides an
example FORTRAN subroutine listing implementing these features. All such
subroutines are readily available for use and can be easily adapted to suit the
model's requirements.
Chapter 15 presents the printable output produced by the code which provides,
among other things, a summary of the problem specification and monitoring
information generated during the calculation.
Chapter 16 explains how pro-STAR can be customised, in terms of user-defined
panels, macros and keyboard function keys, to meet a user’s individual
requirements.
Finally, Chapter 17 covers some of the less commonly used features of
STAR-CD, including the interaction between STAR and pro-STAR and how
various system files are used.

Version 4.02 1
Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Introduction

Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES


Introduction
The aim of this section is to introduce the most important issues involved in setting
up and solving a continuum mechanics problem using a computational continuum
mechanics code. Although the discussion applies in principle to any such code,
reference is made where appropriate to the particular capabilities of the STAR-CD
system. It is also assumed that the reader is familiar with the material presented in
the Methodology volume.
The process of computational mechanics simulation does not usually start with
the direct use of such a code. It is indeed important to recognise that STAR-CD, or
any other CFD, CAD or CAE system, should be treated as a tool to assist the
engineer in understanding physical phenomena.
The success or failure of a continuum mechanics simulation depends not only on
the code capabilities, but also upon the input data, such as:
• Geometry of the solution domain
• Continuum properties
• Boundary conditions
• Solution control parameters

For a simulation to have any chance of success, such information should be


physically realistic and correctly presented to the analysis code.
The essential steps to be taken prior to computational continuum mechanics
(CCM) modelling are as follows:
• Pose the problem in physical terms.
• Establish the amount of information available and its sufficiency and validity.
• Assess the capabilities and features of the STAR-CD code, to ensure that the
problem is well posed and amenable to numerical solution by the code.
• Plan the simulation strategy carefully, adopting a step-by-step approach to the
final solution.

Users should turn to STAR-CD and proceed with the actual modelling only after the
above tasks have been completed.

The Basic Modelling Process


The modelling process itself can be divided into four major phases, as follows:
Phase 1 — Working out a modelling strategy
This requires a precise definition of the physical system’s geometry, material
properties and flow/deformation conditions based on the best available
understanding of the relevant physics. The necessary tasks include:
• Planning the computational mesh (e.g. number of cells, size and distribution
of cell dimensions, etc.).
• Looking up numerical values for appropriate physical parameters
(e.g. density, viscosity, specific heat, etc.).
• Choosing the most suitable modelling option from what is available
(e.g. turbulence model, combustion option, etc.).

Version 4.02 1-1


COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Spatial description and volume discretisation

The user also has to balance the requirement of physical fidelity and numerical
accuracy against the simulation cost and computational capabilities of his system.
His modelling strategy will therefore incorporate some trade-off between these two
factors.
This initial phase of modelling is particularly important for the smooth and
efficient progress of the computational simulation.
Phase 2 — Setting up the model using pro-STAR
The main tasks involved at this phase are:
• Creating a computational mesh to represent the solution domain (i.e. the
model geometry).
• Specifying the physical properties of the fluids and/or solids present in the
simulation and, where relevant, the turbulence model(s), body forces, etc.
• Setting the solution parameters (e.g. solution variable selection, relaxation
coefficients, etc.) and output data formats.
• Specifying the location and definition of boundaries and, for unsteady
problems, further definition of transient boundary conditions and time steps.
• Writing appropriate data files as input to the analytical run of the following
phase.

Phase 3 — Performing the analysis using STAR


This phase consists of:
• Reading input data created by pro-STAR and, if dealing with a restart run, the
results of a previous run.
• Judging the progress of the run by analysing various monitoring data and
solution statistics provided by STAR.

Phase 4 — Post-processing the results using pro-STAR


This involves the display and manipulation of output data created by STAR using
the appropriate pro-STAR facilities.

The remainder of this chapter discusses the elements of each modelling phase in
greater detail.

Spatial description and volume discretisation


One of the basic steps in preparing a STAR-CD model is to describe the geometry
of the problem. The two key components of this description are:
• The definition of the overall size and shape of the solution domain.
• The subdivision of the solution domain into a mesh of discrete, finite,
contiguous volume elements or cells, as shown in Figure 1-1.

1-2 Version 4.02


Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Spatial description and volume discretisation

Figure 1-1 Example of solution domain subdivision into cells

This process is called volume discretisation and is an essential part of solving the
above equations numerically. In STAR-CD both components of the spatial
description are performed as part of the same operation, setting up the finite-volume
mesh, but separate considerations apply to each of them.
Solution domain definition
Through its internal design and construction, STAR-CD permits a very general and
flexible definition of what constitutes a solution domain. The latter can be:
• A fluid and/or heat flow field fully occupying an open region of space
• Fluid and/or heat flowing through a porous medium
• Heat flowing through a solid
• A solid undergoing mechanical deformation

Arbitrary combinations of the above conditions can also be specified within the
same model, as in problems involving fluid-solid heat transfer. The user’s first task
is therefore to decide which parts of the physical system being modelled need to be
included in the solution domain and whether each part is occupied by a fluid, solid
or porous medium.
Whatever its composition, the fundamental requirement is that the solution
domain is bounded. This means that the user has to examine his system’s geometry
carefully and decide exactly where the enclosing boundaries lie. The boundaries can
be one of four kinds:
1. Physical boundaries — walls or solid obstacles of some description that
serve to physically confine a fluid flow
2. Symmetry boundaries — axes or planes beyond which the problem solution
becomes a mirror image of itself
3. Cyclic boundaries — surfaces beyond which the problem solution repeats
itself, in a cyclic or anticyclic fashion
The purpose of symmetry and cyclic boundaries is to limit the size of the
domain, and hence the computer requirements, by excluding regions where
the solution is essentially known. This in turn allows one to model the
problem in greater detail than would have been the case otherwise.
Version 4.02 1-3
COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Spatial description and volume discretisation

4. Notional boundaries — these are non-physical surfaces that serve to


‘close-off’ the solution domain in regions not covered by the other two types
of boundary. Their location is entirely up to the user’s discretion but, in
general, they should be placed only where one of the following apply:
(a) Flow/deformation conditions are known
(b) Flow/deformation conditions can be guessed reasonably well
(c) The boundary is far enough away from the region of interest for boundary
condition inaccuracies to have little effect

Thus, locating this type of boundary may require some trial and error.

The location and characterisation of boundaries is discussed further in “Boundary


description” on page 1-10.
Mesh definition
Creation of a lattice of finite-volume cells to represent the solution domain is
normally the most time-consuming task in setting up a STAR-CD model. This
process is greatly facilitated by STAR-CD because of its ability to generate cells of
an arbitrary, polyhedral shape.
In creating a finite-volume mesh, the user should aim to represent accurately the
following two entities:
1. The overall external geometry of the solution domain, by specifying an
appropriate size and shape for near-boundary cells. The latter’s external faces,
taken together, should make up a surface that adequately represents the shape
of the solution domain boundaries. Small inaccuracies may occur because all
boundary cell faces (including rectangular faces) are composed of triangular
facets, as shown in Figure 1-2. These errors diminish as the mesh is refined.

triangular facet

Figure 1-2 Boundary representation by triangular facets

2. The internal characteristics of the flow/deformation regime. This is achieved


by careful control of mesh spacing within the solution domain interior so that

1-4 Version 4.02


Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Spatial description and volume discretisation

the mesh is finest where the problem characteristics change most rapidly.
Near-wall regions are important and a high mesh density is needed to resolve
the flow in their vicinity. This point is discussed further in “Mesh distribution
near walls” on page 1-7.

Mesh spacing considerations


The chief considerations governing the mesh spatial arrangement are:
• Accuracy — primarily determined by mesh density and, to a lesser extent,
mesh distortion (discussed in “Mesh distortion” on page 1-5).
• Numerical stability — this is a strong function of the degree of distortion.
• Cost — a function of both the aforementioned factors, through their influence
on the speed of convergence and c.p.u. time required per iteration or time
step.

Thus, the user should aim at an optimum mesh arrangement which


• employs the minimum number of cells,
• exhibits the least amount of distortion,
• is consistent with the accuracy requirements.

Chapter 2 of the Meshing User Guide describes several methods available in


STAR-CD, some of them semi-automatic, to help the user achieve this goal.
However, even when suitable automatic mesh generation procedures are available,
the user must still draw on knowledge and experience of computational fluid and
solid mechanics to produce the right kind of mesh arrangement.
Mesh distortion
Mesh distortion is measured in terms of three factors — aspect ratio, internal angle
and warp angle — illustrated in Figure 1-3.

b
θ

a
b/a = aspect ratio θ = internal angle φ = warp angle

Figure 1-3 Cell shape characteristics

When setting up the mesh, the user should try to observe the following guidelines:
• Aspect Ratio — values close to unity are preferable, but departures from this

Version 4.02 1-5


COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Spatial description and volume discretisation

are allowed.
• Internal Angle — departures from 90° intersections between cell faces
should be kept to a minimum.
• Warp Angle — the optimum value of this angle is zero, which can occur only
when the cell face vertices are co-planar.

Any adverse effects arising from departures from the preferred values of these
factors manifest themselves through
• the relative magnitudes of the coefficients in the finite-volume equations,
especially those arising from non-orthogonality, and
• the signs of the coefficients (negative values are generally detrimental).

It is difficult to place rigid limits on the acceptable departures because they depend
on local flow conditions. However, the following values serve as a useful guideline:

Aspect Ratio 10
Internal angle 45°
Warp angle 45°

pro-STAR can calculate these quantities and identify cells having out-of-bounds
values, as discussed in Chapter 3, “Mesh and Geometry Checking” of the Meshing
User Guide.
What is really important in this respect is the combined effect of the various
kinds of mesh distortion. If all three are simultaneously present in a single cell, the
limits given above might not be stringent enough. On the other hand, the effects of
distortion also depend on the nature of the local flow. Thus, the above limits may
be exceeded in the region of
• simple flows such as, for example, uniform-velocity ‘free’ streams,
• wall boundary layers, where cells of high aspect ratio (in the flow direction)
are commonly employed without difficulty.

Generally speaking, non-orthogonality at boundaries may cause problems and


should be minimised whenever practicable.
Mesh distribution and density
Numerical discretisation errors are functions of the cell size; the smaller the cells
(and therefore the higher the mesh density), the smaller the errors. However, a high
mesh density implies a large number of mesh storage locations, with associated high
computing cost. It is therefore advisable, wherever possible, to
• ensure that the mesh density is high only where needed, i.e. in regions of steep
gradients of the flow variables, and low elsewhere;
• avoid rapid changes in cell dimensions in the direction of steep gradients in
the flow variables.

The flexibility afforded by STAR-CD’s unstructured polyhedral meshes facilitates


such selective refinement. An illustration of some of the numerous cell shapes that
may be employed is given in Figure 2-43 and Figure 2-44 of the Meshing User
Guide.

1-6 Version 4.02


Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Spatial description and volume discretisation

Of course, it is not always possible to ascertain a priori what the flow structure
will be. However, the need for higher mesh density can usually be anticipated in
regions such as:
• Wall boundary layers
• Jets issuing from apertures
• Shear layers formed by flow separation or neighbouring streams of different
velocities
• Stagnation points produced by flow impingement
• Wakes behind bluff bodies
• Temperature or concentration fronts arising from mixing or chemical reaction

Mesh distribution near walls


As discussed in Chapter 6, “Wall Boundary Conditions” of the Methodology
volume, wall functions are an economic way of representing turbulent boundary
layers (hydrodynamic and thermal) in turbulent flow calculations. These functions
effectively allow the boundary layer to be bridged by a single cell, as shown in
Figure 1-4(a).

Outer
region

Inner
region
y

(a) Wall function model (b) Two-layer or Low Re models

Figure 1-4 Near-wall mesh distribution

The location y of the cell centroids in the near-wall layer, and hence the thickness
of that layer, is usually determined by reference to the dimensionless normal
distance y + from the wall. For the wall function to be effective, this distance must
be
• not too small, otherwise, the ‘bridge’ might span only the laminar sublayer;
• not too large, as the flow at that location might not behave in the way assumed
in deriving the wall functions.

Ideally, y + should lie in the approximate range 30 to 150. Note that the above
considerations apply to Reynolds Stress models as well as several classes of eddy
viscosity model (see Chapter 3, “Turbulence Modelling”).
Alternative treatments that do not require the use of wall functions are also
available. These are:

Version 4.02 1-7


COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Problem characterisation and material property definition

1. Two-layer turbulence models, whereby wall functions are replaced by a


one-equation k-l model or a zero-equation mixing-length model
2. Low Reynolds number models (including the V2F model), where viscous
effects are incorporated in the k and ε transport equations

For the above two types of model, the solution domain should be divided into two
regions with the following characteristics:
• An inner region containing a fine mesh
• An outer region containing normal mesh sizes

The two regions are illustrated in Figure 1-4(b). As explained in the Methodology
volume (Chapter 6, “Two-layer models”), the inner region should contain at least
15 mesh layers and encompass that part of the boundary layer influenced by viscous
effects.
A more recent development, called the hybrid wall function is also available that
extends the low-Reynolds number formulation of most turbulence models. This
may be used to capture boundary layer properties more accurately in cases where
the near-wall cell size is not adapted for the low-Reynolds number treatment and
thus achieve y + independent solutions.
Moving mesh features
STAR-CD offers a range of moving mesh features, including:
• General mesh motion
• Internal sliding mesh
• Cell deletion and insertion

The first of these is straightforward to employ and the only caution required is the
obvious one: avoid creating excessive distortion when redistributing the mesh. This
caution also applies to the use of the other two features, but they have additional
rules and guidelines attached to them. These are summarised in the Methodology
volume, Chapter 15 (“Internal Sliding Mesh” on page 15-5 and “Cell Layer
Removal and Addition” on page 15-7). Additional guidelines also appear in this
volume, “Cell-layer Removal/Addition” on page 12-14 and “Sliding Meshes” on
page 12-18; hence they are not repeated here.

Problem characterisation and material property definition


Correct definition of the physical conditions and the properties of the materials
involved is a prerequisite to obtaining the right solution to a problem, or indeed to
obtaining any solution at all. It is also essential for determining whether the problem
can be modelled with STAR-CD. The user must therefore ensure that the problem
is well defined in respect of:
• The nature of the fluid flow (e.g. steady/unsteady, laminar/turbulent,
incompressible/compressible)
• Physical properties (e.g. density, viscosity, specific heat)
• External force fields (e.g. gravity, centrifugal forces) and energy sources,
when present
• Initial conditions for transient flows

1-8 Version 4.02


Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Problem characterisation and material property definition

Nature of the flow


It is very important to understand the nature of the flow being analysed in order to
select the appropriate mathematical models and numerical solution algorithms.
Problems will arise if an incorrect choice is made, as in the following examples:
• Employing an iterative, steady-state algorithm for an inherently unsteady
problem, such as vortex shedding from a bluff body
• Computing a turbulent flow without invoking a suitable turbulence model
• Modelling transitional flow with one of the turbulence models currently
implemented in STAR-CD. None of them can represent transitional behaviour
accurately.

Physical properties
The specification of physical properties, such as density, molecular viscosity,
thermal conductivity, etc. depends on the nature of the fluids or solids involved and
the circumstances of use. For example, STAR-CD contains several built-in
equations of state from which density can be calculated as a function of one or more
of the following field variables:
• Pressure
• Temperature
• Fluid composition

In all cases where complex calculations are used to evaluate a material property, the
following measures are recommended:
• The relevant field variables must be assigned plausible initial and boundary
values.
• Where necessary, properties should be solved for together with the field
variables as part of the overall solution.
• In the case of strong dependencies between properties and field variables, the
user should consider under-relaxation of the property value calculations, in
the manner described in the Methodology volume (Chapter 7, “Scalar
transport equations”).
• When required, STAR-CD’s facility for alternative, user-programmable
functions may be used.

Force fields and energy sources


As already noted, STAR-CD has built-in provision for body forces arising from
• buoyancy,
• rotation.

It is important to remember that as the strength of the body forces increases relative
to the viscous (or turbulent) stresses, the flow may become physically unstable. In
these circumstances it is advisable to switch to the transient solution mode.
It is also possible to insert additional, external force fields and energy sources
via the user programming facilities of STAR-CD. In such cases, it is important to
understand the physical implications and avoid specifying conditions that lead to

Version 4.02 1-9


COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Boundary description

physical or numerical instability. Examples of such conditions are:


• Thermal energy sources that increase linearly with temperature. These can
give rise to physical instability called ‘thermal runaway’.
• Setting the coefficient β i in the permeability function K i = α i v + β i to a
very small or zero value. If the local fluid velocity also becomes very small,
the result may be numerical instability whereby small pressure perturbations
produce a large change in velocities.

Initial conditions
The term ‘initial conditions’ refers to values assigned to the dependent variables at
all mesh points before the start of the calculations. Their implication depends on the
type of problem being considered:
• In unsteady applications, this information has a clear physical significance
and will affect the course of the solution. Due care must therefore be taken in
providing it. It sometimes happens that the effects of initial conditions are
confined to a start-up phase that is not of interest (as in, for example, flows
that are temporally periodic). However, it is still advisable to take some
precautions in specifying initial conditions for reasons explained below.
• In calculating steady state problems by iterative means, the initial conditions
will usually have no influence on the final solution (apart from rare occasions
when the solution is multi-valued), but may well determine the success and
speed of achieving it.

Poor initial field specifications or, for transient problems, abrupt changes in
boundary conditions put severe demands on the numerical algorithm when
substituted into the finite-volume equations. As a consequence, the following
special ‘start-up’ measures may be necessary to ensure numerical stability:
• Use of unusually small time steps in transient calculations.
• Use of strong under-relaxation in iterative solutions.

Specific recommendations concerning these practices are given in “Numerical


solution control” on page 1-13. In either case, increased computing times can be an
undesirable side effect.

Boundary description
As stated in “Spatial description and volume discretisation” on page 1-2, boundary
identification and description are intimately connected with the generation of the
finite-volume mesh, since the boundary topography is defined by the outermost cell
faces. Furthermore, correct specification of the boundary conditions is often the
main area of difficulty in setting up a model. Problems often arise in the following
areas:
• Identifying the correct type of condition
• Specifying an acceptable mix of boundary types
• Ascribing appropriate boundary values

The above are in turn linked to the decisions on where to place the boundaries in the

1-10 Version 4.02


Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Boundary description

first instance.
Boundary location
Difficulties in specifying boundary location normally arise where the flow
conditions are incompletely known, for example at outlets. The recommended
solutions, in decreasing degree of accuracy, are to place boundaries
• in regions where the conditions are known, if this is possible;
• in a location where the ‘Outlet’ or ‘Prescribed Pressure’ option is applicable
(see Chapter 5 in the Methodology volume);
• where the approximations in the boundary condition specification are unlikely
to propagate upstream into the regions of interest.

Whenever possible, it is particularly important to avoid the following situations:


1. A boundary that passes through a major recirculation zone.
2. In transient transonic or supersonic compressible flows, an outlet boundary
located where the flow is not supersonic.
3. A mix of boundary conditions that is inappropriate. Examples of this are:
(a) Multiple ‘Outlet’ boundaries — unless further information is supplied on
how the flow is partitioned between the outlets.
(b) Prescribed flow split outlets coexisting with prescribed mass outflow
boundaries in the same domain.
(c) A combination of prescribed pressure and flow-split outlet conditions.

Boundary conditions
Another source of potential difficulty is in boundary value specification wherever
known conditions need to be set, e.g. at a ‘Prescribed Inflow’ or ‘Inlet’ boundary.
The basic points to bear in mind in this situation are:
• All transport equations to be solved require specification of their boundary
values, including the turbulence transport equations when they are invoked
• Inappropriate setting of boundary values leads to erroneous results and, in
extreme cases, to numerical instability

The following recommendations can be given regarding each different type of


boundary:
1. Prescribed flow — Here, care should be taken to:
(a) Assign realistic values to all dependent variables, including the
turbulence parameters, and also to auxiliary quantities, such as density.
(b) Ensure that, if this is the only type of flow boundary imposed, overall
continuity is satisfied (STAR-CD will accept inadvertent mass
imbalances of up to 5%, correcting them by adjusting the outflows. An
error message is issued if the imbalance exceeds this figure).

2. Outlet — The main points to note for this boundary type are:
(a) The need to specify the boundary, where possible, at locations where the
flow is everywhere outwardly directed; also to recognise that, if inflow

Version 4.02 1-11


COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Boundary description

occurs, it may introduce numerical instability and/or inaccuracies.


(b) The necessity, if more than one boundary of this type is declared, of
prescribing either the flow split between them or the mass outflow rate at
each location.
(c) The inapplicability of ‘prescribed split’ outlets to problems where the
inflows are not fixed, e.g.
i) in combination with pressure boundary conditions, or
ii) in the case of transient compressible flows.

3. Prescribed pressure — The main precautions are:


(a) To specify relative (to a prescribed datum) rather than absolute pressures.
(b) If inflow is likely to occur, to assign realistic boundary values to
temperature and species mass fractions. It is also advisable to specify the
turbulence parameters indirectly, via the turbulence intensity and length
scale or by extrapolating them from values in the interior of the solution
domain.

4. Stagnation conditions — It is recommended to use this condition for


boundaries lying within large reservoirs where properties are not significantly
affected by flow conditions in the solution domain.

5. Non-reflecting pressure and stagnation conditions — A special


formulation of the standard pressure and stagnation conditions, developed to
facilitate analysis of steady-state turbomachinery applications

6. Cyclic boundaries — These always occur in pairs. The main points of advice
are:
(a) Impose this condition only in appropriate circumstances.
Two-dimensional axisymmetric flows with swirl is a good example of an
appropriate application.
(b) For axisymmetric flows, make use of the CD/UD blending scheme to
apply the maximum level of central differencing in the tangential
direction (the default blending factor is 0.95; see also on-line Help topic
“Miscellaneous Controls” in STAR GUIde).

7. Planes of symmetry — It is recommended to use this condition for


two-dimensional axisymmetric flows without swirl

8. Free-stream transmissive boundaries — Used only for modelling supersonic


free streams

9. Transient wave transmissive boundaries — Used only in problems involving


transient compressible flows

10. Riemann boundaries — This condition is based on the theory of Riemann


invariants and its application allows pressure waves to leave the solution
domain without reflection

1-12 Version 4.02


Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Numerical solution control

Numerical solution control


Proper control of the numerical solution process applied to the transport equations
is highly important, both for acceptable computational efficiency and, sometimes,
in order to achieve a solution at all. By necessity, the means of controlling the
process depend heavily on the particular numerical techniques employed so no
universal guidelines can be given. Thus, the recommended settings vary with the
particular algorithm selected and the circumstances of application.
Selection of solution procedure
The basic selection should be based on a correct assessment of the nature of the
problem and will be either
• a transient calculation, starting from well-defined initial and boundary
conditions and proceeding to a new state in a series of discrete time steps; or
• a steady-state calculation, where an unchanging flow/deformation pattern
under a given set of boundary conditions is arrived at through a number of
numerical iterations.

PISO and SIMPLE are the two alternative solution procedures available in
STAR-CD. PISO is the default for unsteady calculations and is sometimes preferred
for steady-state ones, in cases involving strong coupling between dependent
variables such as buoyancy driven flows. SIMPLE is the default algorithm for
steady-state solutions and works well in most cases.
SIMPLE is also used for transient calculations in the case of free surface and
cavitating flows, where it is the only option. In most other transient flow problems,
PISO is likely to be more efficient due to the fact that PISO correctors are usually
cheaper than outer iterations on all variables within a time step of the transient
SIMPLE algorithm. However, there are situations in which PISO would require
many correctors or even fail to converge unless the time step is reduced, whereas
SIMPLE may allow larger time steps with a moderate number of outer iterations per
time step. This is the case when the flow changes very little but certain slow
transients are present in the behaviour of scalar variables (e.g. slow heating up of
solid structures in the case of solid-fluid heat transfer problems, deposition of
chemical species on walls in after-treatment of exhaust gases, etc.). In such cases,
transient SIMPLE can be used to save on computing time.
When doubts exist as to whether the problem considered actually possesses a
steady-state solution or when iterative convergence is difficult to achieve, it is better
to perform the calculations using the transient option.
Transient flow calculations with PISO
As stated in “The PISO algorithm” on page 7-2 of the Methodology volume, PISO
performs at each time (or iteration) step, a predictor, followed by a number of
correctors, during which linear equation sets are solved iteratively for each main
dependent variable. The decisions on the number of correctors and inner iterations
(hereafter referred to as ‘sweeps’, to avoid confusion with outer iterations
performed as part of the steady-state solution mode) are made internally on the basis
of the splitting error and inner residual levels, respectively, according to prescribed
tolerances and upper limits. The default values for the solver tolerances and

Version 4.02 1-13


COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Numerical solution control

maximum correctors and sweeps are given in Table 1-1. Normally, these will only
require adjustment by the user in exceptional circumstances, as discussed below.

Table 1-1: Standard Control Parameter Settings for Transient PISO


Calculations

Variable
Parameter
Velocity Pressure Turbulence Enthalpy Mass fraction
Solver
0.01 0.001 0.01 0.01 0.01
tolerance
Sweep limit 100 1000 100 100 100
Pressure under-relaxation factor = 1.0
Corrector limit = 20
Corrector step tolerance = 0.25

The remaining key parameter in transient calculations with PISO is the size of the
time increment δt . This is normally determined by accuracy considerations and
may be varied during the course of the calculation. The step should ideally be of the
same order of magnitude as the smallest characteristic time δt c for convection and
diffusion, i.e.

2
δt c = min ⎛⎜ δL ρδL
------, ------------⎞⎟ (1-1)
U Γ
⎝ ⎠

Here, U and Γ are a characteristic velocity and diffusivity, respectively, and δL is


a mean mesh dimension. Typically, it is possible to operate with δt ≈ 50 δt c and
still obtain reasonable temporal accuracy. Values significantly above this may lead
to errors and numerical instability, whereas smaller values will lead to increased
computing times.
During the course of a calculation, the limits given in Table 1-1 may be reached,
in which case messages to this effect will be produced. This is most likely to occur
during the start-up phase but is nevertheless acceptable if, later on, the warnings
either cease entirely or only appear occasionally, and the predictions look
reasonable. If, however, the warnings persist, corrective actions should be taken.
The possible actions are:
• Reduction in time step by, say, an initial factor of 2 — if this improves
matters, then the cause may simply be an excessively large δt .
• Increase in the sweep limits — if measure 1 fails, then this should be tried,
only on the variable(s) whose limit(s) have been reached. Again, twofold
changes are appropriate.
• Pressure under-relaxation — a value of 0.8 for pressure correction
under-relaxation, using PISO, may be helpful for some difficult cases (e.g. for
severe mesh distortion or flows with Mach numbers approaching 1).
• Corrector step tolerance — this may be set to a lower value but consult

1-14 Version 4.02


Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Numerical solution control

CD adapco first.

Steady-state flow calculations with PISO


When PISO operates in this mode, the inner residual tolerances are decreased and
under-relaxation is introduced on all variables, apart from pressure, temperature and
mass fraction. However, the last two variables may need to be under-relaxed for
buoyancy driven problems. The standard, default values for these parameters and
the sweep limits, which are unchanged from the transient mode, are given in Table
1-2.
.

Table 1-2: Standard Control Parameter Settings for Steady PISO


Calculations

Variable
Parameter
Velocity Pressure Turbulence Enthalpy Mass fraction
Solver
0.1 0.05 0.1 0.1 0.1
tolerance
Sweep limit 100 1000 100 100 100
Relaxation
0.7 1.0 0.7 0.95 1.0
factor
Corrector limit = 20
Corrector step tolerance = 0.25

These settings should, all being well, result in near-monotonic decrease in the
global residuals during the course of the calculations, depending on mesh density
and other factors. If, thereafter, one or more of the global residuals R φ do not fall,
then remedial measures will be necessary. In some instances, the offending
variable(s) can be identified from the behaviour of the global residuals.
The main remedies now available are:
• Reduction in relaxation factor(s) — this should be done in decrements of
between 0.05 and 0.10 and should be applied to the velocities if the
momentum and/or mass residuals are at fault.
• Decrease in solver tolerances — as in the transient case, this may prove
beneficial, especially in respect of the pressure tolerance and its importance to
the flow solution. A twofold reduction should indicate whether this measure
will work.
• Increase in sweep limits — if warning messages about the limits being
reached appear and are not suppressed by measures 1 and 2, then it may be
worthwhile increasing the limit(s) on the offending variables.
• Under-relaxation of density and effective viscosity — use of this method for
density can be advantageous where significant variations occur,
e.g. compressible flows, combustion, and mixing of dissimilar gases.
Effective viscosity oscillations can arise in turbulent flow and non-Newtonian
fluid flow and can be similarly damped by this device.

Version 4.02 1-15


COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Numerical solution control

Steady-state flow calculations with SIMPLE


As noted previously, the control parameters available for SIMPLE are similar to
those for PISO, except that, in the case of the former, a single corrector stage is
always used and pressure is under-relaxed. The standard (default) settings are given
in Table 1-3.
.

Table 1-3: Standard Control Parameter Settings for Steady SIMPLE


Calculations

Variable
Parameter
Velocity Pressure Turbulence Enthalpy Mass fraction
Solver
0.1 0.05 0.1 0.1 0.1
tolerance
Sweep limit 100 1000 100 100 100
Relaxation
0.7 0.3 0.7 0.95 1.0
factor

In the event of failure to obtain solutions with the standard values, then the measures
to be taken are essentially the same as those for iterative PISO, given in the previous
section. However, here, reduction of the pressure relaxation factor is an additional
device for overcoming convergence problems. The problems usually arise either
from a highly distorted mesh, or from highly complex physics (many variables
affecting each other). If the grid is distorted, one should reduce the relaxation factor
for pressure from the beginning of the run (e.g. to 0.1). If convergence problems are
still encountered, a substantial reduction of the under-relaxation factor for velocities
and turbulence model variables should be tried (e.g. to 0.5). If this does not help, the
problem may lie in severe mesh defects or errors in the set-up. Further reduction of
under-relaxation factors may be tried if the grid is severely distorted and cannot be
improved; otherwise, improving the mesh quality can be of much greater help.
Note that the pressure under-relaxation factor needs to be adjusted within the
limits of some range to make the iteration process converge, where the number of
iterations required to reach such convergence is mainly dictated by the
corresponding factors for velocities (and for scalar variables when strongly coupled
to the flow). In the case of well-behaved flows and reasonable meshes, the
relaxation factor for pressure can be selected as (1.0 - relaxation factor for
velocities), e.g. 0.2 for pressure and 0.8 for velocities. Usually, for a given velocity
relaxation factor, the one for pressure can be varied within some range without
affecting the total number of iterations and computing time, but outside this range
the iterative process would diverge. The lower the relaxation factor for velocities,
the wider the range of pressure relaxation factors that can be used (e.g. between 0.05
and 0.8 if the velocity factor is low, say around 0.5). On the other hand, this range
becomes narrower when the mesh is distorted.
The limit to which the velocity relaxation factor can be increased is both
problem- and mesh-dependent. When many similar problems need to be solved, it
is worth trying to work near the optimum as this may save a lot of computing time.
1-16 Version 4.02
Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Numerical solution control

On the other hand, for an one-off analysis, it may be more efficient to use a
conservative setting.
Note that under some conditions, such as those in Tutorial 13.1, a steady-state
solution cannot be achieved due to the inherent unsteady character of the flow. This
is often the case when the problem geometry possesses some form of symmetry but
the Reynolds (or another equivalent) number is high and recirculation zones are
present. In this case the residuals stop falling at some level and then continue to
oscillate. The “solution” at that stage may be far from a valid solution of the
governing equations and should not be interpreted as such unless the residual level
is sufficiently small. An eddy-viscosity turbulence model (such as the standard k-e)
combined with a first-order upwind scheme for convective fluxes may produce a
steady-state solution, while a less diffusive turbulence model (such as Reynolds
Stress and non-linear eddy-viscosity models) combined with a higher-order
differencing scheme (such as central differencing) may not. In such cases, a
transient simulation should be performed; the unsteady solution may oscillate
around a mean steady state, in which case the quantities of interest (drag, lift, heat
transfer coefficient, pressure drop, etc.) can be averaged over several oscillation
periods.
Transient flow calculations with SIMPLE
The use of this algorithm in transient calculations essentially consists of repeating
the steady-state SIMPLE calculations for each prescribed time step. The default
control parameter settings are therefore as summarised in Table 1-4.
.

Table 1-4: Standard Control Parameter Settings for Transient SIMPLE


Calculations

Variable
Parameter
Velocity Pressure Turbulence Enthalpy Mass fraction
Solver
0.1 0.05 0.1 0.1 0.1
tolerance
Sweep limit 100 1000 100 100 100
Relaxation
0.9 0.3 0.7 1.0 1.0
factor
Outer iteration limit = 5

The main difference compared to the PISO algorithm lies in the fact that all
linearizations and deferred correctors are updated within the outer iterations, by
recalculating the coefficient matrix and source term. For this reason, solver
tolerances do not need to be as tight as for PISO; they are actually identical to those
used for steady-state computations. However, since the discretization of the
transient term enlarges the central coefficient of the matrix in the same way as
under-relaxation does, the relaxation factors for velocities and scalar variables can
be increased (the smaller the time step, the larger the values that can be used for
relaxation factors — 0.95 or even more).
The convergence criterion for outer iterations within each time step is by default

Version 4.02 1-17


COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Numerical solution control

the same as for steady-state flows. However, the number of outer iterations is also
set to a default limit of 10; if substantially more iterations are needed to satisfy the
convergence criterion, this is a sign that the time step is too large. In such a case, it
is better to reduce the time step rather than allow more outer iterations for a larger
time step, because this would lead to a more accurate solution at a comparable cost.
On the other hand, if residuals drop below the limit after only a few iterations, one
may increase the time step; experience shows that optimum efficiency and accuracy
are achieved if 5 to 10 outer iterations per time step are performed.
Note also that the reported mass residuals are computed before solving the
pressure-correction equation; after this equation is solved and mass fluxes are
corrected, the mass residuals are more than an order of magnitude lower. For this
reason, one can accept mass residuals being somewhat higher than the convergence
criterion when the limiting number of outer iterations is reached, provided that the
residuals of all other equations have satisfied the criterion. In some cases, an
increase in the under-relaxation factor for pressure (up to 0.8) can lead to a faster
reduction of mass residuals. All these considerations are of course problem-
dependent and if several simulations over a longer period need to be performed, it
may prove useful to invest some time in optimizing the relaxation parameters.
Sometimes, it is necessary to select smaller time steps in the initial phase of a
transient simulation than those at later stages. This is the case, for example, when
starting with a fluid at rest and imposing a full-flow rate at the inlet, or full speed of
rotation (in the absence of a better initial condition). This is equivalent to a sudden
change of boundary conditions at a later time, which would also require that the
time step be reduced. Another possibility of avoiding problems with abrupt starts
from rest is to ramp the boundary conditions (e.g. a linear increase of velocity from
zero to full speed over some period of time).
The transient SIMPLE algorithm allows you to select either the default
fully-implicit Euler scheme or the three-time-level scheme for temporal
discretisation, described in Chapter 4, “Temporal Discretisation” of the
Methodology volume. The latter scheme is second-order accurate but is currently
applied only to the momentum and continuity equations. It should be chosen when
temporal variation of the velocity field is essential, e.g. in the case of a DES/LES
type of analysis. While PISO would normally be the preferred choice for the latter,
under some circumstances (e.g. the existence of very small cells, poor mesh quality
etc.), transient SIMPLE may allow the use of larger time steps than PISO without
loss of accuracy.
Effect of round-off errors
Efforts have been made to minimise the susceptibility of STAR-CD to the effects
of machine round-off errors, but problems can sometimes arise when operating in
single precision on 32-bit machines. They usually manifest themselves as failure of
the iterative solvers to converge or, in extreme cases, in divergence leading to
machine overflow.
If difficulties are encountered with problems of this kind, then it is clearly
advisable to switch to double precision calculations. Instructions on how to do this
are provided in the Installation Manual. As a general rule, however, you should try
to avoid generating very small values for cell volumes and cell face areas by
working with sensible length units. Alternatively, you could re-specify your

1-18 Version 4.02


Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Monitoring the calculations

problem geometry units while preserving relevant non-dimensional quantities such


as Re and Gr.
Choice of the linear equation solver
STAR-CD offers two types of preconditioning of its conjugate gradient linear
equations solvers: one which vectorises fully, and the other, which is numerically
superior to the first one but vectorises only partially. Therefore, the first one (called
‘vector’ solver) is recommended when the code is run on vector machines (such as
Fujitsu and Hitachi computers), and the second one (called ‘scalar’ solver) is
recommended if the code is run on scalar machines (such as workstations).

Monitoring the calculations


Chapter 5 and the section on “Permanent Output” on page 15-1 give details of the
information extracted from the calculations at each iteration or time step and used
for monitoring and control purposes. This consists of:
• Values of all dependent variables at a user-specified monitoring location.
Care should be taken in the choice of location, especially for steady-state
calculations. Ideally, it should be in a sensitive region of the flow where the
approach to the steady state is likely to be slowest, e.g. a zone of recirculation.
In transient flow calculations, the information has a different significance and
other criteria for choice of location may apply. For example, a location may
be chosen so as to confirm an expected periodic behaviour in the flow
variables.
• The normalised global residuals R φ for all equations solved. Apart from
turbulence dissipation rate residuals (see Chapter 7, “Completion tests” in the
Methodology volume), these are used to judge the progress and completion of
iterative calculations for steady and pseudo-transient solutions. In the early
stages of a calculation, the non-linearities and interdependencies of the
equations may result in non-monotonic decrease of the residuals. If these
oscillations persist after, say, 50 iterations, this may be indicative of problems.

Remember that reduction of the normalised residuals to the prescribed tolerance (λ)
is a necessary but not sufficient condition for convergence, for two reasons:
1. The normalisation practices used (see Chapter 7, “Completion tests” in the
Methodology volume) may not be appropriate for the application.
2. It is also necessary that the features of interest in the solution should have
stabilised to an acceptable degree.

If doubts exist in either respect, it is advisable to reduce the tolerance and continue
the calculations.
It follows from the above discussion that strong reliance is placed on the global
residuals to judge the progress and completion of iterative calculations of steady
flows. These quantities provide a direct measure of the degree of convergence of the
individual equation sets and are therefore useful both for termination tests and for
identifying problem areas when convergence is not being achieved.

Version 4.02 1-19


COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1
Model evaluation

Model evaluation
Checking the model
STAR-CD offers a variety of tools to help assess the accuracy and effectiveness of
all aspects of the model building process. In performing the modelling stages
discussed previously, the user should therefore take advantage of these facilities and
check that:
1. The mesh geometry agrees with what it is supposed to represent. This is
greatly facilitated by the built-in graphics capabilities that allow the mesh
display to be
(a) rotated,
(b) displaced,
(c) reduced,
(d) enlarged.

This enables the user to look at the mesh from any viewpoint, with the view
showing the correct three-dimensional perspective. Frequent mesh displays
during the mesh generation stage are very useful for verifying the accuracy of
what is being created and are therefore strongly recommended, particularly
for complex-geometry problems. It is best if such geometries are subdivided
into convenient parts that can be individually meshed and then checked
visually.
2. Materials of different physical properties occupy the correct location in the
mesh. This can be checked visually by using the built-in colour differentiation
scheme. Alternatively, each material’s mesh domain can be plotted
individually. Precise values of specified properties can be checked via the
screen printout.
3. Boundary conditions are correct, by producing special mesh views that show
(a) boundary location,
(b) boundary type,
(c) a schematic of the conditions applied (e.g. inlet velocities).

More complete information on specified boundary values can be obtained


from the screen printout.
4. The initial conditions should also be checked, particularly for transient
problems and initial fields specified through user subroutines, by running the
STAR-CD solver for zero iterations/time steps and plotting the relevant field
variables.

Checking the calculations


Having completed the model preparation, the next task is to run the STAR-CD
solver and check the results of the numerical calculations. These results are
presented in various ways, details of which are given in the Post-Processing User
Guide. Briefly, printouts and/or plots can be produced of the following:
• Field values of all primary variables at interior and boundary nodes.
• Interpolated values of the above quantities at arbitrary, user-specified points
or surfaces within the solution domain.

1-20 Version 4.02


Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES
Model evaluation

• Surface heat and mass transfer coefficients and forces; also values of the
dimensionless coordinate y+ for near-wall mesh nodes.
• Global quantities such as total force components (e.g. drag, lift) on
submerged bodies and their dimensionless counterparts, overall energy
balances, etc.

It is important to examine this information carefully to verify that the calculations


have been properly set up and are producing sensible results. In particular, the user
should ensure that:
• The interior fields are examined for plausibility and similar checks made on
global quantities.
• For turbulent flow calculations, the near-wall node y+ values are within the
recommended range (30-100) in regions where adherence to this constraint is
important. In the case of calculations with a two-layer model, checks should
be made that the mesh is sufficiently dense within the near-wall layer.
• The magnitude of numerical discretisation errors (spatial and, where
relevant, temporal) is assessed and arrangements made for their reduction to
acceptable levels, if necessary.
Of the above tasks, the last is currently the most difficult, for it is not possible to
achieve it by a simple calculation. What is required are the following:
• A reliable means of evaluating the discretisation errors. At present, this is
accomplished by repeating the calculations with finer meshes and smaller
time steps (strictly, these should be done independently) and noting regions of
appreciable change in the solution.
• Strategies for altering the mesh or time step to reduce errors. These
adjustments are made manually.

Ideally, the error correction process should continue until the changes fall to
acceptable levels. In practice, this approach may not be feasible, especially for
three-dimensional problems involving complex geometries, due to the large
preparation and computing overheads.
An alternative way of gaining some insight into the presence of spatial truncation
errors is to change the spatial discretisation scheme and note the effect on the
solution. The second-order options, or blends thereof, available in STAR-CD will
usually produce the lowest numerical errors.

Version 4.02 1-21


Chapter 2 BASIC STAR-CD FEATURES
Introduction

Chapter 2 BASIC STAR-CD FEATURES


Introduction
The main aim of this part of the manual is to provide users, whether experienced or
not in the application of general-purpose computational continuum mechanics
codes, with advice on effective ways of setting up and running a basic continuum
mechanics model using STAR-CD. The reader is, however, expected to have gone
through Chapter 1 and the material in the Methodology volume.
All aspects of user interaction are handled by pro-STAR, the pre- and
post-processing subsystem of the STAR-CD suite. As a pre-processor, pro-STAR
is the means by which the user defines the
• geometry,
• calculation mesh,
• boundary conditions,
• initial conditions,
• fluid and solid material properties,
• analysis controls,

which uniquely determine the problem to be solved. As a post-processor,


pro-STAR can
• read and re-format the various data files produced by the analysis,
• manipulate the data read in,
• produce extensive and easily comprehensible printouts,
• summarise information on the calculated results,
• draw sophisticated 3-D graphical images,
• animate those images,
• draw graphs of various calculated quantities.

Both pre- and post-processing operations are served by an extensive set of plotting
facilities, enabling rapid visualisation of even the largest models, plus on-line
context sensitive help that provides detailed information on usage.
pro-STAR is a combined command-, menu-, and process panel-driven program.
The choice of working interface is entirely up to the user and depends on
• whether the available terminal can accept and display graphical input and
output,
• whether the host computer’s operating system supports a windowed,
graphical user interface (GUI) environment,
• user preference and level of experience with STAR-CD.
GUI facilities are available for UNIX, Linux or Windows implementations of
STAR-CD using the OSF Motif graphics environment. They consist of two basic
types:
1. Graphical tools such as drop-down menus, dialog boxes, push-buttons,
sliders, etc. to assist users in specifying the desired pro-STAR actions. These
facilities are arranged around the main pro-STAR window, or have their
starting point located somewhere on that window. Their purpose and best way
of using them are explained throughout this volume.

Version 4.02 2-1


BASIC STAR-CD FEATURES Chapter 2
Running a STAR-CD Analysis

2. A series of process-oriented panels contained within the STAR GUIde


window. These represent an additional GUI facility, suitable for building
STAR-CD models from scratch. An outline description is given in the section
entitled “The STAR GUIde Environment” on page 2-38. Information on how
to use this environment is provided by an on-line Help system accessed from
within the STAR GUIde window.

Note, however, that:


• In the present release, a number of pro-STAR facilities are not accessible via
either of the GUI systems. Where this is the case, the discussion is in terms of
commands rather than GUI operations.
• For the convenience of users who prefer to work with commands, the
description of every GUI panel and dialog box also includes a list of
commands that have equivalent functionality. A summary of all pro-STAR
commands is given in Appendix B of the Commands volume. A summary of
pro-STAR’s conventions regarding command syntax can be found in this
volume, Appendix A. The same information is also available on line by
choosing Help > pro-STAR Help from the menu bar in the main pro-STAR
window and then selecting item PROGRAM (for command syntax) or
COMLIST (for command summary) in the scroll list at the bottom of the
Help dialog box.
• Details of all available commands and specific aspects of the command-driven
mode of operation are discussed in the Commands volume.

Whichever operating mode is chosen, the same principles of use apply, namely:
• A model is constructed or examined with the aid of numerous functions or
‘tools’, each of them represented by a menu-item choice, a special dialog box,
a STAR GUIde panel or a command.
• Tools are selected as necessary, in a sequence that is sensible for modelling
purposes. The recommended sequence is described in Chapter 1, “The Basic
Modelling Process” and is further elaborated in the Tutorials volume.
• A tool always provides instant feedback so the user can tell immediately if it
was used properly.
• Users can greatly influence the speed with which certain operations are
performed by intelligent use of the available options.

Running a STAR-CD Analysis


A STAR-CD analysis may be performed in one of the following two ways:
• By typing a series of script names in a shell or command prompt, each
designed to help you build a CFD model, obtain a solution and then display
the analysis results. This is the original method of working with STAR-CD
and, for reasonably experienced users, may be the quickest way of getting
results.
• By employing a new utility, STAR-Launch, as an aid to navigating through
the various STAR-CD functions. This method should be particularly
beneficial to novice users.

2-2 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Running a STAR-CD Analysis

Using the script-based procedure


To perform the analysis using scripts, the procedure described below should be
followed in the order indicated:
Step 1
Set up an appropriate environment for your STAR-CD system. The desired
pro-STAR setup is defined by a number of environment variables such as:

STARUSR — path to the location of files PRODEFS (for command


abbreviations) and PROINIT (for pro-STAR initialisation) — see
Chapter 16, “Set-up Files”
MACRO_LOCAL and MACRO_GLOBAL — paths to the local and global macro
locations (see Chapter 16, “Macros”)
PANEL_LOCAL and PANEL_GLOBAL — paths to the local and global
user-defined panel locations (see Chapter 16, “Panel definition
files”)
TMPDIR — path to the location of pro-STAR’s temporary (scratch) files

Further instructions on how to set the STAR-CD environment variables are given
in the Installation and Systems Guide, supplied with the STAR-CD installation
CD-ROM. Note that these settings can usually be made once and for all, at the time
when STAR-CD is first installed on your computer.
Step 2
Create a separate subdirectory for each case to be analysed and give it a descriptive
name. This helps to organise the various files created during a run and makes it
much easier to check or repeat previous work.
Step 3
Move to the appropriate subdirectory and start a pre-processing (model building)
session by typing:

prostar

The system will respond by prompting you to define the pro-STAR variant you wish
to use

Please enter the required graphics driver


Available drivers are:
x, xm, glm, mesa [xm]

where the options refer to the various types of graphics libraries commonly used for
graphical displays in workstations or X-terminals, i.e.

x — X-windows
xm — X-windows using the Motif interface for pro-STAR’s GUI
functions
glm — Motif interface plus the standard OpenGL libraries for
pro-STAR’s GUI functions
mesa — As above but using the Mesa OpenGL library (this option is not
available in Windows ports)

Version 4.02 2-3


BASIC STAR-CD FEATURES Chapter 2
Running a STAR-CD Analysis

The precise list of options displayed by the prompt depends on how the pro-STAR
environment was originally set up on your particular machine. Type in a response
that is appropriate to the workstation or terminal you are using.
Note that pro-STAR automatically searches for the highest depth pseudo colour,
direct colour or true colour visual that exists for your screen and uses it. This may
be overridden by specifying option -c when starting up pro-STAR, as shown
below:

prostar -c

This is an 8-bit pseudo colour setting with shared colour map. The setting causes no
screen flashing but requires sufficient available colours to work.
Once the desired pro-STAR variant has been chosen, an introductory panel
opens up leading you into STAR-CD’s model-building environment, as discussed
in the section on “pro-STAR Initialisation”. From that point on, you may provide
input for setting up your model according to the descriptions given in the remaining
chapters of this manual.
Step 4
When you have finished setting up your STAR-CD model, it is advisable to check
the files created so far in your working directory. These should include:
• File .mdl, containing all user-supplied information about the model
• File .ccm, containing a full description of the model geometry. The
STAR-CD solver operates only in SI units and all dimensions must therefore
be defined in metres. However, it is possible to scale the mesh dimensions by
a scaling factor if non-SI units were used during mesh generation.
• File .prob, containing problem data, such as material properties, boundary
conditions, control parameters, etc.
• File .echo, containing a log (echo) of all instructions issued to pro-STAR
during the session

Depending on the nature of your problem (e.g. whether it requires a special


modelling facility such as Lagrangian multi-phase) additional files may be created.
These are discussed fully in individual chapters of this volume dealing with such
topics. A detailed description of all commonly used data files is given in Chapter
17, “Commonly used files”.
Step 5
If user-defined subroutines are not required, go to Step 6.
Otherwise, create a subdirectory called ufile and place your subroutine files
in it. The most convenient way of doing this is to create both the subdirectory and
the files from within the pro-STAR session (see Chapter 14, “Subroutine Usage”).
Note that these files contain default (dummy) code to start with and you should edit
them as necessary to insert your own code.
Step 6
Based on the geometrical and physical data of the model just created, you are now
in a position to run STAR. This may be done in one of the following ways:
1. Via STAR-GUIde’s “Run Analysis Interactively” panel. Examples of using
this panel are provided in the Tutorials volume. This way, the STAR

2-4 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Running a STAR-CD Analysis

executable will be run automatically and the analysis results (in terms of
solution residuals) will be displayed on a separate window; see “The
StarWatch Utility” on page 17-15 for more details.
2. By exiting from pro-STAR and then running STAR from your session’s shell
or command prompt. For a large number of cases, it will be sufficient to type
one command. For a single-precision run, type:

star

whereas for a double-precision run, type:

star -dp

Please note that it is not necessary to provide the case name of the model you
are running. However, for better bookkeeping, it is still important to keep
every case in its own directory.
In most cases, and based on the model characteristics specified in
pro-STAR, STAR automatically recognises the default run-time requirements
and proceeds with the CCM analysis without further user input. Some cases,
however, require the specification of additional options related to both
run-time resources and/or behaviour. Briefly, the user can control the
operational behaviour of STAR in one of the following areas:
• Job precision (single or double precision)
• Job control (to abort, kill or restart a job)
• Environment (to export environment variables)
• User coding (to control the compilation and/or linking of user-supplied
code)
• Parallel setup (pertaining to domain decomposition variations, data
distribution and parallel communication libraries)
• Resource allocation (to choose which machines to use)

A full list of such options can be obtained by typing:

star -h

or

star -help
The listing will also contain a short description of each option’s purpose. A
more complete description can be found in Appendix F of this manual.

Please note that, in general, one needs to specify the machine (node)
resources for running STAR and this input is automatically used to determine
the type of run required. The following examples illustrate this point:

star Runs sequentially on the local node

star origin Runs sequentially on a host called origin

star 4 Runs in parallel with 4 processes on the local node

Version 4.02 2-5


BASIC STAR-CD FEATURES Chapter 2
Running a STAR-CD Analysis

star origin,16 Runs in parallel with 16 processes on a host called


origin

star cheese,2 pickle,2 curry,2 rice,2


Runs in parallel on a cluster of 4 machines with 2 processes each

Please note that, for parallel cases, the computational domain decomposition
is automatically handled by the star front-end script. The output files
generated during the course of the run will be merged and placed in the case’s
directory. There is then no visible difference between running in sequential
and running in parallel.
Extra options exist to cater for special situations which cannot be detected
automatically. Please refer to Appendix F for a list of such options, their
syntax and their intended purpose.
Step 7
Once the run starts, iteration or time-marching continues until one of the following
conditions is met:
• All the iterations or time steps specified for the current run have been
completed.
• The normalised residual sum drops below a specified value (steady-state runs
only).
• The solution starts to diverge. This occurs when a residual anywhere inside
the solution domain reaches a very high value or a numeric overflow
condition. Divergence is automatically detected by STAR, which then stops
the calculations and writes a file with extension .div. This is identical in
format and content to a normal solution data (.ccm) file and thus enables you
to inspect the residuals and identify the mesh location(s) where numerical
instability has occurred.

Check the condition under which your run has terminated. The parameters involved
in controlling the STAR-CD simulation are set in pro-STAR using the facilities
provided by the “Analysis Controls” folder in STAR GUIde. Additional
information, such as printout of input data, boundary conditions, residual histories
of the inner iterative loops, etc. can also be generated, as described in Chapter 15.
Step 8
At the start of the analysis, STAR will read the following files:
• case.ccm — geometry data (plus solution data for restart runs)
• case.prob — problem data

and, optionally, one or more problem-dependent files such as


• case.vfs — view factors for radiation problems
• case.evn — transient event data
• case.drp — droplet data

On completion of the run, file case.ccm will contain the current analysis results
in a form suitable for post-processing or for starting another STAR run. A number
of additional files will also be present in your working directory, including:

2-6 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Running a STAR-CD Analysis

• case.run — summary of input data plus numerical statistics and (optional)


printout of solution variables
• case.info — STAR warning messages and (optional) additional
numerical statistics
• case.rsi — Solution residuals, in a form that can be displayed graphically.

The above is the minimum number of output files created by a STAR run and you
should confirm that they are all present. Additional files may appear depending on
the nature of the problem. Such cases are discussed and explained individually in
the relevant chapters of this volume. A description of all commonly used output
files appears in Chapter 17, “Commonly used files”.
Note that, at the beginning of every restart run, all current results files (such as
the ones listed above) are automatically saved in a local sub-directory called
RESULTS.xxx, where xxx stands for the run number. These sub-directories thus
contain results obtained at the end of each successive run and are available for future
inspection, or as a backup in case the restart run’s files are corrupted. If the case is
subsequently run from initial conditions, the results of the last run performed are
stored in sub-directory RESULTS.000 and all other RESULTS directories deleted.
The process then repeats itself with the creation of a new RESULTS directory for
each new restart.
Step 9
You should now check the results of the analysis by looking at the run history
(.run) file (see Chapter 15 for more information on its contents). The additional
information (.info) file should also be examined for any signs of numerical
problems. These are normally translated into warning messages. Both these files
may be inspected via a suitable text editor or via panel “Run History of a Previous
Analysis” in STAR-GUIde.
Satisfactory completion of steady-state STAR runs can usually be judged by
observing the following quantities:
• The residual history printed during the run. The sum of the normalised
absolute residuals should diminish steadily.
• The monitoring values of the dependent variables at a critical location within
the solution domain. These should stabilise to the converged solution.

In transient calculations, completion is defined in terms of the elapsed (simulation)


time or establishment of a steady state. In the latter case, information on the global
change and monitoring values can be used in the same way as for a steady state
analysis.
It is important that checks are made regularly during the initial stages of the
analysis to monitor the solution progress. If divergence occurs, the run should be
terminated and appropriate adjustments made to the relevant control parameters
such as under-relaxation factors. Neglecting this can result in costly and
unproductive runs. Note, however, that increases in residuals and oscillations in the
computed variables during the early stages of a run are not uncommon and should
disappear after a few iterations. The run should therefore be given sufficient time to
stabilise before any judgement is made on its progress.
Step 10
Continue with an evaluation of the simulation results (post-processing) using the
Version 4.02 2-7
BASIC STAR-CD FEATURES Chapter 2
Running a STAR-CD Analysis

relevant facilities in STAR-GUIde. If you have previously exited from pro-STAR


and run STAR separately (see Step 6 above), continue by typing

prostar

to re-enter pro-STAR. Reply as before to the initial prompt

Please enter the required graphics driver


Available drivers are:
x, xm, glm, mesa [xm]

and then supply the case name and other input, as described in Step 3.

Using STAR-Launch
STAR-Launch is a graphical interface that provides access to most of the CD-
adapco modelling tools, including pro-STAR, several es-tools and the STAR solver.
Using STAR-Launch eliminates the need to enter multiple script names manually,
as described in the previous section, and also ensures settings can be saved between
sessions and between cases. STAR-Launch is intended to be used with only one
case at a time. There is, however, no limit on the number of STAR-Launch windows
that can be active simultaneously.

Activating STAR-Launch
On Unix/Linux
Either double-click the appropriate icon on your desktop (for systems which support
this), or else type

starlaunch &

in an appropriate X-terminal window. This will display the STAR-Launch main


window shown below:

On Windows
Double-click the appropriate icon on your desktop.

Window layout
The key parts of the STAR-Launch main window are highlighted below. The

2-8 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Running a STAR-CD Analysis

Shortcut Buttons provide quick access to the three main functions of STAR-Launch,
namely:
• Setting the working directory
• Launching a pre-/post-processing tool
• Running the STAR solver

These functions are also accessible through the Main Menubar running along the
top of the window. The current working directory is displayed to the right of the
Shortcut Buttons. This is the directory that will be used when launching a
pre-/post-processing tool or running the STAR solver.

Shortcut Buttons

Main Menubar Current Working Directory

Workspace for Process Output

Run STAR Interactively


Launch Pre-Post Tool
Set Working Directory

Setting the working directory


Choose File > Set Working Directory or click the first shortcut button on the main
window. This will display a directory browser as follows:

Version 4.02 2-9


BASIC STAR-CD FEATURES Chapter 2
Running a STAR-CD Analysis

Navigate to the desired directory and click OK. Note that a path can be entered
manually in the Look In entry box at the top of the browser window. The directory
tree will be updated to reflect any valid path entered here.
The path that will be set on clicking OK is shown along the base of the browser
window.

Starting a pre-/post-processing tool


To start a pro-STAR session, or an equivalent pre-/post-processing tool, select the
appropriate entry in the Pre-Post menu, or click the second shortcut button on the
main window. The tool that will be started from this button is set using the Pre-Post
tab of the Preferences dialog. Only tools available in the current installation will be
listed in the Pre-Post menu.
STAR-Launch will open a new Process Output window as shown below, which
will contain any text generated by the Pre-/Post-processing tool as it starts up.

The STAR-Launch window can be resized as necessary to display more of the text
appearing in the Process Output window.
Only one pre-/post-processing tool can be running at any one time. If an attempt
is made to start another one, a prompt will appear asking if the existing tool should

2-10 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Running a STAR-CD Analysis

be closed. Choosing Yes will kill the existing process, which could result in loss of
any unsaved data.
When a process is active, the ball appearing in the Process Output window tab
will be shaded red. This will change to black when the process is finished.

Running STAR interactively


Selecting Solver > Run Star Interactively, or clicking the third shortcut button,
will display the Run Star Interactively dialog shown below. The dialog provides
several options for running the STAR solver; detailed information on these options
can be found in the STAR-Launch On-line Help, accessed from Help > Online
Manual. When all settings have been made, the solver is started by clicking Run.
STAR output will appear in a new Process Output window, similar to the one
shown above for the Pre-/Post-processing tool. When the STAR solver finishes, the
ball on the tab of the output window will turn black.
Note that only one STAR solver can be run at any one time from a STAR-Launch
session. If multiple solver processes are required, more STAR-Launch sessions
must be opened.

STAR-Launch project files


.starlaunch directory and launcherGlobal.xml
When STAR-Launch is first used, it will attempt to create a hidden directory,
.starlaunch in the users home directory (as given by $HOME). Within this
directory, STAR-Launch will write file launcherGlobal.xml. The file is
normally written on exit from STAR-Launch and contains details of the last
working directory specified by the user. It also stores a flag indicating whether this
stored path is to be used automatically in a new session.

starProject.xml
Another file, starProject.xml, can be written by STAR-Launch if requested
by the user. This stores settings from the Preferences and Run Star Interactively
dialogs. The various File menu options affecting this are explained below:
• File > Open Project — This presents a file browser that should be used to
find the required starProject.xml file. STAR-Launch will be updated to

Version 4.02 2-11


BASIC STAR-CD FEATURES Chapter 2
pro-STAR Initialisation

reflect settings from the new file.


• File > Save Project — This will write a starProject.xml file in the
current working directory. Settings within the file will reflect the current state
of STAR-Launch.
• File > Save As Default — This will write a starProject.xml file in the
hidden .starlaunch directory within the user’s home directory.

STAR-Launch start-up procedure


When STAR-Launch is first started, it will look for the launcherGlobal.xml
file in the hidden .starlaunch directory. This will be read to determine the
initial working directory. If a starProject.xml file is also contained in the
hidden .starlaunch directory, STAR-Launch will read all settings within the
file, and use these to configure the initial state of the GUI. If a
starProject.xml file is also found within the initial working directory,
STAR-Launch will read the settings within that file, and use these to update the
initial state of the GUI. Settings contained in a local starProject.xml file (i.e.
one within the initial working directory) will always take precedence over settings
obtained from a starProject.xml file in the hidden .starlaunch directory.

Preferences dialog
Selecting File > Preferences... will display the Preferences dialog shown below:

The options contained here are explained fully in the STAR-Launch Online Help
(Help > Online Manual). Their state will be saved in the starProject.xml
file.

pro-STAR Initialisation
Once the basic GUI mode of operation has been chosen (x, xm, glm or mesa, see
“Running a STAR-CD Analysis”, Step 3 above, or via the Preferences dialog in
STAR-Launch), the introductory panel shown below appears. The following three
optional inputs may be provided:
1. The desired case name — star is the default name assigned to the current
problem at the start of a pro-STAR session. Overtype this by the correct name
in the Case Name text box. Note that:
(a) If a model already exists in your present working directory, its name will
be picked up automatically by pro-STAR.
(b) If you have more than one model, you may choose the desired one by
clicking on the file selection icon next to the Case Name box. This
2-12 Version 4.02
Chapter 2 BASIC STAR-CD FEATURES
pro-STAR Initialisation

activates a File Selection browser (see page 2-33) that enables you to
choose the desired model, stored in a file of form case.mdl

2. The Resume mode — This can be either a restart from an existing model
definition, via its corresponding model (.mdl) file or a brand new case. Clear
this option if the latter applies.
3. The Append mode — The session’s user input will be appended to an existing
log or echo (.echo) file or a new echo file will be created. Clear this option if
the latter applies.

Refer to the description given in Chapter 17, “Commonly used files” for a definition
of pro-STAR’s model and echo files.
Click on Continue to display the basic pro-STAR GUI windows or Exit to abort
the current session.

Two windows are displayed automatically immediately following the initialisation


stage. These are described in the sections entitled “Input/output window” below and
“Main window” on page 2-15.
Input/output window
This window, shown on the next page, consists of the following three sub-windows,
in top-to-bottom sequence:
1. Command Output — displays the time and date of the run, plus summary
data for the model in hand, if such data were read in from a Restart file at the
initialisation stage. All subsequent output in that window are the echo of
every instruction issued by the user plus pro-STAR’s response to it. The latter
serves as feedback to help determine whether a facility was used properly.
2. Command Input — accepts pro-STAR instructions in the conventional
‘Command keyword plus parameters’ format described in the pro-STAR
Commands volume. Thus, it is possible to work in ‘command’ mode at any
stage of the model building process despite the fact that the GUI version of
the code is active. This is useful when working with facilities that cannot be
activated from a GUI panel or dialog box in the present pro-STAR version.
This sub-window can be re-sized by dragging the control ‘sash’ (the small
square at the top right-hand corner) up and down.
3. Command History — provides a numbered ‘command history’ list that keeps
track of all pro-STAR instructions issued in the current session, either as

Version 4.02 2-13


BASIC STAR-CD FEATURES Chapter 2
pro-STAR Initialisation

choices from a menu in the main GUI window (see “Main window” on page
2-15) or as commands typed in the sub-window above. Menu choices are
translated into their equivalent commands before being added to the list. The
list can be used in the following two ways:
(a) Single-click the command number to copy a command into the Input
window and then edit it.
(b) Double-click the command number for immediate re-execution.

The Command History sub-window can be re-sized by dragging its control


‘sash’ up and down.

Note that:
1. The Command Input sub-window can accept multiple commands by cutting
and pasting from the window of another application (e.g. a text editor). If any
of the imported command text needs editing prior to execution,
(a)click the Pause action button under the window (see the above panel)
(b)paste in the required group of commands
(c)make the necessary changes
(d)click the Pause action button again to allow pro-STAR to begin executing
the commands one by one
2. The Command History sub-window will normally list all commands issued to
pro-STAR, including those generated indirectly via an external command file
(see Chapter 17, “Commonly used files”) or a user macro (see Chapter 16,

2-14 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
pro-STAR Initialisation

“Macros”). It will also list details (e.g. coordinate values) of items such as
vertices, splines, cell faces, etc. that are directly picked from the main window
display with the mouse. Such output may become extremely voluminous and
may thus obscure the record of primary operations performed by the user.
Clicking the Short Input History button will prevent this and will cause
pro-STAR to list only the instructions directly issued by the user.
Main window
The main GUI window, shown below, is used for the following purposes:

• For graphical display of various aspects of the current model.


• As a launch pad for those pro-STAR utilities that are available in GUI form.
The user should click one of the eleven drop-down menus appearing in the
menu bar and select one of the displayed choices. Commonly used functions
affecting the model display in the graphics area are also implemented, in the
form of action buttons. These are distributed along the top and left-hand-side
borders of the window and are described in Chapter 4 of the Meshing User
Guide. Letting the mouse rest on top of any button causes a brief explanatory
legend to appear in a special window provided for this purpose.
• To show messages for the user, such as prompts to supply data, in the space
underneath the graphics area. The default display shows:
(a) The current plot parameters (see “Plot Characteristics” on page 4-3 of the
Version 4.02 2-15
BASIC STAR-CD FEATURES Chapter 2
pro-STAR Initialisation

Meshing User Guide).


(b) A clock display showing the current time and date. This may be turned on
or off by selecting Show Clock or Hide Clock from the Utility menu.
(c) Three status indicators showing the result of processing the latest
command, irrespective of whether it was typed in directly or issued via a
GUI operation. The indicators are arranged as a set of ‘traffic’ lights
whose significance when lit is as follows:
i) Green — the command was executed successfully. The displayed
message is

Command: <Command Name> is Done

ii) Amber — the indicator flashes to signal the presence of warning


messages in the Output window. The displayed message is

Command: <Command Name> has a Warning, check


the output window

iii) Red — the command has failed. The displayed message is

Command: <Command Name> has an Error. Click


on the red light to view the error

Clicking on the red light displays an Error/Warning Summary


pop-up window with more information on what has gone wrong, as
discussed under “Error messages” below.

Note that if a GUI operation generates a series of commands, a message is


issued for each one in turn as soon as it is processed. If all goes well, the
message finally seen on the screen is for the last command that was
executed.
The menu bar
The menu bar items are listed below, along with a reference to chapters containing
a detailed description of their functionality:
1. File
Provides all basic housekeeping utilities, including those related to
input/output operations — see Chapter 17, “File Handling”.
2. Tools
Activates dialog boxes that allow definition and manipulation of basic
pro-STAR entities (cells, vertices, splines, etc.). Most of these are covered in
Chapter 2 of the Meshing User Guide. Another type of tool facilitates
routinely-used, complex operations such as colour selection and mesh surface
lighting effects (see Chapter 4, “Colour settings” in the Meshing User Guide).
3. Lists
Displays lists of all available entities of a certain type (cells, vertices,
boundaries, etc.) as well as those currently grouped into a user-defined set.
4. Modules
Accesses special dialog boxes that set up various STAR-CD model

2-16 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
pro-STAR Initialisation

parameters in connection with


(a) Animation control — see Chapter 12 of the Post-Processing User Guide
(b) Transient condition definition — see Chapter 5, “Load-step based
solution mode”

5. Plot
Contains most of the facilities and options used for mesh plotting operations
— see Chapter 4 of the Meshing User Guide.
6. Post
Displays the results of a STAR run — see Chapter 1 of the Post-Processing
User Guide.
7. Graph
Produces various types of graph — see Chapter 14 of the Post-Processing
User Guide.
8. Utility
Provides miscellaneous utility functions designed to aid model control and
development, such as calculation of cell volumes and distance between
vertices — see Chapter 3, “Mesh and Geometry Checking”in the Meshing
User Guide. It also supports special user-controlled operations, such as the
assignment of user-defined functions to keyboard keys.
9. Panels
Allows you to set up your own screen buttons or panel tools for performing
common pro-STAR operations — see Chapter 16.
10. Favorites (optional)
This menu appears only if you have chosen any ‘favourite’ (i.e. frequently
used) panels in the STAR GUIde tree structure (see “Panel navigation
system”). The relevant panels are listed under this menu, enabling you to
jump to them directly.
11. Help
Displays pro-STAR command help information in a scrolled-text fashion.
Also contains on-line versions of the STAR-CD manuals and tutorials.

A mouse click on any of the above menu names displays a drop-down list. In
general, clicking an item on the list starts up the action indicated, unless the name
is followed by
• an ellipsis (…) which means the item displays a new dialog box, or
• an arrow (⇒) which means the item opens a secondary list with more items to
choose from.

Throughout this manual, the “>” sign denotes successive mouse clicks on menu
names, menu list items, dialog box buttons, etc. For example,

Tools > Cell Tool > Edit Types

means click Tools in the menu bar, then click the Cell Tool item in the drop-down
list, then click the Edit Types button on the displayed Cell Tool dialog.

Version 4.02 2-17


BASIC STAR-CD FEATURES Chapter 2
General Housekeeping and Session Control

General Housekeeping and Session Control


When pro-STAR is initially installed on a computer system, default settings are
provided for the program’s fundamental operating features. These settings,
specified mostly via commands typed in the Command Input window, can be
altered in special circumstances. The following aspects of the program’s operation
are covered:
Basic set-up
These settings are helpful in establishing an appropriate environment for pro-STAR
and for accessing facilities related to the operating system of the host machine. They
are as follows:
1. Operating mode — command BATCH disables pro-STAR’s periodic prompts
to stop or continue displaying long lists of data.
2. pro-STAR size — command SIZE lists the maximum number of cells,
vertices, boundaries, etc. that the code can handle. If any of these values is
inadequate for the model in hand, it may be increased by following the
procedure described in Chapter 17, “Resizing pro-STAR”.
3. Reporting cpu time required to complete a pro-STAR function by typing
command TPRINT.
4. Accessing special, user-written pro-STAR subroutines by typing command
USER. It is advisable to use this facility only after consultation with
CD-adapco.
5. Communicating with the operating system itself. This may be done by first
choosing File > System Command from the menu bar to display the System
Command dialog box shown below and then typing system commands in its
text box.

Command: SYSTEM

This is useful for issuing instructions to the host operating system without
having to exit from the pro-STAR environment.

Screen display control


There are several facilities for controlling the screen display during a session, as
follows:
• Defining the layout and look of the pro-STAR windows. Default settings are
normally used for these but the user can override them at will, as explained in
Chapter 16, “Set-up Files” and also in Appendix D.
• Switching from the terminal’s graphics screen to the text screen via command
TEXT. This is applicable only when running a non-GUI version of pro-STAR

2-18 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
General Housekeeping and Session Control

and is used for controlling terminals that operate entirely either in text or in
graphics mode.
• Setting the number of lines that appear on each ‘page’ of the Command
Output window during lengthy listings using command PAGE.
• Displaying a history of the most recent commands issued during the session
via command HISTORY. Again, this applies only when running non-GUI
versions of pro-STAR since these do not provide a command history window.
• Echoing the user input stream to the same device as the output stream (e.g. the
screen or a disk file) via command ECHOINPUT.
• Reading stored cursor picks from an input file, rather than displaying a
crosshair cursor and reading the user-specified picks off the screen —
command CURSORMODE.
• Providing a descriptive title for the current model that helps to identify each
plot produced subsequently — choose File > Model Title from the menu bar
to display the dialog box shown below. The desired title and up to two lines of
subtitle text should be typed in the text boxes provided.

Command: TITLE

Error messages
pro-STAR issues error messages as a result of receiving incorrect commands or if
it is unable to execute a valid command for whatever reason. Such messages appear
in three places:
• On the standard Output window
• At the bottom of the main pro-STAR window, after the red indicator light (see
page 2-16)
• On the Error/Warning Summary pop-up panel, as in the example shown
below:

Version 4.02 2-19


BASIC STAR-CD FEATURES Chapter 2
General Housekeeping and Session Control

The panel shows a list of all current errors, including their error id and command in
which the error occurred. If you know the cause of the problem, click Clear to close
the panel. Otherwise, select any item in the list to see the error description at the
bottom of the panel.
Command SUCCEED, typed on its own, tells you (in the output window) whether
the previous command produced any errors or warnings. On the other hand, typed
as SUCCEED, QUIT, it will immediately terminate the pro-STAR session. This can
be useful when pro-STAR is run in batch mode, where it is not desirable for the job
to continue after an error.
Error recovery
If mistakes are made during a session, the following operations are useful for error
recovery:
• Re-executing a named range of previously issued commands by typing
command RECALL. This can be most conveniently used in conjunction with
the HISTORY command above.
• Retrieving the state of the model description as it was at the time of the
previous SAVE or RESUME operation — command RECOVER. This is
useful if a mistake is made but the user does not notice it until some time later.
A list of commands issued since the last SAVE or RESUME operation is
displayed, along with a prompt to choose the last command in the list to
re-execute. The chosen command will normally precede the one where the
mistake was made. Once all commands up to that point are re-executed, the
user should type in a correct command and carry on from there.
• Note that the above safety features can be switched off using command
SAFETY. This might speed up pro-STAR execution but at the potential cost of
making any sort of recovery from mistakes nearly impossible. Thus, turning

2-20 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Set Manipulation

off these features should be used with extreme caution.

Session termination
The current pro-STAR session is terminated by choosing File > Quit from the menu
bar. This displays the Quit pro-STAR dialog box shown below, reminding you to
save the results of the session to a .mdl file (in case this has not already been done
explicitly). Alternatively, you may deliberately exit from pro-STAR without saving
the present session’s work, by clicking Quit, Nosave.

Command: QUIT

Set Manipulation
pro-STAR has extensive facilities for collecting and modifying sets of objects.
These are accessible by clicking one of the coloured buttons down the left-hand side
of the main window. The pro-STAR entities serviced by the buttons are:
• C-> — cell sets
• V-> — vertex sets
• S-> — spline sets
• Bk-> — mesh block sets
• B-> — boundary sets
• Cp-> — couple sets
• D-> — droplet sets

Each button offers a wide range of possibilities to select, delete or re-select sets. For
example, selection may be done by picking all objects falling within a given
geometric range in a local coordinate system. Using other criteria, one can collect
together all cells or boundaries connected to the current vertex set (and the reverse).
Selection can also take place by simply using the screen cursor to point to items on
the current plot.
Each button gives direct access to the following set manipulation options:
• All — select the entire set
• None — empty out the current set
• Invert — invert the current set, i.e. select all entities that are not currently
selected and un-select the ones that are
• New — replace the current set with a new set, formed on the basis of a
criterion given in a secondary drop-down list
• Add — add more members to the current set, selected using one of the criteria
in the secondary drop-down list
• Unselect — remove some members from the current set, selected using one

Version 4.02 2-21


BASIC STAR-CD FEATURES Chapter 2
Set Manipulation

of the criteria in the secondary drop-down list


• Subset — select a smaller group of members from those in the current set,
selected using one of the criteria in the secondary drop-down list

In addition, C-> offers one extra option, Surface, which selects all cells lying on
the surface of the most recent mesh plot and makes them the current set. Further
details on the above set selection options are given in Chapter 2 of the Meshing User
Guide, for each of the mesh entities described there.
Note that it is possible to save and restore useful cell, vertex, spline, block,
boundary and couple sets without the need to rebuild them frequently. This is done
by clicking the INFO button at the left-hand side of the main pro-STAR window.
The following operations are possible:
1. To perform a ‘save set’ operation, select INFO > Store Set/Surface/View
and then click the Sets tab to display the dialog shown below:

Commands: SETWRITE SETDELETE

The input required is as follows:


(a) Set File — The name of the set (.set) file that will store the set
definition. If such a file already exists, pro-STAR’s built-in file browser
may be used to help locate it.
(b) Name — An identifier for the set being saved, up to 80 characters long

Click Write to save the set definition.


2. To delete a set definition previously stored, use the same dialog as above and
specify the following information:
(a) Set File — The name of the set (.set) file containing the definition to be
deleted. pro-STAR’s built-in file browser may be used to locate it.
(b) Select Entry — The location of the set to be deleted, as select from the
list.

Click Delete to delete the set definition.

3. To perform a ‘restore set’ operation, select INFO > Recall Set/Surface/View

2-22 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Set Manipulation

and then click the Sets tab to display the dialog shown overleaf:

Command: SETREAD

The input required is as follows:


(a) Set File — The name of the set (.set) file containing the set definition.
pro-STAR’s built-in file browser may be used to help locate it
(b) Select Entry — Select the particular set required by name from the scroll
list. The status of the selected entry is displayed in the box underneath
(c) Choose Data — Specify the type of set to be read in (All, Cells, Vertices,
etc.) by clicking one of the displayed option buttons
(d) Read Option — Specify how the sets to be read in will modify any
existing sets by selecting one of the menu options (Newset, Add,
Unselect or Subset)

Click Recall to recall the selected set.

Note that it is possible to print a summary of all data sets stored so far by
typing command FSTAT.

Selecting sets of various entities has two major uses:


1. To display only items in the currently active set. For example, each time Cell
plot is chosen from the Plot menu, pro-STAR plots only cells in the currently
active cell set. Note that command SETADD causes all newly-defined cells to
be automatically added to the current set. Thus, successive plots of the current
state of the mesh can be made without needing to build a new set after each
new cell definition. SETADD may also be used in the same way for other
kinds of sets, i.e. boundaries, cell couples and splines.
2. To perform almost any modelling or post-processing operation on the

Version 4.02 2-23


BASIC STAR-CD FEATURES Chapter 2
Table Manipulation

currently active set, instead of on individual objects or a range of them. For


example:
(a) Choosing Lists > Cells from the menu bar and clicking the Show Cset
Only option button will list only cells in the current set.
(b) When working with commands, typing
VMOD,VSET,2.5
will modify the X-coordinate of every vertex in the current vertex set.

All set operations can also be performed by typing commands CSET, VSET, BSET,
BLKSET, CPSET, SPLSET and DSET. These are described in detail in the
pro-STAR Commands volume.

Table Manipulation
pro-STAR tables are multi-variable entities akin to spreadsheets and can be used to
store values for up to 100 dependent variables as functions of a combination of
several independent variables. For most commonly used tables, the independent
variables can be the three spatial coordinates, plus time for transient cases or
iteration for steady-state cases. The dependent variables are normally flow field
solution variables but, in principle, they could be anything of relevance to
STAR-CD.
Basic functionality
At present, tables are used principally as a substitute for user subroutines in the
following situations:
• Boundary Conditions — variable conditions along the surface of a boundary
region; see Chapter 4, “Boundary Region Definition”, page 4-7. For most
boundary types, the independent variables may be any combination of spatial
coordinates and, for transient cases, time. The only exception is outlet
boundaries where only time is allowed (i.e. there can be no spatial variation in
outflow conditions along the outlet surface). The permissible dependent
variables vary according to the boundary type considered; a full list is given
under the various boundary type descriptions in Chapter 4, or the
corresponding on-line Help topics for STAR GUIde’s “Define Boundary
Regions” panel.
• Initial Conditions — non-uniform initial distributions of field variables; see
Chapter 4, “Solution Domain Initialisation”. The independent variables may
be any combination of spatial coordinates, for both steady and transient cases.
The permissible dependent variables for fluid materials are listed under topic
“Initialisation”. Scalar variables representing chemical species mass fractions
may also be initialised, as described in a separate topic for scalar
“Initialisation”. Note that:
(a) The applicability of field variable and scalar initialisation tables can be
restricted to a selected domain or a cell type
(b) The only dependent variable allowed for solid materials is temperature

• Source Terms — a description of mass, heat, momentum or scalar species


sources; see Chapter 3, page 3-8. The independent variables may be any

2-24 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Table Manipulation

combination of spatial coordinates and time for transient cases, or iteration


number for steady-state cases. The permissible dependent variables vary
according to the source type considered; a full list is given in the on-line Help
topics for the various sources definable via panel “Source Terms”. Note that,
as with initial conditions, the applicability of source tables can be restricted to
a selected domain or a cell type.
• Rotational Speeds — variable angular velocity in rotating systems, specified
in panel “Rotating Reference Frames”. The independent variable is time for
transient cases, or iteration number for steady-state cases. The dependent
variable is angular velocity, expressed in r.p.m.
• Run Time Controls — variable time step for transient cases, specified in
panel “Set Run Time Controls”. The independent variable is time, the
dependent variable the time step size. Note that STAR assumes a linear
variation in step size between the size values entered at two consecutive time
points. This is illustrated by the example below, showing the desired time step
variation and the table structure needed to achieve it:
0.35

0.3

0.25
DT (sec)

0.2

0.15

0.1

0.05

0.0
0.00 5.00 10.00 15.00 20.00
Time (sec)
Figure 2-1 Example of time step variation

Table 2-1: Time step size table

TIME DT
0.0 0.01
1.0 0.1
5.0 0.1
5.0 0.2
10.0 0.3
20.0 0.3

Version 4.02 2-25


BASIC STAR-CD FEATURES Chapter 2
Table Manipulation

In addition, a special table type is used to enter problem data for Lagrangian
Multi-Phase cases. The following two options are available in this category:
• Mass Flow Rate — injection rate history, specified in panel “Spray Injection
with Atomization” which activates STAR-CD’s built-in spray modelling
facilities. The table is used in transient analyses only and contains injector
mass flow rates vs. time (see also topic “Define Injectors”). The same table
type may also be used in panel “Injection Definition” as part of an explicit
specification of injection characteristics.
• Diameter Distribution Function — a definition of the droplet diameter
distribution function, in terms of spray percentage mass vs. droplet diameter.
This table may also be specified in panel “Injection Definition”.
The table editor
Table data are stored in text files and may be created or modified either via a
suitable text editor or via pro-STAR’s own GUI facilities. Both options are accessed
by clicking the special table editor button

at the bottom left-hand side of the main window. The basic functionality of the
editor is described below.
New tables
To create a new table, click New Table to display the table view shown below:

2-26 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Table Manipulation

Panel Data Entry


To use the dialog facilities directly, the following input is required (reading from
left to right along the panel):
1. Table Title — enter a title up to 80 characters long, including spaces. Note,
however, that only the first 30 characters found up to the first space in the
string are usable by STAR.
2. Coordinate System — specify the coordinate system number to be used for
spatial independent variables (see “Coordinate Systems” on page 2-8 of the
Meshing User Guide). A search button is provided for choosing any of the
currently defined systems from the Coordinate Systems dialog. Depending on
your selection, the three space coordinates are interpreted as follows:

Cartesian Cylindrical Spherical Toroidal


x (X) r (R) r (R) r (R)
y (Y) θ (ΤΗΕΤΑ) θ (ΤΗΕΤΑ) θ (ΤΗΕΤΑ)
z (Z) z (Z) φ (PHI) φ (PHI)

The coordinate names shown above inside parentheses should be used as table
headers when creating a table outside this GUI environment.
3. Out of bound value options — prescribe the action to be taken if needing to
calculate dependent variable values at points lying outside the table range.
Obviously, this does not apply to mass flow rate tables. The available options
are:
(a) Error — issue an error message
(b) Extrapolate — use the closest two data points to calculate an
extrapolated value
(c) Cutoff — use the closest data point as the variable value

4. Select Table Type — choose the basic table type from the list of options
described under “Basic functionality”. The correct type is selected
automatically if you enter the editor indirectly, i.e. by clicking button New in
a STAR GUIde panel that requires the use of tables.
5. Select Dependent Variables — for boundary and source tables, select also the
specific type of boundary or source required from a secondary menu. All valid
variables for the chosen table type are displayed automatically in the adjacent
scroll list. To select an item from this list:
(a) For single items, click the desired variable
(b) For two or more items in sequence, click the first variable, press and hold
down the Shift key, then click the last variable in the group
(c) For a random selection, hold down the Cntrl key and then click each
variable in turn

6. Select Independent Variables — all valid variables for the chosen table type
are displayed automatically as a series of option buttons. Choose those needed
to define your table by clicking the corresponding button.

7. Click Setup to confirm your selections and enter the data input mode, as
shown in the example below.
Version 4.02 2-27
BASIC STAR-CD FEATURES Chapter 2
Table Manipulation

Commands: TBDEFINE TBCLEAR TBWRITE TBGRAPH

The following points should be kept in mind when specifying table data:
• Table values should be entered for each dependent variable selected in step 5
above. Your selection will be automatically reflected in the options shown on
Dependent Variables scroll box. Fill in all required data for the currently
selected variable before scrolling to the next one.
• The left-hand side of the panel will display a number of columns, one for each
independent variable selected in step 6 above. Fill each column with all the
values assumed by that variable in the table, in ascending order.
• Tables containing two or more independent variables are essentially
multi-dimensional and need to be specified as a series of two-dimensional x-y
tables, as in a spreadsheet. Accordingly, a pair of independent variable values
are displayed as row and column headings and the user fills in appropriate
values for the current dependent variable, as shown in the example above.
• To create such two-dimensional tables:
(a) Select the required pair from the Independent Variables menu, noting that
pro-STAR activates only those combinations that correspond to the
choice made in step 6 above. The available pairs for the example shown
(an X, Y, TIME selection) will be X - Y, X - TIME and Y - TIME and
the pair chosen is X - Y.
(b) Fix the other independent variable(s) to a desired value, by clicking the
radio button next to that value in its column on the left-hand side. In the
example, TIME is fixed to 0.
2-28 Version 4.02
Chapter 2 BASIC STAR-CD FEATURES
Table Manipulation

(c) Click the FILL button. This sets up the 2D table and displays the chosen
pair’s values as row and column headings
(d) Fill the table with the required dependent variable values and then click
Save Data.
(e) Fix the other independent variable(s) to a different value and repeat steps
(b) to (d) above as many times as necessary
(f) Select another pair from the Independent Variables menu and fill in
another series of 2D tables. This might happen, for example, if instead of
choosing to enter an X - Y set for a series of fixed TIME’s, you chose
instead to enter X - TIME sets for fixed Y’s followed by Y - TIME sets
for fixed X’s.

• Tables for rotational speeds, run-time controls and Lagrangian multi-phase


specifications always have one independent variable and thus involve filling in
a two-column table. The same also applies to the other tables if only a single
independent variable is specified. A simplified display appears in the editor
panel in these cases.

Once your data input is complete, you may:


1. Check the table contents graphically by plotting them as a pro-STAR graph
(see Chapter 14 of the Post-Processing User Guide). To use this facility:
(a) Select the variable to be checked from the Dependent Variables scroll
box. This will be plotted along the graph’s y-axis.
(b) Go to the graph setup section at the bottom of the panel (which now
displays the chosen variable) and select an independent variable from the
versus scroll box. This will be plotted along the graph’s x-axis.
(c) The names of the remaining independent variable(s) will also be
displayed in the const boxes. For the purposes of the graph, these will be
fixed to the value indicated by the radio button in each variable’s column.
These values will also appear inside the @ boxes.
(d) Click Graph to see the result of your selection.

2. Save your data in a table file. The file name should have extension .tbl and
should be entered in the File Name box at the bottom of the panel.
pro-STAR’s built-in browser may also be used to locate an existing file. Click
Write Table to save your data in this file.

File Data Entry


An alternative method of generating a new table is to import existing tabular data
from an ASCII file created outside pro-STAR. To use this method:
1. After opening the Table Editor dialog, select the Import button situated under
the New Table option. This will display the alternative panel view shown
below.

Version 4.02 2-29


BASIC STAR-CD FEATURES Chapter 2
Table Manipulation

2. Click Instruction to open a special text panel giving detailed information on


how the user file shoud be structured and formatted. Check that your own file
conforms to this standard and modify if necessary.
3. Enter the name of your file in the ASCII Data File Name box, or use
pro-STAR’s built-in file browser to help locate it.
4. Select option space or comma from the Delimiter menu to indicate how the
numerical values in your table are separated from each other.
5. Click Import to import your data into pro-STAR.
6. Enter the remaining table specification items on the right-hand side of the
panel, as described on page 2-27.
7. Check the table contents graphically, if required, and then save them in a
pro-STAR table file, as described on page 2-29.

Existing table display/modification


To read and display the contents of an existing table, click Read Table at the top
left-hand side of the editor and then enter the file name (of form case.tbl) in the
File Name box. pro-STAR’s built-in browser may be used to help locate the file.

Commands: TBREAD TBLIST TBMODIFY TBGRAPH

Once the table has been read, its contents can be checked visually using the graph
function described in the previous section or modified as required. Note that:
1. You cannot add new dependent or independent variables to an existing table
(or delete any that are currently defined)

2-30 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Plotting Functions

2. You may alter both individual values and the number of such values for any
independent variable. Click Save Modified Data to confirm the changes.
3. Changes to existing dependent variable entries are made by over-typing and
confirmed by clicking Save Data.
4. At the end of the editing session, you should always save your updated table
in a named file by clicking Write Table
Useful points
1. Only one table at a time may be loaded into the pro-STAR editor. If you need
to access a second table, you must first save the current one to a named file (if
you have made changes) before reading in the new one.
2. If you change your mind about the contents of your current table and wish to
make drastic change, clicking New Table enables you to erase all entries and
start afresh.
3. The scale factor applied when saving model geometry data (see Chapter 17,
“Data repository file (.ccm)”) is also applied to table coordinate data when
they are accessed by STAR.
4. Apart from the table file itself, table data needed for the next STAR-CD
analysis are also stored in the STAR problem file (see Chapter 17, “Problem
data file (.prob)”) so that they are available to STAR during the run. The user
specifies which tables will be needed as part of the boundary, initial condition
or other model specification requiring the use of tables.
5. You may use command TBSCAN to scan a named .tbl file. Information
about its contents is displayed in the I/O window.

Plotting Functions
Basic set-up
The basic hardware-related plotting features are set by a single command,
TERMINAL. This command sets:
• The display mode of X-based terminals (use option ALTERNATE only for
improving the plotting speed of certain older types of workstation). This
setting may also be accessed from the menu bar by switching between options
Plot > Standard Plot Mode and Plot > Alternate Plot Mode.
• The plot destination — this specifies whether plots are to appear directly on
the screen or written to the neutral plot file (see Appendix B in the
Post-Processing User Guide).
• The operating mode of the plotting device — a choice between raster, vector
or extended (for high-performance workstations). It is also possible to toggle
between raster and extended plot mode by clicking the X / GL button at the
bottom left-hand side of the main window. Note, however, that this option is
available only if you are working with the glm version of pro-STAR (see
“Running a STAR-CD Analysis”, Step 3).

The basic features of devices operating under one of the above modes are:
1. Vector devices, such as pen plotters, can draw lines in one or more colours,
but are not generally capable of filling in closed polygons or erasing parts of
the plot after drawing in them. When this mode is set:
Version 4.02 2-31
BASIC STAR-CD FEATURES Chapter 2
Plotting Functions

(a) All hidden-line plot calculations are done by software.


(b) Large amounts of time may be required for large models.
(c) All contour plots displayed as line contours rather than filled colours.

2. Raster devices, such as most workstation screens, Postscript laser printers,


etc. are capable of filling in polygons quickly and overwriting previously
coloured-in regions with new colours. When this mode is set:
(a) Hidden-line plots are done by hardware.
(b) Contour plots are rendered in filled colours.
(c) VECTOR mode operation is still possible if, for example, the user wants
fringe-style rather than filled-colour contour plots.

3. Extended mode devices offer additional functionality such as true (24-bit)


colour, hardware Z-buffers, double-frame buffering, coordinate
transformation pipelines, Gouraud shading, etc. Machines with these
high-specification graphics attributes can provide:
(a) Real-time rotation, translation and zooming of plots.
(b) Contour plots rendered in smoothly varying colour bands.
(c) Added lighting effects to enhance a user’s perception of the model
geometry.

This style of plot is limited to machines that support the OPENGL standard
and cannot be stored in the neutral plot file at present.

Appendix C in the Post-Processing User Guide lists all currently available


combinations of plot mode and plot characteristics. The same information can also
be listed on line by choosing Help > pro-STAR Help from the menu bar and then
selecting the COMBINAT item from the list shown at the bottom of the pro-STAR
Help dialog.
Advanced screen control
Advanced screen control functions are implemented as follows:
• Background/foreground colour reversal — from the menu bar, select Plot >
Background > Standard (for white lines and text on a black background) or
Plot > Background > Reverse (for black lines and text on a white
background). Alternatively, use command CLRMODE.
• Maximising the graphics area — from the menu bar, select Plot > Maximum
Plot Screen to hide the GUI buttons surrounding the graphics area so as to
make the plot as large as possible. The window is also enlarged to take up
almost the entire screen. This is helpful when making animations since the
largest number of pixels are used, thereby obtaining the highest possible
plotting resolution. Select Plot > Standard Plot Screen to return the window
to its default size and appearance. Alternatively, use command WHOLE.
• Restoration of the original screen settings — command RESET.
• Temporary, on-line storage of complete screen images — command SCROUT.
• On-line retrieval of screen images previously stored with SCROUT —
command SCRIN. This command also provides an elementary animation
facility, by replaying a sequence of screen images in quick succession.

2-32 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
Plotting Functions

• Deletion of screen images previously stored with SCROUT — command


SCRDELETE.
• Customised scaling of text fonts used in pro-STAR — command TSCALE.
• Image display control — command PLTBACK. This enables images to be
created and stored in memory and then popped onto the screen (as opposed to
displaying them as they are being created).

For further details on using the above commands, refer to the pro-STAR Commands
volume.
Screen capture
It is often very useful to be able to save the contents of the graphics screen as a
picture file. The latter can then be pasted into a document created by another, say
presentation or word-processing, application. pro-STAR provides this facility via
the Utility > Capture Screen menu option (or by typing command SCDUMP). The
result of this operation is the creation of a new window containing the picture
currently displayed in pro-STAR’s main graphics area. The picture can be
subsequently saved in a file by choosing Utility > Save Screen As and selecting one
of the following options for the file format:
• XWD (X Window Dump) — X-Motif version of pro-STAR only
• GIF (Graphics Interchange Format)
• PS (PostScript, either Level 1 or Level 2 format)
• EPSF (Encapsulated PostScript, either Level 1 or Level 2 format)

The user needs to make sure that the choice of format is appropriate to the end
application. Selecting any of the above options opens the File Selection dialog
shown below, enabling you to specify the name and destination directory of the
picture file.

If you are working in OpenGL extended graphics mode (see page 2-32), you also
have a choice of saving a high-resolution screen dump (HRSD) of the extended
mode plotting window. This appears as an additional option, High Res. Screen
Dump, in the Utility menu (alternatively, use command HRSDUMP). Selecting this

Version 4.02 2-33


BASIC STAR-CD FEATURES Chapter 2
Plotting Functions

option from the main menu opens the High Resolution Screen Dump dialog shown
below:

The user input is as follows:


1. Select the required file format from the File Type menu as one of
(a) png
(b) gif
(c) ps (PostScript)
(d) eps (Encapsulated PostScript)

2. Enter the file name in the box provided. Clicking the adjacent browser button
opens the File Selection dialog shown above which helps locate the required
file.
3. Clicking the Options button opens a secondary Image Options dialog that
enables you to specify the required image resolution and/or page properties
(for PostScript files). An example for GIF/PNG images is shown below.

It is also possible to use the HRSD facility in batch mode to produce high-quality
plots using OpenGL style graphics (i.e. including translucency, special lighting
effects, etc.). You do not require a special OpenGL graphics card on your machine
to do this; the pictures can be made off-screen using the ‘mesa’ software emulation
of OpenGL as follows:
• Run pro-STAR with mesa graphics in batch mode

prostar mesa -b

• Set extended mode graphics

term,,exte

2-34 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
The Users Tool

• Set up the model as you wish, including any CPLOT/REPLOT operations


needed to display the picture, and then use the HRSD command as follows:

hrsd,png,output.png (write a .png file)


hrsd,ps,test.ps (write a .ps file)

You can also use the various options to change image size, resolution, etc., just as
for interactive mode above (see the HRSD command Help on options for setting
image size, resolution, etc.)

The Users Tool


The Users Tool enables you to create your own customised user interface, by
running a Tcl/Tk script from within pro-STAR by means of a built-in interpreter.
To make use of this tool, you need solid knowledge of Tcl/Tk programming. The
basic idea is that the user builds a dialog box as he/she would for any other
Tcl/Tk-based application, with widget callbacks designed to pass pro-STAR
command strings back to pro-STAR (much as it happens now when you click a
button in STAR GUIde). An introductory panel, shown below, is provided via the
main menu, by choosing Tools > Users Tool. Clicking the left-hand button invokes
the built-in interpreter which then runs your script.

To use this facility, it is important to


• save your Tcl script in a file called STARTkGUI.tcl
• assign the path to this file to an environment variable called
STAR_TCL_SCRIPT

Getting On-line Help


The Help menu in the main pro-STAR window is divided into three parts. There are
three options in the top part, About pro-STAR, Select Item and pro-STAR Help.
Clicking About pro-STAR displays pro-STAR version information, as shown
below:

Version 4.02 2-35


BASIC STAR-CD FEATURES Chapter 2
Getting On-line Help

Clicking pro-STAR displays the pro-STAR Help dialog shown below:

This dialog contains on-line information on:


• Conventions regarding command line syntax
• All valid combinations of plot mode and plot characteristics
• One-line summaries of every pro-STAR command, grouped by command
module and listed in alphabetical order
• A list of all database files available under pro-STAR
• pro-STAR environment variable definitions
• All file extensions used
• A description of pro-STAR’s macro files
• A description of pro-STAR’s user-defined Motif panels
• A tabulation of radiation parameters required for walls and baffles
• Units for all physical quantities used in STAR-CD
• A list of user subroutine names and brief descriptions
• A list of all GUI tools and dialog boxes

Help on any of the above items is obtained simply by selecting the appropriate title
in the scroll list underneath the main information display area.
In addition, the default listing of any user subroutine may be displayed by
selecting item UserSubs from the Module pop-up menu and then choosing the
2-36 Version 4.02
Chapter 2 BASIC STAR-CD FEATURES
Getting On-line Help

required subroutine name from the second scroll list.


Details on the functionality and syntax of every command may be displayed as
follows:
• By typing the command name in the Find Command text box and pressing
Return
• By selecting the appropriate command module from the Module pop-up menu
(see the pro-STAR Commands volume for a description of modules) and then
choosing the required command name in the scroll list
• By searching through the available help text for a keyword, as typed in the
Keyword text box
• In a context-sensitive manner, by choosing option Select Item from the Help
menu. This changes the mouse pointer from an arrow to a ‘hand’ (Help)
pointer with which you can click any part of the main pro-STAR window.
Such an action will automatically display the corresponding command
description for that part of the window.

An example of command help is shown below:

The middle section of the Help menu gives on-line access to every volume in the
STAR-CD documentation set, consisting of Release Notes for the current version,
pro-STAR Commands, Methodology, Tutorials and also the CCM, Meshing and
Post-Processing User Guides. To view these documents, users must make sure that
Adobe’s Acrobat™ Reader is installed on their machine. Instructions on how to do
this are given in the STAR-CD Installation and Systems Guide. There is also a Help
section containing useful information on how to best use Acrobat for viewing
on-line help text corresponding to each panel of the STAR GUIde system described
below.
The last section of the Help menu activates your machine’s web browser and
directs it to useful web sites set up by CD-adapco.

Version 4.02 2-37


BASIC STAR-CD FEATURES Chapter 2
The STAR GUIde Environment

The STAR GUIde Environment


STAR GUIde represents the latest development in easy-to-use GUI tools for
building STAR-CD models. It works by
• dividing the CCM analysis task into groups of major modelling activities;
• displaying pre-defined groups of panels relating to each of the activities so
that the user can specify model parameters and characteristics pertinent to the
current activity;
• guiding the user through the modelling process in a logical sequence so that
no steps of that process are overlooked.

At present, the STAR GUIde panels cover a subset of pro-STAR’s capabilities, i.e.
those that relate to the most common tasks of the modelling process. Additional
capabilities are being continually added and appear in each new version of
STAR-CD.
STAR GUIde may be accessed from pro-STAR’s main window using either of
the following two methods:
1. Selecting Tools > STAR GUIde from the menu bar
2. Clicking the STAR GUIde button at the top left-hand side of the window.

This displays the introductory screen shown below. The screen consists of two
parts:
• On the left is the Navigation Centre (NavCenter), a tool for guiding the user
through the various stages of the model building process. These stages are
represented by panels and are subdivided into logical groups. The panels and
their groups are shown as a tree structure within the NavCenter sub-window.
• On the right is the initial Help screen explaining how STAR GUIde works and
what its function buttons do. This is replaced by the contents of the current
process panel as you go through each stage of model building.

2-38 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
The STAR GUIde Environment

The following points should be borne in mind when using this tool:
1. The NavCenter tree contains a set of yellow folder icons representing each
major modelling activity and acts as the starting point for defining your own
model. A complete STAR-CD simulation can be set up and run by performing
the activities in the folder tree and in the order shown.
2. Click on one of the yellow folder icons (or on the text next to them) to open
and close the folder and to display its constituent process panels and
sub-folders.
3. Click on a grey panel icon (or on the text next to it) to open the panel; its
contents will be displayed on the right-hand side of the STAR GUIde window.
Each process panel enables you to enter or generate data needed to complete
that process.
4. Where appropriate, the input for a given process is distributed amongst
colour-coded, ‘file tabs’. These are brought to the forefront by clicking on the
appropriate tab. The colour coding depends on the entity (block, spline, cell,
etc.) being processed and is consistent with the colour coding used in the main

Version 4.02 2-39


BASIC STAR-CD FEATURES Chapter 2
The STAR GUIde Environment

pro-STAR window.
5. In some instances, clicking a button on a panel activates a separate,
free-floating dialog box. This happens whenever such a dialog provides the
most convenient means of entering the data required.
6. To exit from the STAR GUIde, click the Close STAR GUIde button at the
bottom of the NavCenter sub-window.

Panel navigation system


The set of five buttons at the top right-hand side of the STAR GUIde window are
designed to help you navigate through the system and get more information about
what to do. The function of each button is as follows:

Go Back — returns to the previously selected panel.

Collapse/Expand Navcenter — Closes the left-hand


(NavCenter) side of the STAR GUIde window to make more
space on your screen. The window may be expanded back to its
original size by clicking this button again.

Favorite — enables you to store the names of frequently used


panels so that you may jump to them directly, i.e. without first
opening the STAR GUIde window and then searching through
the NavCenter tree. A ‘favourite’ panel is selected by first
displaying it in STAR GUIde, clicking Favorite and then
choosing the Add to favorites option. The reverse operation is
performed by choosing Remove from favorites. The current
favourites are listed under the Favorites menu in the main
pro-STAR window.

Help — provides concise information on the current panel,


including descriptions of the data required, explanations of the
choices available, suggestions on things to look out for, etc. Help
screens use Adobe’s Acrobat™ Reader system; their contents
therefore appear in a separate window opened by that system.
Information on how to best use Acrobat for reading these screens
is given under the Help menu in the main pro-STAR window
(option Help).

2-40 Version 4.02


Chapter 2 BASIC STAR-CD FEATURES
General Guidelines

Go Fwd — if the Go Back control has already been used, goes


forward to the most recently displayed panel.

STAR GUIde usage


The STAR GUIde panels should be used in conjunction with the facilities (pop-up
menus and action buttons) offered by the main pro-STAR window. The input/output
window should also be displayed to cater for operations that need command input
(see also the “Introduction” section). For maximum ease of use, all three windows
should be displayed side-by-side on your screen, as shown below:

General Guidelines
The following general guidelines should be kept in mind when running STAR-CD
models, including those described in the Tutorials volume:
1. Take advantage of the on-line Help facilities to check the code’s conventions
and, if necessary, the structure and meaning of individual commands. These
facilities are accessed either from the GUI Help menu (see “Getting On-line
Help”) or by typing

HELP, command_name

in the pro-STAR I/O window.

Version 4.02 2-41


BASIC STAR-CD FEATURES Chapter 2
General Guidelines

2. Make frequent use of the File > Save Model option (or command SAVE) to
store the current state of your model description on the pro-STAR model file
(.mdl). This safeguards against unexpected mishaps (power failures, system
crashes, etc.) by enabling you to restart your work from the point where the
last SAVE operation was performed. You should, however, make sure that the
model is in a satisfactory state before saving it.
3. If necessary, split lengthy model-building sessions into several parts, by using
option File > Quit (or command QUIT) at any convenient point in your
current session and then saving your work on the .mdl file. To continue
working on the model, re-enter pro-STAR as discussed in Step 10 on page 2-7
and then perform the next operation. However, remember that for transient
problems the transient data (.trns) file has to be explicitly re-connected to
the pro-STAR session by using the Connect button in the Advanced
Transients dialog (or command TRFILE).
4. Mistakes in pro-STAR can be rectified in two ways:
(a) Use option File > Resume Model (or command RESUME) to go back to
the state of the model saved with the last SAVE operation and start again
from there
(b) Use command RECOVER to play back all commands issued since the last
SAVE operation, re-execute the code up to the one that went wrong, and
continue from there

5. Note that command execution can be terminated half way through in the
following circumstances:
(a) By typing Abort instead of a parameter value while supplying parameter
values to a command in ‘novice’ mode.
(b) By typing Ctrl+C while waiting for a command to finish processing.
Note that the effect of this operation is machine-dependent and therefore
great caution should be exercised in its use; in some machines it will abort
the entire pro-STAR session.

6. Display the relevant STAR GUIde panels frequently to check the settings of
pro-STAR parameters; alternatively use command STATUS. In the latter case,
the screen information relates to the active command module, so make sure
you are in the right module by typing the appropriate keyword (MESH,
PROPERTY, CONTROL, etc.)
7. Remember that all pro-STAR windows can be re-sized using the mouse. It is
recommended that both the I/O and the main window are positioned and sized
so that both are visible simultaneously. This is particularly helpful when you
need to use commands for a particular operation, or if you want to check the
commands that were generated automatically by a particular GUI operation.

2-42 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Introduction

Chapter 3 MATERIAL PROPERTY AND PROBLEM


CHARACTERISATION
Introduction
The physical properties of the fluid and/or solid materials within the model are
typically defined immediately after setting up the mesh and performing a thorough
visual and numerical check on it. STAR-CD can analyse problems containing
arbitrary combinations of
• multi-domain fluids, where there is no mixing of fluid streams,
• porous materials,
• solids materials.

The Cell Table


The process of setting up properties is usually quite simple and relies on the concept
of cell identity and the consequent use of the cell table, as discussed under “Cell
types” on page 2-37 of the Meshing User Guide. The cell table can be defined using
pro-STAR’s Cell Table Editor, accessed by clicking the CTAB button on the
left-hand side of the main pro-STAR window.
All cells in the mesh can be indexed and differentiated in various ways with the
aid of an entry in the cell table. This enables the user to specify a
• cell table index
• cell type
• material number
• colour table index
• porosity index
• spin index
• group number
• surface lighting material index
• processor number
• conduction thickness
• radiation switch
• initial free-surface identifier
• identifying name

for a set of cells, as shown in the dialog below. The meaning of the various
parameters that may be set in this table is described in “Cell properties” on page
2-38 of the Meshing User Guide.

Version 4.02 3-1


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
The Cell Table

Commands: CTABLE CTNAME CTMODIFY CTLIST


CTDELETE CTCOMPRESS

The rules governing the use of the cell table are as follows:
• All entries in the table are identified by an index, listed under the Table #
heading in the editor’s scroll list. A new entry is set up by clicking on the next
available number in the list and then specifying the relevant cell properties.
• Every cell in the model is associated with a cell table index.
• All cells linked via a common index belong to a common Cell Type (Fluid,
Solid, Baffle, etc.), selected from the editor’s pop-up menu.
• Different materials are identified by separate material property numbers,
typed in the Material Number text box.
• The default cell table index is number 1 and is associated with a fluid whose
material number is 1.
• By default, material number 1 refers to air properties at standard conditions.
• Cell indexing normally differentiates the cells’ material type. However, it can
also be used purely for visual and/or selection purposes. Thus, in the diffuser
model shown in Figure 3-1 there is a single material number (no. 1),
corresponding to the one and only domain in the model, but the cells can be
indexed to different colours or different types of surface shading (see Chapter
4 of the Meshing User Guide). This is done by typing different values in the
Color Table Index or Lighting Material text boxes, respectively.
• Colour selection is facilitated by clicking the multi-coloured button next to
the Color Table Index box. This opens a Color Palette panel where the desired
colour is selected by simply clicking the appropriate square. The
corresponding colour number is then automatically entered into the box.
• Another possibility is to index cells on the basis of a common group number,
typed in the Group Number text box. This groups together all cells belonging
to a particular ‘object’, e.g. a distinct portion of the mesh. Such objects might
typically be generated with the help of an external CAD package and

3-2 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
The Cell Table

imported into pro-STAR using IGES or VDA data files. Group numbers are
normally generated automatically as part of the data import function (see
“Importing Data from other Systems” on page 3-1 of the Meshing User
Guide).
• Cell table entries can be further identified by a name, typed in the Name box.

A cell table definition is confirmed by clicking the Apply button.

Cell index 1 Cell index 2 Cell index 3

Colour 2
Colour 3

Colour 4

Figure 3-1 Cell indexing to implement differentiation by cell colour

Cell table entries may be displayed at any stage of the pro-STAR session by clicking
CTAB on the main window. Any identifier, index, or reference number used in a
cell table entry may be changed to a different value simply by selecting the entry in
the Cell Table Editor’s scroll list and making the required changes.
Cell table entries may also be deleted by clicking the Delete button. Note that all
cells indexed to this entry must be deleted or changed to a different index before the
table entry itself can be deleted. Tables that contain deleted (or undefined) entries
such as this may be cleaned up by clicking the Compress button. This removes all
redundant entries and re-numbers the remaining ones.
Cell indexing
Cells are assigned an identity (cell index) using the Cell Tool shown overleaf. This
may be done in two ways:
1. Implicitly, by taking on the index that is active at the moment of their
creation. The active cell type can be changed at any time by highlighting the
type required in the Cell Table list displayed by the Cell Tool and then
clicking the Set Active Type button. The selection is indicated in the list by a
letter ‘A’ against the active type.
2. Explicitly, by collecting together a group of cells and then changing their
identity to the currently-active type. This can be done by:
(a) Pointing at the desired cells with the screen cursor — choose option
Modify Type > Cursor Select. The action is terminated by clicking the
Done button displayed on the plot.
(b) Changing all cells contained within a polygon drawn on the screen with
the screen cursor — choose option Modify Type > Zone. The action is
terminated by clicking on
i) the same point twice to complete the polygon;

Version 4.02 3-3


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
The Cell Table

ii) the Close button displayed, to let pro-STAR complete the polygon;
iii) the Abort button displayed, to abort the selection operation.

(c) Changing all surface cells encountered when searching from a starting
position given by a ‘seed’ vertex (see the description on page 2-49 of the
Meshing User Guide). This can be done by choosing option Modify Type
> Surface (New Edge Vertex Set) (or Surface (Current Vertex Set)).
The ‘seed’ vertex is selected with the screen cursor.
(d) Changing all cells in the current cell set — choose option Modify Type >
Cell Set.

Commands: CTYPE CCROSS CFIND CZONE CTCOMPRESS

Another method of making changes is via the Cell List dialog, shown overleaf. This
may be displayed by clicking the Cell List button on the Cell Tool or choosing Lists
> Cells from the main menu bar. The cell or cell range to be changed must first be
selected on the list. To change the associated cell type, click Change Type, choose
a different cell table index on the displayed Change Cell Table box and then click
Apply.

3-4 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Multi-Domain Property Setting

Command: CMODIFY

The result of the above process can be checked using the Check Tool, option Double
Cells (see “Microscopic checking” on page 3-28 of the Meshing User Guide). This
will verify whether a cell table entry exists for every cell within the range specified.

Multi-Domain Property Setting


The user is free to define as many material types (of the fluid or solid variety) as are
necessary to represent the problem conditions. The most general case, involving
multiple fluid domains in the presence of solids, is illustrated in the example below:

Domain 1

Metal plate

Domain 2

Figure 3-2 Multi-domain flow with solid material domains

Version 4.02 3-5


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
Multi-Domain Property Setting

Setting up models
Step 1
Create an appropriate set of cell types and material indices for your model during
mesh generation, using the procedure described in “The Cell Table” on page 3-1.
The appropriate settings to be supplied via the Cell Table Editor for the example
shown in Figure 3-2 are as follows:

Domain 1

Metal plate

3-6 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Multi-Domain Property Setting

Domain 2

The Material Number indices 11, 12 and 13 above refer to the physical property sets
associated with each fluid domain and with the solid plate domain. Note that
different cell table indices 1, 2 and 4 are also assigned to each of these because each
cell table index can only refer to one material number. In cases with multiple
domains it is recommended that each domain is given a separate material number,
even if domains have identical physical property sets. This is to allow each domain
to have its own initialisation, reference values and residual normalisation.
Step 2
Open the Thermophysical Models and Properties folder in STAR GUIde. For
thermal problems, specify any special thermal transfer conditions (radiation, solar
radiation or solid-fluid heat transfer) prevailing in your model by making the
relevant selection(s) in the “Thermal Options” panel.
Step 3
Set the physical properties of each fluid domain by opening sub-folder Liquids and
Gases and then entering numerical values and/or selecting appropriate options in
the “Molecular Properties” panel. Note that:
• The option chosen for density calculations determines whether the flow is
treated as compressible or incompressible. Special considerations regarding
the analysis of compressible flows are discussed in “Compressible Flow” on
page 3-9 of this chapter.
• Non-Newtonian flow may be simulated by selecting the relevant molecular
viscosity calculation option. The treatment of non-Newtonian fluids is
discussed further in “Non-Newtonian Flow” on page 3-11 of this chapter.

Each domain must be selected in turn via the Material # control at the bottom of the
panel (see also the “Liquids and Gases” Help topic).

Version 4.02 3-7


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
Multi-Domain Property Setting

Step 4
If you have selected the solid-fluid heat transfer option, an additional sub-folder,
Solids, will appear in the STAR GUIde tree structure. Specify the physical
properties of the solid material by entering numerical values and/or choosing
appropriate options in the “Material Properties” panel. If your model contains
multiple solid domains possessing different properties, each domain may be
selected in turn via the Material # control at the bottom of the panel (see also the
“Solids” Help topic).
Step 5
For turbulent fluid domains, choose an appropriate option from the “Turbulence
Models” panel. Further details are given in “Turbulence Modelling” on page 3-12
of this chapter.
Step 6
For thermal problems, turn on the enthalpy equation solver in all fluid domains
using the “Thermal Models” panel. The enthalpy equation solver for solid materials
is activated simply by selecting Solid-Fluid Heat Transfer in the “Thermal Options”
panel. Special considerations regarding the use of this option are discussed in “Heat
Transfer In Solid-Fluid Systems” on page 3-16 of this chapter.
Step 7
Specify initial values for the flow variables in each fluid domain using the
“Initialisation” panel (Liquids and Gases folder). The temperature distribution
inside solid materials is specified via a separate “Initialisation” panel under the
Solids folder.
Step 8
Set the reference quantities (pressure and temperature) and monitoring cell
location(s) for each domain using the “Monitoring and Reference Data” panel
(Liquids and Gases folder). The reference temperature and monitoring cell location
for solids is specified via a separate “Monitoring and Reference Data” panel under
the Solids folder.
Step 9
For buoyancy-driven or any other problems involving body forces, specify the
necessary parameters using the “Buoyancy” panel. Special considerations
regarding the use of this option are discussed in “Buoyancy-driven Flows and
Natural Convection” on page 3-20 of this chapter.
Step 10
If necessary, specify mass sources or additional source terms for the solution of the
momentum, turbulence or enthalpy equation. The type of source is chosen by
selecting the appropriate tab in STAR GUIde’s “Source Terms” panel (sub-folder
Sources):
• Mass — specify mass sources or sinks, i.e. fluid injection or withdrawal, to be
used in the solution of the mass conservation equation (tab “Mass”). Special
considerations regarding the use of subroutine FLUINJ for this purpose are
discussed in “Fluid Injection” on page 3-21 of this chapter.
• Momentum — specify momentum sources, e.g. a fan driving the flow at
some location of your model, where the fan is not explicitly modelled (tab

3-8 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Compressible Flow

“Momentum”).
• Turbulence — specify sources appropriate to the turbulence model used.
These may be additional source terms or, in the case of the k-ε model,
replacements for the existing terms (tab “Turbulence”).
• Enthalpy — specify heat sources or sinks, e.g. radioactive sources in a
nuclear reactor cooling problem (tab “Enthalpy”).

All property and physical model settings in your problem may be inspected by
selecting the relevant panels in the Thermophysical Models and Properties folder.
In sub-folders Liquids and Gases and Solids, open each constituent panel in turn and
scroll through the available materials. Alternatively, type command MLIST to
display a brief or comprehensive listing of properties for any material in the Output
window.

Compressible Flow
The theory behind compressible flow problems and the manner of implementing it
in STAR-CD is given in the Methodology volume (Chapter 16, “Compressible
Flows”). This section contains an outline of the process to be followed when setting
up such problems and important points to bear in mind. Also included are
cross-references to appropriate parts of the STAR GUIde on-line Help system,
containing details of the user input required.
Setting up compressible flow models
Step 1
Go to panel “Molecular Properties” in STAR-GUIde and select each compressible
fluid domain via the slider at the bottom of the panel.
Step 2
Declare the flow as (ideal gas) compressible by selecting option Ideal-f(T,P) from
the “Density” pop-up menu. This effectively switches on the compressibility
calculations by making the density a function of both pressure and temperature.
Step 3
Set up boundary conditions that are appropriate to the type of flow being analysed.
These are as follows:

Subsonic flow (Ma < 1 throughout the solution domain)

Inflow Outflow
Stagnation conditions Pressure
Inlet Pressure
Inlet Outlet (for steady flow, but see point no. 1 below)
Inlet Wave transmissive (for transient flow)

Supersonic flow (Ma > 1 throughout the solution domain)

Inflow Outflow
Inlet Outlet
Inlet Pressure
Version 4.02 3-9
MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
Compressible Flow

Transonic flow (Ma < 1 and Ma > 1 within the solution domain)

Subsonic Inflow Subsonic Outflow


Stagnation conditions Pressure
Inlet Pressure
Supersonic Inflow Subsonic Outflow
Inlet Pressure
Supersonic Inflow Supersonic Outflow
Inlet Pressure
Subsonic Inflow Supersonic Outflow
Stagnation conditions Pressure

The user should refer to the on-line Help text for panel “Define Boundary Regions”
(especially that for “Inlet” boundaries) for a description of how to set up boundary
conditions for this type of flow.
Useful points on compressible flow
1. The combination of inlet and outlet boundary conditions for subsonic flows
presented under Step 3 above does not constitute, strictly speaking, a ‘well
posed’ problem. However, it is offered as an option for use in circumstances
where the pressure is known at the inflow (or at some other point inside the
solution domain) but not at the outflow. In such a case, users should designate
the known pressure as the reference pressure and make sure the corresponding
cell location lies as close as possible to the known location (e.g. the inlet
boundary surface). The success of the simulation will depend on the
magnitude of the Mach number. For the higher Mach numbers (e.g. Ma > 0.7)
very low under-relaxation factors will have to be specified (e.g. 0.001 for
pressure) in order to obtain a converged solution.
2. Special considerations apply to tetrahedral meshes or meshes containing
trimmed (polyhedral) cells. If such meshes contain supersonic inlet
boundaries then, to obtain a stable/convergent solution, it is necessary to
create at least two cell layers immediately next to the boundary (see Figure
4-5 on page 4-23). If pro-STAR’s automatic meshing module is employed for
this purpose, use its built-in mesh generation capabilities. If the mesh is
imported from a package that lacks these facilities, you must extrude the mesh
in a direction normal to the boundary and then shift the boundary location to
the edge of the newly-created, layered structure.
3. In the case of a transonic problem with subsonic inflow, residual
normalisation for momentum (and k, ε if appropriate) is based on the
momentum (and k, ε) flux values at the inlet, as usual. However, because of
the large difference in velocity magnitude between the inlet and the rest of the
flow field, this may place an unnecessarily stringent condition on the built-in
solution convergence criterion (as discussed in Chapter 1, “Monitoring the
calculations”, this is based on the magnitude of the normalised residuals). In
this situation, it could be more appropriate to inspect the convergence history
of, say, mass and enthalpy and terminate the solution process after a
sufficiently large number of iterations.
4. For inviscid flows, it is possible to calculate temperature from a constant

3-10 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Non-Newtonian Flow

stagnation enthalpy relationship rather than the standard enthalpy equation.


To do this, go to panel “Thermal Models” in STAR-GUIde and select option
Stagnation Enthalpy from the Conservation pop-up menu. The appropriate
stagnation temperature should then be typed in the Stagnation Temp. text box.
5. It could be advantageous, even when a steady state is sought, to do a transient
calculation using the “Pseudo-Transient Solution” method. To do this, select
option Pseudo-Transient from the pop-up menu in the “Solution Method”
STAR GUIde panel.
6. In the case of flow through ducts of non-uniform cross-section where
supersonic conditions are expected over the whole or part of the solution
domain, it is sometimes necessary to under-relax the initial velocities. This is
done by activating special flux under-relaxation using panel “Miscellaneous
Controls” in STAR GUIde. This operation affects only the velocity
initialisation.

Non-Newtonian Flow
The theory behind non-Newtonian flow and the manner of implementing it in
STAR-CD is given in the Methodology volume (Chapter 16, “Non-Newtonian
Flows”). This section contains an outline of the process to be followed when
specifying non-Newtonian fluids and includes cross-references to appropriate parts
of the STAR GUIde on-line Help system. The latter contains details of the user
input required.
Setting up non-Newtonian models
Step 1
Decide whether the power law offers an adequate representation of the
non-Newtonian fluid behaviour and what the value of the constants m and n in
equation (1-6) of the Methodology should be. Alternatively, supply a suitable
expression in subroutine VISMOL.
Step 2
Go to panel “Molecular Properties” in STAR-GUIde and select the domain
containing the non-Newtonian fluid via the slider at the bottom of the panel.
Step 3
Use the “Molecular Viscosity” menu to either specify the model parameters m and
n (option NonNewt, text boxes EM and EN) or call subroutine VISMOL (option
User).
Useful points on non-Newtonian flow
1. Bear in mind that constitutive relations for non-Newtonian flow are basically
empirical curve-fitting formulae. It is therefore inadvisable to use them
beyond the range of the available data.
2. The model parameters are functions of temperature, pressure and
composition. They may also be functions of the rate of strain tensor’s range
II s (see equation (1-5) in Chapter 1 of the Methodology volume), over which
the equation is fitted. If any of these effects are significant, they should be
allowed for in user subroutine VISMOL.

Version 4.02 3-11


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
Turbulence Modelling

Turbulence Modelling
The theory behind the currently available models is given in Chapter 2 of the
Methodology manual. A number of methods are also available for implementing the
no-slip boundary conditions for turbulent flow, as follows:
1. Wall functions, applied to cells immediately adjacent to a wall. This method
employs special algebraic formulae (described in Chapter 6, “High Reynolds
number turbulence models and wall functions” of the Methodology volume)
to represent velocity, temperature, turbulence parameters, etc. within the
boundary layer that forms next to the wall; see Figure 3-3(a). The method is
also appropriate for use with one-equation (k-l, Spalart-Allmaras), k-ω and
Reynolds Stress models. An alternative, ‘non-equilibrium’ type of wall
function is also provided for taking pressure gradient effects into account (see
equation (6-17), (6-18) and (6-19) in the Methodology volume) but this is
available only for k-ε models (linear and non-linear).
2. Two-layer models, employed as combinations of a high Reynolds number
(k-ε) model with a low Reynolds number (one-equation or zero-equation)
model. The latter is applied to the near-wall region where the mesh should be
finely spaced, as shown in Figure 3-3(b); see also Chapter 6, “Two-layer
models” in the Methodology volume. You are free to combine the wall
function and two-layer approach within the same problem, provided that a
linear k-ε type model is in use and the two treatments are applied to different
boundary regions. However, care must be exercised at transition points
between the two methods.
3. Low Reynolds number models, in which viscous effects are incorporated in
the k and ε transport equations. No special near-wall treatment (other than an
optional definition of wall surface roughness) is therefore required; see also
Chapter 6, “Low Reynolds number turbulence models”. Both low Re and wall
function treatments may be used in the same problem, but only if they apply
to separate domains.
4. Hybrid wall boundary condition, which offers a special wall treatment for
low Reynolds number models independent of the normalised parameter y + .
For finely spaced meshes, this is identical to the standard low Reynolds
number treatment. For coarser meshes, it provides special algebraic formulae
to represent velocity, temperature, turbulence parameters, etc. similar to
ordinary wall functions (see also Chapter 6, “Hybrid wall boundary
condition”).

The choice of wall treatment (where relevant) is made in the “Near-Wall


Treatment” tab of the “Turbulence Models” panel. If a two-layer model is
employed, you will need to indicate the wall or baffle region to which it applies via
the “Define Boundary Regions” panel.

3-12 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Turbulence Modelling

k - ε model

match location

NWL

y
Low Re model

(a) Wall function model (b) Two-layer models

Figure 3-3 Mesh spacing in the near-wall region

The following points should be borne in mind when considering the effectiveness
or accuracy of a particular turbulence model or near-wall treatment:
Wall functions
1. For reasons of accuracy, the normal distance y from the wall for near-wall
cells (see Figure 3-3) should be such that the dimensionless parameter y + is
kept within the limits 30 < y + < 100 , where:
+ 1⁄4 1⁄2
y ≡ ρ Cµ k y⁄µ

2. It is important to place y outside the viscous sublayer. This can be achieved by


observing the lower limit on the value of y + .
3. The above considerations apply equally to both standard and non-equilibrium
wall functions. The difference between the two is that the latter takes the
pressure gradient into account. This provides more accurate results in terms of
wall shear forces but has little effect on the character of the flow.
4. If the non-equilibrium option is chosen, the normal user inputs for wall
roughness (specified via the Roughness pop-up menu for wall and baffle
boundaries, see panel “Define Boundary Regions”) are not applicable.
Two-layer models
1. These should be preferred for non-equilibrium flows, as they produce
improved friction and heat transfer predictions. Their use, however, will result
in larger meshes within the model and hence significantly higher calculation
times. This is because the near-wall region requires a finer mesh than that
needed by the wall function treatment.
2. In order to resolve properly the distributions of velocity and other variables
within the near-wall region (i.e. at y + ≤ 40 ), it is necessary to ensure that it is
spanned by about 15 mesh nodes. In general, this may require some trial and
error adjustment of the mesh, since the near-wall region thickness is not
known a priori. Once a suitable mesh density is chosen, the value of y + at the

Version 4.02 3-13


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
Turbulence Modelling

node next to the wall should be no larger than ~3 to resolve the velocity
profile, but smaller to resolve the thermal profile.
3. If the prescribed NWL thickness is not sufficiently large to encompass the
near-wall region throughout the domain in question (i.e. the switching
location between high and low Re regions shown in Figure 3-3 lies outside the
NWL in some places), the switching location there is assumed to be at the
edge of the NWL and a warning message is issued on file case.info. In
such cases, it is possible to increase the NWL thickness to a more suitable
value and restart the calculations.
4. There is an additional option for fixing the above switching location to its
current position. If this option is selected from the start of the analysis, its
effect is to make the switching point distance equal to the NWL thickness.
5. During post-processing, the partitioning of the mesh into
(a) near-wall region cells where the one-equation model applies
(b) other cells in the NWL
(c) ordinary cells in the flow field interior

can be inspected by opening panel “Load Data” in STAR GUIde (“Data tab”),
choosing “Cell Data” as the data type and then selecting option Two Layer
from the Scalar Data scroll list. Option FMU allows inspection of the
distribution of a quantity given by
νt
------------------
2
Cµk ⁄ ε

Low Re models
1. These should be preferred for non-equilibrium flows, for the same reasons as
two-layer models. However, their use may require meshes that are even larger
than those for the two-layer approach.
2. The default treatment assumes a smooth wall but the wall surface roughness
may also be specified, if required.
3. In order to resolve properly the distribution of velocity and other variables,
approximately 20 mesh nodes are needed within the near-wall region
( y + ≤ 40 ). The value of y + at the node next to the wall should then be ~1.
Note that this meshing strategy differs from that for two-layer models, where
approximately 15 mesh nodes are needed over the near-wall region. This
means that a mesh designed for two-layer models will not necessarily be
suitable for low Re models.
4. As with two-layer models, computing times are substantially greater than
when using a wall function approach.
5. It is recommended that such models are run in double precision.
Hybrid wall boundary condition
1. The hybrid wall condition is an extension of low Reynolds number boundary
conditions. It applies only to the following low Reynolds number turbulence
models:
(a) k-ε (linear, cubic and quadratic)
(b) k-ω (standard and SST variants)
3-14 Version 4.02
Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Turbulence Modelling

(c) Spalart-Allmaras

2. The approach automatically selects a low Reynolds number wall treatment or


a wall function, depending on the local flow field and near-wall mesh spacing.
It should be preferred in situations where
(a) the normalised parameter y + is unknown, or
(b) large variations in y + create uncertainties as to whether a low Reynolds
number boundary treatment or a wall function is appropriate.
Reynolds Stress models
1. Both the Gibson-Launder and SSG models are high Reynolds number models
so they need to be used in conjunction with wall functions.
2. Since Reynolds Stress models solve additional transport equations for
Reynolds Stress components, they consume a substantially greater amount of
computing time compared to k-ε models.
3. The ‘standard’ wall reflection term used in the Gibson - Launder model is not
suitable for impingement flows. In such circumstances, it will return the
wrong distribution of the stress component normal to the wall. It is therefore
advisable to use the term calculated by the Craft model instead.
DES models
1. A transient analysis setting is required, although the problem being modelled
may in reality be a steady-state one.
2. The 3-time-level temporal discretisation scheme within the transient SIMPLE
algorithm achieves second-order accuracy, but may be computationally
expensive. Choosing the PISO algorithm results in an accuracy comparable to
that of a formally second-order scheme, whilst being computationally
cheaper.
3. Central differencing or automatic blending (see Chapter 2, “Blending
Function” in the Methodology volume) is recommended for the discretisation
of convective terms in the momentum equation; the MARS scheme is
recommended for the turbulence equations.
LES models
1. A transient analysis setting is required, although the problem being modelled
may in reality be a steady-state one.
2. The 3-time-level temporal discretisation scheme within the transient SIMPLE
algorithm achieves second-order accuracy, but may be computationally
expensive. Choosing the PISO algorithm results in an accuracy comparable to
that of a formally second-order scheme, whilst being computationally
cheaper.
3. The time step size should be selected in such a way that the maximum
Courant number does not exceed 0.5
4. Central differencing is recommended for the discretisation of convective
terms in the momentum equation; the MARS scheme, with blending factor
not less than 0.5, can be conveniently used for bounded scalars (e.g. mixture
fractions or enthalpy).

Version 4.02 3-15


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
Heat Transfer In Solid-Fluid Systems

Changing the turbulence model in use


This facility allows you to run a turbulent flow case by restarting from a simulation
done for the same case but with a different turbulence model. No special user input
is required to run such a case.
The table below illustrates the combinations allowed and the conversion formula
adopted when STAR encounters a different turbulence model in the solution file to
the one currently in force:

FROM (Restart field)


Reynolds
k-ω
Spalart- Stress
k-ε type* (Wilcox V2F
Allmaras (GL and
and SST) SSG)

2 2 2
Spalart- k k k k
ν t = C µ ----- ν t = ---- ν t = C µ ----- ν t = C µ -----
Allmaras ε ω ε ε

k-ε type* ε = C µ kω Not needed Not needed


TO (New solution field)

k-ω ε ε ε
(Wilcox ω = --------- ω = --------- ω = ---------
and SST) Cµk Cµk Cµk

Reynolds
Stress Not needed ε = C µ kω Not needed
(GL and
SSG)

V2F Not needed ε = C µ kω Not needed

* k-ε, k-ε Quadratic, k-ε Cubic, k-ε RNG, k-ε CHEN, k-ε Speziale, k-ε Suga Quadratic and Cubic

Heat Transfer In Solid-Fluid Systems


The theory behind this type of heat transfer models and the manner of implementing
it in STAR-CD is given in Chapter 16, “Heat Transfer in Solid-Fluid Systems” of
the Methodology volume. This section contains an outline of the process to be
followed when setting up this type of model and includes cross-references to
appropriate parts of the STAR GUIde on-line Help system. The latter contains
details of the user input required and important points to bear in mind when setting
up problems of this kind.

3-16 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Heat Transfer In Solid-Fluid Systems

Setting up solid-fluid heat transfer models


Step 1
Specify the model regions occupied by the solids and fluids present and define their
physical properties.

Material 1 — steam

Heat flow

Material 3 — steel

Material 2 — hot gas

Figure 3-4 Simple heat exchanger

In terms of the heat exchanger example shown in Figure 3-4, this requires the
following actions (see also “Multi-Domain Property Setting” on page 3-5):
• Set up cell table entries for fluid materials 1,2 and solid material 3
• Assign all cells in the mesh to the appropriate cell type (1, 2, 3) as described
in the section on “Cell indexing” on page 3-3.
• Specify the physical properties of each material
Step 2
Turn on Solid-Fluid Heat Transfer in the “Thermal Options” STAR-GUIde panel.
Note that this also has the effect of switching on the temperature solver in solid
materials.
Step 3
Switch on the temperature solver in each fluid material using the “Thermal Models”
panel.
Step 4
Normally, STAR-CD treats the solid-fluid interface as part of the default wall
region (region 0). However, unlike other parts of this region whose default thermal
condition is adiabatic, the solid-fluid interface is treated as a conducting wall.
Therefore:
• If an additional thermal resistance exists at the interface, define the latter as a
separate region and use the “Define Boundary Regions” panel to specify it as
a conducting wall having the required thermal resistance value (see the STAR
GUIde “Wall” Help topic for more information).
• STAR uses default expressions to calculate heat transfer (film) coefficients at
all solid/fluid interfaces, including those at external walls and baffles. You can
supply alternative expressions for these quantities via subroutine MODSWF
Step 5
If a printout of temperature distribution in the model is required, use command

Version 4.02 3-17


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
Heat Transfer In Solid-Fluid Systems

PRTEMP to specify whether the printed values are absolute or relative to the datum
temperature previously defined (see topic “Reference Data” in the STAR GUIde
on-line Help system).
Heat transfer in baffles
Thermal conduction along the plane of a baffle’s surface is currently neglected (see
the STAR GUIde “Baffle” Help topic for more information). However, this effect
may still be modelled by expanding a baffle into a single layer of solid cells using
command CBEXTRUDE (see also Chapter 2, “Extrusion” in the Meshing User
Guide). The surrounding mesh is automatically adjusted to make room for the solid
cells, as shown in Figure 3-5.
Before After

Ordinary baffle Fully-conducting baffle

Figure 3-5 ‘Fully-conducting’ baffle creation

Note that:
• Special cell shapes (such as prisms) are created at the edges of the solid cell
layer, as shown in the exploded view of the baffle in Figure 3-5. This brings
the baffle thickness down to zero and avoids the need to create coupled cells
in those parts of the mesh.
• The modelling of heat conduction will be slightly in error as a result of the
introduction of the above artificial cell shapes.
• A baffle of the kind described here may be attached directly to an external
boundary or to internal boundaries such as solid-fluid interfaces to model a
conducting fin. In the latter case, you need to make sure that the cell type
assigned to baffle cells is different from that assigned to solid cells at the base
of the baffle.
Alternative treatment for baffle heat transfer
It can be seen that the expansion process described above will create a disturbance
in the fluid cells around the baffle and may result in a highly irregular mesh. In order
to avoid this problem, a facility is provided for specifying a finite baffle thickness
(to be used internally for heat conduction calculations) but without actually
expanding the baffle to that thickness. Thus, the fluid flow calculations are based on
an undisturbed mesh structure.

3-18 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Heat Transfer In Solid-Fluid Systems

To use this facility, the following steps are needed:


Step 1
Using the Cell Table Editor, create a separate baffle cell type and a separate solid
cell type. The latter will be used to represent the ‘conducting baffles’.
Step 2
Create the baffle cells in the appropriate mesh location using the baffle cell type
defined in Step 1.
Step 3
Apply command CBEXTRUDE to the baffle cells created in Step 2 and extrude them
into solid cells using the solid cell type created in Step 1. Note that:
• Upon extrusion, the baffle cells will be removed from the mesh and replaced
by the solid cells that they have been extruded into.
• If no solid cell type identification, ICTID, is supplied in the CBEXTRUDE
command, the solid cell identification will be set as cell type 1.
• If no solid cell thickness, DT, is supplied in the CBEXTRUDE command (this
is the normal practice), the default thickness will be applied, currently set at
0.2 × 10-3 m.
Step 4
Go back to the Cell Table Editor and select the solid cell type defined in Step 1.
Enter the actual conduction thickness in the box labelled Conduction Thickness
Step 5
Turn on Solid-Fluid Heat Transfer in the “Thermal Options” STAR-GUIde panel.
Step 6
Apply the appropriate wall boundary condition to the solid cells created in Step 3.
If none is specified, the default wall boundary condition for region number 0 will
be used. This results in a conducting, no-slip wall.

Note that:
• Conducting baffles of the same thickness DT specified in command
CBEXTRUDE and of the same Conduction Thickness specified in the Cell
Table Editor can share the same cell type.
• Conducting baffles that have a different DT or different Conduction Thickness
must also have a different cell type.
• A conducting baffle that is attached to a solid base must have a different cell
type to that of the solid to which it is attached.
Useful points on solid-fluid heat transfer
1. The On button in the Solid-Fluid Heat Transfer section of the “Thermal
Options” STAR-GUIde panel must always be used to turn on the solution of
the energy equation in solids, even if the entire model is made up of solid
cells.
2. It is usually advisable to run solid-fluid heat transfer simulations in double
precision. This helps to overcome potential convergence problems arising as a
result of a large disparity in thermal conductivity between fluid and solid. The

Version 4.02 3-19


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
Buoyancy-driven Flows and Natural Convection

choice of single or double precision mode can be made when running STAR
(see Chapter 2, “Running a STAR-CD Analysis”, Step 6).
3. A convenient way of modelling thermal contact resistance between two
adjacent solid domains is to define a baffle of suitable properties at the faces
of the appropriate solid cells in one of the domains.
4. In some situations the energy under-relaxation factor in fluid domains has to
be reduced below its default value of 1.0 to aid convergence. In such cases, we
recommend that the corresponding factor for solids is left at 1.0.
5. If your model contains an arbitrary or embedded mesh interface between the
fluid and solid cells, you will need to match cells on either side of the
interface, as described in Chapter 3, “Couple creation” in the Meshing User
Guide.
6. If your model contains scalar variables, the only valid scalar boundary
condition for walls located at the solid-fluid interface is Adiabatic.

Buoyancy-driven Flows and Natural Convection


The theory behind flow problems of this kind and the manner of implementing it in
STAR-CD is given in the Methodology volume (Chapter 16, “Buoyancy-driven
Flows and Natural Convection”). The present chapter contains an outline of the
process to be followed when setting up buoyancy-driven flows and includes
cross-references to appropriate parts of the STAR GUIde on-line Help system. The
latter contains details of the user input required and important points to bear in mind
when setting up problems of this kind.
Setting up buoyancy-driven models
Step 1
Switch on the temperature solver using the “Thermal Models” STAR-GUIde panel
Step 2
Switch on the density solver by selecting one of the following options from the
“Density” pop-up menu in the “Molecular Properties” panel:
• Isobaric — isobaric density variation (normally used for liquids)
• Ideal-f(T) — density variation based on the Ideal Gas Law
• User-f(T) — density variation based on user-defined relationships
Step 3
Set up the problem’s initial conditions using the “Initialisation” panel controls
Step 4
Define the reference pressure and temperature plus the reference pressure cell
location using the “Monitoring and Reference Data” panel
Step 5
Use the “Buoyancy” panel to specify suitable buoyancy parameters for your
problem.
Useful points on buoyancy-driven flow
1. Check the settings in STAR GUIde’s “Gravity” panel (which determine the

3-20 Version 4.02


Chapter 3 MATERIAL PROPERTY AND PROBLEM CHARACTERISATION
Fluid Injection

gravitational body force effects) before starting a buoyancy calculation. Also


note that, if droplets and/or liquid wall films are present in your model,
gravitational effects for these features must be switched on separately.
2. It is usually advisable to run buoyancy-driven flow simulations in double
precision. This is because the body force terms in the momentum equation are
often so small compared to the other terms that they can be masked by the
round-off error of the calculation. The consequences of working in single
precision mode are oscillation in the residual values and non-convergence of
the solution. The choice of single or double precision mode can be made
when running STAR (see Chapter 2, “Running a STAR-CD Analysis”, Step
6).
3. In multi-domain problems, the reference density and datum location should
be defined domain-wise.
4. If you use the option for direct specification of the reference density, the latter
should be assigned a realistic value based on the expected density variation in
the fluid. For simulations without pressure boundaries:
(a) In steady-state calculations, unrealistic values can give rise to a body
force that is out of balance with the piezometric pressure gradient. This
can cause delay in the solution convergence.
(b) In transient calculations, these initial disturbances could also produce
unrealistic initial fields and therefore invalidate the results of the analysis.

5. If convergence problems are encountered, it is advisable to begin the


calculations with a small amount of under-relaxation on both temperature and
density, e.g. 0.9. The desired values may be entered in the corresponding
Relaxation Factor boxes inside panel “Solver Parameters” in STAR GUIde.
This measure often helps to stabilise the solution and promote convergence.
6. In problems of this type, there is very strong coupling between the
temperature, scalar mass fraction and flow fields. It is therefore advisable to
use the PISO algorithm which is more suitable for this type of coupling.
7. If convergence problems are encountered, it may be necessary to run the
model in transient mode. This involves approaching the steady-state solution,
if one exists, by means of time steps. The most convenient way of doing this
is to use the single-transient solution mode (see Chapter 5, “Default
(single-transient) solution mode”), since this way one does not need to set up
load steps.
8. Buoyancy-driven flows with high Grashof number (i.e. Gr > 109) are
sometimes naturally unstable (i.e. time-dependent without a single unique
solution). In such cases, a converged steady-state solution cannot be obtained
and you should opt for the transient approach. A method of calculating the
time step size is given in the Methodology volume (Chapter 16,
“Buoyancy-driven Flows and Natural Convection”).

Fluid Injection
The theory behind flow problems of this kind and the manner of implementing it in
STAR-CD is given in the Methodology volume (Chapter 16, “Local Fluid
Injection/Extraction”). This section contains an outline of the process to be

Version 4.02 3-21


MATERIAL PROPERTY AND PROBLEM CHARACTERISATION Chapter 3
Fluid Injection

followed when setting up fluid injection problems. Also included are cross-
references to appropriate parts of the on-line Help system, containing details of the
user input required.
Setting up fluid injection models
Step 1
Create a set of all cells where fluid injection or removal is to be take place. A
separate cell table index number should be assigned to this set (see “The Cell Table”
on page 3-1).
Step 2
Activate the injection facility using the “Mass” tab in STAR-GUIde’s “Source
Terms” panel.
Step 3
Copy subroutine FLUINJ into the ufile sub-directory of your working directory,
as described in Chapter 14, “Subroutine Usage”.
Step 4
Insert appropriate code in subroutine FLUINJ using a suitable editor. Usually, the
code specifies the mass flux injected or removed (on a per unit volume basis) for
cells of the required type, so that a single value can be used for the entire cell set
selected. An example of this is given in the sample coding supplied in subroutine
FLUINJ. If only the total amount of mass injected is known, the required value may
be obtained by dividing by the total volume of the cell set. Thus, you may need to
calculate this volume first, either by choosing Utility > Calculate Volume > Cell
Set from pro-STAR’s main menu bar or by using command VOLUME.
If mass is being injected, specify all relevant properties of the incoming fluid
(i.e. it is assumed that the fluid is bringing all its properties into the computational
domain). The properties in question may be velocity components (U, V, W),
turbulence parameters (k, ε), temperature and chemical species mass fractions.
If mass is being removed, only the mass flux needs to be specified as the withdrawn
fluid is assumed to possess the (known) properties in its vicinity.

3-22 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Introduction

Chapter 4 BOUNDARY AND INITIAL CONDITIONS


Introduction
The process of defining boundaries in a model can be divided into two major steps:
1. Identify the location of individual, distinct boundaries (i.e. where the
boundaries are).
2. Specify the conditions at the boundaries (i.e. what the conditions are).

It is of the utmost importance that boundaries are chosen and implemented


correctly, since the outcome of the simulation depends on them. Users should have
a good understanding of the physical significance and numerical implications of
different boundary conditions and should apply them correctly to their model. It is
therefore advisable to refer to the relevant sections of the Methodology volume for
guidance.

Boundary Location
The two important geometrical features of boundaries are:
1. They are created on the outer surfaces of the mesh, except for:
(a) so-called baffle boundaries, which are normally positioned at the
interface of two cells;
(b) solid/fluid interface boundaries in heat transfer problems.

2. They are grouped into boundary regions. A boundary region consists of a


group of cell faces that cover the desired boundary surface. Figure 4-1 shows
a boundary region made up of nine cell faces.

Figure 4-1 Boundary region definition

The rules governing the use of boundary regions are as follows:


• Regions are numbered in an arbitrary manner by the user, in order to identify
them.
• The indexing of boundary cell faces (or boundaries, for short) comprising a
region is done automatically by pro-STAR, in a similar manner to the
automatic cell numbering discussed in “Cells” on page 2-37 of the Meshing
User Guide. In the example shown in Figure 4-2, boundary nos. 1 to 9 are
assigned to region 1.

Version 4.02 4-1


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Boundary Location

7 8 9

4 5 6

1 2 3

Figure 4-2 Boundary cell face indexing

Thus, each boundary in the model is identified by a region number (user-defined)


and composed of boundary cell faces that are automatically numbered by
pro-STAR.
pro-STAR offers two methods for setting up boundary regions:
1. Typing commands from the keyboard, as described below
2. Using the facilities of panel “Create Boundaries” in STAR GUIde (“Regions”
tab)
Command-driven facilities
The available functions are as follows:
• Assignment of boundaries to a region using the keyboard — command
BDEFINE. This requires input of the region number, cell number and cell
face number on which the boundary will be created. pro-STAR generates the
boundary number automatically.
• Further boundaries can be created individually or generated from an existing
set, using command BGENERATE. This creates additional boundaries by
applying an offset to the cell numbers of a previously-defined set.
• Modification of the region number assigned to a boundary face — command
BMODIFY.
• Re-assignment of a boundary to a different region graphically — command
BCROSS.
• Conversion of a set of shells into a set of boundaries — command BSHELL.
The starting shells are not deleted by this process.
• Counting the currently defined boundaries — command COUNT. The same
operation can also be executed by choosing Utility > Count > Boundaries
from the menu bar.

For further details on the function and application of boundary commands, refer to
the pro-STAR Commands volume.

4-2 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Boundary Location

Boundary set selection facilities


Boundaries may need to be grouped together for the purposes of mass manipulation
or plotting, thus defining a boundary set. This is done by selecting one of the list
options provided by the B-> button in the main pro-STAR window. The available
options are:
1. All — puts all existing boundaries in the current set
2. None — clears the current set
3. Invert — replaces the current set with one consisting of all currently
unselected boundaries
4. New — replaces the current set with a new set of boundaries
5. Add — adds new boundaries to the current set
6. Unselect — removes boundaries from the current set
7. Subset — selects a smaller group of boundaries from those in the current set

For the last four options, the required boundaries are collected by choosing an item
from a secondary drop-down list, as follows:
• Cursor Select — click on the desired boundaries with the cursor, complete
the selection by clicking the Done button on the plot
• Zone — use the cursor to draw a polygon around the desired boundaries.
Complete the polygon by clicking the right mouse button (or the Done button
outside the display area to let pro-STAR do it for you). Abort the selection by
clicking the Abort button.
• Region (Current) — select all boundaries whose region number is currently
highlighted in the boundary region table
• Region (Cursor Select) — select all boundaries belonging to a given region.
The required region is selected by clicking on a representative boundary with
the cursor.
• Patch (Cursor Select) — select all boundaries containing radiation patches
(see Chapter 7, Step 6). The patches in question are selected by clicking with
the cursor.
• Vertex Set (All) — all constituent vertices of the selected boundaries must be
in the current vertex set
• Vertex Set (Any) — the selected boundaries must have at least one
constituent vertex in the current vertex set
• Attach, Baffle, Cyclic, Degas (Phase Escape boundary condition used in
Eulerian multi-phase problems), Freestream, Inlet, Monitoring,
NonReflective_Pressure, NonReflective_Stagnation, Outlet, Pressure,
Radiation, Riemann, Stagnation, Symplane, Transient, Wall — all
boundaries must be of the type selected, regardless of region number

More boundary set operations are available in the Boundary List dialog (see
“Boundary listing” below) or by typing command BSET (see the pro-STAR
Commands volume for a description of additional selection options).
Boundary listing
Boundary information is displayed in the Boundary List dialog shown below,
obtained by selecting Lists > Boundaries from the main menu bar. Boundary
definitions are displayed in a scroll list in numerically ascending order, in terms of:
Version 4.02 4-3
BOUNDARY AND INITIAL CONDITIONS Chapter 4
Boundary Location

• Boundary serial number


• Parent cell serial number
• Face number of this cell that has been defined as a boundary
• Patch serial number of any radiation patch that has also been created on that
face
• Region number
• Boundary type

There is also a choice of listing all boundaries or just the current set (marked by
asterisks in the Bset column). The choice is made by simply selecting the Show All
Boundaries or Show Bset Only option, respectively.
To select boundaries from the list:
• For single items, click the number of the required boundary.
• For two or more items in sequence, click the first boundary you want to select,
and then press and hold down the Shift key while you click the last boundary
in the group.

Commands: BLIST BDELETE BMODIFY BSET

Once the desired boundaries are selected, the following additional operations are
possible:
1. Addition to (or removal from) the current set — click the Add to Set/Remove
from Set button.
2. Deletion — click the Delete Boundary button.
3. Change of boundary region — click the Change Region button. This
activates an additional dialog, shown below. To change the region type
associated with the selected boundaries, choose a different region number on
the displayed Change Region box and then click the Apply button.

4-4 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Boundary Region Definition

Note that all the above operations have an immediate effect on the boundary
definitions, reflected by immediate changes to what is displayed in the list.
However, any subsequent boundary changes made outside this dialog, e.g. by
issuing commands via the pro-STAR I/O window, will not be listed. To display
these changes, click Update List at the top of the dialog.

Boundary Region Definition


Having specified the location of all boundaries in the model, the next step is to
• define their individual type (i.e. set the boundary condition);
• supply information relevant to that type.

The boundary types available at present are:


1. Inlet
2. Outlet
3. Pressure
4. Non-reflective pressure
5. Stagnation
6. Non-reflective stagnation
7. Wall
8. Baffle
9. Symmetry plane
10. Cyclic
11. Free-stream transmissive
12. Transient-wave transmissive
13. Riemann Invariant
14. Attachment
15. Radiation
16. Monitoring
17. Phase-escape (Degassing)

The extent of the information required to define each boundary properly depends in
many cases on the variables being solved. For example, in problems using the k-ε
model, an inlet boundary needs information concerning the turbulence quantities k
and ε. In most cases, the appropriate variables are activated automatically as a result
of choosing a given modelling option, e.g.

Version 4.02 4-5


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Boundary Region Definition

• In the “Molecular Properties” STAR GUIde panel, the ideal gas option for
density will switch the density solver on
• In the “Turbulence Models” panel, any of the K-Epsilon options will switch
on the k, ε and viscosity solver

Note that:
1. In the case of a variable such as temperature, you need to switch on the
temperature solver explicitly (in panel “Thermal Models”) before proceeding
with region definitions.
2. Specification of alternative sets of variables needed to completely define
boundaries of type ‘Inlet’ or ‘Pressure’ is possible, as discussed in the
sections dealing with such boundaries.
3. It is possible to check for common mistakes in prescribing boundary
conditions (e.g. boundary velocities specified in an undefined local coordinate
system) by using the facilities available within the “Check Everything”
STAR-GUIde panel.
4. Boundary regions may be given an optional alphanumeric name to help
distinguish one region from another more easily.

The easiest way of applying a desired boundary condition to a given region is via
the STAR GUIde system; go to the Define Boundary Conditions folder and open the
“Define Boundary Regions” panel, as in the example shown below:

The number and purpose of the text boxes appearing in the panel and whether they
are active or not depends on
• the type of condition selected;
• which variables are being solved for.

4-6 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Boundary Region Definition

On the other hand, all forms of the panel possess a number of common features,
listed below:
1. New regions are defined by:
(a) Selecting an unused region in the boundary regions scroll list
(b) Choosing the desired boundary condition via the Region Type menu
options. The effect of this is to immediately display input boxes for
supplying boundary values for all flow variables required.
(c) Typing an optional name in the Region Name text box

2. Modification of existing regions is performed in a similar way. The changes


are made permanent by clicking the Apply button.
3. Additional boundary regions with identical properties to a pre-defined base
region set may also be generated by typing command RGENERATE in the
pro-STAR I/O window.
4. Selected region definitions can be deleted by clicking Delete Region.
5. The Compress button eliminates all deleted or undefined regions from the
boundary regions scroll list and renumbers the remaining ones contiguously.
6. All free surfaces in your model that are neither defined as boundaries nor
explicitly assigned to a region will become part of region no. 0 (shown in the
example above). The latter’s properties may be specified in the same way as
for any other region. By default, this region is assumed to be a smooth,
stationary, impermeable, adiabatic wall.
7. Non-uniform or time-varying conditions may be specified for some boundary
types. This is done by choosing one of the following from the Options menu
(the default setting, Standard, means constant and uniform conditions):
(a) User — specify the required conditions in one of the user subroutines
listed below (see also Chapter 14):
i) BCDEFI — Inlet
ii) BCDEFO — Outlet
iii) BCDEFP — Pressure
iv) BCDNRP — Non-reflective pressure
v) BCDEFS — Stagnation
vi) BCDNRS — Non-reflective stagnation
vii) BCDEFW — Wall or Baffle
viii) BCDEFF — Free-stream transmissive
ix) BCDEFT — Transient-wave transmissive
x) BCDEFR — Riemann invariant

The panel also displays a Define user coding button. Click it to store the
default source code in sub-directory ufile, ready for further editing.
(b) Table — use values stored in a table file as boundary conditions. The file
name is of form case.tbl (see Chapter 2, “Table Manipulation”) and
may be entered in the Table Name text box. Alternatively, the file may be
selected using pro-STAR’s built-in browser.
Note that whilst one table can be applied to multiple boundary regions,
multiple tables cannot be applied to the same boundary region. A list of

Version 4.02 4-7


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Boundary Region Definition

valid dependent variable names that may be used in tables is given for
each boundary type in the sections that follow. In addition, the coordinate
system used in a table must be the same as the coordinate system
specified for its associated boundary regions.
Table values are actually assigned to a boundary by the STAR-CD
solver during the analysis. This is done as follows:
i) Table data are mapped onto the appropriate boundary region in the
mesh
ii) Boundary face-centre coordinates are compared with the table
coordinates
iii) Variable values at face centres are calculated from the table data
using inverse distance-weighted interpolation
iv) The resulting values are assigned to the boundary for the whole
duration of the analysis

Figure 4-3 shows an example of using a table to assign boundary


conditions to a computational boundary. The coordinates and
user-supplied values are stored at the nodes of the table data grid and the
STAR flow variables are stored at the boundary face centres. In the
example, boundary values at face centre 1 are calculated as a weighted
average of the table data located at ABCD. Similarly, values at face centre
2 are a weighted average of the table data located at EFGH.

Table data map Boundary mesh Table data node

B A Boundary
face centre
1
F E
C D
2
G H

Figure 4-3 Mapping and interpolation of table data onto a boundary

Please also note the following:


i) It is possible to produce contour or vector plots of the boundary
conditions specified by the table, as a means of checking that the
table values have been entered correctly. To do this, click Plot
Boundary after you have read in the table and then specify which
flow variables you wish to plot.

4-8 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Inlet Boundaries

ii) The use of boundary condition tables is not supported for cases
using the load-step method to define transient conditions (see
Chapter 5, “Load-step based solution mode”)

(c) GT-POWER — set up a link with the GT-POWER engine system


simulation tool (see Chapter 11 of the Supplementary Notes). This
provides automatic updating of boundary conditions at inlet and/or
pressure boundaries during engine simulation runs. Note that this facility
becomes active only after the relevant option has been selected in the
“Miscellaneous Controls” panel.
(d) Rad. Eq. Tip — impose a radial equilibrium condition by specifying the
static pressure at the tip of a turbomachinery case.
(e) Rad. Eq. Hub — impose a radial equilibrium condition by specifying the
static pressure at the rotor hub of a turbomachinery case.

Inlet Boundaries
Introduction
This condition describes an inflow boundary and thus requires specification of inlet
fluxes for
• mass
• momentum
• turbulence quantities
• energy
• chemical species mass fraction

as appropriate. The same boundary type may also be used to specify an outflow
condition (i.e. ‘negative inlet’). Note that boundary values are needed only for
variables pertinent to the problem being analysed (see “Boundary Region
Definition” on page 4-5).
In specifying turbulence quantities, it is possible to select in advance the form in
which the required boundary values will be input. It is also possible to specify how
mass influx is treated under subsonic compressible flow conditions. The choices are
made in the “Define Boundary Regions” panel for inlets, as shown in the example
below, and are fully described in the “Inlet” on-line Help topic.

Version 4.02 4-9


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Inlet Boundaries

Useful points
1. If the Flow Switch and Turb. Switch settings are changed after velocity
components and turbulence boundary conditions have been input, the existing
values are not converted in any way, but are interpreted differently. You
should therefore use “Define Boundary Regions” to correct these values.
2. Boundary values for turbulence in domains using a Reynolds Stress model
may be specified solely in terms of k and ε instead of Reynolds Stress
components. If this option is chosen, turbulence conditions at the boundary
are assumed to be isotropic.
3. At negative inlets, i.e. inlet boundaries with velocity components pointing out
of the solution domain, values for temperature, turbulence quantities and
chemical species mass fractions are ignored.
4. Special considerations apply to tetrahedral meshes or meshes containing
trimmed (polyhedral) cells. If such meshes contain supersonic inlet
boundaries then, to obtain a stable/convergent solution, it is necessary to
create at least two cell layers immediately next to the boundary (see Figure
4-5 on page 4-23). If pro-STAR’s automatic meshing module is employed for
this purpose, use its built-in mesh generation capabilities. If the mesh is
imported from a package that lacks these facilities, you must extrude the mesh
in a direction normal to the boundary and then shift the boundary location to
the edge of the newly-created, layered structure.
5. If boundary conditions are set using a table (see page 4-7), the permissible
variable names that may appear in the table and their meaning is as follows:
(a) U — U-component of velocity

4-10 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Outlet Boundaries

(b) V — V-component of velocity


(c) W — W-component of velocity
(d) TE — Turbulence kinetic energy or intensity, depending on the Turb.
Switch setting
(e) ED — Turbulence kinetic energy dissipation rate or length scale,
depending on the above setting
(f) UU - Reynolds stress component
(g) VV - Reynolds stress component
(h) WW - Reynolds stress component
(i) UV - Reynolds stress component
(j) VW - Reynolds stress component
(k) UW - Reynolds stress component
(l) T — Temperature (absolute)
(m) DEN — Density
(n) Scalar_name — Mass fraction; use the scalar species name, e.g. H2O,
N2 etc. as the scalar variable name(s)

You must also ensure that the coordinate system used is the same as the
coordinate system specified in the “Define Boundary Regions” panel.
6. When running a transient, compressible flow case in which the mass flux is to
be maintained at a constant value (the Flow Switch menu setting is Mass
Flux), you must specify both velocity components and density as inlet
conditions. This applies to all methods of boundary condition input, i.e.,
Standard STAR GUIde panel entry, User coding or Table.
7. Inlet boundaries should only be placed on the external surfaces of a fluid
domain.

Outlet Boundaries
Introduction
This condition should be applied at locations where the flow is outwardly directed
but the conditions are otherwise unknown. There are two types of outlet boundary:
1. Prescribed flow split boundary. The conditions that must be observed are:
(a) The specified split factor f s must be positive.
(b) Flow splits for all outlet regions belonging to a given fluid domain should
sum to unity, i.e.

∑ fs = 1 (4-1)

(c) This type of boundary must not be used in combination with a pressure or
a stagnation pressure boundary within the same fluid domain.
(d) This type of boundary must not be used for transient compressible flow
cases

2. Prescribed mass outflow rate boundary. The conditions that must be observed
in this case are:
(a) The specified outflow rate ṁ out must be positive.
(b) This type of boundary must be used in combination with at least one
Version 4.02 4-11
BOUNDARY AND INITIAL CONDITIONS Chapter 4
Pressure Boundaries

pressure boundary.

The desired boundary type is imposed via the “Define Boundary Regions” panel for
outlets, as shown in the example below, and is fully described in the “Outlet”
on-line Help topic.

Useful points
1. Outlet boundaries of the two basic types described above must not coexist in
the same domain.
2. For solution stability and accuracy, outlet boundaries should be used only far
downstream of strong recirculation areas, where it is reasonable to expect true
outflow everywhere on the boundary.
3. Prescribed mass outflow boundaries are recommended for obtaining fully
developed flow in pipes, channels, etc.
4. The difference between outflow conditions described using negative inlet as
opposed to prescribed mass outflow boundaries is that the former prescribes
both the velocity distribution as well as the mass rate, whereas the latter
prescribes only the mass rate.
5. If boundary conditions are set using a table (see page 4-7), only one variable
name FSORMF, is allowed. The meaning of this variable is either flow split
or mass outflow rate, depending on the Condition pop-up menu setting
described above. Note that the variable must be a function of time only.
6. Outlet boundaries are incompatible with:
(a) Transonic flows
(b) Cavitating flows

7. Outlet boundaries should only be placed on the external surfaces of a fluid


domain.

Pressure Boundaries
Introduction
This condition specifies a constant static pressure or piezometric pressure on a
given boundary. For turbomachinery cases, it is also possible to specify the static
pressure at the tip or hub and impose a pressure distribution that satisfies radial
4-12 Version 4.02
Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Pressure Boundaries

equilibrium. The direction and magnitude of the flow are determined as part of the
solution. Thus,
• if the flow is directed outwards, the values of the other variables are
extrapolated from the upstream direction;
• if the flow is directed inwards, the values are obtained from the supplied
boundary conditions.

In specifying turbulence quantities, temperature, mass fraction or (optional)


tangential velocity components, it is possible to select in advance the way in which
these quantities will be determined. The choices are made in the “Define Boundary
Regions” panel for pressure boundaries, as shown in the example below, and are
fully described in the “Pressure Boundary” on-line Help topic. However, any such
conditions are only applied if the flow direction is towards the solution domain
interior.

Useful points
1. For a given fluid domain, pressure boundaries must not coexist with outlet
boundaries of the ‘Flow Split’ type.
2. Analyses with multiple pressure boundaries inherently converge more slowly
than those where the inlet flow rates and flow splits have been specified.
3. Numerical instability may occur when large or curved surfaces are used as
pressure boundaries.
4. It is advisable to choose a reference pressure that is of the same order as the
pressure values on the boundaries. For example, if the model contains two
boundaries at 10 and 11 bars a reasonable reference pressure would be 10

Version 4.02 4-13


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Stagnation Boundaries

bars. This practice will help to avoid start-up difficulties and to minimise
problems due to machine round-off errors.
5. If the Turb. Switch setting is changed to Zero Grad after turbulence boundary
conditions have been input, the values already supplied are ignored.
6. If the piezometric setting is chosen for problems involving buoyancy driven
flow, you must ensure that the datum level location and density (as specified
in the “Buoyancy” panel) are for a point lying on the pressure boundary itself.
7. In cases where a pressure boundary coexists with another pressure or
stagnation boundary, it is recommended that the user supplies an estimate for
the maximum velocity within the solution domain in the relevant text box of
the “Initialisation” panel (see also “Solution Domain Initialisation” on page
4-42).
8. To obtain a stable/convergent solution for tetrahedral meshes or meshes
containing trimmed (polyhedral) cells, use of the UVW On option (i.e.
explicit velocity specification, see the “Pressure Boundary” STAR GUIde
panel) is recommend.
9. Any type of mesh may be used for problems containing radial equilibrium
boundaries but only one such region must be employed in the model. Note
also that in cases of high circumferential velocity gradients in the radial
direction, the user may change the number of averaging intervals to capture
the problem details more accurately. The default interval value (50) is
however adequate for most cases.
10. If boundary conditions are set using a table (see page 4-7), the permissible
variable names that may appear in the table and their meaning is as follows:
(a) PR — Pressure (relative)
(b) TE — Turbulence intensity
(c) ED — Turbulence length scale
(d) T — Temperature (absolute)
(e) Scalar_name — Mass fraction; use the scalar species name, e.g. H2O,
N2 etc. as the scalar variable name(s)

11. When option Mean On is used (see the “Pressure Boundary” STAR GUIde
panel), the scope for tabular input of pressure is limited. Temporal variations
in pressure may be prescribed through tabular input, but not spatial variations.
12. Pressure boundaries should only be placed on the external surfaces of a fluid
domain.

Stagnation Boundaries
Introduction
This condition is typically used on a boundary lying in a large reservoir where fluid
properties are not significantly affected by flow conditions in the solution domain.
It normally appears in compressible flow calculations, but you may also employ it
for incompressible flows. Information relevant to such a region is supplied in the
“Define Boundary Regions” panel for stagnation boundaries, as shown in the
example below, and is described in the “Stagnation Boundary” on-line Help topic.

4-14 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Stagnation Boundaries

Useful points
1. If a fluid domain contains a stagnation boundary, it must also contain a
pressure boundary.
2. Boundary values for turbulence in domains using a Reynolds Stress model
may be specified solely in terms of k and ε instead of Reynolds Stress
components. If this option is chosen, turbulence conditions at the boundary
are assumed to be isotropic.
3. It is recommended that the user supplies an estimate for the maximum
velocity within the solution domain via the relevant text box of the
“Initialisation” panel (see also “Solution Domain Initialisation” on page
4-42). This will ensure that the calculations start with a reasonable initial
velocity field.
4. For a given fluid domain, stagnation boundaries must not co-exist with outlet
boundaries of the ‘Flow Split’ type.
5. To obtain a stable/convergent solution for tetrahedral meshes or meshes
containing trimmed (polyhedral) cells, it is necessary to create at least two
cell layers immediately next to the boundary (see Figure 4-5 on page 4-23). If
pro-STAR’s automatic meshing module is employed for this purpose, use its
built-in mesh generation capabilities. If the mesh is imported from a package
that lacks these facilities, you must extrude the mesh in a direction normal to
the boundary and then shift the boundary location to the edge of the
newly-created, layered structure.
6. Stagnation boundaries are incompatible with:
(a) Eulerian multi-phase flows

Version 4.02 4-15


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Non-reflective Pressure and Stagnation Boundaries

(b) Free surface flows


(c) Cavitating flows

7. If boundary conditions are set using a table (see page 4-7), the permissible
variable names that may appear in the table and their meaning is as follows:
(a) DCX — Direction cosine for U-component of velocity
(b) DCY — Direction cosine for V-component of velocity
(c) DCZ — Direction cosine for W-component of velocity
(d) PSTAGB — Stagnation pressure (relative)
(e) TSTAG — Stagnation temperature (absolute)
(f) TINTB — Turbulence kinetic energy or intensity, depending on the Turb.
Switch setting
(g) TLSCB — Turbulence kinetic energy dissipation rate or length scale,
depending on the Turb. Switch setting
(h) Scalar_name — Mass fraction; use the scalar species name, e.g. H2O,
N2 etc. as the scalar variable name(s)

You must also ensure that the coordinate system used is the same as the
coordinate system specified in the “Define Boundary Regions” panel.
8. Stagnation boundaries should only be placed on the external surfaces of a
fluid domain.

Non-reflective Pressure and Stagnation Boundaries


Introduction
This type of boundary condition was specially developed for turbomachinery
applications. It may only be used in situations where the working fluid is an ideal
gas and the flow is compressible. Furthermore, it requires the presence of periodic
(cyclic) boundaries in a transverse direction relative to the dominant flow direction,
as illustrated in Figure 4-4 below.

4-16 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Non-reflective Pressure and Stagnation Boundaries

Circumferential
direction

Wall

Cyclic
boundary

Flow (axial)
direction

Cyclic
boundary

Wall

Figure 4-4 Example of non-reflecting boundary mesh structure

Boundaries of this kind are frequently used as non-reflective pressure/stagnation


pairs. The information required for each type represents the average value of the
dependent variables that need to be satisfied by the simulation and is supplied in the
“Define Boundary Regions” panel. The relevant form of this panel for
non-reflecting stagnation boundaries is shown in the example below and is fully
described in the “Non-reflective Stagnation Boundary” on-line Help topic.

Version 4.02 4-17


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Non-reflective Pressure and Stagnation Boundaries

The panel for a non-reflecting pressure boundary is shown below and is fully
described in the “Non-reflective Pressure Boundary” on-line Help topic.

Useful points
1. Non-reflective pressure and stagnation conditions impose a number of
restrictions on the type of mesh employed at the boundary surface:
(a) The boundary must contain only quadrilateral faces, aligned along the
circumferential direction as shown in Figure 4-4.
(b) The cell layer adjacent to the boundary must contain only hexahedral
cells

2. Such conditions cannot be assigned to boundary region no. 0.


3. Each strip of boundary faces along the circumferential direction must be
assigned to a different non-reflective region number. However, these regions
can have the same boundary conditions.
4. The boundary surface must be delimited by cyclic boundaries along the
transverse direction, as shown in Figure 4-4.
5. If N is the number of cells along the circumferential direction, the maximum
number of harmonics to be used by the Discrete Fourier Transform algorithm
is N/2 -1. The minimum number is 0.
6. To ensure that the analysis runs smoothly, it may be necessary to start the
simulation by using standard pressure and stagnation boundary conditions
over a number of iterations. This can then be followed by a restart run where
the non-reflecting boundaries have been applied.
7. For a given fluid domain, non-reflective boundaries must not coexist with
outlet boundaries of the ‘Flow Split’ type.
8. At present, certain physical features must not be present in cases containing
non-reflective boundaries. The excluded features are:
(a) Transient calculations
(b) Chemical reactions and scalar variables
(c) Radiation
(d) Reynolds Stress and V2F turbulence models
(e) Two-phase flow

4-18 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Wall Boundaries

(f) Moving meshes


(g) Liquid films
(h) Free surface and cavitation

9. If boundary conditions are set using a table (see page 4-7), the permissible
variable names that may appear in the table and their meaning is as follows:
(a) Non-reflective pressure boundaries
i) PR — Pressure (relative static)
ii) TE — Turbulence kinetic energy or intensity, depending on the
Turb. Switch setting
iii) ED — Turbulence kinetic energy dissipation rate or length scale,
depending on the above setting

(b) Non-reflective stagnation boundaries


i) DCX — Direction cosine for U-component of velocity
ii) DCY — Direction cosine for V-component of velocity
iii) DCZ — Direction cosine for W-component of velocity
iv) PSTAGB — Stagnation pressure (relative)
v) TSTAG — Stagnation temperature (absolute)
vi) TINTB — Turbulence kinetic energy or intensity, depending on the
Turb. Switch setting
vii) TLSCB — Turbulence kinetic energy dissipation rate or length
scale, depending on the above setting

The user must also ensure that the coordinate system used is the same as the
coordinate system specified in the “Define Boundary Regions” panel.
10. Non-reflective boundaries should only be placed on the external surfaces of a
fluid domain.

Wall Boundaries
Introduction
STAR-CD’s implementation of wall boundaries involves a generalisation and
extension of the no-slip and impermeability conditions commonly used at such
surfaces. Thus, a wall boundary may be defined as:
• Of the no-slip or slip type. The latter is applicable to inviscid flows (in
practice µ is set to 10–30 Pa s). The no-slip boundary conditions for turbulent
flow are implemented using one of the methods discussed in Chapter 3,
“Turbulence Modelling”.
• Smooth or rough.
• Moving or stationary. A wall may move within the surface it defines. If
motion normal to that surface is desired, use the moving mesh features
discussed in Chapter 12, “Moving Meshes”.
• Permeable or impermeable to heat and/or mass flow.
• Resistant or not to heat flux due to a thermal boundary layer or intervening
solid material.
• Radiating or non-radiating (see also Chapter 7).

Version 4.02 4-19


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Wall Boundaries

As with other boundaries, wall boundary values are needed only for variables
pertinent to your problem (see also “Boundary Region Definition” on page 4-5).
These are specified via the “Define Boundary Regions” panel for walls, shown in
the example below, and are fully described in the “Wall” on-line Help topic.

Thermal radiation properties


In thermal radiation problems:
1. Values for thermal emissivity, reflectivity and transmissivity [dimensionless]
are required (see also Chapter 7). These should be typed in the text boxes
provided.
2. The thermal absorptivity is calculated as (1- reflectivity - transmissivity).
3. The defaults are those for a black body (emissivity equal to 1.0, reflectivity
and transmissivity equal to 0.0).

Kirchoff’s law (emissivity = absorptivity) is not enforced by the solver. For your
wall boundary condition to obey Kirchoff’s law, you must enter the condition:
emissivity = absorptivity = 1 - transmissivity - reflectivity
Solar radiation properties
In solar radiation problems:
1. External walls must be declared as Exposed or Unexposed to incident
radiation, by selecting the appropriate option from the Solar Heating pop-up

4-20 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Wall Boundaries

menu. Note that:


(a) This option does not apply to internal ‘walls’ (i.e. baffles and solid/fluid
interfaces)
(b) The external (FASTRAC) DTRM method distinguishes between direct
and diffuse solar radiation — see topic Solar Radiation. However, the
above option affects both of them equally.

2. The thermal resistance of an exposed wall to incident solar radiation is


neglected.
3. Walls can be made transparent to incident radiation, in which case a value of
transmissivity [dimensionless] should be supplied in the text box provided.
Thus, the direct solar radiation received by walls can be
(a) absorbed,
(b) reflected as diffuse radiation, or
(c) transmitted.

4. Direct radiation transmitted through transparent walls (e.g. windows), is


tracked along the angle of solar inclination (specified via the Solar Radiation
option in the “Thermal Options” panel) until it falls on an obstructing surface.
5. The remaining user input depends on the problem conditions:
(a) If only solar radiation is present:
i) The reflected diffuse radiation is neglected
ii) The absorptivity is calculated as (1 – transmissivity)

(b) If both thermal and solar radiation are present:


i) The code treats the two radiation components separately
ii) Values of reflectivity and transmissivity for each component are
supplied in separate text boxes and the corresponding absorptivity
calculated as (1 – reflectivity – transmissivity)
iii) Choosing the internal DTRM method for radiation calculations has
the effect of making the solar transmissivity equal to the thermal
transmissivity.
Other radiation modelling considerations
• The FASTRAC method must be used for thermal/solar radiation problems
with transmissive external walls.
• If the FASTRAC method is used, it is necessary to specify the transmissivity
value prior to the view factor calculation.
• The user input required under the various combinations of thermal and/or
solar radiation conditions may be conveniently summarised in the table
below:

Version 4.02 4-21


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Wall Boundaries

Table 4-1: Summary of radiative surface property requirements

Property
Condition
Emissivity Reflectivity Absorptivity Transmissivity Exposure
Thermal Y Y N (=1-R-T) Y N
Thermal Y Y N (=1-R-T) Y Y
& Solar N (=0) Y N (=1-R-T) Y
Solar N (=0) N (=0) N (=1-T) Y Y

Useful points
1. For stationary mesh cases, only velocities in directions parallel to the wall
surface may be specified, e.g. a planar wall can move only within its own
plane. For moving mesh cases, all velocity components should in general be
specified.
2. Wall function and two-layer models can be used with any kind of mesh.
However, for tetrahedral meshes or meshes containing trimmed (polyhedral)
cells, it is advisable to create at least one cell layer immediately next to the
wall boundary (see Figure 4-5 below). If pro-STAR’s automatic meshing
module is employed for this purpose, use its built-in mesh generation
capabilities. If the mesh is imported from a package that lacks these facilities,
you must extrude the mesh in a direction normal to the boundary so that the
wall is located at the edge of the newly-created, layered structure.
3. The practice recommended above is particularly important for wall
boundaries that strongly influence the character of the flow.
4. If boundary conditions are set using a table (see page 4-7), the permissible
variable names that may appear in the table and their meaning is as follows:
(a) U — U-component of wall velocity
(b) V — V-component of wall velocity
(c) W — W-component of wall velocity
(d) TORHF — Wall temperature (absolute) or heat flux
(e) RESWT — Wall thermal resistance
(f) Scalar_name — Mass fraction at the wall; use the scalar species name,
e.g. H2O, N2 etc. as the scalar variable name(s)
(g) Scalar_name-RSTSC — Wall resistance for a given species, e.g.
H2O-RSTSC, N2-RSTSC, etc.

You must also ensure that the coordinate system used is the same as the
coordinate system specified in the “Define Boundary Regions” panel.
5. Wall boundaries should only be placed on the surfaces of a fluid or solid
domain. They are the only valid boundary type for the interfaces between
solid and fluid domains.
6. Scalar boundary conditions at solid-fluid interfaces must be set to zero flux.

4-22 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Baffle Boundaries

Figure 4-5 Example of tetrahedral plus layered mesh structure

Baffle Boundaries
Introduction
Baffles are zero-thickness cells within the flow field. They represent solid or porous
domains whose physical dimensions are much smaller than the local mesh
dimensions, as shown in the example of Figure 4-6.

Figure 4-6 Example model with baffles: duct bend with turning vanes

Baffle ‘cells’ are normally defined via the Cell Tool, as described in Chapter 2, page
2-48 of the Meshing User Guide and should be placed on cell faces inside a fluid
domain. If no boundary conditions are specified for the baffle surfaces, they are
assumed to be smooth, stationary, impermeable, adiabatic walls. If one needs to
specify any other conditions, it is necessary to define special boundaries (called
baffle boundaries) explicitly on the baffle surfaces. These boundaries can then be
grouped into regions and the “Define Boundary Regions” panel can be used to apply
the desired conditions, as shown in the example below.

Version 4.02 4-23


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Baffle Boundaries

The discussion of porous media in Chapter 6 also applies, in a modified form, to


porous baffles. Thus, it is possible to calculate such a flow by formulating the
porous media equation in terms of a pressure drop, ∆p , across the baffle. The
definition of the baffle resistance coefficients is also adjusted to account for this
change. Obviously, it is now necessary to provide only one pair of such coefficients.
Setting up models
Inputs for baffle regions are very similar to inputs for walls, including a choice
between wall functions and the two-layer model (see the “Baffle” on-line Help
topic). There are a few exceptions which are noted below:
1. It is usually possible to impose different boundary conditions on either side of
the baffle. As shown in the example dialog above, conditions for Side 1 are
supplied first. It is then necessary to click the Apply button, which displays
the Side 2 dialog and a message to enter appropriate parameters for that side.
Once this is done, the process should be completed by clicking Apply a
second time.
2. The numbering of the sides is based on the manner in which the baffle was
defined. Side 1 is the ‘outward normal’ side as defined by the cross product of
two vectors pointing from the first node to the second node and from the first
node to the fourth node, as shown in Figure 4-7.

4-24 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Baffle Boundaries

4 3,3

1
1

Side 2
Side 2
2 2

Side 1 Side 1

Figure 4-7 Numbering convention for various baffle shapes

Another way of determining side numbers is to view the baffle cell and
consult the cell definition. If the ordering of the cell vertices is counter
clockwise, you are viewing Side 1.

3. The fact that the boundary conditions can be designated separately for each
side enables the user to have one side moving and the other stationary or one
side isothermal and the other side adiabatic. The conditions can be mixed in
any combination with two exceptions:
(a) If the thermal boundary condition for Side 1 of the baffle is Conduction,
STAR-CD calculates the one-dimensional heat transfer across the baffle
based on the local temperature and flow conditions on either side. This
choice of boundary condition naturally excludes a different choice for
Side 2 and therefore the Wall Heat pop-up menu is deactivated for that
side. An exception to this rule occurs when thermal radiation is switched
on, in which case radiation properties for both sides of the baffle need to
be supplied.
(b) In a similar way, if the baffle is porous, only one set of resistance
coefficients is needed. The required values are supplied as input for Side
1. Since these naturally apply to the entire baffle, no input is necessary for
Side 2.

Specific input required for baffles is fully described in the STAR GUIde “Baffle”
Help topic. The user should supply values first for Side 1 and then for Side 2 (with
the exceptions noted above).
Thermal radiation properties
In thermal radiation problems:
1. Values for thermal emissivity, reflectivity and transmissivity [dimensionless]
are required (see also Chapter 7). These should be typed in the text boxes
provided.
2. The absorptivity is calculated as (1 – reflectivity – transmissivity).
3. The defaults are those for a black body (emissivity equal to 1.0, reflectivity

Version 4.02 4-25


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Baffle Boundaries

and transmissivity equal to 0.0).

The effect of baffle transmissivity is taken into account during the view factor
calculations. Therefore, any changes in transmissivity during the run (for example,
as part of a transient calculation) will activate beam tracking and a re-calculation of
view factors. Note that use of transparent baffles is restricted to surface-to-surface
radiation only and thus excludes participating media radiation.
Note also that Kirchoff’s law (emissivity = absorptivity) is not enforced by the
solver. For your baffle boundary condition to obey Kirchoff’s law, you must enter
the condition:
emissivity = absorptivity = 1 - transmissivity - reflectivity
Solar radiation properties
In solar radiation problems, user input depends on the problem conditions:
1. If solar radiation only is present:
(a) It is assumed to be completely absorbed by the baffle (i.e. absorptivity =
1)
(b) The reflected diffuse radiation is neglected
(c) As a result, no user input is required

2. If both thermal and solar radiation are present:


(a) Values for emissivity, reflectivity and transmissivity [dimensionless] are
supplied separately for each radiation component
(b) Choosing the internal DTRM method for radiation calculations has the
effect of making the solar transmissivity equal to the thermal
transmissivity. Therefore, only a reflectivity value needs to be supplied
for the solar component.
Other radiation modelling considerations
• If the FASTRAC method is used, it is necessary to specify the transmissivity
value prior to the view factor calculation.
• The user input required under the various combinations of thermal and/or
solar radiation conditions may be conveniently summarised in the table
below:
.

Table 4-2: Summary of radiative surface property requirements

Property
Condition
Emissivity Reflectivity Absorptivity Transmissivity Exposure
Thermal Y Y N (=1-R-T) Y N
Thermal Y Y N (=1-R-T) Y N
& Solar N (=0) Y N (=1-R-T) Y
Solar N (=0) N (=0) N (=1) N (=0) N

4-26 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Symmetry Plane Boundaries

Useful points
1. For stationary mesh cases, only velocities in directions parallel to the baffle
surface may be specified, e.g. a planar baffle can move only within its own
plane. For moving mesh cases, all velocity components should in general be
specified.
2. If boundary conditions are set using a table (see page 4-7), the permissible
variable names that may appear in the table and their meaning is as follows:
(a) U — U-component of baffle velocity
(b) V — V-component of baffle velocity
(c) W — W-component of baffle velocity
(d) TORHF — Baffle temperature (absolute) or heat flux
(e) RESWT — Baffle thermal resistance
(f) Scalar_name — Mass fraction; use the scalar species name, e.g. H2O,
N2 etc. as the scalar variable name(s)
(g) Scalar_name-RSTSC — Baffle resistance for a given species, e.g.
H2O-RSTSC, N2-RSTSC, etc.

You must also ensure that the coordinate system used is the same as the
coordinate system specified in the “Define Boundary Regions” panel.

Symmetry Plane Boundaries


Symmetry boundaries are used for two purposes:
1. To reduce the size of the computational mesh by placing the boundary along a
plane of geometrical and flow symmetry.
2. To approximate a free-stream boundary.

No user input is required beyond definition of the boundary location. The quantities
set to zero at the boundary are:
• The normal component of velocity
• The normal gradient of all other variables

Symmetry boundaries should only be placed on the external surfaces of a fluid or


solid domain. They cannot be used in FASTRAC radiation calculations.

Cyclic Boundaries
Introduction
Cyclic boundaries impose a repeating or periodic flow condition on a pair of
geometrically identical boundary regions, numbers 1 and 2 in the example of Figure
4-8. Selected scalar variables are forced to be equal at corresponding faces on the
two regions. As shown in Figure 4-8, velocity components are also equalised in a
common local coordinate system specified by the user. Such boundaries thus serve
to reduce the size of the computational mesh. This is illustrated by the example of
Figure 4-9, showing a cascade of repeating baffles.

Version 4.02 4-27


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Cyclic Boundaries

U2
V1 U1
V2

Cyclic boundary 2 Cyclic boundary 1


YL
U1 = U2
RL
V1 = V2
W1 = W2
ΘL
XL Local cylindrical system

Figure 4-8 Cyclic conditions defined using a local coordinate system

Cyclic boundary 1

Inlet

Cyclic boundary 2

Figure 4-9 Regular cyclic boundaries with integral match

Setting up models
Cyclic boundaries are defined using STAR GUIde panels in the following multi-
stage process:
1. In panel “Create Boundaries”, use tab “Regions” to set up a pair of regions, of
identical size and shape, and designate them as cyclic
2. In tab “Cyclics”, specify a number of parameters that enable them to be
matched to each other geometrically and which take into account the mesh
characteristics at either end. This involves the following considerations:
(a) Specification of suitable coordinate increments (offsets) that allow one

4-28 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Cyclic Boundaries

member of the pair to be located if one starts at the other member. A local
coordinate system in which the regions are matched is also specified.
(b) Whether the regions form a regular cyclic (as in Figure 4-9) or an
anticyclic pair (as in Figure 4-10). The latter appears in problems where
all flow variable profiles have to be reversed in a specified direction of
the matching coordinate system. This operation also reverses the
coordinate value of each boundary face in that direction before adding the
corresponding offset. Thus, placing the coordinate system origin on an
axis of symmetry and choosing its location carefully can eliminate the
need for offsets, as in the anticyclic system shown in Figure 4-10.

Cyclic boundary 1

Local Cartesian coordinate system

Cyclic boundary 2

Figure 4-10 Partial anticyclic boundaries with integral match

(c) Whether there is a one-to-one correspondence between boundary faces on


either side of the cyclic pair, as in the examples shown in Figure 4-9 and
Figure 4-10. This requires a so-called integral matching operation. If no
such correspondence exists, typically because one side is more finely
meshed than the other (as in Figure 4-11), the system requires an
arbitrary matching operation. The latter is similar to matching cell faces
on either side of an interface between mesh blocks (see Chapter 3,
“Arbitrary connectivity” in the Meshing User Guide). It thus involves
matching of so-called master boundary faces on one side of the cyclic pair
with slave faces on the other side.

3. In tab “Cyclics”, finish up by performing the geometric matching operation


between boundary faces on either side of the pair to form so-called cyclic sets.
Note that the same operation may also be performed manually, whereby each
cyclic set and the boundaries contributed by each cyclic pair member are
named explicitly using command CYCLIC. The list of cyclic pairs can also be
extended with the CYGENERATE command, beginning from a pre-existing
starting set.

Version 4.02 4-29


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Cyclic Boundaries

Figure 4-11 Cyclic boundaries with arbitrary match

4. In panel “Define Boundary Regions”, specify the physical cyclic boundary


conditions that exist between the members of the pair, as shown below:

These can be of two types:


(a) Ordinary cyclic conditions, whereby all flow variable values on one
member are matched with the corresponding values on the other member.
(b) Partial cyclic conditions, whereby the matching process is subject to an
additional constraint of either a prescribed pressure drop or a fixed mass
flow rate across the cyclic pair. An example of a fixed mass flow rate
system, representing one half of a continuous loop flow system, is shown
in Figure 4-10. For thermal problems, the bulk mean temperature on the
inflow side of the cyclic pair is also required.
Useful points
1. One member of the cyclic pair must be designated as an Inflow boundary and
the other as an Outflow boundary

4-30 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Cyclic Boundaries

2. If the partial cyclic condition is specified via the STAR GUIde interface, the
Inflow and Outflow sides are indicated via the Flow Direction menu and the
pressure drop or flow rate must be a positive number. Note that this number
must be the same for both members of the pair.
3. If the partial cyclic condition is specified via the RDEFINE command, the
inflow and outflow sides are distinguished by assigning a pressure drop or
flow rate to one member that is equal in magnitude but of opposite sign to that
for the other member. The sign convention is as follows:
(a) Pressure Drop
+ Inflow
– Outflow

(b) Flow rate


+ Outflow
– Inflow

4. Partial cyclic conditions can only be applied to boundaries matched in


Cartesian coordinates
5. Such conditions are not available for chemical species mass fractions and
cannot be used in variable-density flows
6. Arbitrary cyclic matching (see page 4-29 above) is not allowed for partial
cyclic conditions
7. Cyclic boundaries cannot be used in FASTRAC radiation calculations.
8. Cyclic boundaries should only be placed on the external surfaces of a fluid
domain.
Cyclic set manipulation
All currently defined cyclic sets are shown in the Cyclic Set List below:

Commands: CYLIST CYDELETE CYCOMPRESS

The list may be displayed by choosing Lists > Cyclic Sets from the main menu bar.
The sets are numbered and listed in numerically ascending order, together with their
constituent master and slave boundary numbers for arbitrarily matched regions (see
page 4-29 above). There is a choice of showing all cyclic sets (click button Show
All Cyclic Sets) or just those with at least one member (master or slave boundary)
in the current boundary set (click button Show Cyclic Sets with Boundaries in

Version 4.02 4-31


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Free-stream Transmissive Boundaries

Bset Only). Items in the second category are marked by asterisks in the Bset
column.
To select cyclic sets from the list:
• For single items, click the required set number.
• For two or more items in sequence, click the first set you want to select, press
and hold down the Shift key and then click the last set in the group.

Once the desired sets are selected, the following operations are possible:
• Deletion — click on the Delete button.
• Compression — click on the Compress button. This involves the elimination
of all deleted cyclic sets and renumbering of the remaining ones.

A third operation, for validating arbitrarily matched cyclic boundaries, is


implemented in the “Check Everything” panel. The operation checks that
• all sets in a given range exist and reference arbitrarily-matched cyclic regions;
• there is overlap between boundaries on the two sides of the cyclic set;
• the overlapping areas from either side match up.

All checks are performed to within a specified tolerance.

Free-stream Transmissive Boundaries


Introduction
This type of boundary may be used only in models involving supersonic free
streams where the working fluid is an ideal gas. The facility enables shock waves
generated in the interior of the solution domain to be transmitted, without reflection,
through the boundary to the wider (free stream) space surrounding the domain.
Flow can be out of the solution domain (compression waves) or into the solution
domain (expansion waves). In either case, boundary values of scalar variables are
extrapolated from the solution domain interior. In the case of turbulent inflow
(expansion waves), the turbulence quantities have to be specified as part of the user
input.
To set up boundaries of this kind, you need to:
1. Decide on an appropriate location for the boundary, preferably parallel to the
main (supersonic) stream.
2. Supply values in the “Define Boundary Regions” panel for all free-stream
properties, as shown in the example below. The required input is fully
described in the “Free-stream Transmissive Boundary” on-line Help topic.

4-32 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Free-stream Transmissive Boundaries

The STAR-CD solver calculates the magnitude and direction of the flow at the
boundary as part of the analysis, based on the simple wave theory given in [3] and
[4].
Useful points
1. A value of temperature at the boundary is obligatory. The user must therefore
ensure that temperature calculations are activated, via the “Thermal Models”
panel, before defining the boundary conditions.
2. Boundary values for turbulence in domains using a Reynolds Stress model
may be specified solely in terms of k and ε instead of Reynolds Stress
components. If this option is chosen, turbulence conditions at the boundary
are assumed to be isotropic.
3. To obtain a stable/convergent solution for tetrahedral meshes or meshes
containing trimmed (polyhedral) cells, it is necessary to create at least two
cell layers immediately next to the boundary (see Figure 4-5 on page 4-23). If
pro-STAR’s automatic meshing module is employed for this purpose, use its
built-in mesh generation capabilities. If the mesh is imported from a package
that lacks these facilities, you must extrude the mesh in a direction normal to
the boundary and then shift the boundary location to the edge of the
newly-created, layered structure.
4. Boundary conditions specified in a table will be applied only if fluid is
entering the solution domain from the outside. If this is not the case, i.e.the
flow is parallel to the boundary or crossing it from inside the domain,
boundary values will be extrapolated from interior values and the table data
will not be used.
5. If boundary conditions are set using a table (see page 4-7), the permissible
Version 4.02 4-33
BOUNDARY AND INITIAL CONDITIONS Chapter 4
Transient-wave Transmissive Boundaries

variable names that may appear in the table and their meaning is as follows:
(a) UINF — U-component of velocity
(b) VINF — V-component of velocity
(c) WINF — W-component of velocity
(d) PINF — Pressure (relative)
(e) TINF — Temperature (absolute)
(f) TEINF — Turbulent kinetic energy or intensity, depending on the Turb.
Switch setting
(g) EDINF — Turbulent kinetic energy dissipation rate or length scale,
depending on the Turb. Switch setting

You must also ensure that the coordinate system used is the same as the
coordinate system specified in the “Define Boundary Regions” panel.
6. Free-stream transmissive boundaries should only be placed on the external
surfaces of a fluid domain.

Transient-wave Transmissive Boundaries


Introduction
This type of boundary may be used only in transient, compressible flows where the
working fluid is an ideal gas. It enables transient waves to leave the solution domain
without reflection. STAR-CD uses the simple wave theory to calculate conditions
behind the wave and to specify such conditions at the boundaries.
To set up boundaries of this kind, you need to:
1. Decide on an appropriate location for the boundary
2. Supply values in the “Define Boundary Regions” panel for all dependent
variables, representing conditions outside the boundary (at ‘infinity’). The
required input is fully described in the “Transient-wave Transmissive
Boundary” on-line Help topic.

4-34 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Transient-wave Transmissive Boundaries

STAR-CD calculates the magnitude and direction of the flow at the boundary as
part of the analysis, based on the transient wave theory given in [3] and [4].
Useful points
1. A value of temperature at the boundary is obligatory. The user must therefore
ensure that temperature calculations are activated, via the “Thermal Models”
panel, before defining the boundary conditions.
2. Boundary values for turbulence in domains using a Reynolds Stress model
may be specified solely in terms of k and ε instead of Reynolds Stress
components. If this option is chosen, turbulence conditions at the boundary
are assumed to be isotropic.
3. To obtain a stable/convergent solution for tetrahedral meshes or meshes
containing trimmed (polyhedral) cells, it is necessary to create at least two
cell layers immediately next to the boundary (see Figure 4-5 on page 4-23). If
pro-STAR’s automatic meshing module is employed for this purpose, use its
built-in mesh generation capabilities. If the mesh is imported from a package
that lacks these facilities, you must extrude the mesh in a direction normal to
the boundary and then shift the boundary location to the edge of the
newly-created, layered structure.
4. Boundary conditions specified in a table will be applied only if fluid is
entering the solution domain from the outside. If this is not the case, i.e.the
flow is parallel to the boundary or crossing it from inside the domain,
boundary values will be extrapolated from interior values and the table data
will not be used.
5. If boundary conditions are set using a table (see page 4-7), the permissible

Version 4.02 4-35


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Riemann Boundaries

variable names that may appear in the table and their meaning is as follows:
(a) UINF — U-component of velocity
(b) VINF — V-component of velocity
(c) WINF — W-component of velocity
(d) PINF — Pressure (relative)
(e) TINF — Temperature (absolute)
(f) TEINF — Turbulent kinetic energy or intensity, depending on the Turb.
Switch setting
(g) EDINF — Turbulent kinetic energy dissipation rate or length scale,
depending on the above setting

You must also ensure that the coordinate system used is the same as the
coordinate system specified in the “Define Boundary Regions” panel.
6. Transient-wave transmissive boundaries should only be placed on the external
surfaces of a fluid domain.

Riemann Boundaries
Introduction
This type of boundary is typically employed in external aerodynamics simulations
and may be used only if the working fluid is an ideal gas. It enables weak pressure
waves to leave the solution domain without reflection and is valid for both
steady-state and transient problems.
To set up boundaries of this kind, you need to:
1. Decide on an appropriate location for the boundary
2. Supply values in the “Define Boundary Regions” panel for all dependent
variables, representing conditions outside the boundary (at ‘infinity’). The
required input is fully described in the “Riemann Boundary” Help topic.

4-36 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Riemann Boundaries

STAR-CD calculates the magnitude and direction of the flow at the boundary as
part of the analysis, based on the Riemann invariant theory given in [7].
Useful points
1. Check that the density of the domain to which such a boundary belongs is set
to Ideal-f(T,P)
2. Boundary values for turbulence in domains using a Reynolds Stress model
may be specified solely in terms of k and ε instead of Reynolds Stress
components. If this option is chosen, turbulence conditions at the boundary
are assumed to be isotropic.
3. A value of temperature at the boundary is obligatory. The user must therefore
ensure that temperature calculations are activated, via the “Thermal Models”
panel, before defining the boundary conditions.
4. To obtain a stable/convergent solution for tetrahedral meshes or meshes
containing trimmed (polyhedral) cells, it is necessary to create at least two
cell layers immediately next to the boundary (see Figure 4-5 on page 4-23). If
pro-STAR’s automatic meshing module is employed for this purpose, use its
built-in mesh generation capabilities. If the mesh is imported from a package
that lacks these facilities, you must extrude the mesh in a direction normal to
the boundary and then shift the boundary location to the edge of the
newly-created, layered structure.
5. Boundary conditions specified in a table will be applied only if fluid is
entering the solution domain from the outside. If this is not the case, i.e.the
flow is parallel to the boundary or crossing it from inside the domain,
boundary values will be extrapolated from interior values and the table data

Version 4.02 4-37


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Attachment Boundaries

will not be used.


6. If boundary conditions are set using a table (see page 4-7), the permissible
variable names that may appear in the table and their meaning is as follows:
(a) UINF — U-component of velocity
(b) VINF — V-component of velocity
(c) WINF — W-component of velocity
(d) PINF — Pressure (relative)
(e) TINF — Temperature (absolute)
(f) TEINF — Turbulent kinetic energy or intensity, depending on the Turb.
Switch setting
(g) EDINF — Turbulent kinetic energy dissipation rate or length scale,
depending on the above setting
(h) Scalar_name — Mass fraction; use the scalar species name, e.g. H2O,
N2 etc. as the scalar variable name(s)

You must also ensure that the coordinate system used is the same as the
coordinate system specified in the “Define Boundary Regions” panel.
7. Riemann boundaries should only be placed on the external surfaces of a fluid
domain.

Attachment Boundaries
Attachment boundaries are used for the following two purposes:
1. To define the interface between cells that may be connected or disconnected
from each other (see Chapter 12, “Cell Attachment and Change of Fluid
Type”).
2. To define the interface between mesh blocks that slide past each other, either
in an ‘integral’ or ‘arbitrary’ manner — see “Regular sliding interfaces” on
page 12-18.

Two input parameters are needed:


• A local coordinate system in which the boundaries are to be matched
• An alternate boundary region number

The second parameter is required for cell layer attachment cases and serves to
maintain appropriate boundary conditions in the solution domain if the cells on
either side of the interface become disconnected. The alternate boundary region

4-38 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Radiation Boundaries

must be of ‘wall’ or ‘inlet’ type. Examples of cases requiring attachment boundaries


are given in Chapter 12.
Useful points
1. To obtain a stable/convergent solution for meshes containing trimmed
(polyhedral) cells, it is necessary to create at least two cell layers immediately
next to the boundary (see Figure 4-5 on page 4-23). If the pro-STAR Auto
Mesh module is employed, use its built-in mesh generation capabilities for
this purpose. If the mesh is imported from a package that lacks such facilities,
you must extrude the mesh in a direction normal to the boundary and then
shift the boundary location to the edge of the newly-created, layered structure.
2. Attachment boundary regions must be created in pairs, one on each of the
mesh blocks that are attached to, detached from or sliding past each other.
3. Attachment boundaries should be placed on the surfaces of domains/
subdomains.

Radiation Boundaries
Radiation boundaries are used for the purpose of separating that part of your model
where radiation effects are important from other parts where such effects are
negligible. This type of boundary only influences radiation calculations and is
completely transparent to the fluid flow and non-radiative heat transfer in your
model.

Two input parameters are needed (see also Chapter 7):


1. The boundary radiation temperature [K], normally set to a value close to the
expected temperature in the surrounding area
2. The boundary surface emissivity [dimensionless], normally set to 1.0

The location and properties of such a boundary should be chosen so that:


• Radiant energy passing through it escapes to the outside world with minimal
back-radiation into the sub-domain where it emanated. The escaped radiation
should be low enough not to influence conditions in the outside world.
• Its presence does not adversely affect the accuracy of the calculations inside
the radiative sub-domain(s)
• If a coupled-cell interface exists between the radiative and non-radiative
sub-domains, the boundary must be placed on the cells that are inside the

Version 4.02 4-39


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Phase-Escape (Degassing) Boundaries

radiative sub-domain.
Useful points
1. The use of radiation boundaries is not required in radiation problems
employing the Discrete Ordinates method. Turning radiation off for a
particular cell type is sufficient to exclude radiation calculations in that mesh
region.
2. Radiation boundaries are currently incompatible with FASTRAC radiation
calculations.
3. Radiation boundaries should be placed on cell faces inside a fluid domain.

Phase-Escape (Degassing) Boundaries


This type of boundary appears exclusively in Eulerian multi-phase problems (see
Chapter 10 of this volume) and represents a degassing free-surface bounding a
two-phase system of gas bubbles in a liquid, corresponding to the dispersed and
continuous phases, respectively. The boundary conditions applied to each phase are
as follows:
1. For the continuous phase, the boundary acts like a slip wall, allowing the
liquid to flow parallel to the boundary surface without friction
2. For the dispersed phase, the boundary acts like an opening allowing bubbles
to escape into the surrounding medium, unless retained within the solution
domain by the drag forces acting on them.

No further user input is required on the “Define Boundary Regions” panel. Note that
only one boundary of this type should be present in your model. Degassing
boundaries should only be placed on the external surfaces of a fluid domain.

Monitoring Regions
These are arbitrary surfaces, defined in the same way as ordinary boundaries but
placed on any cell faces within a fluid or solid domain so as to form internal
surfaces. They are used purely for monitoring engineering data such as mass flux
(see panel “Monitor Boundary Behaviour”) so no further user input is required on
the “Define Boundary Regions” panel.
Monitoring regions do not affect the flow field in any way; STAR simply
calculates the monitored data values at the specified region’s surface and stores
them for subsequent display as a function of time or number of iterations (see panel
“Engineering Data”). The same data values are available at monitoring regions as
at open boundary regions, except that:
• Item Heat Flux is not available
• Field values are taken from the neighbouring cell centres and are not
interpolated to the boundary
• Item Enthalpy In/Out is based on convection only, so it will be zero in solid
materials

Each face of a monitoring region “belongs” to a neighbouring cell, such that the
mass flux is defined as being positive when it leaves this cell through the face. This
in turn determines the face’s orientation and, for consistent calculation of the total

4-40 Version 4.02


Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Boundary Visualisation

mass flux through the region, it is important that all its faces are oriented the same
way.
The choice of which cell a monitoring region face belongs to is made when that
face is defined. Commands BFIND, BCROSS, and BZONE do this by picking a cell
face, as do their associated GUI operations, and are therefore suitable for this
purpose. On the other hand, commands BDEF, BGEN and BDX should not be used
to define monitoring region faces as they do not involve the explicit selection of a
cell face and hence the orientation of the monitoring region face is indeterminate.
Caution should also be exercised when generating such regions automatically, for
example by cell refinement. When visualising monitoring regions, their orientation
is indicated using an arrow normal to the boundary and whose direction indicates
the direction of positive flux, as shown in Figure 4-12 below.

Figure 4-12 Monitoring region display

Boundary Visualisation
As described in “Boundary set selection facilities” on page 4-3, boundaries can be
collected into sets. The currently defined set can then be displayed on top of the
calculation mesh by choosing Cell Plot Display Option Bound from the main
window and re-plotting. The cell faces representing the boundaries will be marked
by distinctive fill patterns and colours, characteristic of the boundary type
represented. Boundary faces will be superimposed on any kind of plot already
displayed on the screen other than a section plot. Note that the boundary display
option may also be selected by choosing Plot > Cell Display > Boundaries from
the menu bar. Alternatively, you may type commands BDISPLAY, ON or
CDISPLAY, BREGION in the I/O window.

Version 4.02 4-41


BOUNDARY AND INITIAL CONDITIONS Chapter 4
Solution Domain Initialisation

Solution Domain Initialisation


Steady-state problems
User action depends on whether the solution is to start from the initial state of the
model (initial run) or to continue from a previously computed solution (restart run).
Initial runs
Initial conditions for flow field variables are assigned in STAR-GUIde’s
Thermophysical Models and Properties folder:
• For fluid field variables, use panel “Initialisation” in the Liquids and Gases
sub-folder. Note that there is a separate “Turbulence tab” for initializing
turbulence parameters. If this is done by specifying the initial turbulence
intensity I and length scale l, the turbulence kinetic energy k and dissipation
rate ε are computed as follows:
2 2
k = 1.5U I (4-2)

1.5
k
ε = -------- (4-3)
l

where U is the initial velocity magnitude. For turbulence models other than
the k-ε type, the turbulence scales (ω for k-ω models or ν t for the
Spalart-Allmaras model) are computed automatically.
• For chemically reacting flows, you may also need to use panel “Initialisation”
in the Additional Scalars sub-folder to specify initial mass fractions for
chemical species.
• In conjugate heat transfer problems, another panel also called “Initialisation”
in the Solids sub-folder can be used to specify initial temperatures in solid
materials.
Restart runs
Various options for this operation are available in panel “Analysis (Re)Start” within
the Analysis Preparation/Running folder. If option Standard Restart is chosen, the
solution from a previous run serves as the starting point for the current run. If Initial
Field Restart is chosen in this panel, the STAR-CD solver only corrects the mass
fluxes to satisfy continuity. The Initial Field Restart option should be chosen if
any change has been made to the boundary conditions or reference quantities
(pressure and/or temperature). This option must also be chosen if new scalars have
been defined by selecting additional modelling options such as Lagrangian
multi-phase or chemical reaction.
Special considerations apply to cases where the restart also involves a change in
the mesh configuration, typically a refinement of a coarser starting mesh. These are
covered in Chapter 5, “Solution Control with Mesh Changes”.
Transient problems
In transient problems, all flow field variables should be given the correct values for
the problem at hand. Depending on the physical conditions being modelled, this can
be done in one of the following ways:
1. Specify uniform values — select option Constant in the “Initialisation” panel
4-42 Version 4.02
Chapter 4 BOUNDARY AND INITIAL CONDITIONS
Solution Domain Initialisation

and then type values for each variable in the text boxes provided. Turbulence
parameters, scalar mass fractions and solid temperatures may be initialised as
described in section “Initial runs” above for steady-state cases.
2. Set values through a user-supplied subroutine — select option User in the
Initialization panel and then specify the required distributions in subroutine
INITFI.
3. Read in a previously computed distribution that corresponds to the desired
setting — select an option from the “Analysis (Re)Start” panel (usually
Initial Field Restart plus one of the options in the Initial Field Restart
pop-up menu depending on the problem at hand). Option Standard Restart
must be used for all moving mesh cases and should also be chosen to start a
transient analysis from a previously computed steady-state solution. Note that
such restarts should not be performed for Lagrangian multi-phase cases, as
the meaning of the droplet treatment is different between steady-state and
transient analyses.

Version 4.02 4-43


Chapter 5 CONTROL FUNCTIONS
Introduction

Chapter 5 CONTROL FUNCTIONS


Introduction
At this stage of modelling, the following tasks should have been completed:
1. Mesh set up
2. Material property and continuum mechanics model specification
3. Definition of boundary type and location

The penultimate task before a STAR analysis run is to set the parameters that
control that run. This consists of
• setting various parameters that affect the progress of the numerical solution
algorithm used by STAR;
• specifying the type and amount of run-time output and post-processing data.

The user should also decide whether the problem is steady-state or transient so as to
perform the appropriate operations for the above tasks.

Analysis Controls for Steady-State Problems


Solution controls
Solution control parameters have a strong influence on the progress of the analysis,
so it is important to have a basic understanding of their significance and effect
during a run. You are therefore advised to refer to Chapter 7 in the Methodology
volume for a detailed discussion of under-relaxation and other solution control
topics.
STAR-CD offers two alternatives for solving steady-state problems:
1. An iterative method employing under- relaxation factors
2. A pseudo-transient time marching to the steady-state solution with a
fixed-length time step. Note, however, that an under-relaxation factor (default
value 0.2) is still used on the pressure correction equation.

Incompressible, non-reacting and low Mach number flows usually converge


smoothly and fast in a combination with the inertial under-relaxation shown in
equation (7-14) of the Methodology volume.
If fluid flows which are characterized by travelling waves (e.g. pressure waves
in compressible fluids, or gravity waves in free surface flows) can reach a steady
state, the convergence process is typically much more robust and stable if one can
resolve to some extent the waves travelling during the iteration process. For this, it
is important that waves travel in all cells with the same pseudo time step, and the
pseudo-transient mode is usually better suited to this class of problems.
In other problems (e.g. inviscid flows and where the initial velocity field is zero),
we can have very small or zero values of the central coefficients A P at an early stage
of the iteration process. In such cases, local pseudo time steps become very large
or infinite (see equation (7-19)), which again has an impact on the convergence and
stability of the solution. In flows which exhibit this kind of problem, use of the
pseudo-transient mode is recommended.
The main advantage of the iterative method employing under-relaxation factors
compared with the pseudo-transient mode is that, in the former, the under-relaxation
Version 4.02 5-1
CONTROL FUNCTIONS Chapter 5
Analysis Controls for Steady-State Problems

factors vary between 0 and 1 and a large number of cases run nearly optimally with
default values (e.g. 0.7 for momentum and 0.2 for pressure correction) that are
based on considerable past experience. However, in the pseudo transient approach,
the time step varies between zero and infinity and a suitable value is not always easy
to find. The optimum value can be determined only by numerical experiments. As
a guideline, one should choose a time step such that the Courant number based on
the characteristic velocity and the characteristic mesh size is between 1 and 8. Note
that, as with under-relaxation, lower values of time step are more likely to promote
convergence, while larger ones lead to a faster solution.
The task of setting up solution controls for either of these methods can be divided
into the following steps:
Step 1
Start up the STAR GUIde system and then define the type of problem you are
solving by selecting Steady State from the Time Domain pop-up menu in the
“Select Analysis Features” panel
Step 2
Go to the Solution Controls folder and open the “Solution Method” panel. From the
pop-up menu at the top of the panel select:
• Steady State for conventional steady-state runs. Also choose the numerical
algorithm to be used (see topic “Steady-State Solution”). In every case,
specify the maximum residual error tolerance (i.e. maximum acceptable level
of remaining error in the solution), plus any additional parameters required by
the algorithm you have chosen.
• Pseudo-Transient for pseudo-transient runs (see topic “Pseudo-Transient
Solution”). The maximum residual error tolerance (i.e. maximum acceptable
level of remaining error in the solution) should be specified; the normalised
residuals are displayed on the screen and also saved on file case.run, as in
ordinary steady-state runs.
Step 3
In the “Primary Variables” panel, inspect the solution status for flow variables and
material properties (see topic “Equation Status”) to confirm that the right variables
will be solved for.
Step 4
Check the “Solver Parameters” (under-relaxation factors, number of calculation
sweeps and residual error tolerances for each solution variable).
Step 5
Choose one of the available “Differencing Schemes”. It is suggested that
higher-order differencing schemes such as LUD or MARS should be used if high
spatial discretisation accuracy is required.

Output controls
Having set the solution control parameters, the next task is to choose the type and
volume of output from the forthcoming STAR run. The bulk of this output consists
of solution variable values at cell centroids. Output controls can be applied by going
to the Output Controls folder in the STAR GUIde system and following the steps

5-2 Version 4.02


Chapter 5 CONTROL FUNCTIONS
Analysis Controls for Steady-State Problems

below:
Step 6
Consider whether detailed printout on the solution progress is required and if
necessary specify the appropriate settings in the “Monitor Numeric Behaviour”
panel.
Step 7
Decide whether you want to follow the progress of the analysis by generating
various types of monitoring data at every iteration. If so, go to the Monitor
Engineering Behaviour sub-folder and use one or both of the following panels:
• “Monitor Boundary Behaviour” — select one or more boundary regions and
the type of monitoring information to be generated for them
• “Monitor Cell Behaviour” — select one or more sub-domains, defined in
terms of cell sets, and the type of monitoring information to be generated for
them

The requested data are stored in special files (case.erd and case.ecd for
boundary and cell data, respectively), from where they may be displayed as
pro-STAR graphs at the end of the analysis (see panel “Engineering Data” in the
Post-Processing folder) or read by an external post-processing package.
Step 8
Specify the manner of saving mesh data for use in post-processing and/or restart
runs via the “Analysis Output” panel (“Steady state problems”). If desired, go to the
“Additional Output Data” section to select any wall data to be included in the
solution (.ccm) file. This is important, as these settings will affect the availability
of data for post-processing. You can also select what wall data are to be ‘printed’
(i.e. displayed on your screen) and stored in the .run file at the end of the run.
For both post and print control parameters, it is up to you to check the default
settings and change them, if necessary, according to the type of problem being
analysed.

Other controls
Step 9
Go to the Sources sub-folder and inspect the “Source Terms” panel to see if any
additional information (such as extra source terms for flow variables) is needed to
completely describe your problem. Note that STAR-CD provides special switches
and constants for activating various beta-level features in the code, or for turning on
calculation procedures designed for debugging purposes. These are found in the
“Switches and Real Constants” panel and are normally used only after consultation
with CD-adapco. An alternative way of performing this function is to enter special
debugging instructions into the Extended Data panel, accessible from the Utilities
menu in the main window (or issue command EDATA).
Step 10
Go to the Analysis Preparation/Running folder and open the “Set Run Time
Controls” panel:
• For conventional steady-state runs, enter the maximum number of iterations

Version 4.02 5-3


CONTROL FUNCTIONS Chapter 5
Analysis Controls for Transient Problems

(or calculation loops, see “Steady state problems”).


• For pseudo-transient runs, specify the time step size and the maximum
number of time steps. A variable step magnitude may also be specified via
user subroutine DTSTEP, by selecting option User in the Time Step Option
pop-up menu (see “Steady state problems (Pseudo-transient)”).
Step 11
To complete the controls specification, you need to decide whether the analysis is
to start from initial conditions or restart from a previous run. Set the appropriate
solution controls in the “Analysis (Re)Start” panel.

Analysis Controls for Transient Problems


Transient problems can be divided into three groups:
1. Systems whose flow, thermal and chemical fields are originally in
thermodynamic equilibrium and which are subjected to a set of
non-equilibrium boundary conditions at the start of the calculation. The
system’s response is to gradually approach a new steady state. Such problems
can be analysed in STAR either in the steady-state or transient mode; some
buoyancy driven flows are best run in transient mode (see also Chapter 3,
“Buoyancy-driven Flows and Natural Convection”).
2. Systems whose boundary conditions change in a prescribed fashion, e.g. due
to opening and shutting of flow valves.
3. Inherently unstable systems that never reach a steady state and exhibit either
(a) a cyclic (or periodic) behaviour, as in some vortex shedding problems, or
(b) chaotic behaviour, as in some buoyancy driven flows.

Procedures for solving all of these problem types are described below.
Default (single-transient) solution mode
This procedure, referred to as the ‘single-transient’ solution mode in earlier versions
of STAR-CD, is the quickest and easiest way of setting up transient problems. It is
also suitable for steady-state compressible or buoyancy driven flows that require
close coupling between the momentum, enthalpy, chemical species and density
equations. Other important characteristics are:
• It is fully supported by pro-STAR’s STAR GUIde interface
• It can accommodate problems with time-varying boundary conditions through
the use of tables (see Chapter 2, “Table Manipulation”)
• Changes in boundary region type (e.g. a pressure boundary changing to a wall
boundary) are also possible but require stopping and restarting the analysis at
those times when such changes occur

The single-transient mode provides an alternative to the “Load-step based solution


mode” discussed below, by eliminating the need for a transient history file and
explicit load step definitions. It is in fact equivalent to performing a single load step,
hence the name ‘single transient’. To use this approach, follow the procedure below.

5-4 Version 4.02


Chapter 5 CONTROL FUNCTIONS
Analysis Controls for Transient Problems

Solution controls
Step 1
Start up the STAR GUIde system and then define the type of problem you are
solving by selecting Transient from the Time Domain pop-up menu in the “Select
Analysis Features” panel.
Step 2
Go to the Solution Controls folder and open the “Solution Method” panel. Select the
numerical algorithm to be used (see topic “Transient problems”). Also select the
time differencing scheme required.
Step 3
Display the “Primary Variables” panel and check that the solution parameter
settings are appropriate for your case. If there is any need for alterations, consult
Chapter 1, “Transient flow calculations with PISO” or “Transient flow calculations
with SIMPLE” in this volume for information and advice.

Output controls
The output to be produced by a transient run is chosen in a similar manner to that
for steady-state problems. However, since the volume of data that can be generated
is potentially very large, additional controls are provided to limit the amount to what
is absolutely essential.
Step 4
Open the “Analysis Output” panel (“Transient problems”).
1. In the “Post tab”, specify control parameters for the wall data that will be
written to the solution (.ccm) file and/or printed and saved in the .run file at
the end of the run, in the same manner as for steady-state problems.
2. In the “Transient tab”, specify control parameters for data destined for:
(a) The transient post data (.pstt) file. The difference between this and the
usual solution (.ccm) file is as follows:
i) File case.ccm contains analysis results only for the last time step.
These form a complete set of all cell data relevant to the current
problem and the file can therefore be used to restart the analysis.
ii) File case.pstt, on the other hand, contains user-selected data,
such as cell pressures, wall heat fluxes, etc. written at
predetermined points in time. These are defined by the parameters
entered in the “Transient tab”. The file is therefore suitable for
post-processing runs but cannot be used to restart the analysis.

(b) The data display appearing on your screen at predetermined points in


time (not necessarily the same as the ones specified for the post data).
This information is also saved in the run history (.run) file.

The “Transient tab” control parameters must be used with care since they could
cause excessively large data files to be written. On the other hand, they must not be
used too sparingly as they may fail to record important data. If the analysis is split
into several stages, as is usually the case with large models and/or lengthy

Version 4.02 5-5


CONTROL FUNCTIONS Chapter 5
Analysis Controls for Transient Problems

transients, it is advisable to give the .pstt file produced at the end of each stage a
unique file name. This helps to spread the output produced amongst several files
and thus eases the data management and manipulation processes.
Step 5
Specify any other output controls required, e.g. whether you want to generate
monitoring data at every time step, in the same manner as for steady-state problems
(see “Analysis Controls for Steady-State Problems”, Step 7).

Other controls
Step 6
Specify any other necessary controls in the Sources and Other Controls sub-folders,
in the same manner as for steady-state problems.
Step 7
Go to the “Set Run Time Controls” panel (Analysis Preparation folder) and specify:
1. The analysis run time
2. The method of calculating the time step size and the total number of time
steps, see “Transient problems”
Step 8
To complete the controls specification, you need to decide whether the analysis is
to start from initial conditions or restart from a previous run. Set the appropriate
solution controls in the “Analysis (Re)Start” panel.

Load-step based solution mode


This older procedure allows for all intricacies in the transient problem specification,
including variable boundary conditions. However, it is more complex to set up and
maintain as it requires definition of so-called ‘load steps’ (see “Load step
characteristics” below) and their storage in special transient history files. Other
important characteristics are:
• It is driven by its own special user interface, the Advanced Transients dialog,
accessed by selecting Modules > Transient in pro-STAR’s main menu bar
• Time variations may be specified only in terms of load steps, as described in
the sections to follow; the use of tables is not permissible
• It is part of the recommended procedure for setting up moving-mesh cases
defined via pro-STAR ‘events’ (see Chapter 12, “Moving Meshes”)
• It does not support models containing features introduced in STAR-CD V3.20
or later, such as Eulerian multi-phase and liquid films
Load step characteristics
For problems involving changing boundary conditions, the main considerations are:
• To define the variation in boundary conditions as a series of events which
occur over a period of time. These events, called load steps in pro-STAR
terminology, represent a transition from one state of the boundary conditions
to another with increasing time.
• To divide each load step into several time increments, or time steps.

5-6 Version 4.02


Chapter 5 CONTROL FUNCTIONS
Analysis Controls for Transient Problems

The allowable variations in the boundary conditions are as follows:


1. Step — where the boundary values change discontinuously from one state to
the next, see Figure 5-1(a).
2. Ramp — where the values change linearly between the state at the beginning
of the load step to that at the end, see Figure 5-1(b).
3. Function of time — where the variation is arbitrary and is prescribed via a
user subroutine.

Any combination of load step types can be specified, as shown in Figure 5-1(c)–(d).
Boundary Boundary
condition condition
value value

R R R R R
S S S S

1 2 3 4 Time 1 2 3 4 5 Time
(a) (b)

Boundary Boundary
condition condition
value value

S S R R R R R R R S S

1 2 3 4 5 Time 1 2 3 4 5 6 Time
(c) (d)

Figure 5-1 Representation of boundary value changes by load steps

The difference between the available alternatives is illustrated in Figure 5-2 for load
step number n and a time increment of DT.
The following information is specified every time a load step is defined:
1. The number of time steps to be performed.
2. The boundary values prevailing at the end of the load step.
3. The manner in which the boundary values should vary between the start and
end of the load step. The action of the program is then as follows:
(a) For step settings, the value at the start and at all intermediate times is kept
equal to value at the end time, as specified in stage 2. above.
(b) For ramp settings, the value at the start is made equal to that specified at
the previous load step. All intermediate values vary linearly between the
start and end values, as shown in Figure 5-2.

Version 4.02 5-7


CONTROL FUNCTIONS Chapter 5
Analysis Controls for Transient Problems

(c) For user settings, values between the start and end times vary in an
arbitrary manner, according to what is prescribed in the user-supplied
subroutine (see Figure 5-2).

Boundary condition value

Load step n–1

A
Load step n
User coding
Ramp

Step Load step n+1

DT
Time

Figure 5-2 Types of change in boundary conditions

Some examples of different load step sequences are shown in Figure 5-1 where the
letters S and R denote a step or ramp setting respectively.
Load step definition
The user should bear in mind the following points when defining load steps:
1. Special considerations apply if the very first load step has a ramp setting. This
is because there is no previous load step to fix the value of its starting point.
The problem is resolved by defining an extra, dummy load step which merely
serves to supply the required boundary value. Examples of this situation are
shown in Figure 5-1, cases (b) and (d).
2. At each new load step, the user is free to modify any existing boundary region
definition. For example, boundaries that were previously outlets can now
become walls and vice versa. However, new boundary regions cannot be
added or existing ones deleted, nor can the physical extent of the boundaries
be modified in any way. The user must therefore plan the model’s boundary
region definitions adequately before starting a transient analysis. A step
setting is always imposed at every boundary type change.
3. When the boundary values at the start and end of a load step are identical, the
sole purpose of defining the load step would be to permit subdivision of time
into discrete time steps so as to track the transient behaviour of the flow field.
4. The time step size can vary from one load step to the next to suit the problem
conditions. The size should be small enough to meet the following two

5-8 Version 4.02


Chapter 5 CONTROL FUNCTIONS
Analysis Controls for Transient Problems

targets:
(a) Stability of the numerical solution algorithm, by minimising the
cumulative error in the numerical solution.
(b) Capture of the transient details of the flow.

A good way of testing the sufficiency of the time step size is by calculating the
Courant number Co, a dimensionless quantity given by

v ∆t
Co = ------------ (5-1)
l

where v and l are a characteristic velocity and dimension, respectively. Note that
in compressible flows v should be replaced by v + c , where c is the velocity of
sound. For optimum results, the user should calculate the Courant number in two
ways:
1. Cell-wise, by setting v to an estimated local velocity and l to the
corresponding local mesh dimension (e.g. cell diagonal). The time step should
be chosen such that the maximum Courant number does not exceed 100.
2. Globally, by setting v to the estimated average velocity in the flow field and
l to a characteristic overall dimension of the model (e.g. pipe length in pipe
flow). The time step should be chosen so that it is commensurate with the time
scale of the physical process being modelled. Although precise figures cannot
be given for all cases, a Courant number derived from this criterion is
typically in the range 100 to 500.

The user should inspect the time steps derived in these two ways and select the
smallest one for use in the analysis.
Solution procedure outline
The overall task of setting up parameters for a load-step based transient calculation
can be divided into the following steps:

Solution controls
Step 1
Start up the STAR GUIde system and then define the type of problem you are
solving by selecting Transient from the Time Domain pop-up menu in the “Select
Analysis Features” panel.
Step 2
Go to the Solution Controls folder and open the “Solution Method” panel. Select the
numerical algorithm to be used (see topic “Transient problems”).
Step 3
Display the “Primary Variables” panel and check that the solution parameter
settings are appropriate for your case. If there is any need for alterations, consult
Chapter 1, “Transient flow calculations with PISO” or “Transient flow calculations
with SIMPLE” in this volume for information and advice. Note that this information
is stored for each load step in file case.trns. Therefore, if any changes are
needed to these parameters after your load steps have been defined, you will need
Version 4.02 5-9
CONTROL FUNCTIONS Chapter 5
Analysis Controls for Transient Problems

to retrieve the load step information, make the changes and then save the
information back in the .trns file (see the description of Step 4 and Step 5 below)
Load step controls
Step 4
Choose Modules > Transient from the menu bar to activate the Advanced
Transients dialog shown below. Select option Advanced Transients On by
clicking the action button at the top right-hand side of the dialog.
Type the maximum load step number that will be specified in the text box
provided and then click Initialize to set up a file (case.trns) for storing all
transient history information (i.e. changes in boundary conditions, distribution and
length of time steps, etc.). This is a binary file that works very much like the normal
pro-STAR problem description (.mdl) file, but is used only in transient problems.
The file’s name is entered in the Transient File text box.
For a restart run, click the Connect action button to retrieve existing load step
information. Note that a number of different files can be utilised in a given run, by
first clicking Disconnect to release the current file and then connecting to a new
one, as specified in the Transient File text box. pro-STAR’s built-in file browser
may be used to locate the required file(s). If necessary, a revised maximum load step
number should be typed in the box provided.

Commands: TRFILE LSTEP LSLIST LSSAVE

5-10 Version 4.02


Chapter 5 CONTROL FUNCTIONS
Analysis Controls for Transient Problems

LSCOMPRESS LSRANGE LSGET LSDELETE


MVGRID CPRINT CPRANGE WPRINT
TDSCHEME CPOST SCTRANS WPOST
CDTRANS

Step 5
If the SIMPLE solution algorithm has been chosen, select a time differencing
scheme from the Temporal Discretization pop-up menu. Option Euler Implicit
selects the (default) first-order Euler implicit scheme while Three Time Level
Implicit selects the second-order three-time-level implicit scheme. The latter gives
more accurate solutions but requires more computer time and memory. Your choice
of the time differencing scheme should be confirmed by clicking Apply.
Step 6
Supply in a sequential manner all information needed to completely define each
load step. The current load step should be indicated by highlighting it in the scroll
list with the mouse. The required information depends on the time-varying
character of the problem and can consist of:
1. Basic parameters of the load step — type these in the text boxes underneath
the load step list. The available parameters are:
(a) Load step identifying number.
(b) Number of time steps.
(c) Time increment per time step — if the option button next to this text box
is selected, pro-STAR will look for time increment definitions in user
subroutine DTSTEP. Any number typed in the text box will be available
to the subroutine as a default value.
(d) A choice of step or ramp setting for changes in the boundary conditions
(note that the ramp setting cannot be chosen if the User option is already
selected in step (c) above).
(e) Output frequency of print and post-processing data (see “Output controls”
below).

2. Redefinition of the boundary type, e.g. changing from wall to outlet boundary
conditions and vice versa to simulate the operation of an exhaust valve in a
reciprocating engine — see “Boundary Region Definition” on page 4-5.
3. Modification of selected boundary values, without changing the boundary
type, as shown in Figure 5-1 — see page 4-7 in the section on “Boundary
Region Definition”.
4. Unusual boundary value changes, i.e. other than step-wise or ramp-wise —
see option User in the section on “Boundary Region Definition” on page 4-7.
The desired variation should be calculated in the appropriate user subroutine
(BCDEFI, BCDEFO, BCDEFS, BCDEFP, BCDEFF, BCDEFT, or BCDEFW, see
“Boundary condition subroutines” on page 14-5). These routines should
supply the required values at every time step and for all boundary regions
affected. Any region not covered in this way will take on the usual ramp or
step variation specified during the basic load step parameter setting.

Remember that in cases where the boundary conditions are to vary linearly from the

Version 4.02 5-11


CONTROL FUNCTIONS Chapter 5
Analysis Controls for Transient Problems

start of the calculation, it is necessary to supply the boundary conditions at the start
of the calculations. This is achieved by introducing a dummy first load step with
ramp setting. With the exception of the boundary condition, all other data for such
a load step are ignored.

Output controls
The output to be produced by a transient run is chosen in a similar manner to that
for steady-state problems. However, since the volume of data that can be generated
is potentially very large, additional controls are provided to limit the amount to what
is absolutely essential. These controls are implemented in the Advanced Transients
dialog and can be sub-divided into a number of basic steps as described below. Note
that they are part of the definition for a given load step and can be repeated as
necessary during subsequent load steps to achieve the desired fine control over the
type and volume of output.
Step 7
Decide whether printed output is required. If so, specify:
• The printout frequency (in terms of a time step interval) by typing a suitable
value in the Print Freq. text box.
• The cell variables (e.g. velocities, pressure, temperature, etc.) to be printed —
click the appropriate Cell Print selection button underneath the desired
variable(s).
• The part of the mesh over which the above quantities will be printed — type a
suitable cell range in terms of starting, finishing and increment cell number in
the text boxes provided.
• The wall variables (e.g. shear forces, heat fluxes, etc.) to be printed — click
the appropriate Wall Print selection button underneath the desired
variable(s).

If some of the cell or wall variables to be printed are additional scalar variables such
as chemical species mass fraction, they are specified via the Scalars Select
selection button (see Chapter 13, “Multi-component Mixing”, Step 8).
Step 8
Decide whether post-processing information is required. If so, specify:
• The output frequency (in terms of a time step interval) by typing a suitable
value in the Post Freq. text box.
• The cell variables (e.g. velocities, pressure, temperature, etc.) to be stored —
click the appropriate Cell Post selection button underneath the desired
variable(s).
• The wall variables (e.g. shear forces, heat fluxes, mass fluxes, etc.) to be
stored — click the appropriate Wall Post selection button underneath the
desired variable(s).

If some of the cell or wall variables to be written are additional scalar variables such
as chemical species mass fraction, they are specified via the Scalars Select button
(see Chapter 13, “Multi-component Mixing”, Step 8).

All the above information is written to a special transient post data (.pstt) file.

5-12 Version 4.02


Chapter 5 CONTROL FUNCTIONS
Analysis Controls for Transient Problems

The difference between this and the usual solution (.ccm) file is as follows:
• File case.ccm contains the calculation results of only the last time step.
These form a complete set of all cell data and the file can therefore be used to
restart the analysis.
• File case.pstt, on the other hand, contains user-selected data, such as cell
pressures, wall heat fluxes, etc. written at predetermined points in time
defined by the parameter typed in the Post Freq. text box. The file is therefore
suitable for post-processing runs but cannot be used to restart the analysis.

The Print and Post Freq. parameters above must be used with care since they,
together with their associated print and post file operations, may cause excessively
large data files to be written. On the other hand, they must not be used too sparingly
as they may fail to record important data. If the analysis is split into several stages,
as is usually the case with large models and/or lengthy transients, it is advisable to
give the .pstt file produced at the end of each stage a unique file name. This helps
to spread the output produced amongst several files and thus eases the data
management and manipulation processes.

Other load step and general solution controls


Step 9
Store each completed load step definition in the transient history (.trns) file by
clicking on the Save action button. The parameters of the saved definition are
displayed in the Load Step scroll list.
Step 10
Once all the necessary load steps have been defined, set the total number of load
steps to be performed during the next STAR analysis by typing the starting and
finishing load step number in the text boxes provided. Confirm by clicking the
Apply button.
Note that all the above operations have an immediate effect on the transient
settings, reflected by immediate changes to what is displayed in the dialog box.
However, any subsequent changes made outside this box, e.g by issuing commands
via the pro-STAR I/O window, will not be shown. To display these changes, you will
need to click the Update button at the bottom of the dialog.
Step 11
In addition to the load-step specific information described above, you may also
request additional, detailed information that applies to the run as a whole. This
includes:
• Values of the field variables at a monitoring cell location at each time step.
The desired location is specified in the “Monitoring and Reference Data”
STAR-GUIde panel. One monitoring cell must be selected for each different
material present in the model.
• Various types of engineering data, as selected from the Monitor Engineering
Behaviour panels for specified grid and/or boundary regions. These are also
produced at each time step.
• Input data, boundary conditions and locations, inner iteration statistics, etc.
These options are set in the “Monitor Numeric Behaviour” panel.

Version 4.02 5-13


CONTROL FUNCTIONS Chapter 5
Analysis Controls for Transient Problems

Step 12
Specify any other necessary controls in the Sources and Other Controls sub-folders,
in the same manner as for steady-state problems.
Step 13
The total number of time steps for the run is normally equal to the sum of all time
steps in each load step, as defined in Step 6. However, this total may be set
independently via command ITER, which may effectively stop the run in the
middle of a load step.
Step 14
To complete the controls specification, you need to decide whether the analysis is
to start from initial conditions or restart from a previous run. Set the appropriate
solution controls in the “Analysis (Re)Start” panel.
Other transient functions
Before initiating a transient run, the user is free to review and modify the existing
set of load step definitions. The relevant facilities available in the Advanced
Transients dialog are:
• Modification — highlight the load step to be changed, type values for the
modified parameters and click Save.
• Deletion — highlight the load step to be deleted and click Delete.
• Compression of the transient history file — clicking Compress eliminates all
deleted steps and renumbers the remaining ones.

Additional points to bear in mind about transient problems are:


1. An analysis can most conveniently be performed in stages, using an initial
and several restart runs. When specifying a restart run, you must remember to
(a) read in the state of the model as it was when the last run finished, using
the “Analysis (Re)Start” STAR-GUIde panel (Standard Restart option)
(b) reconnect to the transient history (.trns) file, as described in page 5-10,
Step 4 of this section, if additional load steps are to be specified.

2. Along with time-varying boundary values and boundary conditions, you may
also elect to vary the geometry of his model, e.g. by moving the mesh in a
cylinder-and-piston problem. This can be done by selecting On in the Moving
Grid Option pop-up menu at the top of the Advanced Transients dialog. This
operation also requires either
(a) a user-defined subroutine (NEWXYZ) to calculate the vertex coordinates as
a function of time, or
(b) the use of special commands provided in the EVENTS module (see
Chapter 12, “Moving Meshes”). These permit changes to both vertex
locations and cell connectivities.

The modified vertex coordinates are also written to the transient post data
(.pstt) file and can be loaded and plotted during post-processing.

5-14 Version 4.02


Chapter 5 CONTROL FUNCTIONS
Solution Control with Mesh Changes

Solution Control with Mesh Changes


The discussion so far in this chapter assumes the usual condition of identical mesh
geometry between restart runs. However, it sometimes becomes apparent that
changes in mesh geometry applied part-way through the solution process will
improve the quality of the final result. For example, inspection of the current
solution file may reveal that mesh refinement is needed in some part of the mesh to
resolve the flow pattern adequately. Rather than beginning a new analysis from
scratch with a new, refined mesh, STAR-CD allows redefinition of the mesh and
resumption of the analysis (via a restart run) from the currently available solution.
This requires a special mapping operation, called SMAP, that utilises the existing
solution data in the .ccm file to create new, approximate solution data that
correspond to the re-defined mesh. An example of the result of such an operation is
given in Chapter 10, “Solution Mapping” of the Post-Processing User Guide.
STAR can read this new file and restart the analysis to obtain a proper solution
for the current mesh.
Mesh-changing procedure
A description of the steps necessary for performing a mesh-changing operation
requiring refinement is given below. Note that although restarting with a refined
mesh is typical, the same rules apply to any other mesh re-definition,
e.g. coarsening, changing cell shapes, or even creating a mesh structure that is
physically larger (or smaller) overall than the original configuration.
Step 1
Check the directory of your current (coarse-mesh) model to confirm that a
pro-STAR model file (say, case-coarse.mdl) and a STAR solution file (say,
case-coarse.ccm) exist.
Step 2
Start a pro-STAR session and read in the coarse-mesh model from
case-coarse.mdl. Then:
• Select File > Case Name from the main window, change the case name to,
say, case-fine and click Apply
• Perform whatever mesh refinement operations are necessary (see for example
Chapter 3, “Mesh Refinement” in the Meshing User Guide)
• Select File > Write Geometry File from the main window and save the
refined mesh geometry in file case-fine.ccm

Step 3
Signal to STAR that the next run will restart from a different (mapped) solution still
to be created:
• Go to the Analysis Preparation/Running folder in STAR GUIde and open the
“Analysis (Re)Start” panel
• Select option Initial Field Restart from the Restart File Option menu
• Accept the (default) Restart File name (case-fine.ccm)
• Select option Restart (Smapped) from the Initial Field Restart menus and
click Apply

Version 4.02 5-15


CONTROL FUNCTIONS Chapter 5
Solution Control with Mesh Changes

Step 4
Save all information for the refined mesh, including the restart mode specification
above:
• Select File > Save Model from the main window to save file
case-fine.mdl
• Select File > Write Problem File from the main window to save file
case-fine.prob

Step 5
Restore the original coarse-mesh model as follows:
• Select File > Resume From in the main window
• Input the original mesh by specifying case-coarse.mdl as the model file
and click Apply
• Go to the Post-Processing folder in STAR GUIde and display the “Load
Data” panel.
• Read in the coarse-mesh solution data by specifying case-coarse.ccm as
the input file name and then clicking Open Post File

Step 6
Select those coarse-mesh cells that should be used in the mapping process and put
them in a cell set (see “Cell set selection facilities” on page 2-46 of the Meshing
User Guide). This is because SMAP operates only on cells in the current set. This set
may include both fluid and solid cells and will normally contain all cells in the
model.
The SMAP operation itself is initiated by choosing Utility > Solution Mapping
from the main window menu bar to display the Smap/Tsmap dialog shown below:

Commands: SMAP TSMAP

The required user input is as follows:


1. Input CCM file — The refined mesh file, case-fine.ccm, created in Step

5-16 Version 4.02


Chapter 5 CONTROL FUNCTIONS
Solution-Adapted Mesh Changes

2 above and containing only geometry data at present. If necessary, use


pro-STAR’s built-in file browser to locate the file.
2. Output CCM file — The refined mesh solution file, case-fine.ccm. At
the end of the mapping operation, this will contain (mapped) solution data as
well as geometry data and will therefore be suitable as a restart file for a fine
mesh analysis.
3. Instructions on how to assign flow variable values to any fine-grid cells that
may lie outside the domain defined by the coarse-grid cells. The available
Outside Options are:
(a) Default — use default values, as defined in panel “Initialisation” of
sub-folder Liquids and Gases in STAR-GUIde
(b) Nearest — use values from the nearest cell neighbours
(c) Zero — use a value of 0.0

Note that the default mapping algorithm is selected by the Use Smap button.
Clicking the Use Tsmap button activates a slightly different algorithm that attempts
to enforce global conservation on the fine-grid domain. Other ways in which this
option differs from the standard option are as follows:
1. It is not applicable to polyhedral fluid cells
2. Only two Outside Options are available, Nearest and Zero.
3. The volume made up by the fine-grid cells should be fully contained within
the volume of the coarse-grid cells. This condition may be satisfied within a
tolerance (specified as a volume fraction) entered in the Volume Tolerance
box.
Step 7
To visualise the outcome of the mapping operation, use the “Load Data”panel in
STAR GUIde’s Post-Processing folder. The mapped solution data file just created
may be accessed via the “File(s) tab” and field values loaded via the “Data tab”. The
data may then be checked by plotting contours but note that only “Cell Data” should
be used for this purpose.
Step 8
If the mapped results are deemed satisfactory, terminate the pro-STAR session
without writing a model file (as this would save the original coarse-grid data) and
then run STAR to continue the analysis from the mapped solution.

Other noteworthy points are:


• If option Use Tsmap is selected in moving mesh problems containing
removed cells (see “Cell-layer Removal/Addition” on page 12-14), the cell set
to be mapped should not include removed cells.
• If any baffles are present in the coarse-grid domain to be refined and mapped,
delete the baffles before refinement and redefine them after refinement.
• Do not change the reference temperature in the restart run.

Solution-Adapted Mesh Changes


Section “Solution Control with Mesh Changes” of this chapter shows how to
transfer a solution from one mesh to another. In that section, Step 2 simply states

Version 4.02 5-17


CONTROL FUNCTIONS Chapter 5
Solution-Adapted Mesh Changes

that you need to perform whatever mesh refinement operations are necessary. This
section aims to show how these changes can be made using the solution from a
previous run as a guide.
The most frequently used refinement procedures have been assembled in the
“Adaptive Refinement” panel of the STAR GUIde system. This caters for mesh
refinement based on the results of a previous run. One may employ a refinement
operation based on either
• flow variable gradients, or
• solution residuals.

Both types of data are stored in the solution (.ccm) file and one may then choose
the flow variable and selection method to be employed. A typical refinement
session would consist of the following steps:
Step 1
Go to the Analysis Preparation/Running folder in the STAR GUIde system and
open the “Adaptive Refinement” panel. In the “Refinement Criteria” tab, choose a
criterion by selecting the appropriate sub-tab. The flow variable on which to base
the refinement depends on the application. For flow-dominated problems, the
velocity magnitude or the turbulence kinetic energy have been found to give good
results; for chemical reaction- dominated problems, the temperature might be a
better choice. Note that:
• Using the Percent of Cells selection method allows you to closely control the
number of cells selected for refinement
• You may perform multiple selections based on different variables and
different criteria; the selection results are accumulated into a compound cell.
set
• You may abandon your current selection at any stage and start again by
clicking the New button
Step 2
Go to the “Set Modifications” tab and select set modification options, e.g.
• The Near Wall Cell Options may be used to ensure that near-wall cells are left
unrefined when limitations on the magnitude of y+ need to be observed.
• The final set can be ‘grown’, i.e. expanded to include neighbouring cells, to
account for inaccuracies in the error estimate and to prevent large differences
in refinement level between neighbouring parts of the mesh.

Check the set to be refined visually by plotting it. If necessary, last-minute


modifications can be made to this set using the standard pro-STAR cell set utilities
(see “Set Manipulation” on page 2-21).
Step 3
Once the required cells are finally selected, the “Refine” tab enables you to
• refine them using a simple 2 × 2 × 2 subdivision,
• recreate the cell connectivity,
• prepare the resulting new model for the next run. This last step entails
mapping the old solution to the new geometry, changing the solution mode to
a restart run from the new (mapped) .ccm file and redefining the monitoring
5-18 Version 4.02
Chapter 5 CONTROL FUNCTIONS
Solution-Adapted Mesh Changes

and pressure reference cells, if these were within the area that has been
refined.

Note, however, that there are many other ways to proceed. Consider filling the
volume occupied by the chosen cells with one or more blocks (maybe after a little
padding out) and then specifying block factors to build a mesh with progressive,
concertina-style refinement. You may also choose to fill the volume with a
completely new mesh built by any pro-STAR operation or imported from an
external package (see “Importing Data from other Systems” on page 3-1 of the
Meshing User Guide). The reverse effect, coarsening the mesh, may by achieved via
one of the above methods or by using the CJOIN command.

Version 4.02 5-19


Chapter 6 POROUS MEDIA FLOW
Setting Up Porous Media Models

Chapter 6 POROUS MEDIA FLOW


The theory behind flow problems of this kind and the manner of implementing it in
STAR-CD is given in Chapter 8 of the Methodology volume. The present chapter
contains an outline of the process to be followed when setting up a porous media
problem and includes cross-references to appropriate parts of the on-line Help
system. The latter contains details of the user input required and important points to
bear in mind when setting up problems of this kind.

Setting Up Porous Media Models


Step 1
Index the cells in the area where distributed resistance exists. This requires use of
the cell table (see Chapter 3, “The Cell Table”). As an example, consider the
specification of a filter in the pipe shown in Figure 6-1.

flow in cell index cell index cell index flow out


1 2 1

filter

Figure 6-1 Flow through filter in a pipe

• For the non-filtered sub-domains (using cell index 1, fluid material property
index 12 and porous material index 0) the Cell Table Editor would look as
follows:

• For the filtered sub-domain (using cell index 2, fluid material property index
12 and porous material index 11) the Cell Table Editor would look as follows:

Version 4.02 6-1


POROUS MEDIA FLOW Chapter 6
Setting Up Porous Media Models

The reason for using an identical fluid property index (i.e. 12) is that both
sub-domains are part of the same fluid domain.
Step 2
Supply property values (resistance coefficients and porosity) for the porous
sub-domain using the “Resistance and Porosity Factor” STAR-GUIde panel. If your
model contains multiple porous sub-domains possessing different properties, each
sub-domain may be selected in turn via the Porous Material # control at the bottom
of the panel (see also the “Porosity” Help topic).

Coordinate system 5
(cylindrical) x2
x1 = r x1
x2 = θ 5 x
3
x3 = z
14
z

r θ

12
Coordinate system 1
(Cartesian) x2 y Honeycombs
x1 = x x x
x2 = y 1 x1
3
x3 = z

Figure 6-2 Coordinate system definition in pipe with honeycomb sections

Thus, for the example shown above, the Resistance and Porosity Factor panel
6-2 Version 4.02
Chapter 6 POROUS MEDIA FLOW
Setting Up Porous Media Models

settings for the two honeycomb sections should be as follows:

First honeycomb section


• Porous material index — 12
• Local coordinate system — 1
• Flow is along the x- (x1-) direction, hence the value chosen for the resistance
coefficients (7) is assigned to Alphax1 and Betax1
• The porosity value (0.5) is required only for transient analyses

Second honeycomb section


• Porous material index — 14
• Local coordinate system — 5
• Flow is along the θ- (x2-) direction, hence the value chosen for the resistance
coefficients (7) is assigned to Alphax2 and Betax2
• The porosity value (0.5) is required only for transient analyses

Version 4.02 6-3


POROUS MEDIA FLOW Chapter 6
Useful Points

Step 3
Consider whether, as a consequence of special conditions in your problem,
additional input is required for each porous material. Specifically:
1. If turbulence effects are important, specify the relevant parameters using the
“Turbulence Properties” STAR-GUIde panel.
2. If there is heat transfer present, specify an effective thermal conductivity and
turbulent Prandtl number using the “Thermal Properties” STAR-GUIde
panel.
3. If the problem requires calculation of chemical species mass fractions, the
effective mass diffusivity and turbulent Schmidt number for each species need
to be specified via the “Additional Scalar Properties” STAR-GUIde panel.
4. If you are doing a transient analysis, enter an appropriate value in the Porosity
box (see also page 8-2 of the Methodology volume).

Useful Points
1. All porous media properties can be modified by a user subroutine (PORCON,
PORDIF, PORKEP, POROS1 or POROS2).
2. α and β should always be positive numbers
3. Excessive values of α and β should be avoided. In cases such as honeycomb
structures where cross-flow resistances are much higher than those in the flow
direction, the difference in α and β between one direction and the other
should be limited to four orders of magnitude.
6-4 Version 4.02
Chapter 6 POROUS MEDIA FLOW
Useful Points

4. Avoid setting β = 0 because this can cause K i → 0 as V → 0, leading to a


potentially unstable situation.
5. When calculating resistance coefficients from expressions involving pressure
drops, remember that the pressure drops are based on unit lengths in each
direction.
6. Bear in mind the difference between velocity magnitude V and velocity
component u i in your coefficient calculations.
7. Special considerations apply to modelling systems incorporating porous
baffles (see “Baffle Boundaries” on page 4-23). Note that baffles may also be
used to model a flow resistance at the interface between a fluid and a porous
sub-domain, by placing baffles of suitable properties on the faces of the
appropriate porous cells.
8. In simulations involving moving meshes, porous media must not be used in
areas where there is internal relative mesh motion (cell expansion or
contraction).
9. As a result of the particular method used in STAR-CD to calculate pressure
gradients at cells on either side of the fluid-porous interface, you need to
ensure that porous sub-domains are at least two cell layers thick in any
coordinate direction.
10. Tetrahedral meshes should not be used in porous media cases.
11. For examples of porous media flow, refer to the Methodology volume
(Chapter 8, “Examples of Resistance Coefficient Calculation”) and to Tutorial
3.1, Tutorial 3.2 and Tutorial 3.3 in the Tutorials volume.

Version 4.02 6-5


Chapter 7 THERMAL AND SOLAR RADIATION
Radiation Modelling for Surface Exchanges

Chapter 7 THERMAL AND SOLAR RADIATION


The theory behind problems of this kind and the manner of implementing it in
STAR-CD is given in Chapter 9 of the Methodology volume. The present chapter
contains an outline of the process to be followed when setting up a thermal radiation
model and includes cross-references to appropriate parts of the on-line Help system.
The latter contains details of the user input required and important points to bear in
mind when setting up problems of this kind.

Radiation Modelling for Surface Exchanges


Step 1
Open the “Thermal Options” panel in STAR GUIde and select one of the following
calculation methods from the Radiation menu:
1. Discrete Transfer - Internal VF Calc, making sure that option
Non-Participating is also selected.
2. Discrete Transfer - FASTRAC VF Calc

Continue by entering all necessary modelling parameters, as discussed in topic


“Thermal Radiation”.
Step 2
If present, solar radiation effects can be included by selecting Solar Radiation On
and then entering all necessary modelling parameters, as discussed in topic “Solar
Radiation”. Note that thermal and solar radiation calculations are independent of
each other. A solar-radiation-only analysis may thus be performed without selecting
any options from the Radiation menu mentioned in Step 1 above.
Solar radiation may enter the solution domain through any open boundary, as
well as through transparent walls; see “Solar radiation properties” on page 4-20 for
a description of how the latter are specified.
Step 3
Inspect the Cell Table Editor entries for cell types assigned to the medium lying
between the model’s radiating surfaces and ensure their Radiation option is set to
On.
Step 4
In the Liquids and Gases folder:
1. Assign thermal properties to the fluid domains via the “Molecular Properties”
panel
2. Turn on the temperature solver in the “Thermal Models” panel

Step 5
In the “Define Boundary Regions” panel, specify surface radiative properties for all
boundaries apart from symmetry and cyclic ones. To do this:
1. If only thermal radiation is modelled:
(a) Specify emissivity, reflectivity and transmissivity of all wall, baffle and
solid-fluid interface boundaries, as necessary. The description given in
“Thermal radiation properties” on page 4-20 (for walls) and on page 4-25

Version 4.02 7-1


THERMAL AND SOLAR RADIATION Chapter 7
Radiation Modelling for Surface Exchanges

(for baffles) should be read before entering values in this panel.


(b) Specify the radiation temperature and emissivity at ‘escape’ surfaces, i.e.
boundaries of type “Inlet”, “Outlet”, “Pressure Boundary”, “Stagnation
Boundary”, “Free-stream Transmissive Boundary”, “Transient-wave
Transmissive Boundary” and “Riemann Boundary”. The required values
are entered in the boxes labelled T Radiation and Emissivity.
Note that if the FASTRAC method has been chosen, the T Radiation
value is not used. Instead, the Surrounding environment temperature
specified in the “Thermal Options” panel is used to describe what lies
beyond such open boundaries. Note also that the Emissivity value must be
set to 0.0

2. For problems involving both thermal and solar radiation, as well as the above
parameters, you also need to specify values for the solar reflectivity and
transmissivity. These are required at walls, baffles, or solid-fluid interfaces.
The description given in “Solar radiation properties” on page 4-20 (for walls)
and on page 4-26 (for baffles) should be read before entering such values.
3. For problems involving only solar radiation, the transmissivity of wall
boundaries is the only user input required.

Step 6
Specify radiation patches unless your problem involves only solar radiation. Tab
“Patches” in panel “Create Boundaries” contains most facilities necessary for this
task. If you are using the Internal method, you may also create patches via one of
the following command-driven options:
1. By specifying the face number that defines the boundary face to be included
in the patch — command BDEFINE.
2. By converting a set of shells into a patch — command BSHELL

Please also note that:


• Patches generated for use by the Internal method cannot also be used by the
FASTRAC method.
• The FASTRAC patch specification procedure is different from that for the
Internal method. Moreover, the patches are not generated until after the view
factor calculation procedure has been initiated (see Step 8 below).
• ‘Escape’ surfaces do not need to be patched if the FASTRAC method has
been chosen.
Step 7
Check the patches created using one of the following methods:
1. Select Patch from the Cell Plot Display Options in the main pro-STAR
window
2. Choose Plot > Cell Display > Boundary Patches from the main menu bar
3. Type commands BDISPLAY, PATCH or CDISPLAY, ON, BPATCH in the I/O
window.

The next cell plot will then display boundaries coloured according to patch number
instead of according to boundary type.

7-2 Version 4.02


Chapter 7 THERMAL AND SOLAR RADIATION
Radiation Modelling for Participating Media

Step 8
The action here depends on your choice of view factor calculation method:
• If you have chosen Discrete Transfer - Internal VF Calc, write the
geometry and problem files in the usual way and then run STAR. The view
factor and any solar radiation flux calculations are performed at the start of the
analysis. In moving mesh cases, view factors are re-calculated at every time
step. View factors are saved in a binary file (case.vfs) and are retrieved
from that file in a restart run.
• If you have chosen Discrete Transfer - FASTRAC VF Calc, go to the
Analysis Preparation/Running folder, open the “Run Analysis Interactively”
panel and start up the external program that calculates the view factors. On
completion, the results are stored in file case.nvfs. Subsequent actions are
as for the Internal method but using the .nvfs file instead. Note that a
re-calculation of the view factors is required if either the solar radiation
parameters (Step 2) or boundary transmissivity (Step 5) are altered.

Radiation Modelling for Participating Media


This approach is most commonly used to model the radiative effects of a fluid filling
the space between radiating solid surfaces. However, STAR-CD is also capable of
calculating radiative heat transfer through transparent solid domains, which may
then be treated in a similar manner to the intervening fluid. This enables you to
make a realistic assessment of, for example, the effect of objects such as windows
on the overall heat transfer within an enclosure.
The necessary steps for participating media analysis are as follows:
Step 1
1. Open the “Thermal Options” panel in STAR GUIde and select one of the
following calculation methods from the Radiation menu:
(a) Discrete Transfer - Internal VF Calc, making sure that option
Participating is also selected.
(b) Discrete Ordinates. The participating media radiation option is turned
on automatically.

2. Continue by entering all necessary modelling parameters, as explained in


topic Thermal Radiation
3. If your problem contains solid domains (including transparent ones) turn on
the Solid-Fluid Heat Transfer option

Note that inclusion of solar radiation effects is not currently possible for this type
of analysis.
Step 2
Using the Cell Table Editor:
• If transparent solid cells are present, index them to a separate cell type and
assign a solid material number to them
• Select option On from the Radiation menu for all fluid and transparent solid
cell types in your model

Version 4.02 7-3


THERMAL AND SOLAR RADIATION Chapter 7
Radiation Modelling for Participating Media

Step 3
Go to the Liquids and Gases folder:
1. Assign thermal properties to the fluid domains via the “Molecular Properties”
panel
2. Turn on the temperature solver in the “Thermal Models” panel and click
Show Options. In the Participating Media section, specify bulk radiative
properties (absorption and scattering coefficients) for the fluid lying between
the radiating surfaces. The Conservation and Enthalpy settings in this panel
do not affect the radiation solution.
Step 4
If transparent solids are present, go to the Solids folder:
1. Assign thermal properties to the solid domains via the “Material Properties”
panel
2. Assign radiative properties (absorption and scattering coefficients) to the
solid domains via the “Radiative Properties” panel
Step 5
In the “Define Boundary Regions” panel, specify surface radiative properties for all
boundaries apart from symmetry and cyclic ones. Thus:
• Specify emissivity, reflectivity and transmissivity of all wall, baffle and
solid-fluid interface boundaries, as necessary. The description given in
“Thermal radiation properties” on page 4-20 (for walls) and on page 4-25 (for
baffles) should be read before entering values in this panel.
• Specify the radiation temperature and emissivity at ‘escape’ surfaces, i.e.
boundaries by type “Inlet”, “Outlet”, “Pressure Boundary”, “Stagnation
Boundary”, “Free-stream Transmissive Boundary” and “Transient-wave
Transmissive Boundary”. The required values are entered in the boxes
labelled T Radiation and Emissivity.

Note that:
• All boundaries are assumed to be diffuse (i.e. their radiative properties are not
dependent on the direction of radiation incident on or leaving the surface).
• The absorptivity of the solid-fluid interface (1 - transmissivity - reflectivity)
should be consistent with the absorptivity of the solid material defined in Step
4.
Step 6
If you have chosen the Discrete Transfer - Internal VF Calc method, create
radiation patches for all relevant boundary regions, including external boundaries
of solid cells. This process is as described in “Radiation Modelling for Surface
Exchanges”, Step 6 and 7.
Step 7
Write the geometry and problem files in the usual way and then run STAR. If the
initialization stage completes successfully, you will see an echo of the specified
modelling parameters in the .info and .run files.
• If you have chosen the Discrete Transfer - Internal VF Calc method, the

7-4 Version 4.02


Chapter 7 THERMAL AND SOLAR RADIATION
Capabilities and Limitations of the DTRM Method

view factor calculations are performed at the start of the analysis. In moving
mesh cases, view factors are re-calculated at every time step. View factors are
also saved in a binary file (case.vfs) and are retrieved from that file in a
restart run. Participating media data are stored in another binary file
(case.pgr) and then retrieved from it in a restart run.
• If you have chosen the Discrete Ordinates method, the STAR solver is called
every n iterations during the run to solve the radiative transfer equation
(where n is the value specified in the “Thermal Options” panel). The solver
allocates and frees memory each time, which is reported. In addition, the
solver prints out a residual history for the solution of the radiative transfer
equation, as well as a summary of the computation.
• At convergence, the displayed value for the Imbalance quantity should be
small compared to heat fluxes of engineering interest. This indicates that the
net radiation emission from the medium equals the net absorption into the
boundary. If all boundaries are adiabatic and there are no other energy source
terms, both the net boundary emission and the net media emission will
separately reach very small values.

Capabilities and Limitations of the DTRM Method


1. Lagrangian particle radiation may be modelled by setting Constant 82 to a
non-zero value equal to the particle emissivity. For coal combustion cases,
this operation may be performed via the “NOx/Radiation” panel in STAR
GUIde.
2. Conducting walls (solid-fluid interfaces) should have their transmissivity set
to either 1 or 0, depending on whether radiative heat transfer through the solid
material is to be considered. If radiation in the solid is on, the transmissivity at
the solid-fluid interface must be 1, otherwise it must be 0.
3. The FASTRAC method must be used for thermal/solar radiation problems
with transmissive external walls.
4. At present, the FASTRAC method does not apply to problems containing
symmetry or cyclic boundary regions.
5. ‘Escape’ or open flow boundaries (inlet, outlet, pressure, etc.) require an
assumption regarding the radiation passing through these boundaries and
emitted from outside the solution domain. The Internal method assumes that
this externally emitted radiation is coming from a surface of given
temperature that coincides with the escape boundary surface. The FASTRAC
method assumes that a distant ‘environmental’ black body emits radiation at a
given temperature. These differing assumptions lead to slightly different
results. In addition, the Internal method allows specification of different
radiation temperatures at each open boundary whereas FASTRAC assumes
that all open boundaries "see" the same environmental surface.
6. Radiation patches cannot be applied to boundaries assigned to the default wall
region (region no. 0). If you need to turn on radiation modelling in a problem
containing such boundaries, you will need to re-assign them first to a non-zero
wall region number.
7. The accuracy of the radiation calculations depends on the patch size since
quasi-uniform radiation properties are assumed for a patch. The accuracy of

Version 4.02 7-5


THERMAL AND SOLAR RADIATION Chapter 7
Capabilities and Limitations of the DTRM Method

the view factor calculations depends on both the patch size and the number of
beams emitted per patch. For maximum accuracy:
(a) Patches should be planar.
(b) The aspect ratios of patches should be close to 1.0.
(c) If you are using the Internal view factor calculation method, any
refinement of patches should be followed by an increase in the number of
beams, so that all patches are resolved adequately (see “Patch and beam
definition” on page 9-2 of the Methodology volume for a discussion of
this point).
(d) Patches should not span multiple regions unless the assumption of
quasi-uniform radiation properties is valid over those regions.

However, acceptable results may be obtained even if one or more of the above
conditions are not fully met.
8. If the wrong patch number is assigned to a cell face during the patch
definition process (Internal view factor calculations only), the mistake can be
rectified either:
(a) numerically via the BMODIFY command, or
(b) graphically (using the screen cursor) via the BCROSS command.

9. Patches defined as part of the Internal view factor calculation can be stored in
a file (case.bnd) and read back from it using the normal boundary export
and import facilities provided in panels “Export Boundaries” and “Import
Boundaries”, respectively.
10. The default number of beams used in the Internal view factor calculation
process (100) may be sufficient for coarse patches. In situations where a patch
is created for every boundary cell face, the number of beams may need to be
increased (between 1600 and 2500 for typical radiation problems) in order to
resolve adequately the patches present in the system. The FASTRAC
calculation method uses a fixed number of beams (1,024).
11. The CPU time for Internal view factor calculations increases in proportion to
the number of patches multiplied by the number of beams. The CPU time for
radiation heat transfer calculations increases in proportion to the number of
patches. The FASTRAC view factor calculations are also dependent on the
number of patches but the CPU time required is considerably reduced.
12. User subroutine USOLAR cannot be used for solar radiation problems
employing the FASTRAC method.
13. STAR-HPC runs are not feasible for problems involving participating media
radiation.
14. STAR-HPC runs for problems involving solar radiation are only feasible if the
FASTRAC method has been chosen.
15. Surface-exchange problems using the Internal calculation method can be run
in STAR-HPC mode, but the view factors have to be calculated in
‘single-processor’ mode. To do this, run the case for zero iterations on a single
processor and save the view factor (.vfs) file.
16. As stated on page 7-3, Step 8 above, Internal view factors for moving mesh
cases are re-calculated at every time step. Therefore, in view of the previous
restriction, STAR-HPC runs for problems involving both radiation and a

7-6 Version 4.02


Chapter 7 THERMAL AND SOLAR RADIATION
Capabilities and Limitations of the DORM Method

moving mesh are not feasible. Note also that FASTRAC view factor
calculations cannot be used in moving mesh cases at present.
17. Note that command PATCH generates only shell surfaces. It cannot be used to
create radiation patches.

Capabilities and Limitations of the DORM Method


1. The discrete ordinates model (DORM) does not need radiation patches and
the computational overheads involved in their use. In addition, to facilitate
switching from a discrete transfer (DTRM) to a discrete ordinate (DORM)
model for the same problem, STAR will accept geometry files with or without
patches.
2. Nevertheless, using DORM can still add significantly to the CPU time and
memory needed for a given simulation. For this reason, users are encouraged
to plan their analyses conservatively until they gain experience with the CPU
time and memory requirements of their model. The run-time output for the
DORM calculation will echo the memory requirements (see Step 7 above).
3. The memory requirements of the calculation depend on your choice of
angular discretization. The table below gives a guide to memory usage per
100,000 cells. Note that this holds for single-precision calculations and a grey
medium.
.

Table 7-1: Approximate memory required for DORM analysis

Angular Additional memory per


Ordinates discretization 100,000 cells
8 S2 45 MB
24 S4 55 MB
48 S6 75 MB
80 S8 95 MB

4. The model may be run in the normal way under STAR-HPC. However, the
solution history for a serial run will be different from that for a parallel run.
Although the radiative transfer equation is similar to a normal transport
equation, there is no equivalent of the diffusion term and so the equation is not
elliptic. To solve this equation efficiently, a specialized solver that follows the
directions of each ordinate is used. Thus, in the STAR-HPC environment,
some domains may receive the information about certain directions only after
it has crossed through the other domains. Nevertheless, converged solutions in
serial and HPC calculations are identical.
5. Coupling between the ordinate directions at cyclic and symmetry boundaries
approximates such boundaries as diffuse.
6. DORM is fully compatible with all cell shapes supported by pro-STAR
(polyhedral cells, baffle cells, etc.)
7. DORM can also be used to model surface-exchange problems (i.e.
non-participating media analyses). Since the participating media mode is

Version 4.02 7-7


THERMAL AND SOLAR RADIATION Chapter 7
Radiation Sub-domains

always on, the absorption and scattering coefficients must be set to zero for
such cases.
8. At present, DORM does not support cases involving solar radiation.

Radiation Sub-domains
In some problems, radiation effects are important only within a restricted
sub-domain of the overall solution domain, e.g. when doing a complete continuum
mechanics analysis around a car body, where radiation calculations are only
necessary under the car bonnet.
Under such circumstances, it is possible to confine the radiative heat transfer
treatment to the part of the model where it is relevant, thus avoiding the lengthy
calculations needed for a full radiation analysis. The following steps are then
necessary:
Step 1
In the “Thermal Options” STAR GUIde panel, turn on the radiation calculations by
selecting either the Discrete Transfer - Internal VF Calc or the Discrete
Ordinates option in the Radiation menu and then specify all necessary radiation
parameters.
Step 2
Using the Cell Table Editor:
• Create a separate cell type for all cells occupying the sub-domain that is
subject to the radiative treatment
• For this cell type only, turn the Radiation option On
Step 3
The action here depends on your choice of method in Step 1 above:
• For the Discrete Transfer - Internal VF Calc method (whether
Participating or Non-Participating), create the necessary number of special
‘Radiation’ boundaries so as to completely separate the radiative from the
non-radiative part of the domain.
• For the Discrete Ordinates method, radiation boundaries are not applied.
Step 2 above is all that is required to define the problem properly; no further
action is necessary.
Step 4
Within the radiative sub-domain, use the “Define Boundary Regions” panel to
specify radiation properties for all boundary regions, including the special
boundaries created above (see also Chapter 4, “Radiation Boundaries”).
Step 5
If you have chosen the Discrete Transfer - Internal VF Calc method, create
patches on all boundaries surrounding the radiative sub-domain, including the
radiation boundaries, as described in “Radiation Modelling for Surface Exchanges”.
Step 6
Write the geometry and problem files in the usual way and then run STAR.

7-8 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Introduction

Chapter 8 CHEMICAL REACTION AND COMBUSTION


Introduction
STAR-CD allows for two kinds of chemical reaction:
• Homogeneous — the reaction occurs within the bulk of the fluid
• Heterogeneous — the reaction takes place only at surfaces, such as in
catalytic converters

Heterogeneous reactions are currently implemented via user-supplied subroutines.


Homogeneous reactions are grouped into three distinct types:
1. Unpremixed/Diffusion — reactions of this type occur when the fuel and
oxidant streams enter the solution domain separately, as in a Diesel engine.
The reactions may be sub-divided into the following groups:
(a) Local Source — these include eddy break-up, chemical kinetic, and
hybrid models (see “Local Source Models” on page 8-2 for more details)
(b) Complex Chemistry — these model the reaction system by including the
full reaction mechanism (see “Complex Chemistry Models” on page 8-11
for more details).
(c) Presumed Probability Density Function (PPDF) — these include single
and multiple fuel implementations and the Laminar Flamelet model (see
“Presumed Probability Density Function (PPDF) Models” on page 8-3
for more details)

2. Partially Premixed — combustion of this type is one of the essential features


in Gasoline Direct Injection engines, where combustion occurs in a
non-uniform mixture. The reactions may be sub-divided into the following
groups:
(a) Local Source, of the type mentioned above
(b) Complex Chemistry, of the type mentioned above
(c) Regress Variable, represented by a Flame Area Evolution (FAE) model

3. Premixed — reactions of this type occur when the fluid initially has a
uniform composition, as in a spark ignition engine
(a) Local Source, of the type mentioned above
(b) Complex Chemistry, of the type mentioned above
(c) Regress Variable, represented by various eddy break-up and flame-area
models (see “Regress Variable Models” on page 8-10 for more details)

The theory behind reaction models of the local source, complex chemistry and
PPDF type is described in Chapter 10 of the Methodology volume. Regress variable
models are normally used in engine combustion simulation and are described
separately in Methodology Chapter 11. A set of recently implemented engine
combustion models is discussed in the section on “Setting Up Advanced I.C. Engine
Models” on page 8-22 of this chapter.
In some cases, the model describing the main chemical reaction(s) may need to
be supplemented by subsidiary models that describe:

Version 4.02 8-1


CHEMICAL REACTION AND COMBUSTION Chapter 8
Local Source Models

• Ignition mechanisms
• Emission of pollutants, typically NOx products. Special considerations apply
to modelling NOx-type reactions, as discussed in “NOx Modelling” on page
8-39.
• Application Specific models, such as engine knock

All these models together constitute a so-called chemical reaction scheme. Note
that:
• Chemical schemes are defined and numbered individually
• Chemical scheme definitions can exist independently of any fluid domains or
scalar variables. However, they need to be explicitly assigned to a domain
before they can be used in your simulation.
• Each fluid domain may be associated with only one chemical reaction
scheme. However, this association may be changed by the user to suit
problem requirements or to try out alternative reaction models.
• Special considerations apply to modelling coal combustion; these are
discussed in the section on “Coal Combustion Modelling” on page 8-41.

Local Source Models


The main characteristics of this group of models are as follows:
1. Up to 30 chemical reactions may be defined per scheme
2. The reactions are irreversible
3. Each reaction is associated with a single chemical species designated as the
leading reactant (equivalent to fuel in a combustion reaction). This species
characterises the reaction and is consumed by it. The remaining reacting
species are defined as reactants.
4. The products of a reaction are defined as products. However,
(a) if a product of a reaction participates as a reactant in a second reaction, it
should be specified as a leading reactant or ordinary reactant, as
appropriate;
(b) if a product is transported into the solution domain from an external
source, it also should be specified as a reactant.

5. The distribution of products within the solution domain can be calculated


algebraically, provided that the products are generated only within the
domain.
6. If all incoming streams consist of identical fuel-to-reactant ratios (in transient
cases the initial fields must also have the same ratio), the reaction process is
termed premixed (see “Premixed reaction/homogeneous systems” on page
10-4 of the Methodology volume). If this is not the case, the process is either
of the diffusion or the partially premixed type and the user needs to solve an
additional scalar transport equation for the mixture fraction (total mass
fraction of burned and unburnt fuel, see “Diffusion reaction /
non-homogeneous systems” on page 10-5).
7. STAR-CD automatically sets up mixture fraction scalars for each leading
reactant in diffusion and partially premixed reactions. However, it is the user’s
responsibility to ensure that boundary conditions for both leading reactant and

8-2 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Presumed Probability Density Function (PPDF) Models

mixture fraction are specified correctly and that they are the same for both of
them.
8. The reactions themselves are defined by specifying the amounts (in
kilomoles) of the participating leading reactants, reactants and products. For
example, the input required for the following reaction (combustion of
methane)

CH 4 + 2O 2 → CO 2 + 2H 2 O (8-1)

is

Reaction (1) kmol


Leading reactant (fuel) (1) — CH 4 1
Reactant (1) — O2 2
Product (1) — CO 2 1
Product (2) — H2 O 2

9. pro-STAR includes facilities for checking that mass is conserved for each
reaction.

Presumed Probability Density Function (PPDF) Models


Models of this type are described in Chapter 10, “Presumed-PDF (PPDF) Model for
Unpremixed Turbulent Reaction” in the Methodology volume. These fall into two
main groups:
• Single-fuel PPDF, where only one type of fuel and one type of oxidiser are
present, though each of these may enter the combustion system through more
than one inlet.
• Multiple-fuel PPDF, where two types of fuel and one type of oxidiser are
present.
Single-fuel PPDF
The basic equations solved are for the mean mixture fraction f and its variance g f
(see Chapter 10, “Single-fuel PPDF” in the Methodology volume). There is a choice
between equilibrium chemistry models (these assume a local instantaneous
chemical equilibrium) and a laminar flamelet model that allows for non-equilibrium
effects (such as flame stretch).
Equilibrium models
In these models, the PDF integration may be performed in two ways:
1. By employing a numerical integration technique
2. By expressing all instantaneous values of the variables as polynomials of the
mixture fraction and then doing the integration analytically. Polynomial
coefficients may be
(a) supplied by the user
(b) read in from a built-in database stored in file ppdf.dbs
(c) calculated by the CEA (Chemical Equilibrium with Applications)
program [5, 6]. This is an auxiliary program that computes the chemical

Version 4.02 8-3


CHEMICAL REACTION AND COMBUSTION Chapter 8
Presumed Probability Density Function (PPDF) Models

equilibrium composition of a mixture. This program is included in the


STAR-CD suite and is used in conjunction with the built-in PPDF model.

There is also a choice between adiabatic and non-adiabatic PPDF:


1. For adiabatic PPDF:
(a) The mixture density and temperature are calculated numerically or from
polynomials in f. Note that these polynomials are based on molar
fractions.
(b) Since temperature is calculated independently, the ‘Constant’ specific
heat property option with default values may be used

2. For non-adiabatic PPDF, the density is calculated from the ideal gas law and
the temperature from the enthalpy transport equation.

The mass fractions of all other chemical species related to the reaction are defined
as additional scalar variables and calculated numerically or from the user-supplied
polynomials in f, as above. Up to forty eight such species can be specified by the
user.
Laminar flamelet model
In this model, the PDF integration is always performed numerically and the results
stored in a look-up table which is characterised by its mean mixture fraction,
mixture fraction variance and strain rate. There is also a choice between an adiabatic
and a non-adiabatic model, as above.
The setup procedure for the model is described in the on-line Help for the
“Reaction System” STAR GUIde panel. One part of this procedure is to specify the
reaction mechanism, stored in a reaction definition file in CHEMKIN format. This
is organized in three sections:
• Element data
• Species data
• Reaction data
The basic data are often supplemented by auxiliary data for special reactions such
as third-body reactions.
Element Data
All chemical species in the reaction mechanism must be composed of chemical
elements or isotopes of chemical elements. Each element and isotope must be
declared using a one- or two-character symbol. The purpose of the element data is
to associate the element atomic weights with their character symbol representations.
If an ionic species is used in the reaction mechanism (e.g, OH+), an electron must
be declared as the element E.
Element data must start with the word ELEMENTS (or ELEM) but, following
that, there are minimal restrictions on the formatting of the rest of the section. Any
number of element symbols can be written on any number of lines. The symbols
may appear anywhere on a line, but those on the same line must be separated by
blanks. Any line or portion of a line starting with an exclamation mark (!) is
considered a comment and will be ignored. Blank lines are also ignored.
If an element is in the list below, then only the symbol identifying it need appear

8-4 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Presumed Probability Density Function (PPDF) Models

in the element data. The recognized elements are as follows:

H, HE, LI, BE, B, C, N, O, F, NE, NA, MG, AL, SI, P, S, CL, AR, K, CA, SC, TI,
V, CR, MN, FE, CO, NI, CU, ZN, GA, GE, AS, SE, BR, KR, RB, SR, Y, ZR, NB,
MO, TC, RU, RH, PD, AG, CD, IN, SN, SB, TE, I, XE, CS, BA, LA, CE, PR, ND,
PM, SM, EU, GD, TB, DY, HO, ER, TM, YB, LU, HF, TA, W, RE, OS, IR, PT, AU,
HG, TL, PB, BI, PO, AT, RN, FR, RA, AC, TH, PA, U, NP, PU, AM, CM, BK, CF,
ES, FM, D, E
For an isotope, the atomic weight must follow the identifying symbol and be
delimited by slashes (/). The atomic weight may be given in integer, floating-point,
or “E” format, but internally it will be converted to a floating-point number. For
example, the isotope deuterium may be defined as D/2.014/. If desired, the atomic
weight of an element in the above list may be altered by including the atomic weight
as input just as though the element were an isotope.
An acceptable format for element data specification is shown below:
ELEMENTS H D /2.014/ O N END! END is optional

Species Data
Each chemical species in a reaction must be identified on one or more species
line(s). Any set of up to 16 upper or lower case characters can be used, as for species
names, which are case sensitive. In addition, each species must be composed of
elements that have been identified in the element data section.
Species data must start with the word SPECIES (or SPEC) but, as already
discussed, subsequent formatting of this section is not particularly important. An
acceptable format for species data specification is shown below:
SPEC H2 O2 H O OH HO2 H2O

Reaction Data
The reaction mechanism may consist of any number of chemical reactions
involving the species named in the species section. A reaction may
• be reversible or irreversible;
• be a three-body reaction with an arbitrary third body and/or enhanced
third-body efficiencies;
• have one of several pressure-dependent formulations.

The rate of each reaction is defined by specifying A R , β R and E R from the general
Arrhenius rate equation for the forward reaction, see equation (10-65) in the
Methodology volume.
Reaction data must start with the word REACTIONS (or REAC). On the same
line, you may specify units of the activation energies to follow by including the
word CAL/MOLE, KCAL/MOLE, JOULES/MOLE, KJOULES/MOLE, KELVINS,
or EVOLTS. The default units for E R are cal/mole and the default units for A R are
cm, mole, sec and K. Including the word MOLECULES on the REACTIONS line
changes the units of A R to cm, molecules, sec and K.
The lines following the REACTIONS line contain reaction definitions together
with their Arrhenius rate coefficients, as described in Table 8-1. The description is

Version 4.02 8-5


CHEMICAL REACTION AND COMBUSTION Chapter 8
Presumed Probability Density Function (PPDF) Models

composed of reaction data and optional auxiliary information data.

Table 8-1:

Species Symbols
Each species in a reaction is described by a unique sequence of characters, as
they appear in the species data (e.g. CH4).

Coefficients
A species symbol may be preceded by an integer or real coefficient. The coeffi-
cient’s meaning is that there are that many moles of the particular species
present as either reactants or products; e.g. 2OH, is equivalent to OH + OH.
Non-integer coefficients are allowed, but the element balance in the reaction
must still be maintained.

Delimiters
A plus sign is the delimiter between each reactant species and each
+
product species.
An equality sign is the delimiter between the last reactant and the
=
first product in a reversible reaction.
An equality sign with an angle bracket on the right is the delimiter
=> between the last reactant and the first product in an irreversible
reaction.

Special Symbols
An M as a reactant and/or product stands for an arbitrary third
body. It should appear as both a reactant and a product. In a reac-
tion containing an M, certain species can be specified as having
enhanced third-body efficiencies; in which case auxiliary data
+M
(described below) must follow the reaction line. If no enhanced
third-body efficiencies are specified, all species act equally as third
bodies and the effective concentration of the third body is the total
concentration of the mixture.
An M as a reactant and product surrounded by parentheses indi-
cates that the reaction is pressure-dependent, in which case auxil-
iary information line(s) (described below) must follow the reaction
(+M) to identify the fall-off formulation and parameters. A species may
also be enclosed in parentheses. For example, (+H2O) indicates
that water is acting as the third body in the fall-off region, not the
total concentration M.
An exclamation mark means that all following characters on the
! reaction line are comments. For example, the comment may be
used to give a reference to the source of the reaction and rate data.

8-6 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Presumed Probability Density Function (PPDF) Models

The second field of each reaction line is used to define the Arrhenius rate
coefficients A R , β R and E R , in that order. At least one blank space must separate
the first number and the last symbol in the reaction. The three numbers must be
separated by at least one blank space and be given in integer, floating point, or “E”
format (e.g., 123, 123.0 or 123E1). Their units are as specified in the REACTIONS
line above. An example of reaction data for a simple mechanism is shown below:
REACTIONS CAL/MOLE
H2 + O2 = OH 1.70E+13 0.000 47780
OH + H2 = H2O + H 1.17E+9 1.300 3626
H + O2 = HO2 3.61E+17 -0.720 0.000
O + H2 = OH + H 5.06E+04 2.670 6290
HO2 + H2 = H2O2 + H 1.25E+13 0.000 0
0.5H2O2 + 0.5H2 = H2O 1.60E+12 0.000 3800
! example of real coefficients
END ! END statement is optional;

The basic rules for specifying reaction data are summarised below:
1. The first line must start with the word REACTIONS (or REAC), and may
include units definition(s).
2. The reaction description can begin anywhere on the line. All blank spaces,
except those between Arrhenius coefficients, are ignored.
3. Each reaction description must have = or => between the last reactant and the
first product.
4. Each reaction description must be contained within one line.
5. Three Arrhenius coefficients ( A R , β R and E R ) must appear in order on each
line, separated from each other and from the reaction description by at least
one blank space; no blanks are allowed within the numbers themselves.
6. No more than six reactants or six products are allowed in a reaction.
7. Comments are any characters following an exclamation mark.

Auxiliary Reaction Data


The format of an auxiliary information line is a character-string keyword followed
by a slash-delimited (/) field containing an appropriate number of parameters (in
either integer, floating point, or “E” format). Different types of auxiliary reaction
data are described below, followed by an example:
1. Third-Body and Pressure-Dependent Reaction Parameters
If a reaction contains M as a reactant and/or product, auxiliary information
lines may follow the reaction line to specify enhanced third-body efficiencies
of certain species. The keyword defining an enhanced third-body efficiency is
the species name of the third body, and its single parameter is its enhanced
efficiency factor. A species that acts as an enhanced third body must be
declared as a species.
If a pressure-dependent reaction is indicated by a (+M) or by a species
contained within parentheses, say (+H2O), one or more auxiliary information
lines must follow to define the pressure-dependence parameters. For all
pressure-dependent reactions, an auxiliary information line must follow to
specify either the low-pressure limit Arrhenius parameters (for fall-off
reactions) or the high-pressure limit Arrhenius parameters (for chemically

Version 4.02 8-7


CHEMICAL REACTION AND COMBUSTION Chapter 8
Presumed Probability Density Function (PPDF) Models

activated reactions). For fall-off reactions, the keyword LOW must appear on
the line, with three rate parameters A 0 , β 0 and E 0 . For chemically activated
bimolecular reactions, the keyword HIGH must appear on the line, with the
three rate parameters A ∞ , β ∞ and E ∞ .There are then three possible
interpretations of the pressure-dependent reaction:
(a) Lindemann formulation - no additional parameters are defined
(b) Troe formulation - in addition to the LOW or HIGH parameters, the
keyword TROE followed by three or four parameters must be included in
the following order: a, b, c and d, as defined in equation (10-77) of the
Methodology volume. The fourth parameter is optional and if omitted, the
last term in equation (10-77) is not used.
(c) To define an SRI pressure-dependent reaction, in addition to the LOW or
HIGH parameters, the keyword SRI followed by three or five parameters
must be included in the following order: a, b, c, d and e, as defined in
equation (10-79) of the Methodology volume. The fourth and fifth
parameters are optional. If only the first three are stated, then by default
d = 1 and e = 0 .

2. Landau-Teller Reactions
To specify Landau-Teller parameters, the keyword LT must be followed by
two parameters — the coefficients B R and C R from equation (10-81) in the
Methodology volume. The Arrhenius parameters A R , β R , and E R are taken
from the numbers specified on the reaction line itself. If reverse parameters
are specified in a Landau-Teller reaction via REV (see item 4 below), the
reverse Landau-Teller parameters must also be defined, with the keyword RLT
and two coefficients B R and C R for the reverse rate.
3. Logarithmic Interpolation of Pressure-Dependent Rates
This generalized way of describing the pressure dependence of a reaction rate
is indicated by the PLOG keyword in auxiliary lines. In this case, the reaction
description should not include (+M) in it, although this is used to indicate that
the reaction is pressure dependent in other cases. This particular option for
describing pressure-dependent reactions cannot be combined in any given
reaction with other options for describing pressure dependence. One
supplementary line starting with the PLOG keyword needs to be supplied for
each pressure in the set. The keyword is followed by slash-delimited values
for the pressure (in atmospheres) and the rate parameters for that pressure.
The supplementary lines need to be in order of increasing pressure. If the rate
expression at a given pressure cannot be described by a single set of
Arrhenius parameters, more than one set may be provided. Each of these
should be followed by the keyword DUPLICATE, meaning the sum of the sets
of rates provided for a given pressure will be used. The units of the rate
parameters provided with the PLOG keyword should match the units used for
the overall reaction description. Note that, in this case, although rate
parameters need to be supplied on the main reaction line to prevent an error,
those values are superseded by the ones provided on the supplementary lines.
4. Reverse Rate Parameters
For a reversible reaction, auxiliary data may follow the reaction to specify
Arrhenius parameters for the reverse-rate expression. Here, the three

8-8 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Presumed Probability Density Function (PPDF) Models

Arrhenius parameters ( A R , β R , and E R ) for the reverse rate must follow the
keyword REV. This option overrides the reverse rates that would be normally
computed by satisfying microscopic reversibility through the equilibrium
constant, as described by equation (10-66) in the Methodology volume.
5. Reaction Order Parameters
Auxiliary data may be included to override the reaction order for a species,
using the auxiliary keywords FORD or RORD, for forward and reverse
reaction descriptions, respectively. Each occurrence of these keywords must
be followed by the species name and the new reaction order. This option
overrides the stoichiometric coefficients for the species included in the
auxiliary data.
6. Reaction Units
It is sometimes convenient to specify units for a particular reaction rate fit that
differ from the default units specified for other reaction expressions in the
chemistry mechanism. In such a case, you should employ the auxiliary
keyword UNITS. This keyword must be followed by one or more of the
following unit descriptors: MOLECULE, CAL, KCAL, JOULE, KJOULE,
KELVIN, or EVOLTS. The inclusion of MOLECULE would indicate that the
reaction rate expression is in units of molecules/cm3 rather than mole/cm3.
The remaining unit descriptors specify the energy units in the rate expression.
Note that the temperature units in the rate expression are always in Kelvin.

An example of the use of auxiliary reaction data for a three-parameter Troe fall-off
reaction with enhanced third-body efficiencies is shown below:
CH3+CH3(+M)=C2H6(+M) 9.03E16 -1.18 654
LOW / 3.18E41 -7.03 2762 /
TROE / 0.6041 6927 132. /
H2/2/ CO/2/ CO2/3/ H2O/5/

Multiple-fuel PPDF
1. Four equations are solved, for the progress variables f p (primary fuel mixture
fraction), f s (secondary fuel mixture fraction), g f (primary fuel variance)
and g ξ (variance of variable ξ, see Chapter 10, “Multiple-fuel PPDF” in the
Methodology volume).
2. Only an equilibrium chemistry model is available in this case
3. The PDF integration is always performed numerically

Other noteworthy points about PPDF models are:


1. In order to increase the efficiency of combustion systems by increasing the
temperature of incoming oxidisers, the use of vitiated air containing
combustion products is a viable option. The basic PPDF model, which
assumes that only fuel and air enter the system, cannot be used for this kind of
problem. However, STAR-CD’s implementation has been extended to allow
up to four dilutants to enter the combustion system. The basic setup is the
same as that used for the standard PPDF model. However, additional
transported scalars are defined to represent the dilutants; therefore, additional
boundary conditions need to be defined for them.

Version 4.02 8-9


CHEMICAL REACTION AND COMBUSTION Chapter 8
Regress Variable Models

It is emphasised that PPDF with dilutants can only be used in conjunction


with the single-fuel, equilibrium chemistry model plus the non-adiabatic
PPDF option.
2. The multiple-fuel PPDF option may also be used to model a system
containing only one type of fuel but two different types of oxidiser.

Regress Variable Models


Models in this group solve a transport equation for a regress variable representing
the combustion process and are described in the Methodology volume, Chapter 11.
Their main features are:
1. The regress variable b defined by equation (11-4) in the Methodology volume
is the transported variable and is a passive scalar
2. All physical scalar variables participating in such schemes are linearly related
to b
3. Regress variable models may be classified into two groups:
(a) Flame-area models, discussed in sections “Premixed Combustion in
Spark Ignition Engines” and “Partially Premixed Combustion in Spark
Ignition Engines” of the Methodology volume:
i) “The Weller flame area model” — makes use of the wrinkling
factor Ξ, which is either obtained from an algebraic relationship
given by equation (11-36) or from the solution of a transport
equation
ii) “The CFM-ITNFS model” — employs a transport equation for the
flame area density Σ, given by equation (11-10)
iii) “The Weller 3-equation model” — requires the solution of
equations for both wrinkling factor and mixture fraction
iv) All have their own ignition models

(b) Eddy break-up models, used in a manner similar to that described above
under “Local Source Models”.

4. The one-step reaction representing the combustion process is associated with


a single chemical species designated as the leading reactant (or fuel). This
species characterises the reaction and is consumed by it. The remaining
reacting species are defined as reactants.
5. The reaction is irreversible and is defined by specifying the amounts (in
kilomoles) of the participating leading reactants, reactants and products.
6. pro-STAR includes facilities for checking that mass is conserved.
7. If all incoming streams consist of identical fuel-to-reactant ratios (in transient
cases the initial fields must also have the same ratio), the reaction process is
termed premixed (see “Premixed reaction/homogeneous systems” on page
10-4 of the Methodology volume). If this is not the case, i.e. the process is of
the partially premixed type, an additional scalar transport equation for the
mixture fraction needs to be solved. The only regress variable model that may
be used in partially premixed systems is the Weller 3-equation model.
8. STAR-CD automatically sets up mixture fraction scalars for each leading
reactant in diffusion and partially premixed reactions. However, it is the user’s

8-10 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Complex Chemistry Models

responsibility to ensure that boundary conditions for both leading reactant and
mixture fraction are specified correctly and that they are the same for both of
them.
9. Exhaust gases present in EGR (Exhaust Gas Recirculation) systems are taken
into account by defining active scalars for each exhaust gas species and
solving additional transport equations for their mass fraction (see also
Chapter 11, “Exhaust Gas Recirculation” in the Methodology volume).

Complex Chemistry Models


The complex chemistry model supports two types of format for reaction mechanism
definition. One of them is the CHEMKIN format, the other the STAR-CD native
format described below.
In order to use STAR-CD’s complex chemistry model, a reaction mechanism file
called cplx.inp&& has to be created by the user for each chemical scheme in
which a complex chemistry model is applied. The characters ‘&&’ at the end of the
file name represent the chemical scheme number in which the complex chemistry
model is applied. For example, if such a model is applied in chemical scheme no. 2,
the reaction mechanism file should be called cplx.inp02. STAR will write an
echo file cplx.inp&&-echo for each cplx.inp&& file it has read, so that
users can ensure settings have been correctly applied.
File cplx.inp&& contains the reaction formula, chemical kinetic data and
keywords and extra parameters for special reactions, as outlined below:
Reaction formula definition
The general form of a reaction formula is given by

n 1 ′R 1 m 1 ′ + n 2 ′R 2 m 2 ′ + … = n 1 ″P 1 m 1 ″ + n 2 ″P 2 m 2 ″ + … A β E

Here, n 1 ′ , n 2 ′ , …, n 1 ″ , n 2 ″ , … are the stoichiometric coefficients which could be


integer or real numbers, R 1 , R 2 , …, P 1 , P 2 , … are species names, m 1 ′, m 2 ′, …,
m 1 ″ , m 2 ″ , … are the mass fraction exponentials, A is the pre-exponential factor (in
units of cm-mole-sec-K), β the temperature exponent and E the activation energy of
the Arrhenius rate constant (in cal/mol). If the mass fraction exponentials are equal
to 1, they are not written into the corresponding echo file (cplx.inp&&-echo).
Rules:
• There are no spaces between stoichiometric coefficients n i ′ , n i ″ and species
names. If n i ′ or n i ″ are equal to 1, they can be omitted.
• m i ′ and m i ″ must be separated by at least one space from the species name.
If the value of m i ′ or m i ″ is not specified, it will be assumed that m i ′ = n i ′
or m i ″ = n i ″ .
• Character ‘=’ is used for reversible reactions; ‘⇒’ for irreversible reactions.
• There is no ‘+’ character between the pre-exponential factor, A, and the
nearest species name. A, β and E are separated by at least one blank space.
• Everything following the ‘!’ character is treated as a comment
• The ‘+’ character should not be used in a real number expression. For
example, 1.2E+05 should be written as 1.2E05.
• The maximum number of reactants or products in a single reaction must not
exceed 5
Version 4.02 8-11
CHEMICAL REACTION AND COMBUSTION Chapter 8
Complex Chemistry Models

Three-body reaction definition


To define a three-body reaction, add a line starting with the keyword M after the
reaction formula, i.e.

/M/ A/α 1 / B/α 2 / …

Rules:
• Keyword M must be enclosed by two ‘/’ characters and is not case sensitive
• A, B, … are species names and α 1 , α 2 , … are the corresponding efficiency
factors. They are separated by at least one blank space.

The Landau-Teller reaction


To define a Landau-Teller reaction, add a line starting with the keyword RLT after
the normal reaction formula, i.e.

/RLT / B C

Rules:
• Keyword RLT must be enclosed by two ‘/’ characters and is not case sensitive
• B and C are the Landau-Teller parameters and are separated by at least one
blank space
• If the reaction is a three-body reaction as well, a new line is added starting
with ‘ /M/ ’ and the third body efficiency factors

The Lindemann fall-off reaction


To define a Lindemann fall-off reaction, add a line starting with the keyword LOW
after the reaction formula, i.e.

/LOW / A L β L E L

Rules:
• Keyword LOW must be enclosed by two ‘/’ characters and is not case
sensitive
• A L , β L , and E L are the pre-exponential factor, temperature exponent and
activation energy, respectively, of the low pressure limit and are separated
each from each other by at least one blank space
• The corresponding values for the high pressure limit are assumed to be those
given above as part of the reaction formula definition
• If the reaction is a three-body reaction as well, a new line is added starting
with ‘ /M/ ’ and the third body efficiency factors

The Troe fall-off reaction


To define a Troe fall-off reaction, add two lines starting with keywords LOW and
TROE, respectively, after the reaction formula, i.e.

/LOW / A L β L E L

8-12 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Complex Chemistry Models

/TROE/ a b c d

Rules:
• The definition of keyword LOW is the same as above
• The pre-exponential factor, temperature exponent and activation energy
values for the high pressure limit are assumed to be those given above as part
of the reaction formula definition
• Keyword TROE must be enclosed by two ‘/’ characters and is not case
sensitive
• a, b, c and d are the corresponding Troe parameters (d is optional)
• If the reaction is a three-body reaction as well, a new line is added starting
with ‘ /M/ ’ and the third body efficiency factors.

The SRI fall-off reaction


To define a SRI fall-off reaction, add two lines starting with the keywords LOW and
SRI, respectively, after the reaction formula, i.e.

/LOW / A L β L E L

/SRI / a b c d e

Rules:
• The definition of keyword LOW is the same as above
• The pre-exponential factor, temperature exponent and activation energy
values for the high pressure limit are assumed to be those given above as part
of the reaction formula definition
• Keyword SRI must be enclosed by two ‘/’ characters and is not case sensitive
• a, b, c, d and e are the corresponding SRI parameters and are separated from
each other by at least one blank space.
• If the reaction is a three-body reaction as well, a new line is added starting
with ‘ /M/ ’ and the third body efficiency factors.

The Eddy Break-up reaction


To define an eddy break-up reaction in turbulent combustion, add a line starting
with keywords EBU after the reaction formula, i.e.

/EBU / A ebu B ebu IOP f i

Rules:
• Keyword EBU must be enclosed by two ‘/’ characters and is not case sensitive
• A ebu and B ebu are constansts appearing in the standard eddy break-up model,
see equation (10-8). If B ebu is not zero, the product will be included in the
reaction rate calculation.
IOP is an integer determining which EBU model is being used:
IOP = 1 : Standard EBU model, reaction rate determined by equation
(10-8)

Version 4.02 8-13


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Chemical Reaction Schemes

IOP = 2 : Combined time scale model, reaction rate determined by


equation (10-15)
IOP = 3 : Hybrid kinetic/EBU model, reaction rate determined by
equation (10-11)

For IOP = 2 or 3, the corresponding pre-exponential factor, temperature


exponent and activation energy are defined as usual.
• f i is optional and represents the burnt fuel mass fraction used in ignition
modelling (see Chapter 10, “Ignition” in the Methodology volume)

In an eddy break-up reaction defined as

A1 + A2 + … → B1 + B2 + …

species A1 will be treated as the fuel, A2 as the oxidizer and B1 as the product. The
mass fractions corresponding to these species are denoted in equation (10-8) as
Y F ,Y O ,Y P , respectively. The eddy break-up reaction is also assumed to be
irreversible.

An example reaction mechanism file is shown in Table 8-2.

Table 8-2
H + O2 = OH + O 2.24E4 0. 16795
O + O = O2 2.62E16 –0.84 0
/M/ H2/2.40/ H2O/5.40/ CH4/2.00/ CO/1.75/ CO2/3.60/
HCO = CO 1.2 + H 5.00E12 0. 19208 ! modified
/M/
CO + O = CO2 1.80E10 0. 2385
/LOW / 6.020E14 0. 3000
H + CH2 = CH3 6.0E14 0. 0.
/LOW / 1.04E26 –2.76 1600
/TROE/ .7830 74.0 2941.0 6964
CH4 + 2O2 = CO2 + 2H2O 0. 0. 0.
/EBU / 4. 0. 1 0.
CH + N2 = HCNN 3.1E12 0.150 0.0
/M/ H2/2.0/ H2O/6.0/ CH4/2.0/ CO/1.5/ CO2/2.0/
/LOW / 1.3E25 –3.16 740
/TROE/ 0.667 235.0 2117.0 4536.0

Setting Up Chemical Reaction Schemes


Step 1
Go to the “Select Analysis Features” panel and choose option Chemical Reaction
from the Reacting Flow menu. Click Apply. The Reacting Flow sub-folder will
appear in the NavCenter tree, nested inside folder Thermophysical Models and
Properties.

8-14 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Chemical Reaction Schemes

Step 2
Open the Reacting Flow sub-folder to display a second sub-folder called Chemical
Reactions. This contains all panels needed to fully define a chemical reaction
scheme.
Step 3
Go to the Chemical Reactions sub-folder, open panel “Scheme Definition” and
select a free scheme number using the Chemical Scheme # scroll bar at the bottom
of the panel. You must then:
• Specify the basic reaction type (Unpremixed/Diffusion, Premixed,
Partially Premixed, or Heterogeneous/Surface) by choosing an option from
the Reaction Type menu
• Select the most appropriate reaction model for your problem from the
Reaction Model menu. The menu options depend on the reaction type
specified above.
• For some models, you will also need to specify the form of their
Implementation or the method of calculating the Unburnt Gas Temperature,
as explained in the “Scheme Definition” Help topic.
Step 4
In the “Reaction System” panel, use the on-line help provided to assist you in
specifying the relevant chemical reaction definitions, control settings and model
parameters. pro-STAR associates all chemical species defined in this panel with
additional scalar variables of the same name and also does a stoichiometric check
for every reaction. The required scalars and their properties are retrieved from
pro-STAR’s built-in database. Note that:
• If a species cannot be mapped to a material in a database, a warning is
displayed in the Output window and a fresh scalar of that name (but with
undefined properties) is created and added to the scalars list. You should
therefore go to the “Molecular Properties (Scalar)” panel to specify the
missing properties before proceeding further. It is also important that
definition of all domain (material) properties via panel “Molecular
Properties” has already been completed before any scalar properties are
defined.
• If the mass fraction of a non-reacting species is to be included in the
calculations, assign a scalar variable to the species via the “Molecular
Properties (Scalar)” panel and put it at the end of the existing scalars list.
• The parameters of a reaction can be redefined at any time by selecting its
parent scheme via the Chemical Scheme # scroll bar and then making the
necessary changes.
Step 5
In the “Ignition” panel, choose an ignition model or ignition start-up scheme,
depending on the chemical scheme type defined in Step 3
Step 6
If required, go to panel “Emission” and activate the built-in pollutant emission
models for NOx and/or soot

Version 4.02 8-15


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Chemical Reaction Schemes

Step 7
Some schemes allow the inclusion of knock modelling as part of the overall
chemical reaction simulation process. Parameters for this model may be specified
in panel “Knock”.
Step 8
If the Coupled Complex Chemistry model is in use, go to panel “Solution
Controls” to select the appropriate solution method controls and to perform the
necessary species-to-scalar mapping.
Step 9
Go back to Step 3 and repeat the above process until all schemes have been defined.
Step 10
Assign a reaction scheme to every fluid domain in your model using the “Scheme
Association” panel. Note that it is not necessary to assign every available scheme to
one of the domains. This allows you to define redundant schemes and then
experiment with different schemes for the same domain, by performing separate
analyses for each combination. In multi-domain problems where each domain has
a different scalar composition, the “Additional Scalars” panel (Equation Behaviour
sub-folder) enables you, in effect, to select which scalars exist in what domain.

It is strongly recommended to make use of pro-STAR Constants 64, 89 and 90 when


running combustion cases. Their effect is as follows:
• Setting Constant 64 = 2 will constrain calculated values for all active scalar
mass fractions to the range 0.0 — 1.0. Thus, numerical under- and
over-shoots that can destabilize the solution process may be avoided
• Constant 89 can be assigned to the minimum allowable temperature
calculated by STAR
• Constant 90 can be assigned to the maximum allowable temperature

Useful general points for local source and regress variable schemes
1. You are strongly recommended to perform stoichiometric checks for every
reaction, especially if Step 4 above found missing scalars that were
subsequently defined manually. To do this, click the Check Stoichiometry
button in the Reaction System tab when you have finished setting up the
model and before writing data to the problem (.prob) file.
2. For steady-state problems involving reactions that use a hybrid model,
experience so far has shown that the best practice is to obtain a converged
solution first, using only the eddy break-up model for all reactions. The
chemical kinetic model should then be employed by selecting the
Combined/User option and the analysis continued using the hybrid model
until the final solution is obtained.
3. The steady-state under-relaxation factors for temperature T and all scalar
variables representing transported mass fraction, mixture fraction, etc. should
be identical. The recommended range is 0.3 to 0.7. Note that this factor has no
effect for scalars calculated by other means, e.g. by an internal algebraic
equation.
4. The residual error tolerance for temperature and all scalar variables can be
8-16 Version 4.02
Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Chemical Reaction Schemes

tightened from the default value of 0.01 to 0.001. This will increase the
number of sweeps per PISO iteration but will improve the accuracy.
5. The turbulent Prandtl and Schmidt numbers for all scalar variables should be
identical.
6. For premixed flames, the value of mixture fraction is known and remains
constant throughout the analysis.
7. When defining domain material properties via the “Molecular Properties”
panel in STAR-GUIde, you are recommended to choose option Polynomial
in the “Specific Heat” pop-up menu. This will load suitable polynomials from
the CHEMKIN or CEC thermodynamic databases [1, 2]. A polynomial
variation for molecular viscosity and thermal conductivity can be specified in
the same way. For mass diffusivity, set via the “Diffusivity” panel in
STAR-GUIde, the Constant option is recommended for maximum efficiency,
particularly in the case of turbulent combustion.
8. If the same reaction appears in more than one scheme, user input can be
reduced by employing command RSTATUS to copy the reaction definition
from a previous scheme to the current one.
9. If modelling considerations demand it, individual reactions in multi-step
reaction systems can be turned on or off at appropriate points in the
simulation. This may be done by selecting Off in the Status pop-up menu
corresponding to the reaction concerned.
10. Chemical reactions (especially those for combustion) commonly take place in
a domain where air is the background material. Given that the nitrogen
component is often chemically inert and therefore does not appear in a
chemical reaction equation, it is convenient to include N2 as a separate scalar
to represent the background material. Therefore:
(a) If N2 does not appear in a reaction definition, pro-STAR will
automatically set up an extra active scalar called N2. By default, its
physical properties are those for nitrogen and the solution method is set to
Internal (see panel “Additional Scalars”). The value of the N2 mass
fraction returned by STAR is such as to make the mass fractions at every
cell sum to 1.0
(b) If N2 is present in a reaction definition, N2 will be set up like any other
scalar and its solution method will be set to Transport.

11. If you are modelling an EGR system, the recirculated gases must be explicitly
defined as active transported scalars within STAR Guide’s “Additional
Scalars” folder. These must also be given names that are different from those
of the parent species participating in the chemical reaction and make sure that
their properties (as defined in the “Molecular Properties (Scalar)” panel) are
correct. STAR will then be able to distinguish between species representing
products of the chemical reactions and the ones coming from the EGR stream.
12. Complex chemistry models must be run in double precision.
13. If a regress variable is employed by a combustion model, its initial value must
be set to 1 for correct model operation. You must therefore ensure that the
regress variable scalar in your model is initialised properly before proceeding
with the simulation.

Version 4.02 8-17


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Chemical Reaction Schemes

Chemical Reaction Conventions


The following conventions should be observed when typing reaction definitions in
the “Reaction System” panel:
1. Enter the ‘→’ symbol as two consecutive characters ‘->’
2. Specify the leading reactant as the first chemical substance on the left-hand
side of the reaction equation. Its name will appear in the Leading Reactants
list at the bottom of the panel, once the reaction details are confirmed.
3. Specify up to three ordinary reactants taking part in the reaction(s). Their
names will appear in the Reactant Parameters list, once the reaction details
are confirmed.
4. If a reaction constituent only occurs on the right-hand side of all reaction
equations, it will be assumed to be a product and its name will appear in the
Products list. However, if you wish this constituent to be a reactant (see, for
example, point no. 4. on page 8-2), type the symbol [R] immediately after its
name.
5. In multiple reaction schemes, the normal rule for what may appear as a
product is as follows:
(a) Reaction 1 is allowed to produce leading reactants 2 to 30 as products
(b) Reaction 2 is allowed to produce leading reactants 3 to 30 as products
(c) Reaction 3 is allowed to produce leading reactant 4 to 30 as products
.
.
.

(d) Reaction 29 is allowed to produce leading reactant 30 as a product


(e) Reaction 30 is not allowed to produce any leading reactants

For example, the two equations in the following scheme

CH 4 + 1.5 O2 → CO + 2H 2 O [ r ]
CO + 0.5 O 2 → CO 2

should be defined in the order shown above and not the other way round in
order to satisfy this rule. The system in this example also includes an influx of
H 2 O from an external source so that both O 2 and H 2 O are reactants in this
case. Therefore, the symbol [R] needs to be entered after the latter’s name.
6. Note that, point no. 5 above notwithstanding, STAR will still allow one
reaction only to create a product that has already been defined as the leading
reactant of a previous reaction.

Useful points for PPDF schemes


1. In single-fuel PPDF models, the quantities f and g f are automatically
assigned by STAR-CD as scalar numbers 1 and 2. For the multiple-fuel
model, the quantities f p , g f , f s and g ξ become scalar numbers 1 to 4,
respectively.
2. Any additional variables are assigned to further scalars, beginning with scalar

8-18 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Chemical Reaction Schemes

number 3 (single-fuel) or 5 (multiple-fuel). This can be confirmed by


displaying a STAR-GUIde panel that contains a Scalar list (for example,
“Initialisation” in the Additional Scalars folder).
3. In adiabatic PPDF applications:
(a) Remember that only the quantities given in item 1 above are calculated
from transport equations. Temperature, density and all other variables are
calculated internally. However, if any additional non-reacting scalars are
defined (see “Setting Up Chemical Reaction Schemes”, Step 4) these are
solved in the normal way.
(b) pro-STAR provides a reminder that density is no longer calculated by one
of the normal options. Thus the density setting in the “Molecular
Properties” panel is automatically changed to read PPDF.
(c) Polynomial coefficients should be supplied in terms of molar fractions
(kmol/kmol). However, scalar concentrations for initial and boundary
conditions should be specified as mass fractions.
(d) If the molecular weights of all scalar species are correctly specified,
STAR will output the calculated species concentrations in terms of mass
fractions. However, if all species molecular weights are assigned the same
value, the output will be in terms of species mole fractions.

4. In non-adiabatic PPDF applications, check the information displayed by the


STAR-GUIde interface to ensure that:
(a) Option Active is selected from the Influence pop-up menu for all
chemical species (“Molecular Properties (Scalar)” panel in folder
Additional Scalars)
(b) Option Chemico-Thermal is selected from the Enthalpy pop-up menu
(“Thermal Models” panel in folder Liquids and Gases)
(c) The Ideal-f(T,P) option is used for density (see topic “Density”)
(d) The Polynomial option is used for specific heat (see topic “Specific
Heat”)
(e) The scalar species concentrations are specified in terms of mass fractions

5. If the PDF is to be calculated by numerical integration, a number of control


parameters should be specified. These are illustrated in the Figure below:

Version 4.02 8-19


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Chemical Reaction Schemes

MF
××

}
×

N1 N2 fs Ni Ni+1 NF f
0 1

Figure 8-1 Control parameters for PDF integration

The quantities shown in Figure 8-1 are defined as follows:


(a) f s — stoichiometric mass fraction
(b) N F — mixture fraction points. This is the total number of locations
where chemical equilibrium calculations are performed.
(c) MF — multiplying factor. This is the number of points added between
any two adjacent points, such as N i and N i + 1 . These extra points are
used for improving the resolution of the calculation and their values are
extrapolated from those at N i and N i + 1 . The total number of points Nt
used in the integration is given by
N t = ( MF + 1 ) × ( N F – 1 ) + 1 . (8-2)
(d) P F — integration partition. This parameter represents the percentage of
points used to resolve the region between 0 and f s in the mixture fraction
space, i.e. the number of points in this region is given by ( P F ⁄ 100 ) N t

6. When using the laminar flamelet model, the following points should be borne
in mind:
(a) Each flamelet library refers to a different strain rate. A typical example
might be to have 6 flamelet libraries at strain rates of 0, 25, 50, 200, 400
and 1000 s-1.
(b) Calculating flamelet libraries may be very time consuming. Therefore,
when creating a new library, you should consider restarting the
calculation from the nearest available strain rate wherever possible.
However, if the difference in strain rate is quite large and convergence
becomes difficult, it will be necessary to specify a new set of initial
conditions and start again.
(c) STAR-CD provides an option for either specifying the inlet strain rate or

8-20 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Chemical Reaction Schemes

calculating it via a built-in code. For simplified reaction mechanisms, try


the first alternative combined with the restart option from a previously
converged strain rate. For more complex mechanisms, you may want to
try the second alternative, check what strain rate the code calculates, and
then change the initial conditions accordingly. When the initial conditions
are sufficiently close to the desired strain rate, you may be able to select
the first alternative with a restart option to achieve a solution.
(d) The results of each flamelet library calculation are printed out in a
separate output panel. You should always inspect that panel to ensure the
displayed values are reasonable.
(e) If your problem setup contains multiple reaction scheme definitions, any
laminar flamelet model(s) should appear at the top of the reaction scheme
list.

Useful points for complex chemistry models


1. The distinction between premixed, partially premixed and unpremixed
combustion made in the pro-STAR GUI is irrelevant for complex chemistry
models, since transport equations are solved for all species (or one of them is
calculated as 1 – ΣY i ). Hence, this model is available for all the above
reaction types.
2. The calculation of reaction rates can be very time-consuming. Users may
therefore specify, via Constant 173, a temperature limit below which reaction
rates will not be calculated. The default value of this limit is 300 K but may be
re-set as necessary.
3. The steady-state complex chemistry solver employs an internal sub-timestep
whose default value is 10–5. Users may change this value via Constant 154.
Normally, a very small sub-timestep value will result in the calculation of
large reaction rates, which could in turn make the solution of the steady-state
transport equations unstable. On the other hand, if the value is too large, the
chemistry solver will become very time-consuming.
4. For very stiff problems, the maximum number of sub-timesteps may need to
be increased beyond its default value, currently set at 500. This is done via
Constant 192. Users can also change the chemistry solver’s relative and
absolute convergence tolerance via Constants 123 and 124, respectively. The
default values for these are set at 10–4 and 10–10, respectively.
5. There is a balance between robustness and convergence rate. The latter may
be increased by higher values of the species under-relaxation factor, but users
should be careful that the stability of the solution is not sacrificed at the same
time.
6. For steady-state cases, it is recommended that the initial species distribution
should correspond to a non-combustible mixture, such as air.

Useful points for ignition models


1. Shell and 4-step ignition models: Option Use Heat of Reaction in the
“Reaction System” STAR GUIde panel is valid only when the
pro-STAR-defined specific heat polynomial coefficients are used. When the
reaction is exothermic, the heat of reaction value is negative. For an

Version 4.02 8-21


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Advanced I.C. Engine Models

endothermic reaction, the value is positive.

2. CFM ignition model: In problems where only a part of the solution domain is
being simulated, you need to specify (via Constant 142) a geometrical factor
whose value is the fraction of the flame kernel area in the partial (simulation)
domain relative to the entire domain. For example, for a wedge-shaped
solution domain in a cylindrical system and with the ignition point lying on
the axis, this value should be ∆θ ⁄ 360 , where ∆θ is the angular extent of the
wedge. The default value of the above factor is 1.

Setting Up Advanced I.C. Engine Models


The notes below describe the set up of combustion simulations that employ the
CFM, ECFM, ECFM-3Z or Level Set models. Note that:
• The GUI facilities presented are available only when running the Auto Mesh
version of pro-STAR (prostar -amm).
• Use of ECFM, ECFM-3Z and their attendant ignition models requires a
special licence obtainable from CD-adapco.

Note also that if you are resuming from an .mdl file in which a combustion model
has been defined, it is important to delete this model, its submodels (such as NOx,
Soot and Knock) and the associated scalars before selecting an alternative model.
This requires issuing the following pro-STAR commands (or performing the
equivalent GUI operations):

SCDEL,ALL — delete all scalars


SOOT,n,OFF — turn off soot modelling, where n stands for every
curently-defined chemical scheme number
NOX,n,OFF — turn off Nox modelling
KNOCK,n,OFF — turn off knock modelling
CRDEL,ALL — delete all chemical reaction schemes
CHSCHEME,m,NONE — remove chemical scheme associations with
STAR domains (streams), where m stands for every currently-defined
domain
CHER,OFF — turn off chemical reaction calculations

To set up a case:
Step 1
In the Select Analysis Features panel, select all general features required for an
engine combustion simulation (these parameters will be selected in advance if the
es-ice engine simulation expert system is used):
• Time Domain > Transient (select option Angle and enter values for RPM
and Initial Position)
• Reacting Flow > Chemical Reaction
• Multi-Phase Treatment > Lagrangian Multi-phase (if modelling sprays)
Step 2
Select folder Thermophysical Models and Properties > Liquid and Gases and then
enter appropriate values in the following panels:
8-22 Version 4.02
Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Advanced I.C. Engine Models

• Molecular Properties
• Turbulence Models
• Thermal Models (select Temperature Calculation On and then choose
options Conservation > Static Enthalpy and Enthalpy > Thermal)
• Initialization
• Monitoring and Reference Data
• Buoyancy (where applicable)
Step 3
At this stage, the special IC set-up panel can be used, accessed by selecting
Advanced > IC Setup from pro-STAR’s main menu. The panel shown below will
then pop up:

• The panel will initially display the Analysis setup sub-panel. Check that the
Combustion option is selected and then choose the type of combustion model
required from the drop-down menu underneath.
• Fuel parameters:
Select the desired fuel from the second drop-down menu. Depending on the
model type, i.e. spark or compression, the panel will display a default octane
or cetane number in the text box on the right. You should replace this with an
appropriate value if necessary. The corresponding chemical reaction formula
will also be displayed below the fuel name.
Step 4
Click the Combustion tab button on the left to display the Combustion sub-panel.
Its contents depend on the combustion model selected, as described below.

Version 4.02 8-23


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Advanced I.C. Engine Models

Coherent Flame model (CFM)

The chemical reaction for the chosen fuel is displayed at the top.
• Modelling parameters — enter values for coefficients α and β to be used in
the source term of the Σ equation (see equation (11-10) in the Methodology
volume).
• Ignition parameters:
(a) In the Spark time box, input the time (in degrees crank angle or in
seconds) at which the spark is to be discharged
(b) Enter the ignition Location in terms of X, Y, Z coordinates relative to

8-24 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Advanced I.C. Engine Models

coordinate system no. 1 (in model units)


(c) In the Delay box, enter the time delay (or transition time t1 , see Chapter
11, “Ignition treatment for the CFM-ITNFS model” in the Methodology
volume)
(d) In the Kernel diameter box, enter the appropriate value in mm

• Enter the Mixture fraction for the premixed air-fuel mixture


• Specify whether exhaust gas recirculation (EGR) should be On or Off
• Specify whether emissions (NOx) and/or Knock is to be modelled by selecting
On or Off from their respective drop-down menus.

Apart from the above input, the panel will also execute commands to effect the
following changes (which will overwrite any property settings specified previously
in pro-STAR’s Molecular Properties panel):
• Set up appropriate property definitions for material #1 and change the
molecular viscosity setting to MultiComponent
• Define the specific heat setting of the background fluid as Polynomial
• Create chemical species scalars and assign appropriate physical properties to
them, imported from the built-in property database props.dbs

Note that there is an alternative method of setting up a CFM model using


pro-STAR’s Chemical Reactions panels in STAR GUIde. The main difference
between the two is that
• pro-STAR specifies the ignition location in terms of the centroid of an
‘ignition cell’
• the IC Setup panel specifies the location in terms of its X, Y, Z coordinates
and passes them on to STAR via an Extended Data segment delimited by the
keywords BEGIN SPARK_DATA and END SPARK_DATA and appended to
the .prob file.

It is most important that the two approaches should not be mixed.

Version 4.02 8-25


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Advanced I.C. Engine Models

Extended Coherent Flame model (ECFM)

The chemical reaction for the chosen fuel is displayed at the top.
• Modelling parameters — enter values for coefficients α and β to be used in
the source term of the σ equation (see equation (11-90) in the Methodology
volume).
• The PSDF Moments soot model may be used by selecting On from the Mauss
Soot Model drop-down menu (see “Soot Modelling” on page 8-39)

8-26 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Advanced I.C. Engine Models

• Ignition parameters — there are two ignition model choices:


(a) Standard:
i) In the Spark time box, enter the time (in degrees crank angle or in
seconds) at which the spark is to be discharged
ii) Enter the ignition Location in terms of X, Y, Z coordinates relative
to coordinate system no. 1 (in model units)
(a) Aktim — see “The Arc and Kernel Tracking ignition model (AKTIM)”
on page 8-33

• Enter a Sector angle value if you want to perform a “Sector Mesh” analysis.

The panel also executes additional commands to effect the following changes:
• Turn on the Transient setting and associate the ECFM chemical scheme with
material #1
• Define 17 scalars and their material properties

Version 4.02 8-27


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Advanced I.C. Engine Models

Extended Coherent Flame model 3Z (ECFM-3Z) — spark ignition

The chemical reaction for the chosen fuel is displayed at the top.
• Modelling parameters — enter values for coefficients α and β to be used in
the source term of the σ equation (see equation (11-90) in the Methodology
volume).
• Select the Multiple Cycles option if you wish to run a simulation over multiple
engine cycles

8-28 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Advanced I.C. Engine Models

• The PSDF Moments soot model may be used by selecting On from the Mauss
Soot Model drop-down menu (see “Soot Modelling” on page 8-39)
• Parameter Start of ECFM-3Z is the time at which fuel / air mixing and
combustion will start to be calculated. The user should set this time equal to
the time when fuel injection starts, in degrees crank angle or seconds.
• Ignition parameters — there are two ignition model choices:
(a) Standard:
i) In the Spark time box, enter the time (in degrees crank angle or in
seconds) at which the spark is to be discharged
ii) Enter the ignition Location in terms of X, Y, Z coordinates relative
to coordinate system no. 1 (in model units)
(b) Aktim — see “The Arc and Kernel Tracking ignition model (AKTIM)”
on page 8-33

• Multiple ignition locations — the number of locations can be


increased/decreased by clicking the up/down # location arrows. Coordinates
for each ignition location can be entered by selecting the particular location
from the drop-down menu.
• Specify whether Knock is to be modelled by selecting On or Off from the
drop-down menu
• Enter a Sector angle value if you want to perform a “Sector Mesh” analysis.

In addition, the panel defines 25 appropriate scalars and their material properties.
Extended Coherent Flame model 3Z (ECFM-3Z) — compression ignition

The chemical reaction for the chosen fuel is displayed at the top.
• Modelling parameters — enter values for coefficients α and β to be used in
the source term of the σ equation (see equation (11-90) in the Methodology
volume).
• Select the Multiple Cycles option if you wish to run a simulation over multiple
engine cycles
• The PSDF Moments soot model may be used by selecting On from the Mauss
Soot Model drop-down menu (see “Soot Modelling” on page 8-39)

Version 4.02 8-29


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Advanced I.C. Engine Models

• Parameter Start of ECFM-3Z is the time at which fuel / air mixing and
combustion will start to be calculated. The user should set this time equal to
the time when fuel injection starts, in degrees crank angle or seconds.
• There is an additional option to turn On the Tabulated Double-Delay
Autoignition Model (see “The Double-Delay autoignition model” on page
8-37).
Useful points for ECFM models
1. The information entered in the ECFM and ECFM-3Z panels is passed on to
STAR via an Extended Data segment delimited by the keywords BEGIN
ECFM_DATA and END ECFM_DATA. This segment is created automatically
and may be inspected by selecting Utilities > Extended Data from the main
pro-STAR window’s menu bar but the user does not need to add any further
information to it. The segment is appended to the end of the .prob file when
the later is saved at the end of the current pro-STAR session.
2. All ECFM models must be run in double precision.
3. Although the list of fuels for use in these models is limited, users can supply
their own fuel definition by selecting option User Defined from the fuel
selection drop-down menu and then specifying the number of C and H atoms.
The specific heat ‘cp’ should be changed accordingly.
4. ECFM models cannot be used in conjunction with the k-ω or the
Spalart-Allmaras turbulence models.
5. Additional scalars may be added but they must be Inactive.
6. It is important to remember that species N, O, H, OH cannot be present as part
of EGR gases.
7. Model parameters should be specified via the panels described in this
document. Users should not attempt to supply any parameters via the standard
pro-STAR STAR GUIde panels.
8. If ECFM models are applied in materials other than no. 1, make sure that the
solution method for scalars in these materials is the same as for material 1.
The correct setting is Transport, except for scalar RVB (the progress variable
‘c’) which should be Internal.
9. The ECFM model is currently emulated by the ECFM-3Z model. Tracers for
species NO, CO, H2 and SOOT are turned off, along with O2UM and FUM,
and the corresponding mass fractions are set to 0.0 (i.e. the system is always
“mixed”). Comparisons with results from earlier STAR-CD implementations
will show slightly differences as different sub-models for the post-flame
regime are used in the current version.
10. For unburnt gases, the initial mass fractions of fuel and oxygen tracers (TF,
TO2) plus any other applicable tracers (TCO, TH2, TNO, TSOOT) must be
set equal to the corresponding initial mass fractions of species Fuel and O2
plus, if applicable, species CO, SOOT, NO, H2.

8-30 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Advanced I.C. Engine Models

Level Set model

The chemical reaction for the chosen fuel is displayed at the top.
• Re-initialization method — select an option from the adjacent pop-up menu
(see Chapter 9, “Re-initialisation” in the Supplementary Notes volume)
• Select the Multiple Cycles option if you wish to run a simulation over multiple
engine cycles
• Ignition parameters
(a) In the Spark time box, enter the time (in degrees crank angle or in
seconds) at which the spark is to be discharged
(b) Enter the ignition kernel Location in terms of X, Y, Z coordinates relative
to coordinate system no. 1 (in model units) and ignition kernel Radius (in
metres)
(c) Enter the ignition Duration (in degrees crank angle or in seconds)

• Multiple ignition kernel locations — the number of locations can be


increased/decreased by clicking the up/down # location arrows. Data for each
ignition kernel can be entered by selecting the particular location from the
drop-down menu.

The information entered in the Level Set panel is passed on to STAR via an
Extended Data segment delimited by the keywords BEGIN LEVELSETDATA and
END LEVELSETDATA. This segment is created automatically and may be
inspected by selecting Utilities > Extended Data from the main pro-STAR
window’s menu bar. The user does not normally need to add any further information
to it unless advanced features of the model need to be implemented, as described in
Chapter 9 of the Supplementary Notes volume.
The Extended Data segment is appended to the end of the .prob file when the
later is saved at the end of the current pro-STAR session.

Version 4.02 8-31


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Advanced I.C. Engine Models

Step 5
Click the Write data tab button on the left to display the Write data sub-panel,
shown below.
Write Data sub-panel

In this panel:
• Click Save parameters to save the combustion parameters to a file called
star.ics. This file is important as it contains combustion data and should
be kept together with the .mdl file. The file name can be changed if
necessary.
• Click Write and close to write the problem settings to a model (.mdl) file
and close the panel
• Click Close to close the panel without writing anything to the .mdl file

Step 6
Initialize certain scalars, as shown below. Only those scalars included in a list
should be initialised; all others should have 0 as their initial value.
• For ECFM:
(a) Fuel → set the fuel initial mass fraction
(b) O2 → set the O2 initial mass fraction
(c) CO2 → set the CO2 initial mass fraction
(d) H2O → set the H2O initial mass fraction
(e) N2 → set the N2 initial mass fraction
(f) TF → set the unburnt fuel initial mass fraction
(g) TO2 → set the unburnt O2 initial mass fraction
• For ECFM-3Z:
(a) Fuel → set the fuel initial mass fraction
(b) O2 → set the O2 initial mass fraction
(c) CO2 → set the CO2 initial mass fraction
(d) H2O → set the H2O initial mass fraction
(e) N2 → set the N2 initial mass fraction

8-32 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Advanced I.C. Engine Models

(f) TF → set the unburnt fuel initial mass fraction


(g) TO2 → set the unburnt O2 initial mass fraction
(h) TCO → set the unburnt CO initial mass fraction
(i) TH2 → set the unburnt H2 initial mass fraction
(j) TNO → set the unburnt NO initial mass fraction
(k) TSOOT → set the unburnt soot initial mass fraction
(l) FUM → set the initial mass fraction of the Unmixed Fuel (default value is
0)
(m) O2UM → set the initial mass fraction of the Unmixed Oxygen

Step 7
Define the time step size and any load steps in the appropriate transient settings
panel (where applicable).

The Arc and Kernel Tracking ignition model (AKTIM)


AKTIM may be used only with one of the ECFM options, as follows:
• Select the required ECFM model
• Choose option Aktim from the Ignition menu
• In the Spark time box, enter the time (in degrees crank angle or in seconds) at
which the spark is to be discharged
• Enter the ignition Location in terms of X, Y, Z coordinates relative to
coordinate system no. 1 (in model units)
• If the electrodes are represented as distinct entities in the mesh, select option
Regions from the Electrode Model menu and input the boundary region
numbers corresponding to the anode and cathode. Otherwise, select Area and
temperature from the Electrode Model menu and input the surface area and
temperature of the anode and cathode. The two alternatives are illustrated
below.
• Specify the Diameter of the cathode and anode electrodes
• Input values for parameters E 2 ( t SI ) , L , R in the Secondary Circuit section of
the panel
• Multiple ignition locations — the number of locations can be increased/
decreased by clicking the up/down # location arrows. Coordinates for each
ignition location can be entered by selecting the particular location from the
drop-down menu.

Version 4.02 8-33


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Advanced I.C. Engine Models

8-34 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Advanced I.C. Engine Models

Useful points for the AKTIM model


Use of Extended Data
The information contained in the above panels is passed on to STAR via the ECFM
Extended Data segment (see “Useful points for ECFM models” on page 8-30). The
contents of this segment do not need to be altered by the user except in the case of
two-dimensional (x-y) problems. For such problems:
• Select Utilities > Extended Data from the main pro-STAR window’s menu
Version 4.02 8-35
CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Advanced I.C. Engine Models

bar to display the current ECFM_DATA segment in the Extended Data panel.
• Enter a new line containing the keyword TWODS anywhere within the
segment (typically at the end)
• Save and then Close the Extended Data panel

Entering the above keyword enforces the turbulent fluctuation contribution to the
calculation of the flame kernels’ positions in the z-direction to be zero.
Post and track data
Two new files are written by STAR when AKTIM is in use:
1. casename.strk, the track file for spark particles
2. casename.ktrk, the track file for flame kernel particles
Track files
The .strk and .ktrk files can be used in the same way as .trk files for droplets
(see Chapter 7, “The Particle Track File” in the Post-Processing User Guide). They
can be loaded via the Plot Droplets/Particle Tracks panel by choosing option Track
File in the Load Droplet Data section and then specifying the appropriate file name
and extension. Note that:
• This action will erase all current droplet track data
• Only option Constant is supported in the Droplet Radius menu
• Only option Color is supported in the Fill Color menu
Post files
The .ccm file contains all solution data and can be used to plot spark particles and
flame kernel particles. These can be loaded into pro-STAR via command
GETD,POST,SPARK for spark particles or GETD,POST,FLAME for flame kernel
particles. Note that:
• After the .ccm file is loaded, the wrinkling factor, progress variable, mass
and burnt gas mass in the flame kernel particle can be plotted by selecting
Diameter, Temperature, Mass and Count, respectively, from the Fill Color
menu in STAR GUIde’s Plot Droplets/Particle Tracks panel
• No other options in the Fill Color menu are supported

There are no files equivalent to .pstt or .ccm_timestep for the spark and
flame kernel particles.
HPC issues
For HPC calculations, the electrodes must be meshed or specified in only one
processor domain. In addition, spark particles will not be permitted to pass between
different processor domains.
Mesh quality
It is important to use good quality meshes when the electrodes are resolved. This is
because the source terms might be very high during the ignition process. Thus,
localized and big numerical errors due to poor quality meshes may lead to wrong
results, instabilities and/or divergences.

8-36 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Setting Up Advanced I.C. Engine Models

The Double-Delay autoignition model


This model may be employed by selecting ECFM-3Z, compression and then
turning On the Tabulated Double-Delay Autoignition Model option, as shown
below. A new scalar YIG2 is created to track the primary autoignition progress.
Scalar YIG monitors the progress of the main autoignition.

STAR requires three types of data for this model:


1. The primary ignition delay time tdelay1
2. The secondary ignition delay time tdelay2
3. The burnt fuel mass fraction ffrac

Each of the above is tabulated at nT values of temperature T, nP values of pressure


p, nE values of equivalence ratio E and nX values of residual gas mole fraction X.
Hence, a 4-dimensional array structure is employed for each data item.
By default, the model uses ignition delay data read from pre-computed tables.
Users may also use their own data by storing them in three separate text files called
delay1.dat, delay2.dat and ffrac.dat for the primary delay, secondary
delay and burnt fuel mass fraction, respectively. The files must be placed in the case
working directory.

User data creation procedure


The following steps are necessary to create user-generated data files:
• Select Utilities > Extended Data from the main pro-STAR window’s menu
bar to display the current ECFM_DATA segment in the Extended Data panel.
• Enter a new line containing the keyword LU2DATA in the line after LAUTO2
• Save and then Close the Extended Data panel
• Create the three data files described above in a format suitable for STAR
input. The Fortran program listed below shows how these files should be
written.
Program Example_UserData
c
c nT, nP, nE, nX, Tmin, Tstep, p, E, X, tdelay1, tdelay2, ffrac are user input data.
c Tmin, Tstep are the minimum temperature and the temperature step (K) in the table,
c respectively, so that
c T(1)=Tmin, T(2)=T(1)+Tstep, T(3)=T(2)+Tstep, ...
c p(1),...,p(nP) are the nP pressure (bars) points in the table in ascending order.

Version 4.02 8-37


CHEMICAL REACTION AND COMBUSTION Chapter 8
Setting Up Advanced I.C. Engine Models

c E(1),...,E(nE) are the nE equivalence ratio points in the table in ascending order.
c X(1),...,X(nX) are the nX residual gases points in the table in ascending order.
c tdelay1(iT,iP,iE,iX), tdelay2(iT,iP,iE,iX), ffrac(iT,iP,iE,iX) are user-defined
c 4-dimensional arrays for the primary delay, secondary delay and burnt fuel mass
c fraction.
c
c... File "delay1.dat"
c
Double Precision Tmin,Tstep,p,E,X,tdelay1,tdelay2,ffrac
Dimension p(nP), E(nE), X(nX)
Dimension tdelay1(nT,nP,nE,nX)
Dimension tdelay2(nT,nP,nE,nX)
Dimension ffrac(nT,nP,nE,nX)
c
open(iunit,file=’delay1.dat’)
write(iunit,*) nT, nP, nE, nX
write(iunit,*) Tmin, Tstep
do ip=1,nP
write(iunit,*) p(ip)
end do
do ie=1,nE
write(iunit,*) E(ie)
end do
do ix=1,nX
write(iunit,*) X(ix)
end do
do ix=1,nX
do ie=1,nE
do ip=1,nP
do it=1,nT
write(iunit,*) tdelay1(it,ip,ie,ix)
end do
end do
end do
end do
close(iunit)

c
c... File "delay2.dat"
c
open(iunit,file=’delay2.dat’)
do ix=1,nX
do ie=1,nE
do ip=1,nP
do it=1,nT
write(iunit,*)tdelay2(it,ip,ie,ix)
end do
end do
end do
end do
close(iunit)
c
c... File "ffrac.dat"
c
open(iunit,file=’ffrac.dat’)
do ix=1,nX
do ie=1,nE
do ip=1,nP
do it=1,nT
write(iunit,*)ffrac(it,ip,ie,ix)
end do
end do
end do
end do
close(iunit)
c
end

Note that if any of the above files is not present, STAR will abort the simulation
with warning messages.
For verification purposes, STAR will also output three text files at the beginning
of the simulation containing the parameters and data specified by the user. The files
are called check_data1 for the primary delay, check_data2 for the

8-38 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
NOx Modelling

secondary delay and check_dataf for the burnt fuel mass fraction and are
written into the working directory for sequential runs or into the first processor’s
directory for HPC runs.

NOx Modelling
NOx concentration is usually low compared to other species in combustion systems.
As a result, it is generally agreed that NOx chemistry has negligible influence and
can be decoupled from the main combustion and flow field calculations.
The recommended procedure for performing a NOx analysis is as follows:
Step 1
Set up the combustion model as usual.
Step 2
In the Chemical Reactions folder of STAR GUIde, open the “Emission” panel and
then go the “NOx” section. Select option On from the NOx Model menu to activate
STAR-CD’s built-in NOx subroutines.
Step 3
Turn on the appropriate NOx production mechanism from the Thermal, Prompt or
Fuel menus (see Chapter 10, “NOx Formation” in the Methodology volume).
Option User in any of these menus enables you to perform the necessary
calculations via subroutine NOXUSR. If option On is selected for Thermal NOx,
specify values for the required constants as explained in the on-line help topic for
“NOx”.
Step 4
Check that pro-STAR has created an extra passive scalar variable called NO, by
opening the “Molecular Properties (Scalar)” panel in the Additional Scalars folder
and inspecting the currently defined scalars.
If the problem requires the prediction of fuel NOx (this is only applicable to
nitrogen-containing fuels, e.g. coal), check that an additional passive scalar called
HCN has also been created.
Step 5
If your model provides for the calculation of OH and H mass fractions, their values
will be used in equation (10-84) of the Methodology volume to implement the
extended Zeldovich mechanism.
Step 6
For steady-state problems, make sure that a sufficient number of iterations has been
performed for the solution of NO and (if present) HCN to have converged.

Soot Modelling
The Flamelet Library soot model is applicable only to unpremixed and partially
premixed reactions and is activated via the Emission panel’s “Soot” section in
STAR GUIde. The only user input required is four scaling factors, see equation
(10-120) and (10-121), that determine the magnitude of the contribution from each
source term.
For example, a decrease in the value of the scaling factors for positive source

Version 4.02 8-39


CHEMICAL REACTION AND COMBUSTION Chapter 8
Soot Modelling

terms (surface growth and particle inception) results in slower formation of soot. In
diffusion flames, this can shift the point of maximum soot volume fraction further
downstream.
A typical range for these factors is 0.5 — 10.0 and their default value is 1.
The PSDF Moments soot model may be used in conjunction with any ECFM
combustion model (see “Setting Up Advanced I.C. Engine Models”) and is
accessed via the GUI facilities presented when running the Auto Mesh version of
pro-STAR (prostar -amm). The relevant panel is shown below:

To use this model:


• Select the required ECFM model
• Select option On from the Mauss Soot Model menu
• Choose the number of Moments to be solved for (0, 2, 3, or 4). The effect of
EGR (if present) will be taken into account. If 0 is selected, “The Flamelet
Library method” described in Chapter 10 of the Methodology volume will be
used.
• Enter scaling factors (see “The method of moments” on page 10-29 of the
Methodolgy volume for definitions) for the following quantities:
(a) Surface Growth
(b) Fragmentation
(c) Particle Inception
(d) Oxidation Rate

Use of this method with models other than ECFM is possible only within the
pro-STAR environment. The steps required for set-up in this case are as follows:
1. Select a combustion model that allows Soot to be employed
2. Select option Soot on in the Emissions panel
3. If no moments need to be solved, bypass steps 4, 5 and 6
4. Define 2, 3 or 4 additional passive scalars with names M0, M1, M2, M3,
depending on how many moments are to be solved for. These represent the
quantities Mr ⁄ ρ with Mr the r-th moment and units of [mol/m3]. STAR will

8-40 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Coal Combustion Modelling

identify them as the moments for the soot model.


5. Select option Transport in the Analysis Control > Solution Control >
Equation Behaviour > Additional Scalars panel as the method of solution for
these scalars
6. Deactivate the SOOT scalar, as in this case the soot mass fraction is obtained
from the soot PSDF moments solution
7. If EGR is present (up to 40% in concentration), then:
(a) Define 3 new active scalars with names EGR_CO2, EGR_N2, EGR_H2O.
These distinguish the CO2, N2, H2O arising from the EGR from the CO2,
N2, H2O arising from the combustion
(b) Assign the same molecular properties to them as for scalars CO2, N2,
H2O in the Molecular Properties (Scalar) panel. in addition, the
properties in the Binary Properties panel should be the same as for scalars
CO2, N2, H2O.
(c) Select option Transport in the Analysis Control > Solution Control >
Equation Behaviour > Additional Scalars panel as the method of solution
of the EGR scalars.
Note that EGR can be set up for soot cases even if no moments are solved for.
The EGR for soot calculation is not supported for PPDF models.

If at least two moments are solved for, the following (mass averaged) data will be
produced at each time-step:
Soot Volume, SootMass, Number Density, Mean Diameter, Dispersion Size
Distribution, Variance of the Size Distribution, Surface Density
These are added after the scalar data in the .spd file (see Chapter 9, “Engine
Combustion Data Files” in this volume).

Coal Combustion Modelling


Coal combustion models involve two-phase flow with complex solid and gas phase
chemical reactions. To reduce CPU time, it is recommended to run such a
simulation as a two-stage process using the STAR GUIde system. Thus, initially the
problem is run isothermally. Then, once a reasonably converged solution is
obtained, the problem is re-run with coal combustion turned on. An outline of the
steps involved and recommendations on model set-up at each stage of the process
is given below:
Stage 1
Run the model as an isothermal (non-reacting) problem and obtain a converged
solution which effectively serves as an initial condition for the flow field.
Step 1
Generate a mesh for the problem as usual and check that the steady-state analysis
mode has been chosen in the “Select Analysis Features” panel.
Step 2
Check that the temperature calculation is switched on in the “Thermal Models”
panel (Liquids and Gases sub-folder) and select an appropriate turbulence model in
the “Turbulence Models” panel.

Version 4.02 8-41


CHEMICAL REACTION AND COMBUSTION Chapter 8
Coal Combustion Modelling

Step 3
Define all boundaries and set up boundary conditions throughout, including
appropriate temperature distributions at inlet boundaries.
Step 4
Run the case until a reasonable flow field is established (see Note 1 below).
Stage 2
Turn on coal combustion and generate the final solution as follows:
Step 1
Go to the “Select Analysis Features” panel and choose option Coal Combustion
from the Reacting Flow menu and Lagrangian Multi-Phase from the Multi-Phase
Treatment menu. Click Apply. The Coal Combustion sub-folder will appear in the
NavCenter tree, nested inside folders Thermophysical Models and Properties >
Reacting Flow. An additional sub-folder, Lagrangian Multi-Phase, will also open
in the NavCenter tree. At the same time, pro-STAR will set up your model
automatically for this type of analysis, using the ‘Constant Rate’, ‘1st-Order Effect’
and ‘Mixed-is-burnt’ sub-models as defaults for volatiles, char and gas combustion,
respectively.
Step 2
If radiation effects are to be modelled, go to the “Thermal Options” panel in folder
Thermophysical Models and Properties and select the appropriate radiation model,
as described in Chapter 7 of the CCM User Guide (see also Note 2 below).
Step 3
Go to the Coal Combustion sub-folder and supply or modify data in each of its
panels in turn:
• In the “Coal Composition” panel, enter coal composition data and click
Apply. The data supplied in this panel can be stored in a file called
coal.dbs by entering a coal name and clicking Save to D/B. Alternatively,
you can read the coal composition from an existing file by clicking Open
D/B.
• Data entered in the Proximate analysis tab should be supplied on an air
dried basis.
• Data in the Ultimate analysis tab should be supplied on a ash-free basis,
where C + H + O + N + Chlorine (Cl) + Sulphur (S) = 1. The Cl and S
components are then assumed to be ash. If Cl and S need to be included in
the calculations, these components can be modelled separately via user
subroutine PARUSR at a later point. Details of this method are given in
Chapter 3 of the Supplementary Notes volume describing coal blend
modelling. This chapter also covers all other aspects of this type of
model, such as specifying the components and reaction rates for the
different coals in the blend.
• Supply the coal Q factor, which is an adjustment for volatile matter, in the
Miscellaneous tab. Studies have shown that under certain heating
conditions, a significantly higher amount of volatile matter can be
devolved than that measured by the standard proximate analysis test. This

8-42 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Coal Combustion Modelling

effect is accounted for by the Q factor which is defined by

(V*/VM) = Q (8-3)

where V* is the volatiles yield and VM the proximate analysis matter.


• Also in the Miscellaneous tab, enter the net calorific value of the fuel and
the fraction of total nitrogen in the volatiles. Finally, choose the volatiles
specific heat option (see Note 3 below).

• In the “Sub Models” panel, select the desired models for volatiles, char and
gas combustion.
• To speed up convergence, the Constant Rate scheme should be selected
initially in the Volatiles tab. The initial devolatilisation temperature
should be set to that of the primary inlet flow containing the coal particles
(this will help with initialising the temperature field and instigating
ignition). After the rest of the coal particle parameters have been set, the
problem should be run using an Initial Field Restart (panel “Analysis
(Re)Start”) from the isothermal solution obtained in Stage 1 and run for
several hundred iterations or until a reasonably stable solution is reached.
The devolatilisation temperature should then be raised to a more realistic
level (e.g. 550 K) and the model run once again using a Standard
Restart until a stable solution is reached. The Single Step or
2-Competing steps model should then be chosen. These require values
for pre-exponential factors and activation energy that should be
determined experimentally or taken from the available literature (as is the
case with the default values used by STAR). The solution should then be
re-run with a standard restart (see Notes 4 and 5 below regarding
changing parameters or submodels in this panel).
• In the Char tab, select one of the three char models available. Char rates
can be determined experimentally or taken from the literature.
• In the Gaseous Combustion tab, select either Mixed-is-Burnt (fast
chemistry approach) or the 1-step or 2-step EBU models. When this is
set, pro-STAR will create the appropriate scalars and select the transport
or internal solution method for them, depending on the model chosen.
Char oxidation products can be changed for specific gaseous combustion
models using Constant 120, as explained in “Switches and constants for
coal modelling” below.

• In the “NOx/Radiation” panel, turn on the NOx generation and/or coal


particle radiation options, as required. Note that, if the latter is chosen, you
should already have set up your model for radiation calculations as described
in Chapter 7. Enter the particle emissivity value. The NOx model can be
turned on near the end of the solution as it has only a small effect on the
overall flow field.
• In the “Control/Printout” panel, specify the required solution control and
printout parameters. It is sometimes necessary to initially reduce the
under-relaxation factor of the particle source term (to as low as 0.1) in order
to achieve a stable solution. The factor may be increased later on in the
solution. The iteration number at which particle source term averaging starts

Version 4.02 8-43


CHEMICAL REACTION AND COMBUSTION Chapter 8
Coal Combustion Modelling

should be set to a high value to ensure a converged solution can be reached


(see Note 6 below).
Step 4
Go to Chemical Reactions sub-folder and open the “Emission” panel. Turn on the
appropriate NOx models, i.e. thermal, fuel and prompt NOx. Then go to the
“Scheme Association” panel and click Apply to assign the chemical reaction
schemes defined above to the current fluid domain.
Step 5
Go to the “Initialisation” panel (Additional Scalars sub-folder) and set up an
appropriate initial mass fraction for the carrier fluid (e.g. for air YO2 = 0.233, YN2 =
0.767).
Step 6
Go the “Lagrangian Multi-Phase” folder, check the settings for the Lagrangian
two-phase modelling scheme and make any changes/additions necessary for
defining coal particle initial positions, entrance behaviour and physical properties
(panel “Droplet Physical Models and Properties”):
• Turbulent dispersion can be turned on in the Global Physical Models tab to
predict a realistic particle track behaviour.
• In the Droplet Properties tab, ensure that all values of Hfg in the Component
Properties list are set to 0.
Step 7
Switch off the heat and mass transfer time scale calculation by going to the
“Switches and Real Constants” panel (Other Controls sub-folder) and setting
constants C71 and C72 to 1.0.
Step 8
Go to the “Thermal Models” panel and check that options Static Enthalpy and
Chemico-Thermal have been selected for the enthalpy equation.
Step 9
Go to the “Scalar Boundaries” panel (Define Boundary Conditions folder) and
adjust the scalar mass fractions at the inlet boundaries.
Step 10
Go to the “Analysis (Re)Start” panel (Analysis Preparation/Running folder) and set
up the analysis as a restart run, beginning from the solution obtained in Stage 1.
Step 11
Run the case until the solution converges or reasonably small residuals are
achieved.
Useful notes
1. If the coal model is turned off to run the case in non-reacting mode, it may be
necessary to first turn off /delete the chemical scheme definition already
set-up in the “Scheme Definition” panel.
2. To avoid solution instabilities and reduce computer time in radiation cases, it
is advisable to run the simulation for several hundred iterations with radiation
turned off before switching it back on to complete the simulation.

8-44 Version 4.02


Chapter 8 CHEMICAL REACTION AND COMBUSTION
Coal Combustion Modelling

3. In the “Coal Composition” panel (Miscellaneous tab) you have three options
for setting the volatiles specific heat:
(a) Coal CV (the recommended option) uses the enthalpy of volatiles
calculated from the heat balance of the coal combustion for both char and
volatiles.
(b) CH4 assumes that all volatiles are methane.
(c) Mass Weighted considers the mixture of volatile components and
calculates the overall volatile enthalpy as the sum of the products of mass
fraction and enthalpy for each individual volatile component, i.e.
H = ΣY i h i
4. If you change any parameter in the “Coal Composition” panel, the chemical
reaction scheme is changed and therefore the gas combustion scheme in the
“Sub Models” panel must be reset. When you re-apply the gas composition
scheme, this also resets the scalars involved. Therefore, you must
(a) re-apply the scheme association,
(b) initialise the additional scalars for the background fluid,
(c) reset Hfg = 0 for all components in the Lagrangian “Droplet Physical
Models and Properties” panel, and
(d) define scalar boundary conditions for the inlet.

5. Accurate modelling can be achieved through input of appropriate values for


devolatilisation and char rates. Manipulating these values can increase the
solution accuracy, while changes in the turbulence models employed can lead
to more accurate prediction of the flow aerodynamics.
6. The iteration number at which to begin source term averaging should be set to
a high value so as to ensure that a stable initial flow field has been achieved
(and also to economize on computer time expended). This number has a
default setting of 50 iterations and should be altered to a value suitable for
establishing a stable flow field. This may be done by setting Constant 24 to
the desired value.
7. When discretising the coal particle size distribution, it is important to include
some sub-5 micron particles. This enables a stable flame to be established in
the immediate vicinity of the burner inlets.
8. When starting the coal combustion calculations in Stage 2, it is important to
use the constant rate devolatilisation option for all particles, and to make the
devolatilisation temperature equal to the particles’ initial temperature. This is
the numerical equivalent of ‘lighting up’ the combustion system in a real-life
situation.
Switches and constants for coal modelling
It is sometimes advisable to define some of the following pro-STAR Constants and
Switches when setting up a coal model:
• Constant 24 — see Note 6 above
• Constant 64 = 2 — constrain active scalar values to the range 0.0 — 1.0
• Constant 71 = 1 — deactivates the mass transfer time scale
• Constant 72 = 1 — deactivates the heat transfer time scale
• Constant 82 — specifies the coal particle emissivity in radiation problems
Version 4.02 8-45
CHEMICAL REACTION AND COMBUSTION Chapter 8
Coal Combustion Modelling

• Constant 88 — specifies the maximum coal particle temperature


• Constant 89 — specifies the minimum carrier fluid temperature limit
• Constant 90 — specifies the maximum carrier fluid temperature limit
• Constant 120 — specifies special options for the char reaction, as follows:
(a) Constant 120 = 0 — use the default V3.26 settings
(b) Constant 120 = 1 — the char reaction product is CO
(c) Constant 120 = 2 — the char reaction products are CO and CO2, where

Y CO
- = A exp ⎛ --------- ⎞
–T *
----------- (8-4)
Y CO2 ⎝ T ⎠

and the default values are A = 3.0×108 and T* = 30,193


• Constant 70 — used for changing the value of A in conjunction with Constant
120
• Constant 74 — used for changing the value of T* in conjunction with
Constant 120
• Switch 71 — specifies an implicit calculation of the source terms in the
particle energy equation (can improve algorithm stability)
Special settings for the Mixed-is-Burnt and Eddy Break-Up models
When Constant 120 is used, the following settings are also required depending on
the combustion model that has been chosen:

For Mixed-is-Burnt:
• Constant 120 = 0 — no extra scalars need to be defined; the product of the
char reaction is CO2
• Constant 120 = 1 — two extra scalars are needed, to be defined in the
Additional Scalars > Molecular Properties panel:
(a) MIX_CO — this is a passive scalar representing the CO mixture fraction
and is to be solved by a transport equation. The latter is specified in panel
Analysis Controls > Solution Controls > Equation Behavior > Additional
Scalars panel by choosing option Transport from the Solution Method
menu.
(b) CO — this is an active scalar to be solved algebraically. This is specified
in panel Analysis Controls > Solution Controls > Equation Behavior >
Additional Scalars panel by choosing option Internal from the Solution
Method menu.
• Constant 120 = 2 — three extra scalars are needed, to be defined in the
Additional Scalars > Molecular Properties panel:
(a) MIX_CO — this is a passive scalar representing the CO mixture fraction
and is to be solved by a transport equation. The latter is specified in panel
Analysis Controls > Solution Controls > Equation Behavior > Additional
Scalars panel by choosing option Transport from the Solution Method
menu.
(b) MIX_CO2 — this is a passive scalar representing the CO2 mixture
fraction and is to be solved by a transport equation. The latter is specified
8-46 Version 4.02
Chapter 8 CHEMICAL REACTION AND COMBUSTION
Coal Combustion Modelling

in panel Analysis Controls > Solution Controls > Equation Behavior >
Additional Scalars panel by choosing option Transport from the
Solution Method menu.
(c) CO — this is an active scalar to be solved algebraically. This is specified
in panel Analysis Controls > Solution Controls > Equation Behavior >
Additional Scalars panel by choosing option Internal from the Solution
Method menu.

For Eddy Break-Up:

One-step model — Constant 120 cannot be used because the char reaction product
can only be CO2
Two-step model
• Constant 120 = 0 — no extra scalars are needed; the char reaction product is
CO
• Constant 120 = 2 — no extra scalars are needed; the char reaction products
are CO and CO2

Version 4.02 8-47


Chapter 9 LAGRANGIAN MULTI-PHASE FLOW
Setting Up Lagrangian Multi-Phase Models

Chapter 9 LAGRANGIAN MULTI-PHASE FLOW


The theory behind Lagrangian multi-phase problems and the manner of
implementing it in STAR-CD is given in the Methodology volume, Chapter 12. The
present chapter contains an outline of the process to be followed when setting up a
Lagrangian multi-phase simulation, including details of the user input required and
important points to bear in mind when setting up problems of this kind.

Setting Up Lagrangian Multi-Phase Models


Step 1
Go to panel Select Analysis Features in STAR GUIde and choose option
Lagrangian Multi-Phase from the Multi-Phase Treatment menu. Click Apply.
The Lagrangian Multi-Phase folder will appear in the NavCenter tree, containing a
number of panels that are appropriate to this type of analysis.
Step 2
In the first panel, “Droplet Controls”, set various solution control parameters (see
the on-line Help text for more details).
The same panel also defines how droplet parcel initial conditions (entrance
behaviour and location) are to be specified. The available options are:
• Spray injection with atomization — use one of the built-in nozzle and
atomisation models (see Chapter 12, “Nozzle flow models” and “Atomisation
models” in the Methodology volume). These are especially useful in internal
combustion engine studies.
• Explicitly defined parcel injection — explicit (‘manual’) setting of all
required quantities. This option also allows the use of distribution functions
for the droplet diameters.
• User Subroutine — specify everything via a user subroutine
Step 3
The second panel, “Droplet Physical Models and Properties”, defines dispersed
phase heat, mass and momentum transport mechanisms (including inter-droplet and
wall collisions), plus droplet physical properties. Several different droplet types
may coexist in your model, so properties are specified for each individual type.
Step 4
The folder’s remaining panels relate to splitting droplets into parcels for modelling
purposes and defining the latter’s entrance behaviour (initial velocities and entrance
properties). How this is done depends on the option chosen in Step 2; the folder will
display the appropriate panel for each choice:
1. Spray injection with atomization — opens a single panel, “Spray Injection
with Atomization”, in which you specify the fuel mass flow rate entering the
solution domain through an injection nozzle. The liquid fuel is converted into
droplets whose injection velocity depends on the nozzle model
characteristics. In addition, a number of atomisation models are employed to
determine the distribution of droplet diameters and velocity directions on exit
from the nozzle.
2. Explicitly defined parcel injection — opens the following two panels:

Version 4.02 9-1


LAGRANGIAN MULTI-PHASE FLOW Chapter 9
Setting Up Lagrangian Multi-Phase Models

(a) “Injection Definition” sets up parcel entrance conditions, in terms of


either velocity and rotation components or nozzle parameters
(b) “Injection Points” defines parcel entrance locations

The association between conditions and locations is made by first dividing parcels
into injection groups that share the same entrance conditions. All entrance locations
defined subsequently are then assigned to one of these groups. The concept is
illustrated by the example shown in Figure 9-1 below:

Droplet Type 1 Droplet Type 2


Momentum ON Momentum ON
Heat transfer ON Heat transfer OFF
Properties of Heptane Properties of Water

Injection Group 1 Injection Group 2 Injection Group 3

Injection Definition Injection Definition Injection Definition


Constant Diam. Rosin-Ram PDF Normal PDF
Wi = –5 m/s Vi = 2 m/s Wi = 7 m/s
mT = 0.05 kg/s mT = 0.02 kg/s mT = 0.05 kg/s
2 parcels/point 3 parcels/point 1 parcel/point

Injection Points Injection Points Injection Points


Set 1: Line, 3 pts Set 1: Single point Set 1: Boundary,
Set 2: Circle, 6 pts Set 2: Line, 8 pts 12 pts

Set 2, 6pts

Set 1, 12 pts
Set 2, 8pts
Set 1, 3pts

Single Parcel
Set 1, 1pt Injection Point

Figure 9-1 Illustration of terminology for explicitly defined parcel injection

3. User Subroutine — opens a single panel, “Droplet User Subroutine”, that


calculates all parcel initial conditions through user coding

9-2 Version 4.02


Chapter 9 LAGRANGIAN MULTI-PHASE FLOW
Setting Up Lagrangian Multi-Phase Models

Note that the Spray injection and Explicitly defined options are mutually
exclusive. Thus, if you change your mind about which method to use for specifying
initial conditions, you will need to go back to panel “Droplet Controls”, pick the
other method and overwrite the previous definitions. On the other hand, User
Subroutine may be used in conjunction with either of the above options, i.e. STAR
will take the definitions supplied in subroutine DROICO into account as well as the
spray or explicit definitions.
Step 5
Check the result of the parcel initialisation process graphically by displaying the
parcels in the context of a plot of the domain into which they are launched, as
illustrated in Figure 9-2:

Figure 9-2 Plot of droplet initial conditions

This is done by going to the Post-Processing folder, panel “Plot Droplets/Particle


Tracks” and using the plotting facilities of the “Droplets” tab, as explained in the
on-line Help text. Alternatively, choose Post > Get Droplet Data from the main
window menu bar to display the Load Droplet Data dialog shown below and
perform the same function from there.

Version 4.02 9-3


LAGRANGIAN MULTI-PHASE FLOW Chapter 9
Data Post-Processing

Data Post-Processing
pro-STAR provides special facilities for visualising the results of a Lagrangian
multi-phase flow analysis. These facilities fall into the following two categories:
1. Static displays — these show the location of one or more droplets at a given
point in time. Alternatively, they may also be used to show successive
positions of a given droplet as it progresses through the solution domain. The
droplets are represented by small circles, as shown in Figure 9-3. The circle
size and colour can be made to depend on a variety of local droplet properties.

Figure 9-3 Static display illustration

2. Trajectory displays — these show droplet tracks, either as continuous


trajectories or as animated streaks, whose rate of progress through the
solution domain can be controlled by the user, as illustrated in Figure 9-4.

9-4 Version 4.02


Chapter 9 LAGRANGIAN MULTI-PHASE FLOW
Data Post-Processing

Figure 9-4 Trajectory display illustration

Static displays
Steady-state problems
Step 1
Read the required droplet data from the track (.trk) file generated automatically
by the STAR-CD solver for Lagrangian flow problems. To do this, use panel “Plot
Droplets/Particle Tracks”, tab “Droplets”.
Step 2
If necessary, use command DTIME to specify a time range over which you want
droplet track data to be plotted. The display will then include only locations visited
by droplets during this time interval.
Step 3
Use the “Droplets” tab controls to choose options appropriate to the plot you want
to create. Note that a droplet display may be superimposed on a post data plot by
choosing Plot > Cell Display > Droplets from the main window menu (or by
issuing command CDISPLAY, ON, DROPLET) before the cell plotting operation. If
the plot is a contour plot and the droplet fill colour varies according to a physical
property, a secondary scale will be displayed for that droplet property. If the
droplets are filled with a single arbitrary colour, and droplet velocity vectors are
displayed, the secondary scale will correspond to droplet velocity magnitude, as
represented by the vector colours.

Version 4.02 9-5


LAGRANGIAN MULTI-PHASE FLOW Chapter 9
Data Post-Processing

Step 4
Select a set of parcels whose progress through the solution domain is to be
displayed. The selection procedure is analogous to that described in Chapter 2 of the
Meshing User Guide regarding sets of cells, vertices, splines, etc. Thus, sets may be
selected by
• a coloured button marked D -> on the left-hand-side of the main window
• a similar button labelled Dp in the “Droplets” tab
• typing command DSET in the I/O window. This provides the most extensive
range of selection options.

The set selection facilities available via the D -> or Dp buttons are as follows:
1. All — puts all parcels in the set
2. None — clears the current set
3. Invert — selects all unselected parcels and clears the current set
4. New — replaces the current set with a new set of parcels
5. Add — adds new parcels to the current set
6. Unselect — deletes parcels from the current set
7. Subset — selects a smaller group of parcels from those in the current set

For the last four items, the target parcels may be assembled by choosing an option
from a secondary drop-down list, as described below. In every case, what
constitutes a valid option depends on how droplet data were read into pro-STAR:
1. For all loading choices, option Cell Set selects parcels that are contained
within the physical space occupied by the current cell set. If the choice was
Track File (see Step 1 on page 9-5), all droplet tracks whose initial positions
fall within the current cell set are selected.
2. If the loading choice was Droplet Initial Conditions (see Step 5 on page 9-3)
or Current Post Data File (see Step 2 on page 9-8), the following options are
available:
(a) Cursor Select — click on the desired parcels with the cursor; complete
the selection by clicking the Done button on the plot
(b) Zone — use the cursor to draw a polygon around the desired parcels.
Complete the polygon by clicking the last corner with the right mouse
button (or click Done outside the display area to let pro-STAR do it for
you). Abort the selection by clicking the Abort button.

3. If the loading choice was Current Post Data File, the following options are
available:
(a) Active — select all active parcels
(b) Stuck — select all parcels that have stuck to a wall and become
immobilised

Note that droplet set information is not saved in the restart (.mdl) file on
completion of the post-processing run
Step 5
Display the selected parcels as a series of droplet circles by clicking Droplet Plot
in the “Droplets” tab
9-6 Version 4.02
Chapter 9 LAGRANGIAN MULTI-PHASE FLOW
Data Post-Processing

The locations of the circles represent the points where a parcel intersects cell
boundaries as it travels from the beginning to the end of its path through the mesh.
Step 6
If detailed numerical information is required on the selected parcels, choose Lists >
Tracks from the main window menu bar to open the Particle/Droplet Track Data
dialog. Select the track file and then click Load Data to read in and display all
available information in that file, as shown below:

The required information is displayed by clicking the appropriate parcel number


(shown in the Track column) with the mouse. The same information (but in a
different format) can also be displayed on the I/O window by typing command
PTPRINT.
Special data requirements
In some situations, the user may require the following additional information:
1. The position of a range of parcels at a given point in time, as opposed to a
specified parcel at a series of time points. The data needed for such a display
may be obtained by interpolation of the available data at the time point in
question using command PTREAD. Continue by specifying the appropriate
parcel set and then use the “Droplets” tab in STAR GUIde to display the
required droplet distributions. Note that the time specified in PTREAD is
independent of any time information specified via command DTIME (see Step
2 above)
2. The ‘age’ of all currently-loaded parcels, given by command DAGE. A
parcel’s age is defined as the interval between the time when the first parcel
entered the solution domain and the time when the parcel in question hits a
wall or exits from the solution domain. Age is calculated from data in the

Version 4.02 9-7


LAGRANGIAN MULTI-PHASE FLOW Chapter 9
Data Post-Processing

track (.trk) file and may be used as the basis for selecting a parcel set, via
command DSET. This information may be listed in the I/O window using
command DLIST.

Transient problems
Step 1
Decide which time step is to be inspected and then load the corresponding data
(from file case.pstt), using STAR GUIde’s “Load Data” panel (“File(s) tab”).
If more than one transient file is available, pro-STAR will locate the right one
automatically.
Step 2
Open panel “Plot Droplets/Particle Tracks” (“Droplets” tab) and read the contents
of the transient file by selecting option Current Post Data File from the pop-up
menu at the top.
Step 3
Choose appropriate options in the Droplet Plot Options section of the same tab, as
for “Steady-state problems”.
Step 4
Select the desired parcel set using the most appropriate of the methods described
under “Steady-state problems”.
Step 5
Plot droplets by clicking the Droplet Plot button.
Step 6
Information about a range of parcels at the current time step can also be displayed
in the I/O window using command DLIST. For example,

DLIS,1,50,2,OTHER

will list the density, diameter, mass, droplet count and temperature of every second
parcel between 1 and 50. Information on parcel ‘age’ is also obtainable with this
command (having first executed command DAGE). In transient problems, age is
defined as the interval between the time when the first parcel entered the solution
domain and the current time.

Trajectory displays
Trajectory displays are basically droplet track displays. These are plotted as
continuous trajectories or animated streaks, using the options provided in panel
“Plot Droplets/Particle Tracks” (“Droplets” tab). As for particle tracks generated at
the post-processing phase, the data required for such plots are stored in file
case.trk. This file is generated automatically during the Lagrangian multi-phase
analysis for both transient and steady-state calculations.
Note that:
• It is also possible to print position, velocity and other droplet data stored in
case.trk for each track using command PTPRINT.
• The data in this file will be overwritten if the user generates post-processing

9-8 Version 4.02


Chapter 9 LAGRANGIAN MULTI-PHASE FLOW
Engine Combustion Data Files

particle tracks without first saving the droplet data.

Engine Combustion Data Files


In addition to the normal results files, engine combustion cases also produce
additional output data (.spd) files, written by STAR if the Lagrangian multi-phase
and/or combustion simulations options are in use. One such file is produced for
every fluid domain in your model and contains both fuel droplet data (represented
as globally averaged quantities) and general engine data.
The information in this file may also be displayed in graphical form using the
utilities provided in STAR GUIde’s Graphs folder (see panel “External Data”). The
meaning of the quantities appearing in the file is as follows:

Name Meaning

T-Step Time step number


Time Elapsed time at this time step [s]
Crank_Ang. Crank Angle [degrees]
Average_P Cylinder absolute average pressure [pa]
Average_T Cylinder absolute average temperature [K]
Average_d Cylinder average density [kg/m3]
Cylinder_Mass In-cylinder mass [kg]
Tot_Inj_Lqui Total injected mass [kg]
Cur_mas_Fue Total mass of liquid phase [kg]
Evaporated Total evaporated mass [kg]
Evaprt_% Ratio of total evaporated mass to the total injected mass [%]
Leading_par Unused
Distance Unused
Velocity Unused
V_mag Unused
Idr Unused
Sauter_D Sauter mean diameter [m]
Fluid angular momentum w.r.t. the X-axis of the local coordi-
AngMom_X
nate system used in the model [kg/m2s]
AngMom_Y Fluid angular momentum w.r.t. the Y-axis
AngMom_Z Fluid angular momentum w.r.t. the Z-axis
Mass_Burnt Burnt fuel mass [kg]

Version 4.02 9-9


LAGRANGIAN MULTI-PHASE FLOW Chapter 9
Useful Points

%Evap_Burnt Burnt fuel as a percentage of fuel evaporated


Heat_Release_R
Heat release rate [J/s]
ate
Scalar Mass of scalar no. i [kg]

Note that, depending on the model, some of the above data may have no meaning.

Useful Points
1. The above treatment is strictly valid only for droplets whose physical
dimensions are appreciably smaller than those of a typical mesh cell through
which they travel. It is recommended that the total droplet volume
(i.e. volume of a typical droplet times the number of droplets in the parcel)
should not exceed 40% of this cell volume.
2. If a convergent solution cannot be easily obtained in steady-state models
using the coupled approach, it may be beneficial to start the analysis by
obtaining a solution that does not include the dispersed phase. The latter
should then be introduced into the calculated flow field and the analysis
restarted using the Initial Field Restart option to produce the final, complete
solution.
3. In steady-state models using the uncoupled approach, the computer time
required may again be reduced by obtaining the solution in two stages. First, a
converged solution without the dispersed phase should be calculated. The
dispersed phase should then be introduced and the analysis restarted using the
Initial Field Restart option to obtain the desired solution in one iteration only.
4. In transient analyses involving droplets that move faster than their
surrounding fluid, the Courant number used for estimating a reasonable time
step size (see Chapter 5, “Load step definition”) should be based on the
droplet rather than the fluid velocity.
5. STAR-CD’s default treatment for heat transfer coefficients can be combined
with user-calculated mass transfer coefficients and vice-versa. In practice,
however, the user will most probably want to use the same calculation
procedure for both of them.
6. Complex or unusual physical conditions relating to momentum, heat and
mass transfer between droplets and the continuous phase can be
accommodated by supplying user subroutines DROMOM, DRHEAT and
DRMAST that describe each transfer process, respectively. Similarly, special
conditions relating to the momentum, heat and mass transfer behaviour of
droplets at wall boundaries can be specified by supplying the required
relationship via subroutine DROWBC.

9-10 Version 4.02


Chapter 10 EULERIAN MULTI-PHASE FLOW
Introduction

Chapter 10 EULERIAN MULTI-PHASE FLOW


Introduction
The theory behind problems of this kind is given in the Methodology volume,
Chapter 13. This chapter contains an outline of the process to be followed when
setting up an Eulerian multi-phase analysis. Also included are cross- references to
appropriate parts of the on-line Help system, containing details of the user input
required.

Setting up multi-phase models


Step 1
Switch on the Eulerian multi-phase simulation facility using the “Select Analysis
Features” panel in STAR GUIde:
• Select Eulerian Multi-Phase from the Multi-Phase Treatment menu
• Click Apply. pro-STAR checks if another multi-phase simulation option
(Lagrangian, Free Surface, Cavitation) is already on. If so, it issues a warning
message and turns it off.
• An additional sub-folder called Eulerian Multi-Phase now appears in the
NavCenter tree, within the Thermophysical Models and Properties folder.
Step 2
Set up the mesh and define the boundary region locations as usual. At present, only
part of the full STAR-CD boundary type set is available for this kind of analysis.
The permissible options are:
1. Inlet
2. Outlet
3. Pressure
4. Wall
5. Non-porous baffle
6. Cyclic
7. Symmetry
8. Degassing
9. Attachment
10. Monitoring

Note that:
• The above list contains an additional boundary type, ‘Degassing’, valid only
for Eulerian multi-phase flows. This permits dispersed phase mass to escape
into the media surrounding the solution domain (see also Chapter 4,
“Phase-Escape (Degassing) Boundaries” in this volume). Your problem
should not contain more that one boundary of this type.
• Only the currently available boundary types, as listed above, can be set up via
the “Create Boundaries” panel.
Step 3
Open the Thermophysical Models and Properties folder and use each of its
sub-folders to provide relevant information about your problem. Note that:

Version 4.02 10-1


EULERIAN MULTI-PHASE FLOW Chapter 10
Setting up multi-phase models

• Thermal/solar radiation is not supported in this version of the code.


• Use the Liquids and Gases panels to specify physical properties and special
flow conditions in your model. Note that:
(a) Only a single domain (or material) is allowed at present, so the Material #
slider in each panel remains set to 1.
(b) Where appropriate, data are entered per phase, with the number of phases
currently restricted to two. Of these, no. 1 is treated as the continuous and
no. 2 as the dispersed phase.
(c) “Molecular Properties” — compared to single-phase problems, only a
restricted range of options is available for evaluating physical properties.
The specification process and permissible options are common to both
phases.
(d) “Turbulence Models” — if turbulent flow conditions prevail, specify a
method for calculating the turbulence characteristics of both phases and
also the turbulence-induced drag
(e) “Thermal Models” — if heat transfer is present in the analysis, turn on the
temperature solver for each phase as required
(f) “Initialisation” — specify initial conditions for each phase
(g) “Monitoring and Reference Data” — supply a reference pressure and
temperature and the cell location corresponding to the reference pressure.
The values specified apply to both phases.
(h) “Buoyancy” — if buoyancy effects are important, specify a datum
location and reference density. Again, these values apply to both phases.

• The current version does not support the following features:


(a) Multi-component mixture problems requiring the presence of additional
scalar variables in either phase. Therefore, STAR GUIde does not display
the Additional Scalars sub-folder.
(b) Porous media flow, therefore the Porosity sub-folder is not displayed.
(c) Chemical reactions of any kind, including coal combustion and the
STAR/KINetics package. Therefore, the “Select Analysis Features” panel
does not permit the above options to be turned on.
(d) Liquid films of any kind. Again, the “Select Analysis Features” panel
does not allow this option.
Step 4
In the Eulerian Multi-Phase folder:
• Open the Interphase Momentum Transfer sub-folder to specify appropriate
models and related parameters for this part of the analysis. The information is
supplied in two separate panels:
(a) “Drag Forces” — define a model for calculating drag forces directly or
via the drag coefficient
(b) “Other Forces” — define models for calculating other interphase forces
(e.g. virtual mass and/or lift force)

• If heat transfer is present in the analysis, use the “Interphase Heat Transfer”
panel to specify the method of calculating the Nusselt number (and hence the
heat transfer coefficient).
10-2 Version 4.02
Chapter 10 EULERIAN MULTI-PHASE FLOW
Setting up multi-phase models

• Specify the size of the particles making up the dispersed phase using the
“Particle Size” panel. At present, all particles are assumed to be of equal size.
Step 5
If required by problem conditions, use the “Source Terms” panel in folder Sources
to specify mass sources or additional source terms for the momentum, turbulence or
enthalpy equations of either phase. At present, multi-phase sources may only be
specified via user subroutines.
Step 6
Specify boundary conditions using the “Define Boundary Regions” panel. The
permissible range of boundary types is shown in Step 2. Note that for inlet, pressure,
wall/baffle and cyclic regions, separate boundary conditions are needed for each
phase.
When pro-STAR’s boundary display facilities are used to check the various
boundary region definitions (see Chapter 4, “Boundary Visualisation”), inlet phase
velocities will be displayed according to the setting of the Phase # slider in panel
“Define Boundary Regions”.
Step 7
In the Analysis Controls folder:
• Select Solution Controls > Equation Behavior, open the “Primary
Variables” panel and make any necessary adjustments to the current settings
• If you wish to monitor the value of any flow variable(s), as a function of
iteration or time step, select Output Controls > Monitor Engineering
Behavior and then open panel “Monitor Boundary Behaviour” and/or panel
“Monitor Cell Behaviour”. The choice depends on whether you wish to
monitor values at a boundary region or within a cell set. Note that the choice
of which variables to monitor is phase-dependent.
• If you are running a transient problem, use the “Transient tab” in the
“Analysis Output” panel to select which variables you wish to store in the
transient post data file (.pstt). Note that the choice of such variables is
phase-dependent.
Step 8
Run STAR in double precision mode. There are two reasons for this:
• Solving the volume fraction equation in this manner gives rise to a smaller
truncation error, especially in parts of the mesh where the volume fraction is
close to 1 or 0. This is sometimes essential for convergence of the solution.
• Double precision cases have been more extensively tested

Step 9
Post-processing the analysis results follows the same rules as single-phase
problems. Note that:
• Analysis data are stored in the .ccm file per phase. A phase slider in the
“Data tab” of panel “Load Data” enables you to select the precise data
required.
• Likewise, phase-specific data may be plotted in a graph. The types of graph
available are described in topics “Residual / Monitored History Data”,

Version 4.02 10-3


EULERIAN MULTI-PHASE FLOW Chapter 10
Setting up multi-phase models

“Engineering Data” and “Analysis History Data”.


Useful points on Eulerian multi-phase flow
1. The momentum under-relaxation factors should be the same for both
continuous and dispersed phases. The pressure under-relaxation factor should
also be equal to the volume fraction factor. Suggested values for these
parameters are 0.3 on momentum for both phases and 0.1 on pressure and
volume fraction.
2. To ensure satisfactory convergence for steady and pseudo-transient cases, a
maximum residual error tolerance of 1.0 × 10-6 is recommended.

10-4 Version 4.02


Chapter 11 FREE SURFACE AND CAVITATION
Free Surface Flows

Chapter 11 FREE SURFACE AND CAVITATION


Free Surface Flows
The theoretical description of free-surface flow models is given in Chapter 14 of the
Methodology volume. This section contains an outline of the procedure to be
followed when setting up free-surface flow problems. Also included are
cross-references to appropriate parts of the on-line Help system, which contains
details of the user input required.
Setting up free surface cases
Step 1: Define the mesh and boundary regions
Set up an appropriate mesh and define its boundary regions in the usual way. All
standard STAR mesh features are applicable to free-surface flows but this is not also
the case for all types of boundary region. The boundary types currently supported
in free-surface problems are:
• Inlet
• Outlet
• Slip and no-slip impermeable walls
• Symmetry planes
• Static and piezometric pressure (with both environmental and mean options
deactivated)
• Baffles
• Cyclic boundaries (except for partial cyclics in which the mass flow rate is
specified)
• Attachment boundaries
• Monitoring boundaries

Step 2: Activate the free surface model


Turn on the free-surface option using the “Select Analysis Features” panel of the
STAR GUIde system:
• Select On from the Free Surface menu
• Select option Transient from Time Domain menu. Free-surface flows have to
be computed in a time-marching manner, even if the final solution is steady. In
the latter case, one can choose larger time steps or only one iteration per time
step to save on computing time, as described in Step 7 below.
• Click Apply. An additional folder called Free Surface will now appear in the
NavCenter tree.

Note that:
1. A passive scalar named VOF is required for free-surface problems. pro-STAR
will automatically define such a scalar (if it has not been defined already) on
pressing the Apply button in the Select Analysis Features folder. This scalar
stores the volume fraction of the ‘heavy’ fluid in the solution domain (see
“Mathematical model” on page 14-2 of the Methodology volume) and
requires definition by the user of appropriate boundary conditions.
2. Certain combinations of the free surface model with other STAR-CD features

Version 4.02 11-1


FREE SURFACE AND CAVITATION Chapter 11
Free Surface Flows

are not currently supported. Such features are:


(a) Eulerian and Lagrangian multi-phase flow
(b) Reacting flow
(c) Radiative heat transfer
(d) Aeroacoustic analysis

pro-STAR will issue a warning message if an attempt is made to switch on


any of the above features and the free surface model will then be switched off.

Step 3: Define model control parameters


In the Free Surface folder, open the “Controls” panel and specify appropriate
settings for the following parameters:
1. Differencing Scheme — this defines the differencing scheme to be used for the
solution of the VOF transport equation:
(a) The default scheme is HRIC, which stands for ‘High-Resolution Interface
Capturing’. As suggested by the name, this scheme should be employed if
a sharp interface between the heavy and light fluids is to be resolved.
There is also a blending factor associated with the scheme. The default
value for this factor is appropriate for most situations; higher values
provide a sharper interface but there is a danger of interface alignment
with grid lines under unfavourable flow direction. In such a case, the
blending factor should be reduced.
(b) The Upwind scheme will not provide a sharp interface but it may be used
on coarse or poor-quality meshes or when a sharp resolution of the
interface is not an issue.

2. Surface Tension — determines whether the surface tension effect across the
heavy-light fluid interface is to be included in the calculations. The effect is
excluded by default as it plays an important role only in small-scale problems,
where the mesh is fine enough to resolve the interface curvature on a scale
that results in an appreciable pressure difference. The latter is proportional to
σ/R, where σ is the surface tension coefficient and R the radius of interface
curvature. Note that the HRIC scheme must be selected if you choose to
include surface tension in your model.

Step 4: Define material properties


The fluid medium in a free-surface problem is defined as a single fluid material
possessing two components: a ‘heavy’ and a ‘light’ one. To define their respective
material properties, go to the Free Surface folder and open the “Molecular
Properties” panel. For each of the heavy and light components, fill in the relevant
property values or select materials from pro-STAR’s built-in property database and
then click Apply.

Step 5: Define thermophysical models


In the Thermophysical Models and Properties folder:
1. If your application involves solid-fluid heat transfer:

11-2 Version 4.02


Chapter 11 FREE SURFACE AND CAVITATION
Free Surface Flows

(a) Open the “Thermal Options” panel and then select Heat Transfer On in
the Solid-Fluid Heat Transfer section
(b) Click Apply
(c) As a result of the above, a sub-folder called Solids will appear in the
STAR GUIde tree. Use this sub-folder to define solid material properties.
More than one solid domains may be defined in such models.

Note that radiative heat transfer is not currently supported in free-surface


problems.
2. If gravitational effects are important in your application, open the “Gravity”
panel and define the gravitational acceleration and its direction with respect to
the coordinate system of the solution domain.
3. The overall flow conditions should be specified by entering the Liquid and
Gases sub-folder. This is designed to supply relevant information for the
following aspects of the model:
(a) Open the “Turbulence Models” panel and choose an appropriate
turbulence model for your case, or select the Laminar flow option if
applicable.
(b) If thermal effects are important, open the “Thermal Models” panel and
select the Temperature Calculation On option. In the Show Options
section, choose the enthalpy formulation and transport equation to be
solved for it. Please note that the following are not supported in
free-surface cases:
i) Stagnation Enthalpy option in the Conservation menu
ii) Chemico-Thermal option in the Enthalpy menu

(c) Initialise the flow field and turbulence quantities in the “Initialisation”
panel. Only the Constant and User options are supported for free-surface
cases.
(d) Use the “Monitoring and Reference Data” panel to specify the locations
of the monitoring and reference cells, as well as the reference pressure
and temperature.
(e) Use the “Buoyancy” panel to specify whether buoyancy effects are to be
included in the calculation. Select the On button if this effect is
important. It is advisable to use a Datum Density value corresponding to
the ‘light’ fluid density and, if possible, to choose the Datum Location at
a cell that is likely to be always occupied by the ‘light’ fluid.

4. Define ‘active’ and ‘passive’ scalars in the Additional Scalars sub-folder. A


passive scalar named VOF should already be defined at this stage. Note that:
(a) When the Cavitation option is turned On (see “Cavitating Flows” on
page 11-5), an active scalar named CAV also needs to be defined
(b) Apart from CAV, no other active scalar can be defined for free-surface
problems
(c) You may define as many ‘passive’ scalars as are necessary for your
model
(d) The diffusion term in the transport equation for all scalars defined in a
free-surface model will be switched off

Version 4.02 11-3


FREE SURFACE AND CAVITATION Chapter 11
Free Surface Flows

5. In the Additional Scalars sub-folder, initialize the distribution of the VOF


scalar by opening the “Initialisation” panel and choosing one of the available
options. In most cases, either Constant values, User coding, or assignment
according to cell type is applied. In the latter case, click CTAB on the main
pro-STAR window to open the Cell Table Editor and use it to set the Initial
Free Surface Material switch to Light (for those cell types initially occupied
by the ‘light’ fluid) or to Heavy (for cell types initially occupied by the
‘heavy’ fluid).
6. For free-surface problems involving porous media, use the Porosity sub-folder
to define properties for the porous materials in the normal way.
7. Use the Sources sub-folder to define external source terms for momentum,
turbulence, enthalpy and scalars. Note that:
(a) User Coding is the only supported option in this case
(b) The VOF transport equation does not accept additional source terms

Step 6: Define boundary conditions


Go to the Define Boundary Conditions folder
• Open the “Define Boundary Regions” panel and specify appropriate boundary
conditions in the usual manner
• Open the “Scalar Boundaries” panel and specify boundary conditions for the
VOF scalar

Valid boundary types for free-surface problems are listed under Step 1. If a pressure
boundary condition is specified, make sure that both the Envir Press (environmental
pressure) and Mean (pressure profile mean value) options are set to Off.

Step 7: Define analysis control parameters


Go to the Analysis Controls folder:
1. Select the Solution Controls sub-folder and then open the “Solution Method”
panel to set/adjust the solution algorithm parameters. Note that:
(a) Only the SIMPLE algorithm is applicable to free-surface problems
(b) Both CG and AMG solvers are applicable and the desired one may be
selected from the Solver Type menu. AMG is recommended since it
usually leads to shorter computing times.
(c) Select option Euler Implicit from the Temporal Discretisation menu.
This is the only option supported for this type of problem.
(d) If a steady-state solution is expected, one can limit the number of outer
iterations per time step (see topic “Transient problems” in the STAR
GUIde on-line Help) to 1, in which case a pseudo-transient marching
towards the steady state is obtained. This is applicable to both cavitating
and free-surface flows, but one needs to be certain that a steady-state
solution can be reached.

2. Select the Equation Behavior sub-folder and then open the “Primary
Variables” panel. Make any necessary adjustments to the current or default
settings in the Equation Status, Solver Parameters and Differencing Schemes

11-4 Version 4.02


Chapter 11 FREE SURFACE AND CAVITATION
Cavitating Flows

tabs.
3. In the “Additional Scalars” panel:
(a) Adjust the under-relaxation factor for the VOF scalar, if necessary (the
default value is 0.8).
(b) Ignore the section concerning the differencing scheme because the latter
has already been set in the free-surface “Controls” panel.
(c) Select an appropriate differencing schemes and other control parameters
for all scalars other than VOF.
(d) Click Apply before changing to another scalar or exiting from the panel.

4. Select the Output Controls sub-folder and use the “Monitor Numeric
Behaviour” panel to print additional information such as convergence
residuals and conservation checks (optional).
5. Select the Monitor Engineering Behavior sub-folder and, if you wish, use the
“Monitor Cell Behaviour” panel to save selected cell data for subsequent
plotting against iteration or time step number.
6. The Analysis Output sub-folder enables you to specify the frequency of
saving solution results in the .ccm file.
7. Use the “Switches and Real Constants” panel in the Other Controls sub-folder
to set switches and constants for any ‘non standard’ practices. Please check
carefully the meaning of each switch and constant and use it only when
absolutely necessary.

Step 8: Define the time step size and run duration


There are two ways to define the time step size and the run time length:
1. Go to the Analysis Preparation/Running folder and open the “Set Run Time
Controls” panel. Fill in appropriate values for simulation time and time step
size in the relevant boxes. This is the recommended way of defining time
steps.
2. If your application involves a moving mesh defined by an events file, you will
need to use the Advanced Transients panel by choosing Modules > Transient
from the main pro-STAR window. In this panel, you can define load steps,
each of which contains the time step size and number of time steps to be used
for each load step. Please note that you need to go through Step 1 to Step 7
before defining load steps using the Advanced Transients panel.

Cavitating Flows
The theoretical description of cavitation models is given in Chapter 14 of the
Methodology volume. This section contains an outline of the procedure to be
followed when setting up cavitating flow problems. Also included are
cross-reference to appropriate parts of the on-line Help system, which contains
details of the user input required.
Setting up cavitation cases
Step 1: Define the mesh and boundary regions
Set up an appropriate mesh and define its boundary regions in the usual way. All
standard STAR mesh features are applicable to cavitation but this is not also the

Version 4.02 11-5


FREE SURFACE AND CAVITATION Chapter 11
Cavitating Flows

case for all types of boundary region. The boundary types currently supported in
cavitation problems are:
• Inlet
• Slip and no-slip impermeable walls
• Symmetry planes
• Static and piezometric pressure (with both environmental and mean options
deactivated)
• Baffles
• Cyclic boundaries (except for partial cyclics in which the mass flow rate is
specified)
• Attachment boundaries
• Monitoring boundaries

Step 2: Activate the cavitation model


Turn on the cavitation option using the “Select Analysis Features” panel of the
STAR GUIde system:
• Select On from the Cavitation menu
• Select option Transient from Time Domain menu. Cavitating flows have to
be computed in a time-marching manner, even if the final solution is steady. In
the latter case, one can choose larger time steps or only one iteration per time
step to save on computing time, as described in Step 7 below.
• Click Apply. An additional folder called Cavitation will now appear in the
NavCenter tree.

Please note:
1. An active scalar named CAV is required for cavitation problems. pro-STAR
will automatically define such a scalar (if it has not been defined already) on
pressing the Apply button in the Select Analysis Features folder. This scalar
stores the volume fraction of vapour generated during the cavitation process
(see “Mathematical model” on page 14-6 of the Methodology volume) and
requires definition by the user of appropriate physical properties and
boundary conditions.
2. Certain combinations of the cavitation model with other STAR-CD features
are not currently supported. Such features are:
(a) Eulerian and Lagrangian multi-phase flow
(b) Reacting flow
(c) Radiative heat transfer
(d) Aeroacoustic analysis

pro-STAR will issue a warning message if an attempt is made to switch on


any of the above features and the cavitation model will be switched off.
3. Combinations of the cavitation and free surface models are supported. This
typically occurs in applications requiring resolution of a sharp interface
between a cavitating liquid and a gas, in which case you may select both the
Free Surface and Cavitation options. If both are selected, a passive scalar
called VOF is defined automatically by pro-STAR (unless this definition
already exists in the model).

11-6 Version 4.02


Chapter 11 FREE SURFACE AND CAVITATION
Cavitating Flows

4. When the cavitation and free surface options are combined, physical
properties for the gas component are defined using the “Molecular Properties”
panel in the Free Surface folder. You will also need to define the differencing
scheme for the VOF scalar, as discussed in Step 3 of the “Free Surface Flows”
section.

Step 3: Define material properties


The fluid medium in cavitating flows is defined as a single fluid material consisting
of two or three components. Thus, for cavitation without a free surface, there is a
heavy component and a vapour component; for cavitation with a free surface, there
is a heavy component, a light component and a vapour component.
• For cases involving cavitation only, go to the Cavitation folder and open the
“Molecular Properties” panel. Two tabs labelled Light Fluid and Heavy Fluid
will appear, of which the Light Fluid one is always inactive. Use the Heavy
Fluid tab to define molecular properties for the cavitating liquid. Please note
that displayed values for surface tension and contact angle will not be used for
cases involving only cavitation.
• When cavitation is combined with a free surface, the definition of molecular
properties for the heavy and light fluids is identical to that for free surface
flows, as described in the previous section.
• Vapour molecular properties are defined using the “Molecular Properties
(Scalar)” panel, as for any other active scalar.

Step 4: Define model parameters


Go to the Cavitation folder and open the “Cavitation Model” panel. The STAR-CD
default is currently the Rayleigh model but you may also define your own model by
choosing the User option from the Model Selection menu.
Three parameters are needed for the Rayleigh model: the Saturation Pressure,
Average Nuclear Radius and Number of Nuclei contained in 1 m3 of liquid. Of these,
the saturation pressure may be either constant or user-defined in subroutine
CAVPRO; the other two parameters are constants.
Please note that the number of nuclei per m3 of liquid has a strong influence on
the amount of vapour generated and therefore requires your own knowledge as to
its likely value. Although only limited measurement data are available, it is well
known that liquid purity (affected by filtering, degassing and possibly other
treatment) strongly affects the cavitation process. The following recommendations
can be used in the absence of more specific information:
• For small-scale, high-pressure systems such as engine injectors, a value in the
range 1011 — 1014 was found to be adequate.
• For large-scale, low-pressure systems such as ship propellers and large
pumps, smaller values in the range 106 — 1010 may be more appropriate.
However, the best choice is always the one based on your own experience.

Step 5: Define thermophysical models


In the Thermophysical Models and Properties folder:
1. If your application involves solid-fluid heat transfer:

Version 4.02 11-7


FREE SURFACE AND CAVITATION Chapter 11
Cavitating Flows

(a) Open the “Thermal Options” panel and then select Heat Transfer On in
the Solid-Fluid Heat Transfer section
(b) Click Apply
(c) As a result of the above, a sub-folder called Solids will appear in the
STAR GUIde tree. Use this sub-folder to define solid material properties.
More than one solid domains may be defined in such models.

Note that radiative heat transfer is not currently supported in cavitation


problems.
2. If gravitational effects are important in your application, open the “Gravity”
panel and define the gravitational acceleration and its direction with respect to
the coordinate system of the solution domain.
3. The overall flow conditions should be specified by entering the Liquid and
Gases sub-folder. This is designed to supply relevant information for the
following aspects of the model:
(a) Open the “Turbulence Models” panel and choose an appropriate
turbulence model for your case, or select the Laminar flow option if
applicable.
(b) If thermal effects are important, open the “Thermal Models” panel and
select the Temperature Calculation On option. In the Show Options
section, choose the enthalpy formulation and transport equation to be
solved for it. Please note that the following are not supported in cavitation
cases:
i) Stagnation Enthalpy option in the Conservation menu
ii) Chemico-Thermal option in the Enthalpy menu

(c) Initialise the flow field and turbulence quantities in the “Initialisation”
panel. Only the Constant and User options are supported for cavitation
cases.
(d) Use the “Monitoring and Reference Data” panel to specify the locations
of the monitoring and reference cells, as well as the reference pressure
and temperature.
(e) Use the “Buoyancy” panel to specify whether buoyancy effects are to be
included in the calculation. Select the On button if this effect is
important.
4. Define ‘active’ and ‘passive’ scalars in the Additional Scalars sub-folder. An
active scalar named CAV should already be present at this stage, defined
automatically by pro-STAR. Note that:
(a) If the Free Surface option is turned On as well (see “Free Surface
Flows” on page 11-1), a passive scalar named VOF is also defined
automatically by pro-STAR. This tracks the distribution of the
‘heavy’ fluid volume fraction.
(b) Apart from CAV, no other active scalar can be defined in cavitation
problems.
(c) The default molecular properties of the CAV scalar are those of water
vapour. You may therefore need to define alternative properties if
your vapour corresponds to a different fluid.
(d) You may define as many ‘passive’ scalars as are necessary for your
11-8 Version 4.02
Chapter 11 FREE SURFACE AND CAVITATION
Cavitating Flows

model
(e) The diffusion term in the transport equation for all scalars defined in
the cavitation model will be switched off

5. In the Additional Scalars sub-folder, initialize the distribution of the VOF


scalar by opening the “Initialisation” panel and choosing one of the available
options. In most cases, either Constant values, User coding, or assignment
according to cell type is applied. In the latter case, click CTAB on the main
pro-STAR window to open the Cell Table Editor and use it to set the Initial
Free Surface Material switch to Light (for those cell types initially occupied
by the ‘light’ fluid) or to Heavy (for cell types initially occupied by the
‘heavy’ fluid).
6. For cavitation problems involving porous media, use the Porosity sub-folder
to define properties for the porous materials in the normal way.
7. Use the Sources sub-folder to define external source terms for momentum,
turbulence, enthalpy and scalars. Note that:
(a) User Coding is the only supported option in this case
(b) The VOF transport equation does not accept additional source terms
(c) When the default Rayleigh model is used, you cannot define additional
source terms for the CAV scalar.

Step 6: Define boundary conditions


Go to the Define Boundary Conditions folder
• Open the “Define Boundary Regions” panel and specify appropriate boundary
conditions in the usual manner
• Open the “Scalar Boundaries” panel and specify boundary conditions for the
CAV and (if applicable) VOF scalars, the latter representing the heavy fluid
volume fraction.

Valid boundary types for cavitation problems are listed under Step 1. If a pressure
boundary condition is specified, make sure that both the Envir Press (environmental
pressure) and Mean (pressure profile mean value) options are set to Off.

Step 7: Define analysis control parameters


Go to the Analysis Control folder:
1. Select the Solution Controls sub-folder and then open the “Solution Method”
panel to set/adjust the solution algorithm parameters. Note that:
(a) Only the SIMPLE algorithm is applicable to cavitation problems
(b) Both CG and AMG solvers are applicable and the desired one may be
selected from the Solver Type menu. AMG is recommended as it usually
leads to shorter computing times.
(c) For transient cases, select option Euler Implicit from the Temporal
Discretisation menu. This is the only option supported for this type of
problem.
(d) If a steady-state solution is expected, one can limit the number of outer
iterations per time step (see topic “Transient problems” in the STAR
GUIde on-line Help) to 1, in which case a pseudo-transient marching

Version 4.02 11-9


FREE SURFACE AND CAVITATION Chapter 11
Cavitating Flows

towards the steady state is obtained. This is applicable to both cavitating


and free-surface flows, but one needs to be certain that a steady-state
solution can be reached.

2. Select the Equation Behavior sub-folder and then open the “Primary
Variables” panel. Make any necessary adjustments to the current or default
settings in the Equation Status, Solver Parameters and Differencing Schemes
tabs.
3. In the “Additional Scalars” panel, select a differencing scheme and
under-relaxation parameter for the CAV scalar and for scalars other than VOF
(the differencing scheme for the latter is set in the free-surface “Controls”
panel; however, you may want to adjust its under-relaxation parameter).
4. Select the Output Controls sub-folder and use the “Monitor Numeric
Behaviour” panel to print additional information such as convergence
residuals and conservation checks (optional).
5. Select the Monitor Engineering Behavior sub-folder and, if you wish, use the
“Monitor Cell Behaviour” panel to save selected cell data for subsequent
plotting against iteration or time step number.
6. The Analysis Output sub-folder enables you to specify the frequency of
saving solution results in the .ccm file.
7. Use the “Switches and Real Constants” panel in the Other Controls sub-folder
to set switches and constants for any ‘non standard’ practices. Please check
carefully the meaning of each switch and constant and use it only when
absolutely necessary.

Step 8: Define the time step size and run duration


There are two ways to define the time step size and the run time length:
• Go to the Analysis Preparation/Running folder and open the “Set Run Time
Controls” panel. Fill in appropriate values for simulation time and time step
size in the relevant boxes. This is the recommended way of defining time
steps.

• If your application involves a moving mesh defined by an events file, you will
need to use the Advanced Transients panel by choosing Modules > Transient
from the main pro-STAR window. in this panel, you can define load steps,
each of which contains the time step size and number of time steps to be used
for each load step. Please note that you need to go through Step 1 to Step 7
before defining load steps using the Advanced Transients panel.

11-10 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Rotating Reference Frames

Chapter 12 ROTATING AND MOVING MESHES


The theory behind rotating and moving mesh problems and the manner of
implementing it in STAR-CD is given in the Methodology volume, Chapter 13. The
present chapter contains an outline of the process to be followed when setting up a
rotating or moving mesh simulation, including details of the user input required and
important points to bear in mind when setting up problems of this kind.

Rotating Reference Frames


Models for a single rotating reference frame
Step 1
Go to the “Select Analysis Features” STAR GUIde panel and select option On from
the Rotating Reference Frame Status pop-up menu. This activates an additional
folder in the NavCenter tree called Rotating Reference Frames.
Step 2
In the above folder, open the “Rotating Reference Frames” panel and select option
Single Frame. This enables you to define spin parameters for the material in your
model. The required parameters are angular velocity and a local coordinate system
whose Z-axis defines the axis of rotation, see Figure 12-1.

ω = 200 rpm

Figure 12-1 Solid body rotation

Useful points on single rotating frame problems


1. The angular velocity can vary with time, with the variation specified in
(a) user subroutine UOMEGA, or
(b) a user-defined table, or
(c) by giving it a different value at each load step of a transient run (see
Chapter 5, “Load-step based solution mode”).

2. The boundaries of the rotating domain are also assumed to be rotating. To


model stationary walls, it is necessary to specify an equal and opposite spin
velocity in the Omega text box of the Boundary Region dialog for walls (see
the STAR GUIde “Wall” Help topic). Similarly, to model axial inflow, it is
necessary to specify a spin velocity in the dialog for Inlet regions.
3. When a stagnation boundary condition is used, an option is provided to
specify whether the direction cosines are based on relative or absolute
velocities. Stagnation quantities are also defined using either relative or
Version 4.02 12-1
ROTATING AND MOVING MESHES Chapter 12
Rotating Reference Frames

absolute velocities.
4. When turbulence is specified as an intensity (at inlet or pressure boundaries),
the turbulence kinetic energy is computed on the basis of static coordinate
frame velocities. For stagnation boundaries, the specified intensity uses the
same velocity as the stagnation quantities.
5. Boundary velocities are computed in the local rotating coordinate system.
This is important in interpreting the information passed to the user
subroutines.
6. When post processing results, you may view velocities in either the relative or
the absolute reference frame (see the “Coord System tab”, located in the
“Load Data” STAR GUIde panel).

Models for multiple rotating reference frames (implicit treatment)


Step 1
Go to the “Select Analysis Features” STAR GUIde panel and select option On from
the Rotating Reference Frame Status pop-up menu. This activates an additional
folder in the NavCenter tree called Rotating Reference Frames.
Step 2
• Decide how many reference frames are required to model the problem
adequately. For example, the two-dimensional mixer problem shown in
Figure 12-2 requires two rotating frames.
• Generate the mesh.

Baffle

r = 15 cm

ω = 0 rpm Sub-domain 2
Spin index = 2

Sub-domain 1
r = 10 cm Spin index = 1

ω = 500 rpm

Baffle
r = 5 cm

Figure 12-2 Multiple rotating frame illustration

12-2 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Rotating Reference Frames

Step 3
Display the Cell Table Editor by clicking the CTAB button on the main pro-STAR
window. Define cell index numbers to correspond to each of the rotating mesh
blocks (sub-domains) (see “The Cell Table” on page 3-1). Assign different spin and
colour table indices to each cell type, as shown below, for the two rotating
sub-domains of Figure 12-2. Note that the table entries for both sub-domains have
the same material property reference number since the sub-domains belong to the
same fluid domain.

Sub-domain 1

Sub-domain 2

Version 4.02 12-3


ROTATING AND MOVING MESHES Chapter 12
Rotating Reference Frames

Step 4
Assign all cells within a sub-domain in turn to each of the cell types created above
(see “Cell indexing” on page 3-3).
Step 5
In the Rotating Reference Frames folder, open the “Rotating Reference Frames”
panel and select option Multiple Frames - Implicit. This enables you to specify spin
parameters (angular velocities and axes of rotation) for each of the spin indices
already defined. In terms of the example of Figure 12-2, zero rotational speed needs
to be assigned explicitly to sub-domain no. 2 since its local coordinate system is
used in transforming velocities across the sub-domain interface.
Useful points on multiple implicit rotating frame problems
1. When modelling multiple rotating reference frame (m.r.f.) problems, it is
advisable to check the results carefully and see if they are reasonable and
within the limitations of this approach. If this is not the case, one may need to
resort to moving mesh methods, such as those described in the section on
“Regular sliding interfaces”.
Note, however, that a result obtained via the m.r.f. method can always be
used as an initial field for a transient moving mesh simulation. This will
reduce the time needed to reach a periodic state solution.
2. It is important to ensure that the interface between the different m.r.f.
sub-domains is a smooth surface (i.e. a constant-radius surface). This point
needs particular attention in all-tetrahedral mesh cases.
3. An angular velocity can vary with time, with the variation specified in
(a) user subroutine UOMEGA, or
(b) a user-defined table, or
(c) by giving it a different value at each load step of a transient run (see
Chapter 5, “Load-step based solution mode”).

4. The boundaries of a rotating domain are also assumed to be rotating. To


model stationary walls, it is necessary to specify an equal and opposite spin
velocity in the Omega text box of the Boundary Region dialog for walls (see
the STAR GUIde “Wall” Help topic). Similarly, to model axial inflow, it is
necessary to specify a spin velocity in the dialog for inlets.
5. When a stagnation boundary condition is used, an option is provided to
specify whether the direction cosines are based on relative or absolute
velocities. Stagnation quantities are also defined using either relative or
absolute velocities.
6. When turbulence is specified as an intensity (at inlet or pressure boundaries),
the turbulence kinetic energy is computed on the basis of static coordinate
frame velocities. For stagnation boundaries, the specified intensity uses the
same velocity as the stagnation quantities.
7. In cases where the mesh structure changes across the interface between two
sub-domains (for example, between two axial turbomachinery stages, with the
blades swept in opposite directions):
(a) Build each sub-domain separately with its own ‘best fit’ mesh structure,
and cell types with different spin indices

12-4 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Rotating Reference Frames

(b) Create a continuous mesh by coupling together the cells layers on either
side of the interface using the Couple tool (Create Couples option; see
also Chapter 3, “Couple creation” in the Meshing User Guide).
(c) Use the Couple tool’s Couple Transform option to replace the coupled
cells with polyhedral cells that have a one-to-one cell face
correspondence at the interface.

8. Boundary velocities are computed in the local rotating coordinate system.


This is important in interpreting the information passed to the user
subroutines.
9. When post processing results, you may view velocities in either the relative or
the absolute reference frame (see the “Coord System tab”, located in the
“Load Data” STAR GUIde panel).
10. The present version of STAR-CD does not support the use of rothalpy (see
“Rothalpy” on page 1-5 of the Methodology volume) in combination with the
implicit solution technique.
Models for multiple rotating reference frames (explicit treatment)
Step 1
Go to the “Select Analysis Features” STAR GUIde panel and select option On from
the Rotating Reference Frame Status pop-up menu. This activates an additional
folder in the NavCenter tree called Rotating Reference Frames.
Step 2
• Decide how many rotating frames of reference are required to model the
problem adequately, and the locations of the interfaces.
• Generate the mesh. The interface between adjacent rotating mesh blocks is
defined by pairs of adjacent (but spatially coincident) boundaries, as shown in
Figure 12-3. The coincident boundaries are first defined as independent
boundary regions using separate sets of vertices and then coupled together as
described in Step 7 below. Note that the interface must be either a plane
perpendicular to the axis of rotation or a conical section, i.e. a surface
generated by rotating a straight line around that axis.

Version 4.02 12-5


ROTATING AND MOVING MESHES Chapter 12
Rotating Reference Frames

Boundary Regions
no. 5 no. 6 no. 7 no. 8
circumferential (pressure) (inlet) (pressure) (inlet)
direction
4 ω = 100 rpm 36 40 ω = 500 rpm 64 68 ω = 1000 rpm 100
8 12 16 20

3 35 39 63 67 99
7 11 15 19

2 34 38 62 66 98
6 10 14 18

1 IMAT = 1 33 37 IMAT = 2 61 65 IMAT = 3 97


5 9 13 17

cell number boundary number

(a)

1134 1135

134 135

33 1034 1035
37

34 35

(b)
Figure 12-3 Coupled boundary illustration

Step 3
Display the Cell Table Editor by clicking CTAB on the main pro-STAR window.
Define cell index numbers to correspond to each of the rotating domains (see “The
Cell Table” on page 3-1). Assign different material property and colour table
indices to each cell type but ignore the spin index. In the above example, cell and
material indices 1, 2 and 3 are defined to correspond to each domain.
Step 4
Assign all cells within a domain in turn to each of the cell types created above (see
“Cell indexing” on page 3-3). Also ensure that separate monitoring cell and
reference pressure locations are specified for each domain.
Step 5
Go to panel “Create Boundaries” in STAR GUIde, open tab “Regions” and use its
facilities to create separate boundary regions at either side of each interface between
domains, as shown in Figure 12-3.
Step 6
Specify boundary conditions for both sides of an interface using panel “Define
Boundary Regions” (only inlet and pressure boundary types are allowed). Example
dialog boxes for boundary regions 5 and 6, making up the first interface in the above
example, are shown below:

12-6 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Rotating Reference Frames

Step 7
Go back to panel “Create Boundaries” and use tab “Couples” to join the interface

Version 4.02 12-7


ROTATING AND MOVING MESHES Chapter 12
Rotating Reference Frames

boundaries together. In doing so, you also need to:


1. Specify whether to join individual boundaries from each region on a
one-to-one basis, or to couple the two regions to each other as a whole. If the
latter is chosen, the value to be imposed on the couple’s pressure boundary is
found by an averaging process. For example, the average of the values
assigned to boundary region no. 5 in Figure 12-3 is

8 8
⎛ ⎞ ⎛ ⎞
P region 5 = ⎜ ∑ p i s i⎟ ⁄ ⎜ ∑ s i⎟ (12-1)
⎝i = 5 ⎠ ⎝i = 5 ⎠

where p is the pressure and s the area of each boundary face.


2. If necessary, place region couples (as defined above) into separate groups.
This enables you to identify boundary faces across which mass must be
conserved and is only necessary in problems that have only inlet boundary
couples. Such domains are recommended for solving closed loop problems
where the flow rate needs to be determined as part of the solution. The groups
to balance are specified in the “Rotating Reference Frames” panel (see Step 8
below).
Step 8
In the Rotating Reference Frames folder, open the “Rotating Reference Frames”
panel and select either option “Multiple Frames - Explicit” or option “Multiple
Frames - NR-Explicit”. This enables you to specify:
1. Spin parameters (angular velocities and axes of rotation) for each of the mesh
domains already defined. In the above example, domains 1, 2 and 3 have
angular velocities of 100, 500 and 1000 r.p.m., respectively. The spin axis is
normally common to all domains.
2. Control parameters required by the explicit solution algorithm and, if
required, the coupled region groups mentioned in Step 7 above.
Useful points on multiple explicit rotating frame problems
1. When modelling multiple rotating reference frame (m.r.f.) problems, it is
advisable to check the results carefully and see if they are reasonable and
within the limitations of this approach. If this is not the case, one may need to
resort to moving mesh methods, such as those described in the section on
“Regular sliding interfaces”.
Note, however, that a result obtained via the m.r.f. method can always be
used as an initial field for a transient moving mesh simulation. This will
reduce the time needed to reach a periodic state solution.
2. It is important to ensure that the interface between the different m.r.f. domains
is a smooth surface (i.e. a constant-radius surface). This point needs particular
attention in all-tetrahedral mesh cases.
3. An angular velocity can vary with time, with the variation specified in
(a) user subroutine UOMEGA, or
(b) a user-defined table, or
(c) by giving it a different value at each load step of a transient run (see

12-8 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Moving Meshes

Chapter 5, “Load-step based solution mode”).

4. The boundaries of a rotating domain are also assumed to be rotating. To


model stationary walls, it is necessary to specify an equal and opposite spin
velocity in the Omega text box of the Boundary Region dialog for walls (see
the STAR GUIde “Wall” Help topic). Similarly, to model axial inflow, it is
necessary to specify a spin velocity in the dialog for inlets.
5. When a stagnation boundary condition is used, an option is provided to
specify whether the direction cosines are based on relative or absolute
velocities. Stagnation quantities are also defined using either relative or
absolute velocities.
6. When turbulence is specified as an intensity (at inlet or pressure boundaries),
the turbulence kinetic energy is computed on the basis of static coordinate
frame velocities. For stagnation boundaries, the specified intensity uses the
same velocity as the stagnation quantities.
7. Interfaces between differentially-rotating mesh domains are best placed at
positions that do not lie inside recirculating flow fields.
8. Caution should be exercised when using this approach because of the explicit
coupling at the special boundaries. The method is most suitable for problems
involving strong outflow across the coupled interface.
9. The NR-Explicit option should be chosen over the Explicit option for
configurations where the turbomachinery blades are closely packed and/or if a
shock wave is expected to hit either of the two coupled boundaries at the
interface.
10. Boundary velocities are computed in the local rotating coordinate system.
This is important in interpreting the information passed to the user
subroutines.
11. When post processing results, you may view velocities in either the relative or
the absolute reference frame (see the “Coord System tab”, located in the
“Load Data” STAR GUIde panel).

Moving Meshes
Basic concepts
The moving mesh feature is activated by command MVGRID. Changes in mesh
geometry can be specified either by pro-STAR commands (i.e. the Change Grid
operation in the EVENTS command module), or by user coding included in
subroutine NEWXYZ. In this subroutine, the user can vary the geometry of a model
by defining vertex coordinates as a function of time. The deformed coordinates are
written to the transient post data (.pstt) file and can be loaded and plotted during
post-processing.
As an alternative, the Change Grid (CG) operation can be used to alter the vertex
positions with time. Its distinguishing features are as follows:
• The operation is initiated at an ‘event step’ specified by the user and remains
active at all subsequent time steps, until the CG operation is explicitly turned
off by a termination event, or a new set of CG commands are provided as part
of another event step.
• The main body of the operation consists of a set of pro-STAR commands that

Version 4.02 12-9


ROTATING AND MOVING MESHES Chapter 12
Moving Meshes

are used while STAR is running (as part of a STAR/pro-STAR interaction


process).
• The above commands utilise a set of both program-defined and user-defined
parameters that can store anything that is of relevance to the problem
description.
The parameters used by the CG command set are:
1. Program-defined
(a) ITER — current time step number
(b) TIME — current solution time
(c) LSTP — current load step (see Chapter 5, “Load step definition”)
(d) EVEX — last executed event number
(e) EVNO — event number to be executed next
(f) ETIM — time at which the next event is scheduled
(g) YPST — piston position; a special parameter for piston engine problems,
calculated on the basis of other parameters supplied by command
EVPARM (see “Setting up models” on page 12-15).
2. User-defined
These are specified by the user in subroutine UPARM to provide additional
parameters. They are of two kinds:
(a) Integer parameters in the range 0-999
(b) Real parameters in the range 0-999

Note that pro-STAR restricts the number of active parameters to 99.


The CG operation uses all the standard pro-STAR facilities and is therefore more
flexible and powerful for mesh geometry changes than user coding supplied in
subroutine NEWXYZ. Note that STAR-CD also provides other special operations
related to moving meshes, as follows:
• Cell removal/addition — (see “Cell-layer Removal/Addition” on page 12-14)
• Sliding mesh — (see “Sliding Meshes” on page 12-18)
• Conditional cell attachment and change of fluid type — (see “Cell
Attachment and Change of Fluid Type” on page 12-22)
Setting up models
The main steps for setting up a moving mesh model are outlined below. For more
detailed information, refer to Tutorial 11 in the Tutorials volume.
Step 1
Generate the mesh at time t = 0 and issue the following command:

TIME,TRANS (turn on the transient solution option)

followed by either

MVGRID,ON (turn on the moving-grid option, when


using subroutine NEWXYZ only)

or

12-10 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Moving Meshes

MVGRID,ON,EVENT,PROSTAR (turn on the moving-grid option, when


using the EVENTS command module)

Step 2
(Skip this step only if mesh changes are input through the user subroutine NEWXYZ)
Define an event step data file, e.g.

EVFILE,INITIAL,case.evn (initialise the events file)


EVSTEP,1,TIME,0.0 (define an event)
EGRID,READ,case.cgrd (get the description of mesh operations
from file case.cgrd, in coded form)
EVSAVE,1 (save this information as event no. 1)

The contents of file case.cgrd mentioned above for the problem shown in Figure
12-4 are as follows:

! Comments like this are allowed by starting the line with “!”

VSET,NONE (clear the vertex set)


VSET,ADD,VRANGE,1,2,1 (add vertices 1 and 2 to the set)
*SET,YBOT,TIME (set parameter YBOT equal to the current
time)
VMOD,VSET,F,YBOT (change the y-coordinates of the vertex
set so that they follow the bottom bound-
ary movement)
VFILL,1,11,4,3,2,2,1 (re-position the mesh vertices between
the two boundaries)

11 12

9 10

7 8

5 6
Y

3 4
1 m/s

X 1 2

Figure 12-4 Moving mesh illustration

Version 4.02 12-11


ROTATING AND MOVING MESHES Chapter 12
Moving Meshes

Note that:
1. An event step can be
(a) deleted, if necessary, with command EVDELETE and remaining event
steps re-numbered via command EVCOMPRESS;
(b) modified with command EVGET;
(c) listed on the screen with command EVLIST.

2. Command EVUNDELETE restores a previously deleted event step.


3. User-specified offsets can be applied to the actual event time via command
EVOFFSET.

Step 3
• If using the method described in Chapter 5, “Load-step based solution mode”,
define the load step for the transient run.
• Check the validity of specified events and prepare the events data file for
subsequent use via command EVPREP.
• Save the problem’s data files using commands GEOMWRITE,
PROBLEMWRITE, etc. or their equivalent GUI operations accessible from the
File menu.

Note that the events data file can be


• written in coded form to a (.evnc) file with command EVWRITE, typically
in order to transfer data to another computer
• read in coded form from a (.evnc) file with command EVREAD, typically
when transferring data from another computer

Step 4
Exit from pro-STAR and then run STAR from your session’s X-window, as
described in Chapter 2, “Running a STAR-CD Analysis”, Step 6.

Step 5
Post-process the data. For example, the commands needed to process time step no.
10 are:

SUBTITLE
Results at time step 10
Velocity field
EVFI CONN case.evn (connect the event file)
TRLOAD case.pstt (load the transient post data file)
STORE ITER 10 (the appropriate events are loaded and
executed automatically)
GETC ALL (get the cell data)
POPT VECT
PLTY NORM
CSET NEWS FLUID
CPLOT
QUIT,NOSAVE
12-12 Version 4.02
Chapter 12 ROTATING AND MOVING MESHES
Moving Meshes

Be very careful not to save problem information to file case.mdl as the current
geometry corresponds to the state of the mesh at time step no. 10.
Useful points
1. STAR can be run in ‘mesh preview’ mode only, which is very useful for
checking out the mesh set-up. To do this, a hidden switch has to be set up in
pro-STAR as follows:

RCONSTANT, 4, 1. (set constant number 4 to 1.)

The message “MESH PREVIEW RUN” should appear both on the screen and
in the run-time output (.run) file when running STAR. Note that this facility
is not available for parallel runs.
2. Moving grid events normally describe a continuous motion and will therefore
remain operational throughout the run. If, however, the grid motion needs to
be stopped for whatever reason, this can be done via a termination event as
follows:

EGRID,NONE

3. The transient post data (.pstt) file is usually very large, so care must be
taken when specifying the post data output frequency. If the analysis is split
into several stages, it is also advisable to give the .pstt file produced at the
end of each stage a unique filename. This helps to spread the output produced
amongst several files and thus ease the data management and manipulation
processes.
4. Porous media should not be used in areas of the mesh where there is relative
internal movement (i.e. cell expansion or contraction).
5. You are strongly advised to set the pressure correction under-relaxation factor
to a value less than 1.0 (e.g. 0.8) before starting the analysis.
6. Flow boundary conditions on boundaries that have moving vertices may result
in mass flux into / out of the domain, caused by the displacements of the
boundaries.
7. The only valid option for restart runs is Standard Restart (see the “Analysis
(Re)Start” panel in STAR GUIde.

Automatic Event Generation for Moving Piston Problems


pro-STAR provides a special command, MMPISTON, which may be used in an
engine model to automatically generate the moving part of the piston mesh, the
Change Grid (.cgrd) command file and the event (.evn) file. More specifically,
the moving mesh commands accomplish the following:
Starting from a basic mesh, they create the cells, vertices, boundaries, events and
moving grid commands to completely specify the mesh motion for the STAR
solver. These commands are designed to be used in sequence, so that all entities are
created at the end of the current model and do not compromise earlier events. As the
output is standard pro-STAR events and EGRID commands, an advanced user can
readily modify these to suit specific problems.

Version 4.02 12-13


ROTATING AND MOVING MESHES Chapter 12
Cell-layer Removal/Addition

Cell-layer Removal/Addition
Basic concepts
A cell is removed by collapsing intervening faces between two opposite sides in a
given direction. This is done by moving together the vertices making up the faces.
Cells can be collapsed at the beginning of a given time step or prior to the start of
the calculations. The latter case is treated as a special mesh set-up operation and
does not affect the solution in any way. Normally, entire layers of cells are removed
at a given event step. However, it is also possible to remove part of a layer, in which
case cells at the edge of the retained section collapse into prisms. A cell layer (or
partial layer) has the following properties:
• It is defined as a group of cells that is one cell thick in the collapsing
direction.
• The faces which collapse must be quadrilaterals, but those forming the upper
and lower surfaces of the layer may be quadrilateral or triangular.
• The collapsing cell faces on the outer perimeter of the group form boundaries.
• Either the upper or lower surface of a layer may coincide in whole or part with
a boundary, but not both surfaces simultaneously.
• No more than one layer may be removed at each event step.
• The layer must not be composed of tetrahedral cells.
• Trimmed (polyhedral) cells can only be collapsed if they have been formed by
extruding another cell in the direction of collapse.

The reverse operation, adding a cell layer, is achieved by expanding the removed
layer in the direction it was collapsed. This means that layers to be added must have
been removed first. Thus, all restrictions on cell removal also apply to cell addition
so that:
• Only one entire layer (or partial layer) may be restored at each event step.
• When cells are restored, they reappear next to the neighbours they had at the
time of their collapse.
• If any of their faces were boundaries, those boundaries are also restored.
• Cell layers must be restored in the reverse order in which they were removed.

The cells to be removed or added, and the time at which to do this (i.e. event step
and event time) are specified in the EVENTS command module. A cell removal or
addition event is executed when the current simulation time equals the time
specified by the event step, within a given tolerance.
Note that cell removal or addition changes only the cell connectivity within the
mesh. The actual change of mesh geometry has to be specified explicitly through a
moving mesh operation of the kind described in “Moving Meshes” on page 12-9. In
the event of cell removal, the user has to ensure that:
• The mesh geometry changes in a way that reflects the fact that cells have been
removed.
• Cells remain collapsed until they are restored. This means that vertices
belonging to the removed cells must move with the moving boundary for all
subsequent time steps.

12-14 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Cell-layer Removal/Addition

Setting up models
Cell Removal or Addition operations should always be combined with either
• Change Grid operations in the EVENTS command module, or
• the user subroutine NEWXYZ.

The main steps for setting up a model of this kind are outlined below.
Step 1
Generate the mesh at time t = 0. The layers to be removed can be given different cell
index numbers using command CTABLE.
.

7 19 20 21

6 16 17 18

5 13 14 15
Cell index Cell number

4 10 11 12

Y (2) 3 7 8 9

2 4 5 6

1 1 2 3
X (1)

Figure 12-5 Cell layer removal illustration

Referring to the example of Figure 12-5 the relevant commands would be:

CTAB,1,Fluid
RP7,1
*SET,CTY,1,1
*SET,C1,1,3
*SET,C2,3,3
*DEFINE
CTYPE,CTY
CSET,NEWS,CRANG,C1,C2,1
CMOD,CSET
*END
*LOOP,1,6,1

Step 2
Issue the following commands:

TIME,TRANS (turn on the transient solution option)


Version 4.02 12-15
ROTATING AND MOVING MESHES Chapter 12
Cell-layer Removal/Addition

MVGRID,ON,EVENT,PROSTAR (turn on the moving-grid option)

Step 3
• Define an event step data file, e.g.

EVFILE,INITIAL,case.evn (initialise the events data file)

• Turn on the Change Grid operation at time t = 0

EVSTEP,1,TIME,0.0
EGRID,READ,case.cgrd (get the description of mesh operations
from file case.cgrd, in coded form)
EVSAVE,1 (save this information as event no. 1)

• Specify cell layer removal via the cell type

EVSTEP,2,TIME,0.05
EDDIR,LOCAL,1,2 (remove cells in direction no. 2 in the
local coordinate system)
EDCELL,ADD,CTYPE,1 (remove cells with index no. 1)
ECLIST,DEACTIVATED,ALL (list removed cells)
EVSAVE,2

• Specify cell layer removal via a cell range

EVSTEP,3,TIME,0.08
EDDIR,LOCAL,1,2
EDCELL,ADD,CRAN,4,6,1
EVSAVE,3

• Specify cell layer addition, assuming the last cell layer removed had index no.
2

EVSTEP,4,TIME,0.2
EACELL,ADD,CTYPE,2 (add all cells with index 2)
ECLIST,ACTIVATED,ALL (list added cells)
EVSAVE,4

Note that:
1. The event time can also be specified using global parameters. For example

EVPARM PISTON 1000. 0.04 0.13 0.015 COMP 0.1015


↑ ↑ ↑ ↑ ↑ ↑
piston rotating crank length initial piston
engine speed radius of con. piston location
(rpm) rod position at TDC

12-16 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Cell-layer Removal/Addition

EVSTEP 1 PCOMP 0.02


↑ ↑ ↑
event step compression stage piston position

2. An event step can be


(a) deleted, if necessary, with command EVDELETE and remaining event
steps re-numbered via command EVCOMPRESS;
(b) modified with command EVGET;
(c) listed on the screen with command EVLIST.

3. Command EVUNDELETE restores a previously deleted event step.


4. User-specified offsets can be applied to the actual event time via command
EVOFFSET.

Step 4
• If using the method described in Chapter 5, “Load-step based solution mode”,
define the load step for the transient run.
• Check the validity of specified events and prepare the events data file for
subsequent use via command EVPREP.
• Save the problem's data files using commands GEOMWRITE,
PROBLEMWRITE, etc. or their equivalent GUI operations accessible from the
File menu.

Note that the events data file can be


• written in coded form to a (.evnc) file with command EVWRITE, typically
in order to transfer data to another computer
• read in coded form from a (.evnc) file with command EVREAD, typically
when transferring data from another computer

Step 5
Exit from pro-STAR and then run STAR from your session’s X-window, as
described in Chapter 2, “Running a STAR-CD Analysis”, Step 6.

Step 6
Post-process the data. For example, the commands needed to process time step no.
10 are:

SUBTITLE
Results at time step 10
Velocity field
EVFI CONN case.evn (connect the event file)
TRLOAD case.pstt (load the transient post data file)
STORE ITER 10 (the appropriate events are loaded and
executed automatically)
GETC ALL (get the cell data)

Version 4.02 12-17


ROTATING AND MOVING MESHES Chapter 12
Sliding Meshes

POPT VECT
PLTY NORM
CSET NEWS FLUID
CPLOT
QUIT,NOSAVE
Be very careful not to save problem information to file case.mdl as the current
geometry corresponds to the state of the mesh at time step no. 10.
Useful points
1. You are strongly advised to identify cell layers intended for removal/addition
by assigning a unique cell index to each of them.
2. Cell layers can be removed at negative event times. This is useful, for
example, in reciprocating piston engine models where simulation starts with
the piston at top dead centre. In such cases the previously removed cell layers
can thus be added at positive event times.
3. You are advised to first run the model in ‘mesh preview’ mode in order to
check whether the intended cell removal/addition and mesh movement are
carried out correctly. This can be done by issuing the following command in
pro-STAR:

RCONSTANT, 4, 1. (set constant number 4 to 1)

The message “MESH PREVIEW RUN” should appear both on the screen and
in the run-time output (.run) file when running STAR.
4. It is very important to ensure that the locations chosen for reference pressure
and field variable monitoring (via commands PRESSURE and MONITOR,
respectively) correspond to cells that will never be removed.
5. If the simulation includes combustion modelling and the definition of ignition
regions (see Chapter 8, “Setting Up Chemical Reaction Schemes”, Step 5),
make sure that no cells corresponding to these regions have been removed
during the time that ignition takes place.
6. You are strongly advised to set the pressure correction under-relaxation factor
to a value less than 1.0 (e.g. 0.8) before starting the analysis.
7. For STAR-HPC runs, you need to ensure that the removed cell layers do not
collapse towards the inter-processor boundaries. In another words, the
removed cell layers and the inter-processor boundaries should always be
perpendicular to each other. This can be achieved through manual
decomposition.

Sliding Meshes
Regular sliding interfaces
One way of implementing sliding meshes is the regular sliding interface method.
This enables the interface cells to progressively change their connectivity during the
solution.
The change of cell connectivity is activated through a ‘cell attachment’
operation. Cell pairs to be attached and the time of attachment (i.e. event step and

12-18 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Sliding Meshes

event time) are specified by the user in the EVENTS command module. The cell
attachment event is executed when the current simulation time equals the time
specified by the event step within a given tolerance.
Setting up models
The regular sliding interface method combines both the Cell Attachment and the
Change Grid operation in the EVENTS command module. The main steps for
setting up a case are outlined below.

Step 1
• Generate the mesh at time t = 0. The sliding interface is defined as two
coincident boundaries, one for the stationary and one for the moving part of
the mesh. Thus, two sets of coincident vertices must be defined at that
location. The two coincident boundaries have to be defined as different
boundary regions and declared as attachment boundaries using the RDEFINE
command:

RDEF,1,ATTACH (define boundary region no.1 as an


attachment boundary)

1 0
↑ ↑
local coordinate system alternate wall system
(see “Cell Attachment and Change of Fluid Type” on
page 12-22
for an explanation of this parameter)

RDEF,2,ATTACH
1,0

• Issue the following commands:

TIME,TRANS (turn on the transient solution option)


MVGRID,ON,EVENT,PROSTAR (turn on the moving-grid option)

Step 2
• Define an event step data file.

EVFILE,INITIAL,case.evn (initialise the events data file)

• Perform an initial attachment operation for the relevant boundary pairs


(otherwise they will be treated as detached).

EVSTEP,1,TIME,0.0 (event step 1 occurs at time t = 0.0)

followed by either

EAMATCH,1,2 (match regions 1 and 2)

Version 4.02 12-19


ROTATING AND MOVING MESHES Chapter 12
Sliding Meshes

or

EATTACH,6,1 (attached boundaries 6 and 1)


RP5,1,1 (attach the rest of the boundary pairs)
EALIST,ALL (list out all attached boundary pairs)

• Turn on the Change Grid operation at time t = 0.

EGRID,READ,case.cgrd (get the description of mesh operations


from file case.cgrd, in coded form)
EVSAVE,1 (save this information as event no. 1)

• Specify subsequent attachment operations, e.g.

EVSTEP,2,TIME,0.02
EATTACH,6,2 (attach boundaries 6 and 2)
EAGENERATE,4,1,1,1,1 (EAGENERATE works similarly to
CGENERATE, see “Command-driven
facilities” on page 2-44 of the Meshing
User Guide)
EATTACH,10,1 (attached boundaries 10 and 1)
EVSAVE 2 (save event no. 2)

Note that:
1. The attached boundary set definitions in an event step can be
(a) deleted, if necessary, with command EADELETE and remaining
definitions re-numbered via command EACOMPRESS;
(b) listed on the screen with command EALIST.

2. An event step can be


(a) deleted, if necessary, with command EVDELETE and the remaining event
steps renumbered via command EVCOMPRESS;
(b) modified with command EVGET;
(c) listed on the screen with command EVLIST.

3. Command EVUNDELETE restores a previously deleted event step.


4. User-specified offsets can be applied to the actual event time via command
EVOFFSET.

Step 3
• If using the method described in Chapter 5, “Load-step based solution mode”,
define the load step for the transient run.
• Check the validity of specified events and prepare the events data file for
subsequent use via command EVPREP.
• Save the problem’s data files using commands GEOMWRITE,
PROBLEMWRITE, etc. or their equivalent GUI operations accessible from the
File menu.

12-20 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Sliding Meshes

Note that the events data file can be


• written in coded form to a (.evnc) file with command EVWRITE, typically
in order to transfer data to another computer
• read in coded form from a (.evnc) file with command EVREAD, typically
when transferring data from another computer

Step 4
Exit from pro-STAR and then run STAR from your session’s X-window, as
described in Chapter 2, “Running a STAR-CD Analysis”, Step 6.

Step 5
Post-processing the data. For example, the commands needed to process time step
no. 10 are:

SUBTITLE
Results at time step 10
Velocity field
EVFI CONN case.evn (connect the event file)
TRLOAD case.pstt (load the transient post data file)
STORE ITER 10 (the appropriate events are loaded and
executed automatically)
GETC ALL (get the cell data)
POPT VECT
PLTY NORM
CSET NEWS FLUID
CPLOT
QUIT,NOSAVE

Be very careful not to save problem information to file case.mdl as the current
geometry corresponds to the state of the mesh at time step no. 10.
Useful points
1. At time t = 0, cell pairs are detached. They become attached only when an
event containing EATTACH or EAMATCH commands is executed. Once
attached in this way, they remain attached until another EATTACH or
EDETACH command references them, or they are deactivated.
2. When the model’s mesh is being created, it is very useful to set up a regular
boundary numbering scheme at the interface, because this simplifies the
specification of cell attachment.
3. At the initial stages of the analysis, the solution can be accelerated by using
pure sliding only (i.e. without shearing), which in turn allows larger time
steps. In terms of Figure 15-1 in Chapter 15 of the Methodology volume, this
is equivalent to going from Stage 1 to Stage 4 in a single time step. If this is
the case, the time step dt should be made equal to dtsl, where, for cylindrical
systems

dtsl = cell face angle at interface / rotating speed (12-2)

Version 4.02 12-21


ROTATING AND MOVING MESHES Chapter 12
Cell Attachment and Change of Fluid Type

In general, the time step dt should equal dtsl divided by an integer.


If accuracy is not at a premium, one may also slide the mesh by more than
one cell width (e.g. two cell widths) in a single time step.
4. In cylindrical systems, periodic results are usually reached after about seven
revolutions.
5. The transient post data (.pstt) file is usually very large, so care must be
taken when defining the output frequency of post-processing data. If the
analysis is split into several stages, it is also advisable to give the .pstt file
produced at the end of each stage a unique filename. This helps to spread the
output produced amongst several files and thus ease the data management and
manipulation processes.
6. It is advisable to first run the model in ‘mesh preview’ mode in order to check
whether the intended cell sliding and mesh movement are carried out
correctly. This can be done by issuing the following command in pro-STAR:

RCONSTANT,4,1. (set constant number 4 to 1.)

The message “MESH PREVIEW RUN” should appear both on the screen and
in the run-time output (.run) file when running STAR.
7. EATTACH commands are allowed only between active cells.
8. All boundaries belonging to a given region must couple only to boundaries
belonging to a (different) unique region. For example, it is illegal for some
boundaries from region 1 to couple to boundaries from region 2, while other
boundaries from region 1 couple to boundaries from region 3.
9. If one cell of an attached pair is deactivated, the other side reverts to the
alternate wall region.
10. If both cells of an attached pair are deactivated simultaneously and then
reactivated, the EATTACH command must be re-issued.
11. For STAR-HPC runs, you need to ensure that the sliding part of the mesh
resides completely on one processor. This can also be achieved through
manual decomposition.

Cell Attachment and Change of Fluid Type


Basic concepts
Cell attachment permits the following situations to be modelled:
1. The connection of unconnected neighbouring cells in different fluid domains,
say on the basis of local flow conditions. This can be used, for example, to
model leaf valves which pop open when the pressure difference across them
exceeds a given value.
2. The complete disconnection of neighbouring cells. This situation necessitates
two kinds of operation:
(a) A ‘Cell Attachment/Detachment’ operation.
(b) A ‘Change Fluid Type’ operation.

The latter enables a fluid domain to become completely cut off from the rest of the
flow field. Once cut off, the flow solution in such a domain can have its own
reference pressure and temperature. A special type of boundary (‘Attachment’ type)

12-22 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Cell Attachment and Change of Fluid Type

must also be declared at the interface where cell attachment and detachment is to
take place. STAR performs a cell detachment by connecting the detached cells to an
appropriate wall or inlet region.
Cell attachment/detachment operations are specified in the EVENTS command
module. The connection/disconnection event is initiated when the current
simulation time equals the time specified by the event step within a given tolerance.
The same also applies to the ‘Change Fluid Type’ operation. However, when the
designated time for connecting cells is reached, the operation may not necessarily
be carried out immediately. Instead, the precise connection/disconnection time is
determined by the flow solution. All conditions defined for a particular event are
maintained in the next event unless disabled explicitly. Thus, once a boundary pair
is attached, it remains attached until it is explicitly detached.
Setting up models
The main steps for setting up a cell attachment and change of fluid type case are
outlined below.
Step 1
• Generate the mesh at time t = 0. This requires a boundary interface to be set
up separating the (presently or potentially) different fluid domains. The
interface is defined as two coincident boundaries made up of two sets of
coincident vertices. The two boundaries must be first specified as different
boundary regions and then declared as attachment boundaries (see Figure
12-6) using command RDEFINE:

RDEF,1,ATTACH (define boundary region no. 1 as an


attachment boundary)

(boundary region no. 8


1 8 is a dummy region)
↑ ↑
local alternate
coordinate wall or
system inlet region

RDEF,8,inlet (could also be of type wall)

The alternate wall or inlet region is specified in order to enable the code to
assign appropriate (wall or inlet) properties to the attachment boundaries, if
they happen to be detached.

RDEF,2,ATTACH
1,8
RDEF,3,ATTACH
1,8
RDEF,4,ATTACH
1,8

• Issue the following commands:

Version 4.02 12-23


ROTATING AND MOVING MESHES Chapter 12
Cell Attachment and Change of Fluid Type

TIME,TRANS (turn on the transient solution option)


MVGRID,ON,EVENT,PROSTAR (turn on the moving-grid option)

1, 2, 3, 4, Boundary
5, 6, 7, 8 numbers
IMAT = 2 IMAT = 3

Cell numbers 151, 152


96, 97
2 3 4 4 7 8
1 1 2 3 5 6

Boundary region
1, 2, 3, 4
numbers

Y (2) IMAT = 1 Cell numbers


IMAT = 1: cells 1-100
IMAT = 2: cells 101-150
IMAT = 3: cells 151-200

X (1)

Figure 12-6 Outline of conditional cell attachment operation

Step 2
• Assign a material property reference no. to each fluid domain using command
CTABLE. For the model shown in Figure 12-6:

For domain no. 1 (IMAT = 1)

CTAB 1 FLUID 3 0 1 1
↑ ↑ ↑ ↑ ↑ ↑
cell cell colour porosity material group
index type index reference property number
number reference
number

CSET,NEWS,CRAN,1,100 (collect together all cells with property


ref. no. 1)
CTYPE,1 (change the currently active cell type to
1)
CMOD,CSET

For domain no. 2 (IMAT = 2)

CTAB,10,FLUID,4,0,2,2

12-24 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Cell Attachment and Change of Fluid Type

CSET,NEWS,CRAN,101,150 (collect together all cells with property


ref. no. 2)
CTYPE,10
CMOD,CSET
For domain no. 3 (IMAT = 3)

CTAB,20,FLUID,5,0,3,3
CSET,NEWS,CRAN,151,200 (collect together all cells with property
ref. no. 3)
CTYPE,20
CMOD,CSET

• Define the monitoring cell and pressure reference for each material type using
the MONITOR and PRESSURE commands:

For domain no. 1

PMAT 1
MONI,20 (define the monitoring cell)
PRES,1.0E05,10 (define the reference cell and reference
pressure)
STATUS

For domain no. 2

PMAT 2
MONI,120
PRES,1.0E05,110
STATUS

For domain no. 3

PMAT 3
MONI,170
PRES,1.0E05,180
STATUS

Step 3
• Define an event step data file using the EVFILE command (see Figure 12-6):

EVFILE,INITIAL,case.evn (initialise the event data file)

• Perform an initial Attachment and Change Fluid operation for relevant


boundary pairs (otherwise they will be treated as detached and the attachment
boundary type will become equivalent to a wall). For example, to connect
region nos. 1 and 2:

EVSTEP,1,TIME,0.0 (event step no. 1 occurs at time t = 0)


EAMATCH,1,2 (connect regions 1 and 2)

Version 4.02 12-25


ROTATING AND MOVING MESHES Chapter 12
Cell Attachment and Change of Fluid Type

• Change the fluid material property reference number in region 2 to that in


region 1

EFLUID,1,ADD,CRANGE,101,150
(or EFLUID,1,ADD,CTYPE,10)
(or EFLUID,1,ADD,GROUP,2)
• List the latest definitions and save the information supplied

ECLIST,CFLUID,ALL (list all cells of type ‘Change Fluid’)


EVSAVE,1 (save this information as event no. 1)

• If, at time t = 1., region no. 2 is to be cut off from the rest of the flow, issue the
following commands:

EVSTEP,2,TIME,1.
EDETACH,ADD,REGION,1 (add region no. 1 to the ‘detach’ set)
(or EDETACH,ADD,BRAN,1,2)
EDLIST,ALL (list all detached boundary pairs)
EFLUID,2,ADD,CTYPE,10
EVSAVE,2

Note that the detached boundary set definitions in an event step can be
deleted, if necessary, with command EDDELETE and remaining definitions
re-numbered via command EDCOMPRESS.

Step 4
• If it is to be assumed that the valve between boundary regions 3 and 4 opens
when the average pressure in region 4 is greater than that in region 3, set up a
conditional event as follows:

EVCND,3
EAMATCH,3,4 (attach region nos. 3 and 4)
EFLUID,1,ADD,CTYPE,20 (change all cells with cell id. 20 to fluid
no. 1)
EVSAVE,3

• Enable conditional attachment in an actual event

EVSTEP,4,TIME,2.
ECONDITIONAL,3,ENABLE (enable conditional event no. 3)
EVSAVE,4

Step 5
• Define all other events required. Note that:
1. An event step can be
(a) deleted, if necessary, with command EVDELETE;
(b) modified with command EVGET;

12-26 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Cell Attachment and Change of Fluid Type

(c) listed on the screen with command EVLIST.

2. Command EVUNDELETE restores a previously deleted event step.


3. User-specified offsets can be applied to the actual event time via command
EVOFFSET.

Step 6
• If using the method described in Chapter 5, “Load-step based solution mode”,
define the load step for the transient run.
• Check the validity of specified events and prepare the events data file for
subsequent use via command EVPREP.
• Save the problem’s data files using commands GEOMWRITE,
PROBLEMWRITE, etc. or their equivalent GUI operations accessible from the
File menu.

Note that the events data file can be


• written in coded form to a (.evnc) file with command EVWRITE, typically
in order to transfer data to another computer
• read in coded form from a (.evnc) file with command EVREAD, typically
when transferring data from another computer

Step 7
Exit from pro-STAR and then run STAR from your session’s X-window, as
described in Chapter 2, “Running a STAR-CD Analysis”, Step 6.
Step 8
Post-processing the data. For example, the commands needed to process time step
no. 10 are:

SUBTITLE
Results at time step 10
Velocity field
EVFI CONN case.evn (connect the event file)
TRLOAD case.pstt (load the transient post data file)
STORE ITER 10 (the appropriate events are loaded and
executed automatically)
GETC ALL (get the cell data)
POPT VECT
PLTY NORM
CSET NEWS FLUID
CPLOT
QUIT,NOSAVE

Be very careful not to save problem information to file case.mdl as the current
geometry corresponds to the state of the mesh at time step no. 10.
Useful points
1. At time t = 0, cell pairs are detached. They become attached only when an
event containing EATTACH or EAMATCH commands is executed. Once
Version 4.02 12-27
ROTATING AND MOVING MESHES Chapter 12
Mesh Region Exclusion

attached in this way, they remain attached until another EATTACH or


EDETACH command references them, or they are deactivated.
2. When the model’s mesh is being created, it is very useful to set up a regular
boundary numbering scheme at the interface as this simplifies the
specification of cell attachment.

Mesh Region Exclusion


Basic concepts
A group of cells can be excluded from the solution domain by defining an ‘exclude’
event and issuing command EECELL. Note that:
• This is possible only if the cells in the group are not connected to any other
cells in the model. Thus, the group must first be detached from the rest of the
model using a cell detachment event, as described in the section on “Cell
Attachment and Change of Fluid Type”.
• Only active cells can be excluded.
• There are no other restrictions on the cells that may be excluded (e.g. more
than one adjacent layers may be removed at a time).

An important difference with respect to cell deactivation, discussed in the section


on “Cell-layer Removal/Addition”, must also be noted. The mass contained in
excluded cells is removed from the solution; by contrast, the mass in the deactivated
cells is ‘squeezed out’ into the neighbouring cells.

Moving Mesh Pre- and Post-processing


Introduction
The various mesh motions and connectivity changes caused by the execution of
event-type commands can be visualised and verified using special pro-STAR
facilities. These help both in setting up the events (pre-processing) and in examining
the results of the analysis (post-processing). The same facilities can also be used
during the actual solution run, in combination with mesh changes caused by event
execution. Note that mesh changes can be classified into
• geometry changes
• connectivity changes

Geometry changes should occur only as a result of the EGRID event. All other
events can only cause connectivity changes.
Event processing is useful at three different stages of flow modelling and serves
the following requirements:
1. Pre-processing
Here the emphasis is on:
(a) Testing out different event combinations.
(b) Checking out commands read in by EGRID.
(c) Making corrections as needed and re-executing the events.
(d) Working with incomplete events.
(e) Testing out parts of events, e.g. to see if cells to be attached are adjacent

12-28 Version 4.02


Chapter 12 ROTATING AND MOVING MESHES
Moving Mesh Pre- and Post-processing

to each other.
(f) Using events to generate future events, e.g. use EGRID commands to
move the mesh and then EAMATCH to define the attach pairs.

2. Solution run
Here, STAR calls up pro-STAR to alter the grid in some way.
3. Post Processing
By this stage, the mesh geometry applicable to any given point in time is
available from the actual solution. Therefore, the goal here is to generate
vertex data for various flow variables (via command CAVERAGE), display
them using the correct surface and edge plotting options and create particle
tracks. Some error checking capabilities are also needed to detect event errors
which may have previously gone unnoticed. These detected errors are
highlighted in the plots.
Action commands
Commands EVLOAD and EVEXECUTE belong to this category.
EVLOAD is used to ‘load’ all events up to a specified point in time. There are two
basic components involved in this operation:
• Creation of internal tables defining the current status of each cell. These tables
can then be used by command CSET via keywords ACTIVE, DEACTIVE or
ATTACHED. For example, CSET,NEWSET,ACTIVE creates a cell set of the
currently active cells.
• Execution of any grid-changing commands read in by EGRID.

Note that, in general, application of EVLOAD results only in changes to the mesh
geometry and not to the mesh connectivity. The various options of the EVLOAD
command deal with different ways of specifying the current time. There is also a
‘reset’ option which restores the geometry to the ‘original state’, as defined below.
The first time EVLOAD is called, the ‘original state’, i.e. the vertex, cell and
boundary definitions of the model, are saved. Command EVLOAD,RESET restores
the model to this original state. If the model is changed at this point, the next
EVLOAD command will create fresh ‘original state’ files that correspond to the
changes.
Command EVEXECUTE should be used only after a successful EVLOAD
operation. This command applies the current status, stored in the internal tables
mentioned above, to the mesh. Thus:
• Cells marked as ‘deactivated’ are deleted (equivalent to command CDELETE)
and vertex numbers on adjacent cells are changed to reflect their new
connectivity.
• Cells marked as having changed material type are changed to a different cell
type.
• Vertices on the common face between two cells marked as ‘attached’ will be
merged.

The end result of the above is changes to cell connectivity due to cell removal.
Using option OFF with command EVEXECUTE restores the model connectivity to
the ‘original state’ defined by EVLOAD. The internal status tables also retain their

Version 4.02 12-29


ROTATING AND MOVING MESHES Chapter 12
Moving Mesh Pre- and Post-processing

original setting. A succeeding EVLOAD command also implicitly performs an


EVEXECUTE,OFF operation.
Status setting commands
Commands EVFLAG, EVCHECK and PLATTACH belong to this category. EVFLAG
and EVCHECK modify the behaviour of EVLOAD. Any subsequent plotting is
controlled by the PLATTACH options.
Command EVFLAG can be used to selectively turn on or off different types of
events loaded by EVLOAD. It contains two groups of parameters that can be set
independently, one for pre-processing and the other for post-processing. The option
specified with command EVCHECK (PREP or POST) determines which of the two
groups is to be set. The EVLOAD components that can be selectively turned on or
off are:
1. COND — executes enabled conditional events
2. UPARM — calls user subroutine UPARM
3. GRID — processes grid change commands
This option is essential if EVLOAD is to be used for changing the mesh
geometry when pro-STAR is called by STAR. For example, suppose the
following commands are read in by EGRID:

.....
EVFLAG,PRE,OFF,GRID (if the GRID flag is not set to OFF, the
EVLOAD command that follows will
cause EGRID commands to be executed
repeatedly and ad infinitum)
EVLOAD,UPTO,TIME,TIME (Note the use of the predefined parameter
TIME)
CSET NEWS ACTIVE
VSET NEWS SURFACE
VSMOOTH
.....

4. NEWXYZ — calls user subroutine NEWXYZ


5. DEACTIVE — checks that deactivated cells have zero volume. If they do not,
the error is reported and EVLOAD is stopped.
6. ACTIVE — checks that active cells have non-zero volume. If they do not, the
error is reported and EVLOAD is stopped.
7. ATTACH — checks that cell faces to be attached have coincident vertices. If
they do not, the error is reported and EVEXECUTE is stopped. Note that this
particular option only applies to EVEXECUTE.
8. NEWSET — creates a set of cells which fail any tests during EVLOAD.
9. SCDEF — creates scratch files containing the initial mesh state. This option
may be turned off whenever there is no need to backtrack in time, for example
when EVLOAD is called from STAR. This saves CPU time and disk space,
which may be considerable for large models.

Finally, command PLATTACH controls the plotting of attached faces. When it is set
to ON, attached faces are treated like internal faces and thus are not displayed on
any surface plots.
12-30 Version 4.02
Chapter 13 OTHER PROBLEM TYPES
Multi-component Mixing

Chapter 13 OTHER PROBLEM TYPES


Multi-component Mixing
The theory behind flow problems of this kind and the manner of implementing it in
STAR-CD is given in the Methodology volume (Chapter 16, “Multi-component
Mixing”). The present chapter contains an outline of the process to be followed
when setting up problems involving multiple species and includes cross-references
to appropriate parts of the on-line Help system. The latter contains details of the user
input required and important points to bear in mind when setting up problems of this
kind.
Setting up multi-component models
Step 1
Go to the Thermophysical Models and Properties folder in the STAR-GUIde
system and open the “Additional Scalars” sub-folder. Set up a scalar variable for
each species participating in the fluid mixture. The properties of each scalar are
specified in the “Molecular Properties (Scalar)” panel, in two ways:
1. By choosing option Define scalar material and then typing in values
yourself. Clicking Defaults instructs pro-STAR to fill the remaining boxes
with default values (those of air).
2. By choosing option Select scalar from database (see topic “Fluid Property
Database”). pro-STAR then fills in all the required values using data stored in
file props.dbs.

It is important that definition of all material (stream) properties via panel


“Molecular Properties” has already been completed before any scalar properties are
defined.
In multi-stream flow problems, a scalar can be present in some streams but not
in others, or it can be present in more than one stream. The allocation of scalar
variables to streams is entirely up to the user, subject to the following conditions:
• Each scalar must be defined only once.
• Some scalar physical properties are stream-independent and must be set
when the scalar is first defined. These include molecular weight, specific heat,
molecular viscosity and thermal conductivity.
• Other properties such as diffusivity and turbulent Schmidt number are
stream-dependent and must be set on a stream-wise basis (see Step 3 below)
Step 2
Once all scalars are defined, scroll through them one by one via the Scalar # scroll
bar at the bottom of the panel to
• check all property values in the “Molecular Properties (Scalar)” panel
• modify a current value by overtyping in the relevant text box; the change is
made permanent by clicking Apply
• delete an unwanted scalar by clicking Delete Scalar.

Step 3
Specify the stream-dependent (or material-dependent) scalar properties using the
Version 4.02 13-1
OTHER PROBLEM TYPES Chapter 13
Multi-component Mixing

“Binary Properties” panel. Once the settings for all scalars in a given stream are
complete, click Apply and then move on to the next stream in your model.
Step 4
Specify values for the initial mass fraction of each scalar in each stream using the
“Initialisation” panel.
Step 5
If the stream incorporates porous media sub-domains (see Chapter 6 in this
volume), specify the effective mass diffusivity and turbulent Schmidt number for
each additional scalar present in your model using the “Additional Scalar
Properties” panel (“Porosity” sub-folder).
Step 6
Specify scalar boundary conditions using the “Scalar Boundaries” panel (Define
Boundary Conditions folder).
Step 7
Go to the “Analysis Controls” folder and specify solution control parameters for all
currently defined scalars using the “Additional Scalars” panel (Equation Behaviour
sub-folder). In multi-stream problems where each stream has a different scalar
composition, this panel enables you, in effect, to select which scalars exist in what
stream.
Step 8
If a transient analysis is to be performed, use the “Analysis Output” panel
(“Transient tab”) to specify whether cell and/or wall data for selected scalars need
to be printed or written to the transient post file.
For transient problems defined in terms of load steps, go instead to the Advanced
Transients dialog (see Chapter 5, “Load step controls”) and click one of the Scalars
Select buttons. The button to click depends on whether cell or wall data are needed
and whether these are to be printed or written to the transient post file. The scalars
to be printed or post-processed are selected in the Transient Scalar Selection dialog
shown below, by clicking the option button corresponding to the desired scalar
number.

Command: SCTRANS

13-2 Version 4.02


Chapter 13 OTHER PROBLEM TYPES
Aeroacoustic Analysis

Note that this process should be repeated for every load step in the transient setup.
Step 9
If the stream incorporates additional sources for any of the scalars, specify the
source strength and distribution using the Scalar tab in the “Source Terms” panel
(sub-folder Sources).
Useful points on multi-component mixing
1. The under-relaxation factors for all scalar transport equations should be set to
the same value. Note that this factor has no effect for scalars calculated by an
internal method or by user coding.
2. For thermal problems, the scalar under-relaxation factors should equal that for
the energy equation. For combusting or reacting flows, the recommended
range is 0.3 to 0.7.
3. For efficient utilisation of computer memory, it is recommended that scalar
variable numbers are continuous and start at 1.
4. For problems involving large changes in temperature, it is recommended that
the specific heat of both background fluid and active species is defined as a
polynomial function of temperature (see reference [1]). For scalars, this can
be done in the “Polynomial Function Definition (Viscosity and Conductivity)”
dialog that opens from the “Molecular Properties (Scalar)” panel. A
polynomial variation for molecular viscosity and thermal conductivity can be
specified in the same way. An ideal-gas variation for the density is also
recommended, if necessary with a compressible setting.
5. pro-STAR allows new scalar species to be added to its built-in property
database (see topic “Fluid Property Database” in the on-line Help system).
6. Details of existing scalar definitions can be saved to a file of form case.scl
for use in other problems. To do this, issue command CDSCALAR from
pro-STAR’s I/O window. Note that the scalar data are written in the form of
appropriate pro-STAR commands (SC, SCPROPERTIES, SCCONTROL,
etc.). Thus, it is possible to read them back into a model by executing an
IFILE command (see “File manipulation” on page 17-9).
7. STAR uses default wall functions for calculating heat and mass transfer at
wall boundaries. Users can supply alternative expressions for heat and mass
transfer coefficients in subroutine MODSWF, activated via the “Miscellaneous
Controls” STAR-GUIde panel.

Aeroacoustic Analysis
The theory behind aeroacoustic analysis and the manner of its implementation in
STAR-CD is given in the Methodology volume (Chapter 16, “Aeroacoustic
Analysis”). The present section contains an outline of the process to be followed
when setting up a problem of this type. Also included are cross- references to
appropriate parts of the on-line Help system, containing details of the user input
required.
Setting up aeroacoustic models
Step 1
Switch on the aeroacoustic modelling facility using STAR GUIde’s “Select

Version 4.02 13-3


OTHER PROBLEM TYPES Chapter 13
Aeroacoustic Analysis

Analysis Features” panel:


• Select On from the Aeroacoustic Analysis menu
• If a transient analysis mode has already been selected, a pop-up panel will
appear, warning you that the model must be run in steady-state mode. Click
Yes to confirm your choice and proceed with the analysis. Note that the
displayed option in the Time Domain menu will automatically change to
Steady State.
• Click Apply. Note that an additional folder called Aeroacoustic Analysis will
now appear in the NavCenter tree.
Step 2
Open the “Aeroacoustic Analysis” panel. By default, the Aeroacoustic Equation
Sources switch is turned On. The default control parameters required for the
numerical solution algorithms are also set and are explained by the on-line Help
text. If you wish to make any changes, enter the required values in the panel and
then click Apply.
Step 3
Perform the usual model setup in the Thermophysical Models and Properties folder:
In particular, make sure that:
• A density option appropriate to incompressible flow is selected in the
“Molecular Properties” panel
• A two-equation, k-ε type turbulence model has been selected in the
“Turbulence Models” panel
Step 4
Specify initial conditions, boundary conditions and control parameters and then run
STAR as normal, making sure that the analysis has converged. The aeroacoustic
results will be automatically stored in the solution (.ccm) file as an extra scalar
variable called AALS (Aeroacoustic Lilley Source). If the maximum number of
iterations is reached without convergence, it is important to restart the analysis and
run it to convergence.
Step 5
Use the facilities of the Post-Processing folder to load and display the distribution
of the AALS variable, using only cell-based or vertex-based values
Useful points on aeroacoustic analyses
1. If you require an initial solution without the overheads of calculating
aeroacoustic source terms at the last iteration, simply turn the Aeroacoustic
Equation Sources switch Off, click Apply, and then perform the analysis as
usual. You will then need to restart the analysis, turn the switch On and
perform one iteration to obtain the aeroacoustic results.
2. Note that STAR-CD returns the logarithmic values of the aeroacoustic
sources. If you want to display the actual values, you will first need to
calculate the antilogarithm of the stored scalar using the facilities of the Post
Register Operations dialog (see Chapter 13, “The OPERATE utility” in the
Post-Processing User Guide).

13-4 Version 4.02


Chapter 13 OTHER PROBLEM TYPES
Liquid Films

Liquid Films
The theory behind the liquid film model and details of its implementation in
STAR-CD is given in Chapter 16, “Liquid Films” of the Methodology volume. This
section contains an outline of the steps to be followed when setting up a liquid film
simulation.
Setting up liquid film models
The liquid film model can only be used in transient cases. Simulations employing
this feature typically involve droplet deposition on wall boundaries, formation of
liquid films and film interaction with the surrounding fluid and walls.
The basic steps for setting up such a model are as follows:
Step 1
Open the “Select Analysis Features” panel in STAR GUIde and turn On the Liquid
Films option. A pop-up panel may appear, warning you that the model must be run
in transient mode. In such a case, click Yes to confirm your choice. The Time
Domain menu setting will then change to Transient.
Click Apply. The Liquid Films folder will appear in the NavCenter tree,
containing the necessary panels for liquid film analysis.
Step 2
If necessary, allow for the presence of droplets in your model by selecting option
Lagrangian Multi-Phase from the Multi-Phase Treatment menu and clicking
Apply. The Lagrangian Multi-Phase folder will then appear in the NavCenter tree,
containing panels needed for specifying droplet parameters (see Chapter 9, “Setting
Up Lagrangian Multi-Phase Models” in the CCM User Guide). This option should
be selected if either
• droplets are injected into the solution domain and their behaviour needs to be
modelled as part of the analysis, and/or
• droplets are generated by the film itself through a stripping process.

Step 3
The Liquid Films folder will contain a set of four panels called Film Controls, Film
Physical Models and Properties, Film Initialization and Film Boundaries. Note that
film property specifications under the second panel of the above set must be
supplied even if there are no films initially present in the problem.
• The “Film Controls” panel sets up the basic film modelling parameters. The
panel also includes a Liquid Film Creation facility that enables you to specify
which (wall or baffle) boundary regions cannot support liquid films.

• The “Film Physical Models and Properties” panel activates the liquid film
model for specified film materials and sets up a property table for each of
them. Note that:
(a) There is a one-to-one correspondence between film materials defined in
the “Film Models” tab and fluid domain materials defined in the
“Molecular Properties” panel. Films created for and corresponding to
(gaseous) materials in different domains are topologically separate and

Version 4.02 13-5


OTHER PROBLEM TYPES Chapter 13
Liquid Films

their liquid contents do not mix with each other.


(b) In the “Film Properties” tab, you can use the Evaporates to Scalar entry
to specify which liquid film component evaporates to/condenses from
which gas component. To determine which droplet component becomes
which liquid film component when a droplet hits a wall, the following
rules are used:
i) If a liquid film component and a droplet component evaporate to the
same scalar, these components are assumed to exchange mass.
ii) If the Evaporates to Scalar setting for a liquid film component is
NONE, then this component name is compared against each droplet
component name (for single-component droplets, the droplet name
is taken as the component name). If the component names match,
the matched components are assumed to exchange mass.

• In multi-component liquid film simulations:


(a) The specified single value of binary diffusivity is assumed to apply to all
components in the film mixture
(b) For problems involving evaporation from the film surface, the partial
pressure of each component on the gas side of the interface should be
calculated using subroutine LQFPRO (see the “Multi Component” on-line
Help topic)

• The “Film Initialization” specifies film initial conditions for each boundary
region that can support films. If no initialization is specified, the film
thickness on that region is assumed to be zero.

• The “Film Boundaries” panel sets up film boundary conditions. Each


boundary condition is applied to the edges shared by a film and a non-film
region. The currently available boundary condition types are Outlet and
Inlet. In the latter case, if no boundary conditions are specified for a given
variable, the cell value is used as an inlet value (i.e. a Neumann condition
applies).

Step 4
Specify initial conditions, boundary conditions and solution control parameters for
the domain material (normally gas) surrounding the film and then run the STAR
solver as normal. Due to internal parameter settings and having to work with
possibly very small numbers such as film thickness, it is recommended that the
solver be run in double precision.
Step 5
Analysis results pertaining to films are treated by pro-STAR as wall data. Such data
items appear in the scroll lists of panel “Analysis Output”, in both the “Post tab” tab
and the “Transient tab” tab, so that you may select what is to be included in the
.ccm and .pstt files, respectively. pro-STAR assigns names such as LFTHK
(film thickness) and LFT (film temperature) to film variables. A complete list can
be obtained by issuing the PLIST command from the I/O window. Assuming that
the contour plot mode is already selected, a typical pro-STAR macro to plot a scalar
film variable is:

13-6 Version 4.02


Chapter 13 OTHER PROBLEM TYPES
Liquid Films

trlo,,
store last
clrw
getb lfthk
cset news shell
cset subs name wall
wplot

In conjugate heat transfer cases, the following commands may be useful for
selecting only liquid film cells on the fluid side of the interface:

cset newset fluid


cset add atsh
cset subs name wall

To load film velocity components (LFU, LFV, LFW) as a vector, use

getb lu,lv,lw

Film stripping
This process can be modelled in two ways:
1. Via user subroutine FDBRK. If active, the subroutine will be called at all wall
faces containing films, at a point just before the first droplet tracking stage in a
new time step. The user code must provide all necessary information
regarding the new (stripped) droplets leaving the film, including initial
injection velocity and global position coordinates.
2. Via an internal stripping model, currently available as a beta feature (see
Chapter 13 of the Supplementary Notes volume).

If droplets are generated solely by the stripping process, it is still necessary to define
droplet properties in advance, as for normal injected droplets. The new (stripped)
droplets must have a type associated with them, which has previously been defined
in pro-STAR. Obviously, droplet properties should be consistent with those of the
parent film.

Version 4.02 13-7


Chapter 14 USER PROGRAMMING
Introduction

Chapter 14 USER PROGRAMMING


Introduction
This chapter describes how the user can modify or supplement some of the standard
features and operations of STAR, such as physical properties, boundary conditions,
additional sources of momentum, energy, etc. via user-supplied FORTRAN
subroutines. The latter are collectively referred to as UFILE routines. The full set
of currently available user programming inputs comprises:
1. Boundary conditions
2. Density (equation of state)
3. Molecular viscosity (including non-Newtonian flow)
4. Specific heat
5. Temperature to enthalpy conversion and vice versa
6. Thermal conductivity
7. Molecular diffusivity for chemical species
8. Properties of distributed resistance
9. Thermal and mass diffusion within distributed resistance sub-domains
10. Effective viscosity and turbulence length scale
11. Turbulence model parameters (including two-layer models)
12. Turbulence characteristics within distributed resistance sub-domains
13. Local injection or removal of fluid
14. Momentum, enthalpy and turbulence sources
15. Solar and gaseous radiation properties
16. Free surface and cavitation models and properties
17. Heat, mass and momentum transfer in two-phase Lagrangian flow
18. Droplet initial conditions and physical properties
19. Droplet behaviour near walls
20. Inter-droplet collision modelling
21. Eulerian multi-phase drag, turbulence and heat transfer
22. Chemical reaction rates and chemical species mass fractions
23. Chemical species and thermal NOx sources
24. Parameters for sliding mesh and rotating reference frame problems
25. Moving mesh coordinates
26. Cell layer removal or attachment
27. Initial conditions
28. Formation and behaviour of liquid films on walls and baffles
29. Wall functions for momentum, heat and mass transfer
30. Time-step size for transient problems
31. Special post-processing
32. Variation of blending factor for higher-order discretisation schemes

Subroutine Usage
To use UFILE routines you must execute the following steps:

Step 1
Create a subdirectory called ufile under your present working directory as

Version 4.02 14-1


USER PROGRAMMING Chapter 14
Subroutine Usage

follows:
• Choose File > System Command from the menu bar to display the System
Command dialog
• Type ufiles in the command text box
• Click Apply and then Close

Step 2
Select the User option in the appropriate STAR GUIde panel or pro-STAR
command, depending on the special feature that needs to be modelled, as discussed
in “Description of UFILE Routines” on page 14-5.
Step 3
Before a user routine can be used, it must be copied into its own individual file
within the ufile directory created earlier. If you are doing this from scratch, it is
convenient to start by copying a skeleton (dummy) version of the relevant
subroutine into ufile.
• If you want to do this immediately, click Define user coding in your current
panel. A file of the right name containing the right dummy subroutine will be
created automatically.
• If you want to inspect the dummy subroutine listing before proceeding
further, go to the main pro-STAR window and select Utility > User
Subroutines from the menu bar. This activates the User Subroutines dialog
shown below. The dialog box is made up of two sub-windows. The lower one
lists all subroutine names, their description and the pro-STAR command that
activates them. Selecting any line with the mouse displays the default
(dummy) code for that subroutine in the upper part of the box. The relative
size of the two sub-windows can be adjusted by dragging the control ‘sash’
(the small square on the right-hand side) up and down.

14-2 Version 4.02


Chapter 14 USER PROGRAMMING
Subroutine Usage

Command: USUBROUTINE

The required subroutine(s) may be copied into the ufile directory in one of the
following ways:
1. Automatically — click the Write Auto button. This copies all subroutines
already selected implicitly via the User option in the various STAR GUIde
panels (or via the corresponding pro-STAR commands). Such subroutines are
also marked in the above list with an asterisk. Note that if more selections are
made after the above dialog box has been opened, it is necessary to update the
display of selected routines by clicking the Update List button.
2. Explicitly — click the Write File button. This copies the subroutine that is
currently on view.

In Unix systems, the subroutine file names are of the form Usubname.f. If a file
of the same name already exists in the ufile subdirectory, a new file will be
created called Usubname.f.new. Note that generating a subroutine file in this
way is necessary only if
• the subroutine is to be set up for the first time, or
• an existing subroutine is to be replaced, or
• you are updating user code from an earlier version of STAR-CD.

Version 4.02 14-3


USER PROGRAMMING Chapter 14
Subroutine Usage

Step 4
Edit the existing or newly-created subroutine file as required, for example by using
pro-STAR’s built-in file editor (see the section on “File manipulation” on page
17-9). This is necessary in order to either
• utilise some or all of the existing example coding (by removing the comment
character, C, from the beginning of the line), or
• add other coding, as appropriate.

Step 5
The version of a subroutine that is to be used in the current run should always be
located in a file called Usubname.f within the ufile subdirectory. Older files
bearing the same name should either be overwritten or renamed.

Once the above process is complete, the required user routines are automatically
passed on to the STAR-CD system in source form. They are then compiled and
linked to the main program modules (see Chapter 17, “pro-STAR environment
variables”). Note that STAR will issue a warning message if it does not find any of
the required subroutines but will carry on with the run all the same.
Useful points
As a general rule, user routines should be written with due care. You should ensure
that results produced by user code are reasonable and physically meaningful, by
implementing suitable checks and by printing appropriate diagnostic messages
whenever necessary. Default user routines for all modelling functions listed in the
“Introduction” are supplied, containing sample coding. It should be noted that:
1. Most routines are called for every cell, boundary, or droplet (as appropriate
for the routine and model in hand), so a penalty is paid in terms of execution
time when they are active. However, the increase in CPU time may be
minimised through efficient programming, while keeping the source coding
as brief and simple as possible.
2. Each routine has appropriate input data, described in a nomenclature text
stored in file nom.inc in the ufile directory.
3. Each routine includes a file called comdb.inc, designed to ensure that the
routine uses the same precision as STAR itself. This is done by exploiting the
IMPLICIT typing construct present in FORTRAN. According to this, a
variable is given a type based upon its initial letter, those beginning with the
letters A through H and O through Z being REAL variables, while those
beginning with I through N are INTEGER variables. Thus, TIME, ANGLE and
SPEED are real but NUMI, IVAL and JUNK are integer.
The IMPLICIT typing above can be overruled by an explicit declaration
of type, e.g. REAL ITIME makes ITIME real and INTEGER ZVAL will
make ZVAL an integer. It is also possible to change the scope of the
IMPLICIT typing. This is in fact what comdb.inc does:
(a) When STAR is used in single-precision runs, the file contains a single line

C IMPLICIT DOUBLE PRECISION (A-H,O-Z)

14-4 Version 4.02


Chapter 14 USER PROGRAMMING
Description of UFILE Routines

which is just a comment, thus preserving the standard implicit typing of


real and integer variables.
(b) When STAR is used in double-precision runs, the file reads

IMPLICIT DOUBLE PRECISION (A-H,O-Z)

This means that the IMPLICT typing has been overruled to use
double-precision real variables.

The implication for users is that to make sure a routine works correctly,
variables should be named according to the IMPLICT typing shown above.
That way the routine will be compiled with the correct precision.

Typical input data for a subroutine includes the following:


• Cell number
• Global Cartesian or user-defined local coordinates of the cell centroid
• Cell table numbers as defined in pro-STAR
• Material numbers
• Porous media sub-domain numbers
• Iteration number
• Time
• Nodal values of the field variables

For more information on input data for the UFILE routines, see the nomenclature
file (nom.inc). The variables in the argument list are never passed uninitialised:
they always have a sensible value, which is usually the value from the previous
iteration/time step, if applicable, or more generally the “default” value from the
pro-STAR panel.
A brief description of each subroutine and how it is activated from pro-STAR is
given in the next section.

Description of UFILE Routines


Boundary condition subroutines
The first ten of the subroutines listed below (all those with names starting with BCD)
are activated from the Options menu in the Define Boundary Regions panel, or by
command RDEFINE. They specify spatial variations of the boundary conditions at
various boundary types. In order to use them, the boundaries comprising the region
are first defined in the usual way, including the local coordinate system for the
velocity components, the rotational speed of the coordinate frame and any default
boundary values that become input values for the subroutines. The coordinates
passed to the subroutine are defined in the local coordinate system of the boundary
and u, v, w are the corresponding velocities. The latter will be in a rotating frame if
this was originally specified. The transformation to the global Cartesian coordinate
system is done by STAR.

BCDEFI Specifies distributions for all dependent variables that vary spa-
tially over an inlet boundary.

Version 4.02 14-5


USER PROGRAMMING Chapter 14
Description of UFILE Routines

BCDEFO Can be used to specify variations in flow split or mass outflow at


outlet boundaries (e.g. in a transient run).

BCDEFP Specifies boundary conditions at pressure boundaries, i.e. pres-


sure, turbulence intensity, length scale, temperature and species
mass fractions.

BCDNRP Specifies boundary conditions at non-reflective pressure bounda-


ries

BCDEFS Specifies boundary conditions at stagnation boundaries

BCDNRS Specifies boundary conditions at non-reflective stagnation bounda-


ries

BCDEFW Specifies variations in wall boundary conditions, including mov-


ing wall velocities in local coordinates and in a rotating reference
frame. In addition, wall temperature, chemical species mass frac-
tion and heat and mass fluxes, can all be varied over the specified
region.

BCDEFF Specifies non-uniform boundary conditions at free-stream trans-


missive boundaries, e.g. velocity components, pressure and tem-
perature.

BCDEFT Specifies boundary conditions at transient wave transmissive


boundaries, e.g. velocity components, pressure and temperature.

BCDEFR Specifies boundary conditions at Riemann invariant boundaries,


e.g. velocity components, pressure and temperature.

ROUGHW Activated from the Roughness menu in the Define Boundary


Regions panel for walls and baffles, or by command RDEFINE. It
specifies a user-supplied wall roughness model, in problems where
wall functions are used for modelling flow near the wall. STAR
will default to the smooth-wall behaviour should you activate this
subroutine but provide no code for it.

Material property subroutines

CONDUC Activated from the Conductivity menu in the Molecular Properties


(Liquids and Gases) panel or Material Properties (Solids) panel,
or by command CONDUCTIVITY. It specifies the thermal conduc-
tivity within a material in heat transfer problems. The thermal con-
ductivity can vary both spatially and with temperature.

14-6 Version 4.02


Chapter 14 USER PROGRAMMING
Description of UFILE Routines

CONVET Activated from the Specific Heat menu in the Molecular Proper-
ties (Liquids and Gases) panel or by command SPECIFICHEAT.
The activation works in an exclusive manner, i.e. choosing this
option excludes use of subroutines CONVTE, COTEET and
SPECHT. It supplies the variation of temperature T with enthalpy h
and any other scalar variable, i.e. T ( h , m 1 , m 2 , … ) , in any way
chosen by the user (e.g. analytically or by means of a table). The
returned values are valid over a specified temperature range. If the
relationship involves other scalar variables, it is necessary to sup-
ply values for the partial derivatives ∂T ⁄ ∂h and ∂T ⁄ ∂m k . STAR
also requires the inverse relationship, h ( T , m 1 , m 2 , … ) , for inter-
nal calculation purposes and inverts T automatically, using an effi-
cient iterative technique.

CONVTE Activated from the Specific Heat menu in the Molecular Proper-
ties (Liquids and Gases) panel or by command SPECIFICHEAT.
The activation works in an exclusive manner, i.e. choosing this
option excludes use of subroutines CONVET, COTEET and
SPECHT. It supplies the variation of enthalpy h with temperature T
and any other scalar variable, i.e. h ( T , m 1 , m 2 , … ) , in any way
chosen by the user (e.g. analytically or by means of a table). The
range of validity of the relationship should be specified in terms of
a corresponding range in the values of T. If enthalpy is dependent
on a scalar variable, it is also necessary to supply the relevant par-
tial derivatives ∂h ⁄ ∂m k . STAR needs the inverse relationship,
T ( h , m 1 , m 2 , … ) , for internal calculation purposes and inverts h
automatically using an efficient iterative technique. It is helpful
(but not essential) to assist the iteration process by supplying
∂h ⁄ ∂T .

COTEET Activated from the Specific Heat menu in the Molecular Proper-
ties (Liquids and Gases) panel or by command SPECIFICHEAT.
The activation works in an exclusive manner, i.e. choosing this
option excludes use of subroutines CONVET, CONVTE and
SPECHT. It supplies two relationships:
(a) The variation of enthalpy h with temperature T and any other
scalar variable, i.e. h ( T , m 1 , m 2 , … ) , and
(b) the variation of temperature T with enthalpy h and any other
scalar variable, i.e. T ( h , m 1 , m 2 , … ) .

These should be valid over a given temperature range. Obviously,


the two relationships must be consistent. If additional scalar varia-
bles are involved, it is also necessary to supply the relevant partial
derivatives ∂h ⁄ ∂m k . The COTEET option should be used if the
user wants to bypass STAR’s internal calculation procedure for the
inverse temperature/enthalpy relationship (see the CONVET,
CONVTE description above) in favour of a supplied relationship.

Version 4.02 14-7


USER PROGRAMMING Chapter 14
Description of UFILE Routines

DENSIT Activated from the Density menu in the Molecular Properties


(Liquids and Gases) panel or by command DENSITY. It supplies
equations of state for density calculations that are not included in
the standard options. For compressible flow cases where density is
a function of pressure, the routine must also specify the partial
derivative ∂ρ ⁄ ∂ p and return it in parameter DENDP.

DIFFUS Activated from the Material Mass Diffusivity menu in the Binary
Properties (Additional Scalars) panel or by command DIFFU-
SIVITY. It supplies the molecular diffusivity of the background
material in multi-component mixing problems.

PORCON Activated from a menu in the Thermal Properties (Porosity) panel


or by command POREFF. It supplies functions for the calculation
of effective thermal conductivity and turbulent Prandtl number
within a distributed resistance sub-domain.

PORDIF Activated from a menu in the Additional Scalar Properties (Poros-


ity) panel or by command SCPOROUS. It supplies functions for
the calculation of effective mass diffusivity and turbulent Schmidt
number within a distributed resistance sub-domain.

PORKEP Activated from a menu in the Turbulence Properties (Porosity)


panel or by command PORTURBULENCE. It specifies non-uni-
form distributions of turbulence intensity and dissipation length
scale within a distributed resistance sub-domain.

POROS1 Activated from the Resistance Coefficients menu in the Resistance


and Porosity Factor panel or by command POROSITY. It defines
spatially varying coefficients α and β within a distributed resist-
ance sub-domain. The user can also specify them in terms of a
local coordinate system.

POROS2 Activated from the Resistance Coefficients menu in the Resistance


and Porosity Factor panel or by command POROSITY. It defines
the resistance components ( k 1 , k 2 , k 3 ) directly instead of via the
resistance coefficients α and β. This facility is a useful alternative
way of specifying a non-linear variation of porous resistance with
velocity. For this purpose, the global Cartesian velocity compo-
nents are supplied to the subroutine.

14-8 Version 4.02


Chapter 14 USER PROGRAMMING
Description of UFILE Routines

SPECHT Activated from the Specific Heat menu in the Molecular Proper-
ties (Liquids and Gases) panel or Material Properties (Solids)
panel, or by command SPECIFICHEAT. The activation works in
an exclusive manner, i.e. choosing this facility excludes use of
subroutines CONVTE, CONVET and COTEET. The subroutine sup-
plies the variation of fluid or solid mean specific heat with temper-
ature and other quantities, at constant pressure. It is particularly
useful in modelling combusting or reacting flows exhibiting sub-
stantial variation in the value of this property. STAR calculates the
temperature T from the iterative expression

(n) h
T = ---------------------
(n – 1)
- (14-1)
(cp)

where n is the iteration number and c p is the mean specific heat.

THDIFF Specifies a user-supplied method of calculating the thermal diffu-


sion coefficient for chemical species scalars (see Chapter 5, “Sub-
routine THDIFF Set-up” in the Supplementary Notes volume)

VISMOL This subroutine is activated from the Molecular Viscosity menu in


the Molecular Properties (Liquids and Gases) panel or by com-
mand LVISCOSITY. It can specify an arbitrary distribution of
molecular viscosity, but its principal use is for supplying functions
that describe non-Newtonian viscous behaviour.

Turbulence modelling subroutines

LSCALE Activated automatically when the k-l model is selected via menu
option k-l in panel Turbulence Models (Turbulence tab). It can also
be activated by command TURBULENCE. The subroutine supplies
the spatial variation of dissipation length scale (l) required by the
k-l model.

TWLUSR Activated from the Two-Layer Model menu in the Turbulence


Models panel (Near-Wall Treatment tab) or by command
TLMODEL. It defines the user’s own formulation of turbulent
behaviour in problems using a two-layer model.

VISTUR This subroutine is activated from panel Turbulence Models (Tur-


bulence tab) or by command TURBULENCE. The subroutine spec-
ifies the turbulent viscosity distribution for a turbulent flow
calculation.

Version 4.02 14-9


USER PROGRAMMING Chapter 14
Description of UFILE Routines

Source subroutines

FLUINJ Activated from the Define Source menu in the Source Terms panel
(Mass tab). Alternatively, use command RSOURCE. The subrou-
tine initiates fluid injection or removal at specified cells and at a
prescribed rate (in units of kg/s/m3). In the case of injection, the
properties of the injected fluid, i.e. velocity components, turbulence
parameters, temperature, etc. must also be prescribed. This is not
required when fluid is removed.

SORENT Activated from the Define Source menu in the Source Terms panel
(Enthalpy tab) or by command RSOURCE. It specifies additional
enthalpy sources or sinks due, for example, to electric or magnetic
fields, chemical or nuclear reaction and thermal radiation. It can
also fix the temperature value within a cell by making
S1P=GREAT* T fix and S2P=GREAT, where T fix is the desired
fixed temperature value and GREAT is a large number used inter-
nally by pro-STAR.

SORKEP Activated from the Define Source menu in the Source Terms panel
(Turbulence tab) or by command RSOURCE. It allows the user to
redefine the source term components for the k and ε equations, e.g.
to account for special effects due to streamline curvature, magnetic
fields, etc. The subroutine can also be used to fix the value of k.
Note that the quantities S1P and S2P in the example code are the
‘standard’ source and sink terms given in the Methodology vol-
ume. Thus the user, in modifying or supplementing the standard
expressions, effectively replaces the built-in source terms.

SORMOM Activated from the Define Source menu in the Source Terms panel
(Momentum tab) or by command RSOURCE. It enables the model-
ling of additional momentum source terms, for example due to
magnetic or electric fields. The source terms must be specified per
unit volume and linearised as S1P-S2P* φ P , where φ P is the value
of the velocity component in question at node P (see the Method-
ology volume for details). The two components S1P and S2P must
be separately specified for the U, V and W momentum equations.
The cells in which to insert these sources can be selected by their
indices IP, global Cartesian coordinates XP, YP, ZP or the cell
table number ICTID.

14-10 Version 4.02


Chapter 14 USER PROGRAMMING
Description of UFILE Routines

SORSCA Specifies additional source terms for the scalar variable equations
and is activated from one of the following locations:
(a) The Define Source menu in the Source Terms panel (Scalar
tab) or by command SCSOURCE. The source terms might
consist of, for example, the chemical kinetics and rate expres-
sions of a combustion process.
(b) The Model Selection menu in panel Cavitation Model or by
command CAVITATION. In this case the source terms are
used to specify a special cavitation model.
The mass fraction value at selected cells can also be fixed via the
source terms, in the same manner as that described above for
enthalpy.

Radiation modelling subroutines

RADPRO Activated from the Radiative Properties menu in panel Thermal


Models (Liquids and Gases) when radiation with participating
media is turned on. May also be activated from the Radiative
Properties (Solids) panel if solid-fluid heat transfer is turned on.
Alternatively, use command RADPROPERTIES. It specifies
non-uniform distributions of absorptivity and scattering coeffi-
cients within the medium filling the space between radiating
boundaries.

RADWAL Specifies a user-supplied method of calculating radiative proper-


ties for solid walls (see Chapter 6, “Surface Properties” in the Sup-
plementary Notes volume)

USOLAR Activated from the Define Parameters menu in the Thermal


Options panel (Solar Radiation section) or by command SOLAR.
In transient problems, it enables specification of solar angle and
intensity at every time step of the analysis.

Free surface / cavitation subroutines

CAVNUC This subroutine is required only in cavitation problems using the


bubble two-phase model. It is activated from the Parameters for
BTF Model menu in panel Cavitation Model or by command
CAVNUCLEI. It specifies the number of bubble nuclei per cubic
metre and a functional relationship between equilibrium radius
and cell pressure.

Version 4.02 14-11


USER PROGRAMMING Chapter 14
Description of UFILE Routines

CAVPRO This subroutine is needed in cavitation or free surface problems


requiring variable properties. It is activated from one of the follow-
ing locations:
(a) The Saturation Pressure menu in panel Cavitation Model or
by command CAVPROPERTY. It then specifies the speed of
sound in the current material (for both the liquid and vapour
phases) and the saturation vapour pressure.
(b) The Saturation Property Variation menu in panel Mass Trans-
fer (Free Surface folder) or by command VAPORIZATION. It
then specifies the vaporisation properties of the current mate-
rial (saturation temperature and vapour pressure plus latent
heat of vaporisation).

COMDEN Calculates species density and its derivative with respect to pres-
sure and temperature for compressible free-surface flows (see
Chapter 1 of the Supplementary Notes volume)

FSEVAP Activated from the Vaporization Rate menu in the Mass Transfer
panel (Free Surface folder). Alternatively, use command VAPOR-
IZATION. It calculates the vaporization rate in problems involv-
ing mass transfer by evaporation across a free surface.

FSTEN Activated from the Additional Properties menu in the Heavy Fluid
Molecular Properties panel (Free Surface or Cavitation folders).
Alternatively, use command STENSION. It calculates values for
surface tension coefficient and contact angle in free surface and
cavitation problems.

Lagrangian multi-phase subroutines

COLLDT Activated from the Collision Model menu in panel Droplet Physi-
cal Models and Properties (tab Global Physical Models) or by
command DCOLLISION. It specifies the method of detecting
inter-droplet collisions in transient Lagrangian flow problems.

COLLND Activated from the Collision Model menu in panel Droplet Physi-
cal Models and Properties (tab Global Physical Models) or by
command DCOLLISION. It specifies the method of calculating
the droplet number density used for collision modelling in tran-
sient Lagrangian flow problems.

DRAVRG Activated from the Droplet Averaging menu in the Droplet Con-
trols panel or by command DRAVERAGE. It supplies information
about average droplet properties calculated while tracking a drop-
let parcel through the solution domain.

14-12 Version 4.02


Chapter 14 USER PROGRAMMING
Description of UFILE Routines

DRHEAT Activated from the Heat Transfer menu in panel Droplet Physical
Models and Properties (tab Droplet Physical Models) or by com-
mand DRHEAT. It enables the user to define the heat transfer proc-
ess between droplets and the surrounding carrier fluid in
two-phase Lagrangian flow problems.

DRMAST Activated from the Mass Transfer menu in panel Droplet Physical
Models and Properties (tab Droplet Physical Models) or by com-
mand DRMASS. It enables the user to define the mass transfer
process between droplets and the surrounding carrier fluid in
two-phase Lagrangian flow problems.
This subroutine can also be used for specifying mass transfer
between a droplet component and multiple scalars in the surround-
ing medium. This is done by first selecting the component in the
scroll list of the Droplet Properties tab and then typing the key-
word User in the Evaporates to Scalar box. Alternatively, use
command DRCMPONENT.

DROBRK Specifies a user-supplied droplet break-up model (see Chapter 10


of the Supplementary Notes volume)

DROICO Activated from the Droplet User Subroutine (Lagrangian


Multi-Phase) or by command DRUSER. The subroutine enables
the user to specify droplet initial conditions for two-phase,
Lagrangian flow problems. In transient problems, the subroutine
sets the initial conditions for any calculation time step at which
parcels are released.

DROMOM Activated from the Momentum Transfer menu in panel Droplet


Physical Models and Properties (tab Droplet Physical Models) or
by command DRMOMENTUM. It enables the user to calculate
momentum transfer between droplets and the surrounding carrier
fluid in two-phase Lagrangian flow problems.

DROPRO Enables the user to specify any physical property appearing in


panel Droplet Physical Models and Properties (tab Droplet Prop-
erties). It is activated by selecting the Subroutine Usage button
next to any of the properties displayed on the tab, or by command
DRPROPERTIES.

DROWBC Activated from the Droplet-Wall Interaction menu in panel Drop-


let Physical Models and Properties (tab Droplet Physical Models)
or by command DRWALL. It enables the user to calculate momen-
tum, heat, and mass exchange between droplets and wall bounda-
ries.

Version 4.02 14-13


USER PROGRAMMING Chapter 14
Description of UFILE Routines

Liquid film subroutines

FDBRK Specifies the method of calculating liquid film stripping by the


carrier fluid. It is activated by selecting User in the Stripping and
Re-entrainment menu of the Film Controls panel or by command
LFSTRIP.

LQFBCD Specifies boundary conditions at liquid film inlets. It is activated


by selecting User in the Film Boundaries panel or by command
LQFBC.

LQFINI Specifies initial conditions for all liquid film variables at a given
boundary region. It is activated by selecting User in the Options
menu of the Film Initialization panel or by command LQFINI-
TIAL.

LQFPRO Specifies liquid film physical properties or liquid film component


partial pressures. It is activated by selecting the Subroutine Usage
button next to any of the properties displayed on the Film Proper-
ties tab (panel Film Physical Models and Properties), or by com-
mand LQFPROPERTY.

LQFSOR Modifies the source terms of the mass, momentum and enthalpy
equations for liquid films. It is activated by selecting User in the
User Defined Source Term menu of the Film Controls panel or by
command LFQSOR.

Eulerian multi-phase subroutines

UEDRAG This subroutine is used in Eulerian multi-phase problems to calcu-


late the total drag force, per unit volume of the computational cell.
It is activated from the main menu in the Drag Forces panel (Eule-
rian Multi-Phase folder) or by command EDRAG.

UETURB This subroutine is employed in Eulerian multi-phase problems to


calculate the response coefficient C t . The latter is used to derive
the dispersed phase turbulence characteristics from those of the
continuous phase. It is activated from the Ct Model menu in the
Turbulence Models panel (Multiphase Options tab) or by com-
mand ETURB. Note that the drag force per unit volume referred to
above is supplied as an input variable since it is often a parameter
in C t formulations.

14-14 Version 4.02


Chapter 14 USER PROGRAMMING
Description of UFILE Routines

UEHEAT This subroutine is employed in Eulerian multi-phase problems to


calculate the Nusselt number. The latter is then used in the calcula-
tion of the mean interface heat transfer coefficient, which in turn is
used to compute the interphase heat transfer when solving for
energy for either phase. The subroutine is activated from the Inter-
phase Heat Transfer panel (Eulerian Multi-Phase folder) or by
command EHTRANSFER.

Chemical reaction subroutines

COALC Specifies a user-supplied char combustion model in coal combus-


tion cases (see Chapter 3, “User Coding” in the Supplementary
Notes volume)

COALV Specifies a user-supplied volatile evolution model in coal combus-


tion cases (see Chapter 3, “User Coding” in the Supplementary
Notes volume)

FULPRO Specifies user-defined fuel physical properties and chemical reac-


tion parameters for use with the Shell ignition and knock models.
It can be activated in two ways:
(a) From the Ignition Reaction Based On menu in panel Ignition
(folder Chemical Reactions). Alternatively, type command
IGNMODEL.
(b) From the Knock Reaction Based On menu in panel Knock
(folder Chemical Reactions). Alternatively, type command
KNOCK.

NOXUSR Activated by the Thermal NOx, Prompt NOx, or Fuel NOx menus
in panel Emissions (Chemical Reactions folder), or by command
NOX. It contains user coding for the calculation of thermal, prompt
or fuel NOx sources.

PARUSR Specifies a user-supplied particle component evolution model in


coal combustion cases (see Chapter 3, “User Coding” in the Sup-
plementary Notes volume)

RATUSR Specifies a user-supplied method of calculating turbulence effects


in complex chemistry models (see Chapter 4, “RATUSR User
Subroutine” in the Supplementary Notes volume)

REACFN Activated from the Rate Model menu in the Reaction System
(Chemical Reactions) panel when option Combined/User is cho-
sen as the current reaction model. Alternatively, type command
RRATE. It specifies a user-supplied reaction rate for chemical
reactions of any type.

Version 4.02 14-15


USER PROGRAMMING Chapter 14
Description of UFILE Routines

REACUL Specifies a user-supplied reaction rate for the Coupled Complex


Chemistry model. It is activated from panel Reaction System, hav-
ing previously selected option User from the Reaction Rate Calcu-
lated by menu in panel Scheme Definition (folder Chemical
Reactions). Alternatively, type command CRMODEL.

SCALFN In some circumstances, chemical species mass fractions can be


calculated from user-prescribed algebraic relationships, e.g. stoi-
chiometric relationships, rather than from finite-volume transport
equations. These algebraic relationships can be specified in this
routine, activated from the Solution Method menu in panel Addi-
tional Scalars (Solution Controls > Equation Behavior
sub-folder). Alternatively, use command SCPROPERTIES.

Rotating reference frame subroutines

UOMEGA Calculates values of angular velocity (omega) for problems involv-


ing rotating reference frames. It is activated by the User Option
menu in the Rotating Reference Frames panel or by command
SPIN.

UPOSTM Generates post-processing data at coupled boundaries. It is used in


problems with multiple rotating frames of reference that are solved
explicitly. The subroutine is called automatically in the Rotating
Reference Frames panel if option Multiple Frames - Explicit is
selected from the Reference Frame Treatment menu. Alternatively,
use command MFRAME.

Moving mesh subroutines

NEWXYZ Activated by selecting Modules > Transient from the main


pro-STAR menu to open the Advanced Transients dialog, and then
selecting On in the Moving Grid Option menu. Alternatively, use
command MVGRID. The subroutine specifies the cell vertex coor-
dinates at a new time. The old time level coordinates are available
in the VCORN array and must be overwritten with new coordi-
nates. The sample coding supplied describes a moving mesh that is
linearly expanding and contracting between a reciprocating piston
and a fixed cylinder head; the piston is driven by a rotating crank
mechanism.

UASI Specifies the time-varying offsets used in matching arbitrary slid-


ing interface (ASI) boundaries. It is called automatically if a
model employing sliding events is defined using command EVS-
LIDE.

14-16 Version 4.02


Chapter 14 USER PROGRAMMING
Description of UFILE Routines

UBINIT Specifies initial conditions for cells that are re-incorporated into
the solution domain via an INCLUDE event. It is called automati-
cally if command EICOND is issued in a model employing such
cells.

UPARM Generates parameters required for moving meshes. It is called


automatically if a moving mesh model is defined using commands
in the EVENTS module; i.e. if command

MVGRID,ON,EVENT,PROSTAR

is issued.

Miscellaneous flow characterisation subroutines

INITFI Activated from the Values menu in the Initialization panel (Liquids
and Gases or Solids folders). Alternatively, use command INI-
TIAL. It initialises flow field variables to user-specified values.
These values override any constant values also appearing in those
panels. During an initial field restart, the subroutine can also be
used to selectively reset some of the variable values in the field.
Note that the subroutine returns velocities in a local coordinate
system. STAR transforms them to a stationary global Cartesian
system. Velocities in this system will differ from the velocities
produced by the subroutine because of this transformation and,
when that feature is active, the transformation from a rotating ref-
erence frame.

MODSWF Activated by a button labelled Heat and Mass Transfer in the Mis-
cellaneous Controls (Other Controls) panel, or by command
HCOEFF. It modifies or supplies new wall functions for heat and
mass transfer. This is useful, for example, in problems involving
strong natural convection where the standard formulae for the
transfer coefficients might be inaccurate. One such example is
included in the sample coding. Mean temperatures and mass frac-
tions for all fluid materials are made available through the parame-
ter list.

PORHT2 Specifies user-supplied coefficients for a quasi-linear relationship


between porous solid and fluid temperatures in problems involv-
ing conjugate heat transfer in porous media (see Chapter 16,
“Inter-phase heat transfer term” in the Supplementary Notes vol-
ume).

Version 4.02 14-17


USER PROGRAMMING Chapter 14
Description of UFILE Routines

VARPRT Enables specification of either a variable Prandtl number for


enthalpy or a variable Schmidt number. It can be activated in two
ways:
(a) From the Prandtl(Enth). menu in panel Turbulence Models
(Turbulence tab). Alternatively, type command COKE.
(b) Via the Schmidt Number pop-up menu in the Binary Proper-
ties (Additional Scalars) panel, or by typing command
SCPROPERTIES. In this case, the subroutine should supply
special functions for calculating the turbulent Schmidt
number of chemical species in multi-component mixing prob-
lems.

Solution control subroutines

DTSTEP Enables the user to specify a variable time step for transient, sin-
gle-transient or pseudo-transient simulations. It can be activated in
three ways:
(a) For single-transient cases, select option User in the Time Step
Method menu of the Set Run Time Controls panel (Analysis
Preparation/Running folder). Alternatively, use command
DELTIME
(b) For pseudo-transient cases, select option User in the Time
Step Option menu of the Set Run Time Controls panel (Analy-
sis Preparation/Running folder). Alternatively, use command
TIME.
(c) For transient cases, open the Advanced Transients dialog,
select the appropriate load step, and then click the User Flag
button in front of the time step (Delta Time) box. Alterna-
tively, use command LSTEP.

The subroutine can be used, for example, in fire and smoke move-
ment simulations that involve a large, concentrated heat source.
The time step can be adjusted in terms of the number of PISO cor-
rectors and maximum Courant number. Note that STAR does not
alter the number of time steps in a load step, so your code must
ensure that the time step lengths are such that the length of the
load step is correct.

14-18 Version 4.02


Chapter 14 USER PROGRAMMING
Sample Listing

POSDAT Activated by the User subroutine button in the Analysis Output


(Output Controls) panel or by command PRFIELD. It performs
special post-processing operations. For example:
(a) Variable values at several monitoring locations can be written
to user-designated output files for subsequent processing.
(b) A bulk averaging scheme can be prescribed for selected flow
variables and printed at specified intervals.
(c) Calculation of lift and drag coefficients.

This subroutine may be called both at the beginning and at the end
of every time step or iteration. The place from which it is called is
distinguished by the value of parameter LEVEL (=1 — beginning,
=2 — end)

VARBLN Activated by the Blending Method pop-up menus in the Primary


Variables panel (Differencing Schemes tab). It can be used to vary
the blending factor for higher-order discretisation schemes over
the computational domain. Alternatively, use command
DSCHEME.

Sample Listing
The listing for subroutine CONDUC is given below as an example of the default
source code available in STAR-CD. Users wishing to inspect the contents of any
other subroutine should start a pro-STAR session and then activate the User
Subroutines dialog, as explained in “Subroutine Usage” on page 14-1.
C*************************************************************************
SUBROUTINE CONDUC(CON,CKNX,CKNY,CKNZ)
C Conductivity
C*************************************************************************
C--------------------------------------------------------------------------*
C STAR-CD VERSION 4.00.000
INCLUDE ’comdb.inc’
C
COMMON/USR001/INTFLG(100)
C
INCLUDE ’usrdat.inc’
DIMENSION SCALAR(50)
EQUIVALENCE( UDAT12(001), ICTID )
EQUIVALENCE( UDAT11(001), CP )
EQUIVALENCE( UDAT11(002), DEN )
EQUIVALENCE( UDAT11(003), ED )
EQUIVALENCE( UDAT11(006), P )
EQUIVALENCE( UDAT11(007), T )
EQUIVALENCE( UDAT11(008), TE )
EQUIVALENCE( UDAT11(009), SCALAR(01) )
EQUIVALENCE( UDAT11(059), U )
EQUIVALENCE( UDAT11(060), V )
EQUIVALENCE( UDAT11(061), W )
EQUIVALENCE( UDAT11(062), VISM )
EQUIVALENCE( UDAT11(063), VIST )

Version 4.02 14-19


USER PROGRAMMING Chapter 14
New Coding Practices

EQUIVALENCE( UDAT11(067), X )
EQUIVALENCE( UDAT11(068), Y )
EQUIVALENCE( UDAT11(069), Z )
C
C-----------------------------------------------------------------------
C
C This subroutine enables the user to specify thermal conductivity.
C
C CON - Isotropic conductivity or MAX(CKNX,CKNY,CKNZ) if the
C conductivity is anisotropic
C CKNX - Anisotropic conductivity in x-direction
C CKNY - Anisotropic conductivity in y-direction
C CKNZ - Anisotropic conductivity in z-direction
C
C
C STAR calls this subroutine for cells and boundaries.
C
C ** Parameters to be returned to STAR: CON,CKNX,CKNY,CKNZ
C
C-----------------------------------------------------------------------
C
C Sample coding: To specify thermal conductivity for a group of
C cells with cell table numbers 2 and 11 as a function
C of temperature
C
C IF (ICTID.EQ.2.OR.ICTID.EQ.11) CON=4.3+0.001*T
C-------------------------------------------------------------------------C

RETURN
END
C

New Coding Practices


Most standard STAR-CD V3.2X user coding will work without any modification.
The table below explains the differences between V3.2X and V4.00 user routines.
Note that STAR-CD V3.2X / V3.1X common blocks / variable names should not
be used.

User routine Difference


Freedom to modify AL, AV, TSAT and HVAP
CAVPRO
has been removed
For the Free Surface model, freedom to modify
DENSIT the heavy/light density has been removed.
IFLUTYP and DENDT are no longer present

14-20 Version 4.02


Chapter 14 USER PROGRAMMING
New Coding Practices

a/ The DRCOMP common block needs to be


included to the source, e.g.
COMMON /DRCOMP/ NDRCOM_max
b/ The DEMUCO array needs to be dimen-
DROWBC sioned as DEMUCO(NDRCOM_max,*)

For more information, please consult the


default source code for DROWBC supplied
with the STAR-CD installation.
POSDAT Data structure is different
For the Free Surface model, only the mixture
SPECHT Cp needs to be specified. IFLUTYP is no
longer present
For the Free Surface model, only the mixture
VISMOL viscosity needs to be specified. IFLUTYP is no
longer present

Users are advised to consult subroutine POSDAT to gain familiarity with


STAR-CD’s new face-based data structure. Commonly used V3.2X data items that
are now accessed differently in V4.00 are listed below.

V3.2X data item Equivalent V4.00 data item


See material/and loop in posdat.f.
KEY doma(nd)%mattyp.eq.FLUID test tells
you whether you have a fluid or a solid.
LQ, LCU, LCY, LCO, LSI, LX lfc
S, SB, SBSI sv
WF, WFCU, WFCY, WFSI w
VOLF, VOLCU, VOLCY, VOLSI svol
AC, COCU, COCY, COSI,
af
COCO, ACB
F, FB, FBSI fl
KEYSUB doma(nd)%sd(nsd)%irot
T(1,IP) t(ip)
T(scalar index +1, IP) c(IP,scalar index), see posdat.f

Version 4.02 14-21


USER PROGRAMMING Chapter 14
User Coding in parallel runs

User Coding in parallel runs


If user coding is present in a parallel run, it is possible that some of the required
operations need access to flow field values that are distributed throughout the
various computational domains. In such cases, it is necessary to collect such values
prior to manipulation and to do this, the supplied coding needs to use special
message passing routines.
The example shown below is an extract from user subroutine NEWXYZ and it
employs a parallel function called IGSUM to find the global number of active cell
layers in an engine simulation problem.

NLIVE=0
ICELL1=15904
ICELLEND=62209
NCOF=1029
DO I=ICELL1,ICELLEND,NCOF
CALL LIVCLL(I,ISTAT)
IF (ISTAT.EQ.1) NLIVE=NLIVE+1
ENDDO
c. NHPC > 1 if parallel run
IF(NHPC.GT.1) NLIVE=IGSUM(NLIVE)

A synopsis of the available message passing-routines is given in Appendix E. These


routines should only be called if required; they are not necessary for sequential runs.
To aid diagnostics, four variables are provided via file usrdat.inc to user
subroutines:

IHPC — this is the local process number (1 ≤ IHPC ≤ NHPC)


IHPC = 1 for a sequential analysis
IHPC = 1 for the ‘master’ process in STAR HPC
IHPC > 1 for the ‘slave’ process in STAR HPC

NHPC — Number of processes (NHPC = 1 for a sequential analysis)

NHHPC — Number of fluid ‘halo’ cells on the local process


NTHHPC — Total number of ‘halo’ cells on the local process (fluid plus solid)

14-22 Version 4.02


Chapter 15 PROGRAM OUTPUT
Introduction

Chapter 15 PROGRAM OUTPUT


Introduction
Run-time screen output from STAR provides a summary of the input specification
for the problem being solved and also allows monitoring of the calculation progress.
It is therefore important for users to understand this information and examine it
regularly to ensure that
• the problem has been correctly set up;
• the calculations are proceeding satisfactorily.

The amount of information displayed is largely up to the user, apart from a core of
information that is always produced. The various checks and outputs which are
specially activated from pro-STAR’s STAR GUIde environment (Output Controls
folder) are described below, along with the permanent output.

Permanent Output
The core-level screen information from STAR can be divided into two sections:
1. An echo of the input data provided by the user
2. A display of analysis results and information on the progress of STAR
calculations

Input-data summary
As can be seen in Table 15-1 on page 15-4, the input summary begins with the
STAR-CD version number and the date and time of the run. This is followed by a
table of essential model data for checking that all important user-defined inputs are
correct. All listed data reflect the values stored in the problem (.prob) file.
The table is divided into distinct sections, as follows:
General Data
This section provides general information on the problem at hand, including:
• The case name
• Number of cells
• The model’s overall physical dimensions
• The run precision (single or double)

This section of the table also summarises:


• The character of the flow (i.e. steady or transient)
• The starting iteration number for the calculations
• The frequency of solution data and screen output
• The solution algorithm and linear equation solver selected
• The residual tolerance used for convergence tests
• The maximum number of iterations or time steps specified

A sample output can be seen in Section A of Table 15-1.

Version 4.02 15-1


PROGRAM OUTPUT Chapter 15
Permanent Output

Fluid Properties
This section comprises one or more tables containing the properties of all fluid
domains (streams) included in the model. For each domain, the information
supplied includes:
• The variables calculated, including the turbulence model selected and the
appropriate characteristic length
• The physical properties specified, such as density, viscosity, specific heat, and
conductivity
• The pressure/temperature reference locations for the domain
• The reference pressure and temperature (when appropriate)
• Any fixed boundary fluxes included in the model
• The specified initial field values
• The specified boundary conditions

An example of the output for a multi-domain case appears in Table 15-1, Section B.
This shows data for two fluid domains with different physical properties.
Solid Properties
The fluid domain tables are followed, in the case of solid-fluid heat transfer
problems, by tables of properties for solid domains such as density, specific heat
and conductivity. This can be seen in Table 15-1, Section C. The reference
temperature, initial field values and boundary conditions are also included here.
Additional Features
In this part of the table, information is provided on any additional features that are
active in the model, such as:
• Radiation
• Free surface
• Run-specific system settings

The sample output of Table 15-1, Section D, indicates use of memory-based scratch
files and platform-specific solver optimization.
User FORTRAN Coding
This section of the table only appears when user-defined FORTRAN coding is
active during the calculations. The sample case presented in Table 15-1 does not use
this option.
Solution Parameters
This last section of the table deals with the settings for the control parameters used
by the numerical algorithm, such as
• relaxation factors,
• type of differencing scheme used,
• the corresponding blending factors,
• residual normalising factors for each fluid and solid domain,
• solver tolerances,
• sweep limits.

For transient PISO runs the printout of relaxation factors is suppressed as irrelevant,
except for the pressure correction relaxation factor. A typical printout of the above

15-2 Version 4.02


Chapter 15 PROGRAM OUTPUT
Printout of Field Values

parameters can be seen in Table 15-1, Section E.


Run-time output
The run-time output that provides information on the progress of the calculations at
each iteration or time step can be seen in Table 15-2 on page 15-6 and is arranged
in two sections:
1. The left-hand section contains the global absolute residual histories for each
group of transport equation solved (momentum, mass, turbulence, etc.).
2. The right-hand section contains values of the corresponding dependent
variables (velocity magnitude, pressure, turbulent viscosity, etc.) at a
pre-defined monitoring location in domain no. 1.

In steady-state runs, satisfactory progress of the calculations should show


• a steady reduction in the global absolute residuals from iteration to iteration;
• stabilisation of the values of flow field variables at the monitoring location.

However, residuals do not always decrease from iteration to iteration and, in some
cases, oscillations can be observed. These can be ignored as long as the overall
residual levels are reduced over a reasonable number of iterations.
Information on total CPU and elapsed times is also given. This output appears on
the screen during an interactive session and is also saved in the run-time output
(.run) file. Any warning messages generated during the course of the calculations
are stored in the run-time optional output (.info) file and should be inspected by
the user separately. The .run file also contains a reminder to the effect that
warning messages have been produced.

Printout of Field Values


The printout of field values for the solution variables is optional and, if present,
follows the analysis history output. The output quantity and frequency is left up to
the user and may be set using various options in the “Analysis Output” STAR
GUIde panel — see the “Print Cell Range” section for steady-state problems. There
is also a similar “Print Cell Range” section in the transient-problem version of this
panel (“Post tab”). An example of such a printout can be seen in Table 15-3 on page
15-7.

Optional Output
All additional outputs are optional and, if requested, will appear in the .info file.
Output of additional data is activated by various options in the “Monitor Numeric
Behaviour” STAR GUIde panel. Table 15-4 on page 15-8 shows typical information
appearing in this file (in this case up to and including data for iteration no. 1). Note
that velocity component data at the flow field’s extrema are given in the local
coordinate system. This is in contrast to the data described in Chapter 17, “Data
repository file (.ccm)”, which are always stored in the global Cartesian coordinate
system.

Version 4.02 15-3


PROGRAM OUTPUT Chapter 15
Example Output

Example Output
Table 15-1:
|-------------------------------------------|
| STAR-CD VERSION 4.00.000 |
| THERMOFLUIDS ANALYSIS CODE |
| Operating System: Linux |
| Stardate: 6-DEC-2005 Startime: 14:32:39 |
|-------------------------------------------|

|-----------------------------------------------------------|
| STAR Copyright (C) 1988-2005, Computational Dynamics Ltd. |
| Proprietary data --- Unauthorized use, distribution, |
| or duplication is prohibited. All rights reserved. |
|-----------------------------------------------------------|

|-------------------------------------------------------------------------------------------|
| ---------------------------- PROBLEM SPECIFICATION SUMMARY ---------------------------- |
|-------------------------------------------------------------------------------------------|
| CASE TITLE .................. => |
| NUMBER OF CELLS ............. => 192 |
| MESH DIMENSIONS XMIN XMAX YMIN YMAX ZMIN ZMAX |
| (IN METRES) ............ => 0.0E+00 6.0E-01 0.0E+00 8.0E-01 0.0E+00 5.0E-02 |
| MESH QUALITY ................ => |
| Expansion factor .......... => Aver = 1.00, Max = 1.00 (CVs: 43, 44) |
| Non-orthogonality (deg).... => Aver = 0.00, Max = 0.00 (CVs: 0, 0) |
| RUN PRECISION ............... => Single |
| STEADY ANALYSIS ..............=> START FROM ITERATION = 0 |
| INITIALISATION .............. => WILL NOT BE EMPLOYED |
A
| DATA DUMP (FILE.ccm) ........ => EVERY 10 ITERATIONS |
| SOLUTION PROCEDURE .......... => SIMPLE |
| RESIDUAL TOLERANCE .......... => 1.00E-03 |
| MAX. NO. OF ITERATIONS ...... => 100 |
| RESTART DATA ................ => WILL BE SAVED ON out.ccm |
| SURFACE DATA ................ => WILL NOT BE SAVED |
| CONVERGENCE DATA ............ => WILL BE PRINTED ON FILE.info |
| FIELD DATA .................. => WILL BE PRINTED |
| LIN. ALG. EQU. SOLVER ....... => Conjugate gradient with Incompl. Choleski precond. |
|-------------------------------------------------------------------------------------------|
|-> DOMAIN 1: FLUID ------------------------------------------------------------------|
|-------------------------------------------------------------------------------------------|
| SOLVE ....................... => U, V, P, TE, ED, |
| FLUID FLOW .................. => TURBULENT INCOMPRESSIBLE |
| TURBULENCE MODEL ............ => HIGH RE K-EPS MODEL |
| CONSTANTS ................. => C_mu=0.09, C_1=1.44, C_2=1.92, C_3=1.44, C_4=-0.33 |
| => cappa=0.419, Pr_k=1.00, Pr_eps=1.219, Pr=0.90 |
| REFERENCE PRESSURE .......... => PREF = 1.000E+05 Pa |
| DENSITY ..................... =>
| MOLECULAR VISCOSITY ......... =>
IDEAL GAS: MOLW = 2.891E+01
CONSTANT - MU = 1.810E-05 Pas
|
|
B
| |
| INITIAL FIELD VALUES ........ => u v w p |
| => 0.0E+00 0.0E+00 0.0E+00 0.0E+00 |
| => Tur.In. Len.Sc. |
| => 2.6E-02 1.0E-01 |
| BOUNDARY CONDITIONS ......... => |
| Reg. 0 Wall: U = 0.000E+00 V = 0.000E+00 W = 0.000E+00 Om = 0.000E+00 in C.Sys. 1 |
| Elog = 9.000E+00 |
| Reg. 1 Inlet: U = 0.000E+00 V = 5.000E+01 W = 0.000E+00 Om = 0.000E+00 in C.Sys. 1 |
| TI = 5.000E-02 TLS = 5.000E-03 |
| Reg. 2 Constant piezomet. pressure: P = 0.000E+00 |
| TI = 5.000E-02 TLS = 5.000E-03 |
| Reg. 5 Symmetry plane |
|-------------------------------------------------------------------------------------------|
|-> DOMAIN 2: FLUID ------------------------------------------------------------------|
|-------------------------------------------------------------------------------------------|
| SOLVE ....................... => U, V, P, |
| FLUID FLOW .................. => LAMINAR INCOMPRESSIBLE |
| TURBULENCE MODEL ............ => |
| PRESSURE REF. CELL .......... => 145 |
B
| REFERENCE PRESSURE .......... => PREF = 1.000E+05 Pa |
| DENSITY ..................... => CONSTANT - RHO = 1.000E+03 kg/m3 |

15-4 Version 4.02


Chapter 15 PROGRAM OUTPUT
Example Output

| MOLECULAR VISCOSITY ......... => CONSTANT - MU = 1.000E-03 Pas |


| |
| INITIAL FIELD VALUES ........ => u v w p |
| => 0.0E+00 0.0E+00 0.0E+00 0.0E+00 |
| BOUNDARY CONDITIONS ......... => |
| Reg. 0 Wall: U = 0.000E+00 V = 0.000E+00 W = 0.000E+00 Om = 0.000E+00 in C.Sys. 1 |
| Elog = 9.000E+00 |
| Reg. 3 Inlet: U = 0.000E+00 V = 1.000E-01 W = 0.000E+00 Om = 0.000E+00 in C.Sys. 1 |
| Reg. 4 Outlet: Flow split = 1.000E+00 |
| Reg. 5 Symmetry plane |
|-------------------------------------------------------------------------------------------|
|-> DOMAIN 3: SOLID ------------------------------------------------------------------|
|-------------------------------------------------------------------------------------------|
| SOLVE ....................... => T |
| REFERENCE TEMPERATURE ....... => TREF = 2.730E+02 K |
| DENSITY ..................... => CONSTANT - rho = 9.000E+03 kg/m3 |
| SPECIFIC HEAT ............... => CONSTANT - c = 3.800E+02 J/kgK |
| CONDUCTIVITY ................ =>
| INITIAL FIELD VALUES ........ =>
CONSTANT -
T
k = 3.800E+02 W/mK |
|
C
| => 2.9E+02 |
| RELAX. FACT. IN SOLID ....... => URF = 1.00 |
| BOUNDARY CONDITIONS ......... => |
| Reg. 0 Wall: U = 0.000E+00 V = 0.000E+00 W = 0.000E+00 Om = 0.000E+00 in C.Sys. 1 |
| Reg. 5 Symmetry plane |
|-------------------------------------------------------------------------------------------|
|-> ADDITIONAL FEATURES USED --------------------------------------------------------------|
|-------------------------------------------------------------------------------------------|
| PLUG AND PLAY |
| RAMFILES OPTION ENABLED
| TURBO OPTION ENABLED
|
|
D
|-------------------------------------------------------------------------------------------|
|-> SOLUTION CONTROL PARAMETERS |
|-------------------------------------------------------------------------------------------|
| VARIABLE | Mome Mass Turb -- -- -- |
|-------------------------------------------------------------------------------------------|
| RELA. FAC. | 7.000E-01 2.000E-01 7.000E-01 -- -- -- |
| DIFF. SCH. |
| DSCH. FAC. |
UD
0.000E+00
CD
1.000E+00
UD
0.000E+00
--
--
--
--
--
--
|
|
E
| SOLV. TOL. | 1.000E-01 5.000E-02 1.000E-01 -- -- -- |
| SWEEP LIM. | 100 1000 100 -- -- -- |
|-------------------------------------------------------------------------------------------|

Version 4.02 15-5


PROGRAM OUTPUT Chapter 15
Example Output

Table 15-2:

Iter. I--------------- GLOBAL ABSOLUTE RESIDUAL ------------------I I-------- FIELD VALUES AT MONITORING POINT 63 ----------I
No Mome Mass Turb -- -- -- -- Vel Pres TurVis -- -- -- --
1 1.00E+00 1.00E+00 9.98E-01 -- -- -- -- 5.66E+01 6.20E-01 1.60E-03 -- -- -- --
2 6.78E-02 1.36E-01 3.01E-01 -- -- -- -- 5.46E+01 4.04E+00 8.05E-03 -- -- -- --
3 2.36E-02 3.95E-02 1.56E-01 -- -- -- -- 5.38E+01-2.15E+00 1.48E-02 -- -- -- --
4 1.26E-02 4.50E-02 9.22E-02 -- -- -- -- 5.32E+01-4.45E+00 1.94E-02 -- -- -- --
5 8.15E-03 3.10E-02 5.23E-02 -- -- -- -- 5.25E+01-3.93E+00 2.19E-02 -- -- -- --
6 6.02E-03 2.11E-02 2.85E-02 -- -- -- -- 5.18E+01-2.62E+00 2.32E-02 -- -- -- --
7 4.54E-03 1.64E-02 1.48E-02 -- -- -- -- 5.11E+01-1.32E+00 2.38E-02 -- -- -- --
8 3.41E-03 1.33E-02 7.31E-03 -- -- -- -- 5.06E+01-5.78E-01 2.40E-02 -- -- -- --
9 2.56E-03 1.13E-02 3.30E-03 -- -- -- -- 5.01E+01-4.11E-01 2.41E-02 -- -- -- --
10 1.90E-03 9.15E-03 1.79E-03 -- -- -- -- 4.98E+01-1.98E-01 2.40E-02 -- -- -- --
11 1.39E-03 7.66E-03 1.12E-03 -- -- -- -- 4.96E+01 2.35E-01 2.40E-02 -- -- -- --
12 1.00E-03 6.56E-03 7.23E-04 -- -- -- -- 4.94E+01 4.78E-01 2.39E-02 -- -- -- --
13 7.22E-04 5.37E-03 5.25E-04 -- -- -- -- 4.94E+01 6.69E-01 2.39E-02 -- -- -- --
14 5.25E-04 4.33E-03 3.95E-04 -- -- -- -- 4.93E+01 9.90E-01 2.38E-02 -- -- -- --
15 3.83E-04 3.48E-03 2.69E-04 -- -- -- -- 4.93E+01 1.08E+00 2.38E-02 -- -- -- --
16 2.81E-04 2.86E-03 1.78E-04 -- -- -- -- 4.93E+01 1.29E+00 2.38E-02 -- -- -- --
17 2.09E-04 2.39E-03 1.06E-04 -- -- -- -- 4.93E+01 1.34E+00 2.38E-02 -- -- -- --
18 1.56E-04 2.01E-03 6.37E-05 -- -- -- -- 4.92E+01 1.38E+00 2.38E-02 -- -- -- --
19 1.18E-04 1.73E-03 3.72E-05 -- -- -- -- 4.92E+01 1.41E+00 2.38E-02 -- -- -- --
20 9.05E-05 1.46E-03 2.25E-05 -- -- -- -- 4.92E+01 1.44E+00 2.38E-02 -- -- -- --
21 7.07E-05 1.22E-03 1.28E-05 -- -- -- -- 4.92E+01 1.45E+00 2.38E-02 -- -- -- --
22 5.52E-05 1.01E-03 8.80E-06 -- -- -- -- 4.92E+01 1.46E+00 2.38E-02 -- -- -- --
23 4.36E-05 8.36E-04 5.53E-06 -- -- -- -- 4.92E+01 1.50E+00 2.38E-02 -- -- -- --
&&&& --------------------------------------------------------------
--------------------------------------------------------------

**************************************************
* THERE ARE WARNINGS IN FILE out.info
**************************************************

*** CALCULATIONS TERMINATED - CONVERGENCE CRITERION SATISFIED

15-6 Version 4.02


Chapter 15 PROGRAM OUTPUT
Example Output

Table 15-3:
I-------------------------------------------- FIELD VALUES AT ITERATION NO 23--------------------------------------------I

CELL NO U VEL V VEL PRESS TUR EN DISSI VISCO DENSI


1 -3.697E-03 5.000E+01 1.635E+01 8.578E+00 8.004E+02 9.818E-03 1.187E+00
11 -1.021E-02 5.018E+01 6.546E+00 4.465E+00 2.369E+02 8.987E-03 1.187E+00
21 -2.619E-02 5.007E+01 1.337E+01 6.329E+00 4.550E+02 9.403E-03 1.187E+00
31 -2.700E-02 5.026E+01 1.697E+00 3.718E+00 1.673E+02 8.825E-03 1.187E+00
41 -4.814E-02 5.015E+01 9.038E+00 5.222E+00 2.980E+02 9.775E-03 1.187E+00
51 -5.854E-02 4.989E+01 1.610E+01 1.116E+01 5.849E+02 2.275E-02 1.187E+00
61 -5.772E-02 4.935E+01 3.980E+00 1.223E+01 6.708E+02 2.381E-02 1.187E+00
71 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 9.000E+03
81 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 9.000E+03
91 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 9.000E+03
101 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 9.000E+03
111 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 9.000E+03
121 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 0.000E+00 9.000E+03
131 3.276E-05 9.994E-02 -1.737E-03 0.000E+00 0.000E+00 0.000E+00 1.000E+03
141 2.712E-05 9.966E-02 -1.239E-02 0.000E+00 0.000E+00 0.000E+00 1.000E+03
151 2.454E-05 1.001E-01 -6.395E-03 0.000E+00 0.000E+00 0.000E+00 1.000E+03
161 6.541E-06 1.000E-01 -9.026E-04 0.000E+00 0.000E+00 0.000E+00 1.000E+03
171 1.097E-05 1.001E-01 -9.892E-03 0.000E+00 0.000E+00 0.000E+00 1.000E+03
181 6.675E-06 1.000E-01 -4.241E-03 0.000E+00 0.000E+00 0.000E+00 1.000E+03
191 -1.134E-05 1.001E-01 -9.797E-03 0.000E+00 0.000E+00 0.000E+00 1.000E+03

I--------------------------------------------| Wall data at iteration No 23 |--------------------------------------------I

Region No: 0
Cell No Y-PLUS NORM DIST WALL TEMP HTRAN HFLUX
49 2.858E+03 2.500E-02 2.930E+02 3.047E+01 0.000E+00
50 2.939E+03 2.500E-02 2.930E+02 3.124E+01 0.000E+00
51 2.999E+03 2.500E-02 2.930E+02 3.181E+01 0.000E+00
52 3.043E+03 2.500E-02 2.930E+02 3.223E+01 0.000E+00
53 3.075E+03 2.500E-02 2.930E+02 3.253E+01 0.000E+00
54 3.098E+03 2.500E-02 2.930E+02 3.274E+01 0.000E+00
55 3.114E+03 2.500E-02 2.930E+02 3.289E+01 0.000E+00
56 3.125E+03 2.500E-02 2.930E+02 3.300E+01 0.000E+00
57 3.132E+03 2.500E-02 2.930E+02 3.306E+01 0.000E+00
58 3.136E+03 2.500E-02 2.930E+02 3.311E+01 0.000E+00
59 3.139E+03 2.500E-02 2.930E+02 3.313E+01 0.000E+00
60 3.140E+03 2.500E-02 2.930E+02 3.314E+01 0.000E+00
61 3.140E+03 2.500E-02 2.930E+02 3.314E+01 0.000E+00
62 3.139E+03 2.500E-02 2.930E+02 3.313E+01 0.000E+00
63 3.137E+03 2.500E-02 2.930E+02 3.311E+01 0.000E+00
64 3.135E+03 2.500E-02 2.930E+02 3.310E+01 0.000E+00
129 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
130 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
131 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
132 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
133 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
134 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
135 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
136 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
137 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
138 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
139 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
140 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
141 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
142 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
143 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
144 0.000E+00 2.500E-02 2.930E+02 1.172E+04 0.000E+00
65 0.000E+00 0.000E+00 0.000E+00 1.172E+04 0.000E+00
81 0.000E+00 0.000E+00 0.000E+00 1.172E+04 0.000E+00
97 0.000E+00 0.000E+00 0.000E+00 1.172E+04 0.000E+00
113 0.000E+00 0.000E+00 0.000E+00 1.172E+04 0.000E+00
80 0.000E+00 0.000E+00 0.000E+00 1.172E+04 0.000E+00
96 0.000E+00 0.000E+00 0.000E+00 1.172E+04 0.000E+00
112 0.000E+00 0.000E+00 0.000E+00 1.172E+04 0.000E+00
128 0.000E+00 0.000E+00 0.000E+00 1.172E+04 0.000E+00

END OF EXECUTION - STAR


CPU time is 0.36 Elapsed time is 0.92

Version 4.02 15-7


PROGRAM OUTPUT Chapter 15
Example Output

Table 15-4:
|-------------------------------------------|
| STAR-CD VERSION 4.00.000 |
| THERMOFLUIDS ANALYSIS CODE |
| Operating System: Linux |
| Stardate: 6-DEC-2005 Startime: 14:32:39 |
|-------------------------------------------|

CASE NAME : out

*** Warning: Molecular weight of material in domain 2 is


set to 28.96.

*** Reduction of solution matrix bandwidth enabled ***


______________________________MEANING OF PRINTOUT QUANTITIES________________________________________
---- NOMENCLATURE (mass balance) -------
FVIN - total flow in through inlet boundaries, kg/s
FVOUT - total flow out through outflow boundaries, kg/s
FPIN - total flow in through pressure,stagnation,free-stream and transient-wave boundaries, kg/s
FPOUT - total flow out through pressure,stagnation,free-stream and transient-wave boundaries, kg/s
FCYIN - total flow in through partial cyclic boundaries, kg/s
FCYOT - total flow out through partial cyclic boundaries, kg/s
FLOUT - total flow out through outlet boundaries, kg/s
SDRDT - mass accummulation by density change in time, kg/s
SDVDT - mass accummulation by volume change in time, kg/s
FLINJ - mass injection, kg/s
MSDRO - mass transfer from dispersed phase (droplets) to continuous phase, kg/s
FDIFF - mass balance kg/s
SUM - sum of mass sources
RESP - sum of absolute mass sources
RES0 - starting residual in the solver

ITERATION NUMBER = 1
--------------------------

*** Warning: Residuals in eq. UMOM have reached round-off


error limit in iteration 1 and trying to reduce
them further can result in the solver divergence. Because
of this further iterating is stopped.

______________________________BALANCE DATA________________________________________________

DOMAINWISE MASS BALANCE (kg/s)


MAT. NO. PHASE NO. FDIFF TOTAL_FLOW_IN TOTAL_FLOW_OUT MSDRO (FVIN) (FPIN ) (FLOUT) (FVOUT)
(FPOUT)
1 1 1.2811E-03 5.9337E-01 5.9209E-01 0.0000E+00 5.9337E-01 0.0000E+00 0.0000E+00
0.0000E+00 5.9209E-01
2 1 0.0000E+00 1.0000E+00 1.0000E+00 0.0000E+00 1.0000E+00 0.0000E+00 1.0000E+00
0.0000E+00 0.0000E+00

------------ BOUNDARY REGIONWISE ------------


REGION NO. TYPE FLOW-IN(kg/s) FLOW-OUT(kg/s)
1 INLET 5.9337E-01 0.0000E+00
2 PRESSURE 0.0000E+00 5.9209E-01
3 INLET 1.0000E+00 0.0000E+00
4 OUTLET 0.0000E+00 1.0000E+00

______________________________FIELD DATA_________________________________________________
*** FOR FLUID STREAM *** 1

Field Extrema:
Umax Vmax Wmax VMAGmax Pmax TKEmax EPSmax Tmax RHOmax
4.9233E+00 6.2337E+01 0.0000E+00 6.2349E+01 5.5077E+00 6.5615E+00 2.6307E+02 2.9300E+02 1.1867E+00
Umin Vmin Wmin VMAGmin Pmin TKEmin EPSmin Tmin RHOmin
-2.1391E+00 3.5014E+01 0.0000E+00 3.5014E+01 2.2318E-01 3.9212E-02 1.3236E-01 2.9300E+02 1.1867E+00

Field Volume-Averages:
Pvav RHOvav Tvav TKEvav EPSvav
2.8710E+00 1.1867E+00 2.9300E+02 1.3542E+00 2.2631E+01

Field Mass-Averages:

15-8 Version 4.02


Chapter 15 PROGRAM OUTPUT
Example Output

Pmav RHOmav Tmav TKEmav EPSmav


2.8710E+00 1.1867E+00 2.9300E+02 1.3542E+00 2.2631E+01

Field Totals:
Mass Volume TKE EPS
9.4939E-03 8.0000E-03 1.2857E-02 2.1485E-01

*** FOR FLUID STREAM *** 2

Field Extrema:
Umax Vmax Wmax VMAGmax Pmax TKEmax EPSmax Tmax RHOmax
1.4449E-02 1.5644E-01 0.0000E+00 1.5644E-01 0.0000E+00 0.0000E+00 0.0000E+00 2.9300E+02 1.0000E+03
Umin Vmin Wmin VMAGmin Pmin TKEmin EPSmin Tmin RHOmin
-7.3840E-03 5.7720E-02 0.0000E+00 5.8799E-02 -2.7495E-02 0.0000E+00 0.0000E+00 2.9300E+02 1.0000E+03

Field Volume-Averages:
Pvav RHOvav Tvav TKEvav EPSvav
-1.4465E-02 1.0000E+03 2.9300E+02 0.0000E+00 0.0000E+00

Field Mass-Averages:
Pmav RHOmav Tmav TKEmav EPSmav
-1.4465E-02 1.0000E+03 2.9300E+02 0.0000E+00 0.0000E+00

Field Totals:
Mass Volume TKE EPS
8.0000E+00 8.0000E-03 0.0000E+00 0.0000E+00

*** FOR SOLID domain *** 3


Temperature data:
Tmax = 2.9300E+02; Tmin = 2.9300E+02; Tvav = 2.9300E+02

Field Solver information:


NSU = 1 NSV = 2 NSW = 0 NSP = 14 NSTE = 1 NSED = 1 NST = 0
CPU time is 0.13 Elapsed time is 0.33
__________________________________________________________________________________________

Version 4.02 15-9


Chapter 16 PRO-STAR CUSTOMISATION
Set-up Files

Chapter 16 pro-STAR CUSTOMISATION


pro-STAR provides four means by which users can customise the way they work
with the program:
• Set-up files
• Panels
• Macros
• Function keys

All are geared towards making problem data input faster and more flexible and can
be used in combination with each other. The choice of which ones to use is largely
a matter of user preference and the requirements of the model being built.

Set-up Files
These files are read automatically as part of the pro-STAR start-up process and are
used in creating a suitable pro-STAR environment for the problem in hand. The
files have standard names, given below, and are located in a directory chosen by the
user. On Unix systems, the path to this directory is stored in an environment variable
(STARUSR) specified outside pro-STAR using the appropriate Unix environment
setup command (see Chapter 17, “pro-STAR environment variables”). The
available set-up files are as follows:
1. PROINIT — contains pro-STAR commands that are read and executed as the
first action in the current session. This provides a convenient way of setting up
(initialising) pro-STAR in a standard way (regarding, for example, plot type,
viewing angle, etc.) every time a session begins. Some pro-STAR commands
are in fact best used from within the PROINIT file. For example:
(a) Command OPANEL — typically used to open a set of tools (standard
pro-STAR GUI dialogs or user-defined panels) that the user wants on
screen at the start of every new session.
(b) Command SETFEATURE — reports or changes the byte ordering format
of binary files to suit machines such as the Compaq Alpha range. This
facility replaces settings previously made through environmental
variables.

2. PRODEFS — this file is created automatically if the *ABBREVIATE


command is used during the session. *ABBREVIATE enables one or more
frequently used commands and their parameters to be joined together and
executed in sequence, simply by associating them with an abbreviation name.
The command group comes into action every time an existing ‘abbreviation’
is typed in the I/O window.
File PRODEFS stores all current abbreviation definitions and, once created,
may be used in all subsequent pro-STAR sessions. The file itself may be
edited with any suitable text editor to add/modify/delete any particular
abbreviation, as needed.
3. .Prostar.Defaults — a hidden system file containing definitions of
function keys (see “Function Keys” on page 16-9), panel size and location
(see “Panel definition files” on page 16-5) and ‘favourite’ panels (see “Panel

Version 4.02 16-1


PRO-STAR CUSTOMISATION Chapter 16
Panels

navigation system” on page 2-40).

If the set-up file directory is not defined through STARUSR, pro-STAR creates
default set-up files automatically in your current working directory.

Panels
Panels are user-definable tools capable of simplifying the use of pro-STAR
operations that are either not available in the existing GUI menus and dialog boxes
or require additional functionality. Panels are often employed to facilitate the use of
Macros, which are groups of commands that are saved in a separate file (see
“Macros” on page 16-6). Macros can be assigned to Panel buttons so that a large
number of commands can be executed simply by clicking such a button.
Panel creation
Panels can be created or modified by choosing Panels > Define Panel from the
main menu bar to display the Define Panel dialog box shown below.

New panels are created by entering a name in the text box of the Define Panel dialog
box and then clicking on the New action button. This results in the panel name being
added to the list above the text box. Once this is done, the panel itself can be opened
by
• double-clicking on its name in the list, or
• selecting the name in the list and then clicking on the Open action button, or
• clicking on Panels in the main menu bar and selecting the panel name from
the drop-down list.

Any of the above actions will display a panel such as the one shown below.

16-2 Version 4.02


Chapter 16 PRO-STAR CUSTOMISATION
Panels

Once the new panel has been opened, the user can specify its layout and define its
buttons and menu items. The Panel Layout dialog box can be opened by selecting
File > Layout from the panel’s menu bar.

The above dialog box allows definition of the number and layout of the panel
buttons (a maximum of 100).
Users may also specify menus for panels by selecting File > Menus from the
panel’s menu bar. This opens the Define User Menus dialog box, shown below,
where one can define up to six menus, their names and the pro-STAR commands
that will be executed upon selecting a particular menu item. By default, a single
menu called User 1 is defined containing a single menu item called Replot which
executes command REPLOT.

Version 4.02 16-3


PRO-STAR CUSTOMISATION Chapter 16
Panels

Panel button names and definitions are assigned by first selecting a button, and then
entering a new name or definition into the appropriate text box. A button’s
definition is the pro-STAR command(s) that will be executed when the button is
pushed.
The following three examples illustrate the way in which frequently repeated
operations may be simplified by assigning them to panel buttons:

Example 1
Select a number of cells with the screen cursor and then refine them by a factor of
2 in all directions. Assign to option button CCREF.

Example 2
Select a range of fluid cells by drawing a polygon around them, change them to solid
cells and then plot the mesh. Assign to option button CZMOD.

16-4 Version 4.02


Chapter 16 PRO-STAR CUSTOMISATION
Panels

Example 3
Display vertex coordinates in local coordinate system 2 by pointing at the required
vertex with the cursor. Assign to option button VCOOR2.

To select a button without executing the corresponding button definition, move the
mouse pointer to the button and press (but do not release!) the mouse button. Next,
move the mouse pointer clear of the button and then release the mouse button. This
sequence will set the newly selected panel button as the active button, but will not
execute the button function.
Note that selecting File > Reload from the panel’s menu bar will cancel out any
changes made to the panel definition since it was last saved.
Panel definition files
A panel’s button and menu settings as well as its size and location are saved in a
panel definition file when File > Save is selected from the panel’s menu bar. This
file is created using the panel name specified by the user in the Define Panel dialog
box and the suffix .PNL. The file location depends on its name. If the name entered
was prefixed with the letter L or G (note that a space must be typed after each letter),
the file will be placed in directory PANEL_LOCAL or PANEL_GLOBAL, otherwise
it will be put in your current working directory.
On Unix systems, the local and global directory names are stored in environment
variables that can be set outside pro-STAR using the appropriate Unix environment
setup command (see Chapter 17, “pro-STAR environment variables”). The
environment variables can also be set within pro-STAR by selecting Panels >
Environment from the main menu bar. This displays the Set Environment dialog
box shown below, which allows entry of local and global directory names in the
corresponding text boxes.

Version 4.02 16-5


PRO-STAR CUSTOMISATION Chapter 16
Macros

Command: SETENV

Note that a list of available panels can be viewed by opening the Define Panel dialog
box. Panels found in your current working directory are shown in the list with a ‘.’
before the panel name. Any panel definitions found in the directories specified by
the PANEL_LOCAL and PANEL_GLOBAL variables are shown in the list with an L
or G prefix before the panel name, respectively. Once added to the list, a panel can
be opened in a number of ways, as described in “Panel creation” on page 16-2. Note
that panels can also be opened from the pro-STAR input/output window by typing
OPANEL, PANEL but this command is more typically issued from within the
PROINIT set-up file (see “Set-up Files” on page 16-1).
In addition to the panel definition file, a panel’s size and location are also saved
in a hidden system file called .Prostar.Defaults (see “Set-up Files” on page
16-1). Definitions stored there have priority over the size and location information
stored in the panel definition file. This enables you to override such information if
the panels are located in a directory for which you do not have write permission.
Panel manipulation
The Define Panel dialog box provides additional facilities for manipulating panels,
as follows:
• The Re-Scan button recreates the list of available panels. Those that were
removed from the list will re-appear, while those created via the New button
but never saved will disappear.
• The Copy button creates new panels by copying an existing panel definition
file to another file whose name must be typed in the text box.
• The Rename button changes the name of a panel definition file to another
name typed in the text box.
• The Delete button allows you to remove panels from the list but does not
delete the corresponding definition files. The latter can only be deleted outside
pro-STAR by using the appropriate operating system command.

Macros
A macro is a set of user-defined commands that can be executed at any stage of the
pro-STAR session. The constituent commands must be stored in a special file,
identified by a ‘.MAC’ extension and included within
16-6 Version 4.02
Chapter 16 PRO-STAR CUSTOMISATION
Macros

• the current working directory, or


• a pre-defined local macro directory, or
• a pre-defined global macro directory.

As with panel directories, the local and global directory names are stored in
environment variables MACRO_LOCAL and MACRO_GLOBAL that can be set
outside pro-STAR using the appropriate Unix environment setup command (see
Chapter 17, “pro-STAR environment variables”). The environment variables can
also be set within pro-STAR by selecting Panels > Environment from the main
menu bar. This displays the Set Environment dialog box which allows entry of local
and global macro directory names in the corresponding text boxes.
Macros can be created, renamed, copied, and deleted in the Define Macro dialog
box in the same way that panels are in the Define Panel dialog box. The Define
Macro box, shown below, is opened by choosing Panels > Define Macro from the
main menu bar. The name of a new macro must be typed in the text box. An existing
macro can be selected and displayed, by double-clicking its name in the macro list.
Several macros can be displayed simultaneously in multiple windows, by
highlighting them in the list with the mouse and then clicking the Open button.
pro-STAR looks for macro files in three places. Macros found in the user’s current
working directory are shown in the list with a ‘.’ in front of the macro name. Those
found in the directories specified by the MACRO_LOCAL and MACRO_GLOBAL
environment variables are shown with an L or G prefix before the macro name,
respectively.

Clicking the Open or New button in the Define Macro box opens a macro editor to
display the macro file(s) that has been selected in the macro list (or a blank sheet for
new macros), as shown below. The user can then type in the required pro-STAR
commands or amend existing ones. Command PROMPT, which displays messages
in the area underneath the plotting window (see “Main window” on page 2-15) is
particularly useful inside a macro as it can prompt the user to, say, supply required
data or to click an appropriate menu item with the mouse.

Version 4.02 16-7


PRO-STAR CUSTOMISATION Chapter 16
Macros

The macro editor facilities are arranged under three menus in the editor’s menu bar:
1. File
(a) Open — open another macro
(b) Save — save the current changes
(c) Save As — save the current changes to a different macro file
(d) Clear All — clear the editor window
(e) Quit — terminate the editing session

2. Edit
(a) Find — find a character string typed in the dialog box shown below:

(b) Mark Selection — mark the selected characters for subsequent searches
(c) Find Selection — find the selected characters in the macro body
(d) Find Again — repeatedly find the selected characters
(e) Replace — find a character string and replace it with another string. Both
strings are typed in the dialog box shown below:

16-8 Version 4.02


Chapter 16 PRO-STAR CUSTOMISATION
Function Keys

3. Execute
(a) Execute Macro — execute the whole macro
(b) Execute Selection — execute only the highlighted lines in the editor
window

As with panels, the Define Macro dialog box provides additional facilities for
manipulating macros, as follows:
• The Execute button executes the selected macro.
• The Re-Scan button recreates the list of available macros. Those that were
removed from the list will re-appear, while those created via the New action
button but never saved will disappear.
• The Copy button creates new macros by copying an existing macro file to
another file whose name must be typed in the text box.
• The Rename button changes the name of a macro file to another name typed
in the text box.
• The Delete button allows users to remove macros from the list of available
macros but does not delete the corresponding files. The latter can only be
deleted outside pro-STAR by using the appropriate operating system
command.

Note that panel buttons are often used to execute macros, by setting the button
definition to issue command

*macro,exec

This assignment can be made as follows:


• Open the Define Macro dialog box and highlight a macro in the list.
• Open the Define Panel dialog box, select a panel from the list and display it by
double-clicking it.
• Click on a free button in the panel.
• Select Assign from the panel’s Macro menu. This assigns the macro name to
the button and generates the appropriate *MACRO command.
• If necessary, select Edit from the panel’s Macro menu to open the macro text
editor discussed above and type in any further changes
• Save all changes by selecting Save from the File menus of both macro and
panel editors before closing their corresponding dialog boxes

Function Keys
Users can program the keyboard function keys (F2 - F12) to execute pro-STAR

Version 4.02 16-9


PRO-STAR CUSTOMISATION Chapter 16
Function Keys

commands or macros. This is done by choosing Utility > Function Keys from the
menu bar to display the Edit Function Keys dialog box shown below.

Any valid pro-STAR command (or set of commands if a $ character is used to


separate them) can be mapped to individual function keys by typing it in the
appropriate text box. Command parameters such as ‘VX’ or ‘CX’ may be used and
will be interpreted in the normal way. Command strings are limited to 80 characters
in length.
In addition to standard pro-STAR commands, the function keys can also be used
to repeat the last executed command and to open dialog boxes. Thus:
• Command repeat will literally repeat the last command executed, including
parameters such as ‘VX’ or ‘CX’.
• Command string open dialog1,dialog2,... will open the dialog
boxes or tools specified. Available items are:

16-10 Version 4.02


Chapter 16 PRO-STAR CUSTOMISATION
Function Keys

Name Description
ANIM Animation Module
BLIS Boundary List
BLLI Block List
BLOC Block Tool
CELL Cell Tool
CHEC Check Tool
CHEM Chemical Module
CLIS Cell List
COLO Colour Tool
CONT Control Module (unsupported panel)
COUP Couple Tool
CSYS Coordinate Systems panel
DROP Droplets panel
FORE Convert Foreign Formats panel
GENE Convert Generic panel
GRAP Graph Tool
GRDI Graph Module
GRLO Load Graph Registers panel
GRRE Graph Registers panel
POST Post Register Data List
PROP Property Module (unsupported panel)
SPLI Spline Tool
SPLL Spline List
STAR Convert STAR panel
TRAN Transient Module
VERT Vertex Tool
VLIS Vertex List

The default function key definitions are:

F5 – repeat
F6 – replot
F7 – cplot
F8 – zoom,off $replot

Note that the F1 key is reserved for displaying context-sensitive, on-line Help
information on pro-STAR commands (see “Getting On-line Help” on page 2-35)
Any changes to the function key definitions are saved in a file called
.Prostar.Defaults (see “Set-up Files” on page 16-1) at the location specified
by environment variable STARUSR (or in your current directory, see page 16-1).
This file can be modified either through the Function Keys dialog box within
pro-STAR or outside it via any suitable editor. Users may find it useful to keep a
single .Prostar.Defaults file in the STARUSR location so that the particular
setup that they define is available for any pro-STAR session.

Version 4.02 16-11


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
Introduction

Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS


Introduction
This chapter describes some of the less commonly used features and controls in
STAR-CD and covers the following topics:
• File organisation, naming conventions and general utilisation
• Special pro-STAR and STAR features and settings
• The StarWatch utility
• Hard copy production

File Handling
Naming conventions
At every session, pro-STAR creates a set of files whose names are based on a
user-supplied model name or case name. Each file name is of the form
case.xxxx, where xxxx is a three- or four-character filename extension. Thus,
if the case in question is called test, then all its associated files will be called
test.ccm, test.mdl, etc. and will be used for the appropriate input/output
operation during the model building and numerical solution processes. You should
always supply a case name at the beginning of a pro-STAR session (see “pro-STAR
Initialisation” on page 2-12).
A case name may be overridden at any time during a pro-STAR session by
choosing File > Case Name from the menu bar. This displays the Change Case
Name dialog shown below:

Command: CASENAME

Supply a new case name (up to 70 characters long) in the text box provided. This
changes the default file name but does not affect any files that are already open. It
also determines which files will be used during subsequent file operations. Note that
the names of the input and output restart (.ccm) files will be reset by this operation.
Commonly used files
A few key files are always read and/or written to by pro-STAR, whereas the
majority are opened and accessed only in response to a command or a GUI
operation. These key files are described below:

Version 4.02 17-1


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
File Handling

Echo file (.echo)


Used exclusively by pro-STAR and is always opened. It holds a copy of every
command typed by the user or, for GUI operations, their command equivalents, as
generated automatically by pro-STAR during the session. The file can be:
• Reviewed
• Used for recovery purposes (see “Error messages” on page 2-19)
• Copied to a temporary file which can be subsequently edited to make changes
to the recorded commands (see item 12 on page 17-10). Once the editing
process is complete, the modified command file can be replayed into
pro-STAR using the editor’s file execution facilities (or by typing command
IFILE).

Model file (.mdl)


Used exclusively by pro-STAR. Choosing option File > Save Model from the menu
bar instructs pro-STAR to write a full description of your model to this file, using
the specified case name as the file name. It is advisable to save data regularly during
a session so as to minimise the chance of losing large amounts of information due
to user error or system failure. Note that every time you save the model file, its
previous version (i.e. the model you started out with before making any changes) is
also automatically stored as a backup, in a file of form case.bak
If you need to save the .mdl file under a name other than the case name, choose
option File > Save As from the menu bar. This displays the Save As dialog, shown
below, which allows the name to be typed exactly as required. Alternatively, an
existing file may be selected by utilising pro-STAR’s built-in file browser facilities
(see page 17-9).

Command: SAVE

Option File > Resume Model performs the reverse operation, i.e. it instructs
pro-STAR to read a model description from an existing .mdl file. If you need to
resume from a .mdl file that does not have the same name as the case name, choose
option File > Resume From from the menu bar. This displays the Resume From
dialog shown below, which allows the name to be typed exactly as required.
Alternatively, the file may be selected by clicking the browser button provided and
utilising pro-STAR’s built-in file browser facilities (see page 17-9).

17-2 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
File Handling

Command: RESUME

Problem setup (command) file (.inp)


Although the model file is normally saved in binary form, there may be occasions
when you need to write the model data in text (coded) form. Examples of such
instances are:
• To allow you to quickly produce a set of coded pro-STAR input files that will
re-create the case as defined in the model file. This is especially useful if you
want to set up several runs with parametric changes and then submit the job in
batch.
• To enable you to find out which commands would activate certain features
present in your model.
• To facilitate testing of models that were created with a previous version of
pro-STAR.

To write model data in text form, choose File > Save As Coded from the menu bar
to display the CDSave dialog shown below:

Command: CDSAVE

The dialog uses default file names with extensions .inp, .cel, .vrt, and .bnd
for four files that will contain problem set-up, cell, vertex and boundary
information, respectively. Alternative names for any of these files may be entered
in the boxes provided. For moving mesh cases, event definitions (see “Moving
Meshes” on page 12-9) can also be written to file .evnc. For cases containing
droplets, an additional droplet data file (.drpc) is created.
Once the files have been copied to a suitable directory, the model may be
re-activated by choosing File > Read File from the menu bar. This will display the

Version 4.02 17-3


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
File Handling

Input Coded File dialog shown below:

Command: IFILE

Check that the file shown in the File Name window is correct and then click Apply.
All data in files .cel, .vrt, .bnd, .drpc (if present) and, for moving mesh
cases, .evnc will be read in automatically.

Transient history file (.trns)


This is used exclusively by pro-STAR for transient problems specified by means of
load steps (see “Load-step based solution mode” on page 5-6) and contains all
additional information (changes to boundary conditions, number and length of time
steps, etc.) needed for such problems. You must make this file available to your
current session before changing or adding data concerning the analysis. This is done
via the Advanced Transients dialog (see “Load step controls” on page 5-10), or by
typing command TRFILE.
The file is normally written in binary form but a facility also exists for writing it
instead in text (coded) format and to a file with extension .trnc. This is done by
selecting Modules > Transient from the menu bar to display the Advanced
Transients dialog, specifying the file name in the box provided at the bottom of the
dialog and then clicking Apply. Alternatively, use command CDTRANS. If an
existing file needs to be used, pro-STAR’s built-in file browser can help locate it.

Plot file (.plot)


This is used exclusively by pro-STAR and is always open to receive neutral plot
information, i.e. machine-independent representations of a set of plots. The file may
be written in either binary or text (coded) format. CD-adapco supply source code
for several decoding programs that drive hard-copy devices in a variety of formats
(e.g. Postscript), or screen output devices (e.g. X-window workstations). These
programs can also serve as templates for constructing plot drivers for other,
unsupported devices. To make use of the neutral plot facility:
• Specify the plot file name (if other than case.plot) and type (if not
CODED) using command NFILE.
• Switch the plot output from the terminal or workstation to the plot file by
choosing item Plot > Plot To File from the menu bar (or use command
TERMINAL in the form

TERMINAL,,FILE

• Perform the plotting operations required, as normal. Graphical output is now


17-4 Version 4.02
Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
File Handling

diverted to the file instead of being displayed on the screen.


• To restore normal operation, choose option Plot To Screen from the Plot
menu.

Details of data representation in the neutral plot file can be found in the
Post-Processing User Guide, Appendix B.

Data repository file (.ccm)


This file has a special format that facilitates different sets of information to be stored
in it. It is written and read by both pro-STAR and STAR in the following ways:
1. pro-STAR saves all cell topology and model geometry information in the file
once mesh building is complete. Cells are defined as a collection of faces, i.e.
a general polyhedral cell definition is used regardless of the actual cell shape.
The file must be rewritten whenever
(a) the mesh geometry is modified
(b) boundaries are added, subtracted, or assigned to different boundary
regions
(c) the cell type definitions are changed

The file is created by selecting File > Write Geometry File from the menu
bar to display the Geometry File Write dialog shown below:

Command: GEOMWRITE

The input required is:


(a) File Name — enter a name in the text box provided or click the adjacent
button to select an existing file using pro-STAR’s built-in file browser
(see page 17-9)
(b) File Type — select the file format according to the solver (CCM,
CEDRE, BAE) for which the geometry file is intended
(c) Geometry Scale Factor — an optional scale factor applied to all
dimensions of the problem’s geometry
(d) Write Backup File — define the action to be taken if the specified
repository filename already exists in your current directory:
i) Backup — the existing file is renamed as casename.ccm.bak

Version 4.02 17-5


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
File Handling

(or casename.ccm.bak1 if casename.ccm.bak already


exists, or casename.ccm.bak2 if casename.ccm.bak1
already exists, etc.).
ii) No Backup — the geometry information is overwritten

2. STAR’s action is one of the following:


(a) For initial runs, STAR reads the geometry information contained in the
file and then appends the solution results at the end of the analysis.
(b) For restart runs, STAR reads in addition the results of a previous (partially
converged, interrupted or transient) analysis before starting the new
solution. The new analysis results then overwrite the previous ones at the
end of the run.
(c) If solution residual values are required as part of the analysis, STAR also
stores them in this file.

3. Apart from storing problem geometry data, pro-STAR will also


(a) read the file for post-processing purposes, i.e. to make contour, vector or
graph plots of any variable calculated by STAR.
(b) write mapped solution data when an existing mesh is refined (see Chapter
5, “Solution Control with Mesh Changes” in this volume)

Each set of data stored in the repository file is called a “state”. The table below lists
the names of the states and the data stored in them. The available states in a given
file may be displayed using command CCLIST.

STAR-CD 3.2X
State name Type of data
equivalent file
default Problem geometry for .geom
sequential runs
geom_par Problem geometry for .geom
parallel runs (decomposed)
Restart_1 Solution (Restart) data .pst
smap Mapped Solution .smap
(Restart) data
Residue_n Field residuals .rpo
n=1,2,3,...

Problem data file (.prob)


This is written by pro-STAR and read by STAR. It contains information on what
kind of analysis is to be performed and what data are to be printed or saved for post
processing. It also contains all material property values, solution control settings,
boundary conditions and initial conditions. It is written independently of the
geometry file and should be rewritten every time any of the above model parameters
is modified.
The file is created by selecting File > Write Problem File from the menu bar to
display the dialog shown below. Note that a filename other than the default

17-6 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
File Handling

(casename.prob) may be entered if necessary in the text box provided.

Command: PROBLEMWRITE

Transient post data file (.pstt)


This is written by STAR and contains selected transient analysis data at pre-defined
points in time (see “Output controls” on page 5-12). It is used by a subsequent
pro-STAR post-processing run to make contour or vector plots based on the
selected data. Note that the file holds only part of the available information on the
model, so it cannot be used for restarting the analysis; that function can be
performed only by using the data repository (.ccm) file.
File relationships
The use and relationship between files in the STAR-CD environment is illustrated
by Figure 17-1. Appendix B in this document contains a complete list of all files that
can be written or read by either pro-STAR or STAR. The same information may
also be displayed on-line in the Help dialog (choose Help > pro-STAR Help from
the menu bar, select Misc. from the Module pop-up menu, and then highlighting
item FILE). For the great majority of problems, however, only the files shown
below are ever needed.

Version 4.02 17-7


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
File Handling

case.trns
Transient
history data case.mdl
Model data

case.echo
pro-STAR Command
echo

case.plot
case.prob Neutral plot
Boundary conds. case.ccm
Solution params. Geometry

case.pstt
STAR Transient
output data
case.ccm
Solution data

Figure 17-1 STAR-CD file use

In addition to solution and transient post data files, pro-STAR provides a utility for
converting solution monitoring and droplet track data files to coded (text) format
and vice versa. This is useful, for example, in manipulating and displaying the data
outside the pro-STAR environment or for checking the validity of the file contents.
The utility allows conversions between a variety of formats and is accessed by
selecting Tools > Convert > Post from the menu bar. This activates the Post
Convert dialog shown below:

Commands: SMCONVERT PTCONVERT

You may then


• select option Solution Monitoring or Particle/Droplet Track depending on
the file type you wish to convert. The first option deals with residual or
solution monitoring data conversion (see Chapter 5, “Output controls”), the
second with droplet track data conversion (see “Trajectory displays” on page
9-8) or particle track data conversion (see “The Particle Track File” on page
7-6 of the Post-Processing User Guide)
• enter the name of the file containing the data to be converted (Input File with

17-8 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
File Handling

extension .rsi or .trk) or select it using pro-STAR’s built-in file browser


(see page 17-9)
• choose the file type (normally Binary) from the available options in the
adjacent pop-up menu
• enter the name of the file that will store the converted data (Output File with
extension .rsic or .trkc)
• choose the file type (normally Coded) from the available options in the
adjacent pop-up menu

The above operation may also be performed in reverse, i.e. converting the text file
back to binary format, using the same dialog but with Input now being Coded and
Output being Binary, plus a reversal of the file name extensions.
In the course of a session pro-STAR also opens several scratch files. These are
opened automatically and deleted at the end of the session. Their use is normally
transparent except when their size exceeds the amount of free space on your disk.
While some scratch space is used for hidden-line plotting, the largest amount is
needed while the geometry (.ccm) file is being written. The space used varies
linearly with the number of vertices present and the maximum number of cells
connected to any single vertex.
File manipulation
The file-manipulation related capabilities of pro-STAR are as follows:
1. Finding files — If you are not sure of the exact location or name of an existing
file, use pro-STAR’s file browser facility. This is activated by clicking the
browser button

included in numerous GUI dialogs. The button displays the File Selection
dialog shown below:

The scroll lists and filters included in the above dialog allow easy navigation
through various levels of sub-directories until the required file is located.

Version 4.02 17-9


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
File Handling

2. Switching program input from a terminal (or standard input) to any disk file
(of form case.inp) containing pro-STAR commands, and vice versa. This
can be done at any time during a session, using either the pro-STAR editor’s
Execute menu options (see item 12 below) or by typing command IFILE. In
the latter case, input switches back to the terminal automatically at the end of
the specified file. The command also supports a ‘nesting’ capability, i.e. the
new input stream can itself contain IFILE commands that will direct input to
yet another source file and so on.
3. Switching output from a terminal screen (or standard output) to a disk file (of
form case.out), and vice versa. This can be done at any time during a
session using command OFILE. Using parameter NONE with this command
turns the output off completely. The facility enables you to save lists of
various pro-STAR items, for use in other programs or for later review.
4. Writing the geometry file (see “Data repository file (.ccm)” above) by
choosing File > Write Geometry File from the menu bar.
5. Writing the problem data file (see “Problem data file (.prob)” above) by
choosing File > Write Problem File from the menu bar.
6. Restoring a previously created model from a saved model file (see “Model file
(.mdl)” above) by choosing File > Resume From... from the menu bar. When
used for the first time in a pro-STAR session, RESUME will also automatically
read and execute commands stored in a special file called PROINIT (see
“Set-up Files” on page 16-1). This provides a very convenient way of setting
up pro-STAR in a standard way (regarding, for example, plot type, viewing
angle, etc.) every time a session starts.
7. Saving the current model description in binary format to file .mdl, as
described above, by choosing File > Save Model from the menu bar.
8. Saving the model description in text (coded) format, as described above, by
choosing File > Save As Coded from the menu bar.
9. Repositioning a previously used file (including a pro-STAR macro file) to its
starting point by typing command REWIND.
10. Closing a previously used file by typing command CLOSE. The command
may also close all currently open files.
11. Printing a summary of all currently open files by typing command FSTAT.
12. File editing via pro-STAR’s built-in editor — This is activated by choosing
File > Edit File from the menu bar to display the panel shown below. Files
that may be conveniently manipulated using this editor are:
(a) Command files — these allow execution of a set of pre-recorded
pro-STAR commands. As noted in the section on “Commonly used files”
on page 17-1, a common source for them are echo files from previous
pro-STAR sessions. To avoid problems, however, an echo file should be
copied and renamed before using it as a command file.
(b) User subroutine files — these contain special user-supplied FORTRAN
code and are discussed in detail in Chapter 14.

17-10 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
File Handling

The available facilities are arranged under three menus in the editor’s own
menu bar, as follows:

File
(a) Open — open a specified file. This activates the File Selection dialog
shown on page 17-9, enabling the required file to be located.
(b) Save — save the current changes.
(c) Save As — save the current changes to a different file. The dialog box
above re-appears to aid specification of the destination file location.
(d) Clear All — clear the editor window.
(e) Quit — terminate the editing session.

Edit
(a) Find — find a character string, typed in the dialog box shown below.

(b) Mark Selection — mark the selected characters for subsequent searches.
(c) Find Selection — find the selected characters in the file body.
(d) Find Again — repeatedly find the selected characters.
(e) Replace — find a character string and replace it with another string. Both
strings are typed in the dialog box shown below.

Version 4.02 17-11


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
Special pro-STAR Features

Execute (Command files only)


(a) Execute All — execute all commands in the file. This is equivalent to
typing command IFILE in pro-STAR’s Input window.
(b) Execute Selection — execute only the highlighted lines in the editor
window.

In addition, the usual keyboard- or mouse-driven cut, copy and paste functionality
is also available with the editor window.

Special pro-STAR Features


pro-STAR environment variables
pro-STAR uses the values of several environment variables. Some specify the path
to various system directories while others control the operation of the system. You
should ensure that these values are correctly set before using STAR-CD.
The syntax for setting environment variables depends on the shell program you
are using (if in doubt type the command echo $SHELL). The current list of such
variables is as follows:
MACRO_LOCAL and MACRO_GLOBAL
Paths to the local and global pro-STAR macro directories, respectively (see
“Macros” on page 16-6)

PANEL_LOCAL and PANEL_GLOBAL


Paths to the local and global pro-STAR panel directories, respectively (see “Panel
definition files” on page 16-5)

STARBROWSER (not needed for Windows ports)


Path to the user’s choice of Internet browser (Netscape or IE) that will be launched
from the pro-STAR Help menu (see page 2-37). The user’s search path must be
amended to include the directory defined by this variable. The default is to run
Mozilla from your current working directory.

STARFONT0 / starfont0
Font name and size to use for plot title, plot legend, graph title and main axes label
(see the description of command TSCALE in the Commands volume)

STARFONT1 / starfont1
Font name and size to use for the contour and vector scales (see the description of
command TSCALE in the Commands volume)

STARFONT2 / starfont2
Font name and size to use for the secondary contour and vector scales (for droplets
and particle ribbons; see the description of command TSCALE in the Commands
volume)

STARFONT3 / starfont3
Font name and size to use for entity numbers (NUMBER command), x and y tick
labels on graphs and local coordinate system axes (see the description of command

17-12 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
Special pro-STAR Features

TSCALE in the Commands volume)

Note: Variables STARFONT 0-3 described above apply only to X-window


plotting. They have no effect on OpenGL based plotting as the fonts system there is
entirely different.

STAR_TCL_SCRIPT
Path to the location of file STARTkGUI.tcl, containing a user-supplied Tcl/Tk
script (see “The Users Tool” on page 2-35)

STARUSR
Path to pro-STAR files PRODEFS (abbreviations), PROINIT (initial set up) and
.Prostar.Defaults (see “Set-up Files” on page 16-1)
Resizing pro-STAR
pro-STAR is a dynamic-memory executable code and requires a file called
param.prp to be present in your current working directory. The file contains a list
of parameters that determine the data size of the executable on start-up. If this file
is missing, incomplete, or out-of-date, pro-STAR will automatically write a new
local param.prp based on the values in the model (.mdl) file being read, and
also on any values that could be read from an existing param.prp. This happens
the first time pro-STAR is run using the prostar script described in Chapter 2,
“Running a STAR-CD Analysis”.
It is almost always necessary to resize the pro-STAR executable to cater for
special problems (such as moving mesh problems) or to accommodate cases with a
larger number of cells, vertices, etc. (or a smaller number, if you are having
problems with available memory in your machine).
In any of the above situations, file param.prp should be modified but this
should never be done using a text editor. Rather, a new version of the file containing
parameters of the right magnitude must be created in one of the following ways:
1. By running the prosize script. This is accessed by typing

prosize

The script first asks whether you want to modify some of the parameters in
the current file or create a brand new param.prp. You may also exit here
without modifying or creating any files. If continuing, prosize asks:

Is your mesh primarily hex or tet? (Answer H or T)


(The T option should be chosen for wholly or predominantly tetrahedral
meshes; H is appropriate for all others, including meshes containing trimmed
cells)

After this, the script prompts you to specify the values of the parameters to be
stored in param.prp. A carriage return instructs the script to use the
indicated default value, while entering -1 will terminate the script and use the
remaining defaults to write param.prp. The most usual variation from the
default values is in the maximum number of cells (MAXCEL) and vertices
(MAXVRT). Otherwise, the default values suggested by prosize should be
sufficient for most cases.
Version 4.02 17-13
OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
Special pro-STAR Features

2. By issuing command MEMORY from within the pro-STAR session.


This command can be used only to increase the parameter sizes. If during the
session it is found that the value of any sizing parameter(s) is insufficient, a
warning message will appear in the I/O window. pro-STAR will sometimes be
able to adjust the parameter value(s) automatically and then continue.
However, in most cases you will be prompted to enter an appropriate new
value for the indicated parameter(s) using MEMORY, after which you may
continue as normal. Either way, the parameter values are changed internally
without changing the param.prp file. To use the new values in future
pro-STAR sessions, you will need to save them explicitly via a
MEMORY,WRITE instruction. This will rename the existing param.prp file
as param.bak and write the new parameters into a new param.prp file.

Note that, after running pro-STAR with a given model, it is possible to clear all
model parameters (i.e. delete all cells, vertices, boundaries, etc.) but leave the
current memory size intact. This is done using command WIPEOUT and is useful if
you want to abandon the current model and start a new one from scratch without
exiting from pro-STAR. Furthermore, option MEMORY of this command will also
reset the pro-STAR executable back to the size given in the param.prp file.
Special pro-STAR executables
On occasion, you may need to use a user-defined subroutine file, user1.f. This
option refers to subroutines that work in conjunction with pro-STAR, not STAR,
and is not supported in Windows ports at present. In such a case, the required special
pro-STAR executable may be created using script prolinkl. This is accessed by
typing

prolinkl

The script looks for a file named user1.f in the current directory. That file will
be compiled into object code (user1.o) and converted into a dynamically-loaded
shared object (.so or .sl or .dll depending on the operating system). The
directory with the shared object must be added to the shared object library path
(usually LD_LIBRARY_PATH) in order to be found and used by any subsequent
pro-STAR runs. prolinkl will advise the user on how to create this path for the
given operating system.
Use of temporary files by pro-STAR
Choosing the location of temporary files
You can control the location of most pro-STAR temporary files for POSIX-
compliant computers. You should ensure temporary files reside where there is
sufficient capacity and where they can be accessed quickly. In practice, this means
on a fast hard disk on the same computer as that doing the calculations (rather than
on a remote disk accessed through a local area network). Note that the usual location
for Unix temporary files, a directory called /tmp, often has insufficient capacity for
pro-STAR’s temporary files.
You select the location of temporary files by setting an environment variable,
named TMPDIR, to the path name of the directory where pro-STAR should write
the temporary files.

17-14 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
The StarWatch Utility

Deleting temporary files


Take care not to delete pro-STAR’s temporary files during a calculation; it will
crash if you do. pro-STAR may leave temporary files behind if it crashes or you halt
its execution. For POSIX-compliant systems, the operating system automatically
deletes most temporary files if pro-STAR halts or crashes. For other systems, you
might have to manually delete abandoned temporary files after a crash or halt.

The StarWatch Utility


This is a free-standing utility that enables you to monitor the progress of a selected
STAR job running anywhere in your computer network. The monitoring is done
from a special window opened by StarWatch, as shown below.

Specific advantages of StarWatch are:


• You can monitor progress of a number of separate STAR jobs
• The jobs may be running on any machine in your network, including your
own
• You may select the variables whose solution progress you wish to monitor
• You may adjust the display characteristics (e.g. scaling) of the monitored
variables
Running StarWatch
By default, StarWatch uses ports 6200 to 6206 to establish communication between
the STAR executable, the StarWatch daemon (a communication program) and
StarWatch, the display program that runs on your screen. If ports 6200 — 6206 are
acceptable, then no further setup is required. If they are not, perhaps because they

Version 4.02 17-15


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
The StarWatch Utility

conflict with other programs using those ports, they can be set to any ports that the
user (or more likely) system administrator wants to use. The only proviso is that if
the ports are changed on one system, the same change must be made for all systems
for which StarWatch communication is required. If the defaults are not acceptable,
then an administrator must edit the /etc/services file and add the following
lines:
star-chartd 6200/tcp # Star/Stripchart client/server daemon
star-chart1 6201/tcp # Local Stripchart 1
star-chart2 6202/tcp # Local Stripchart 2
star-chart3 6203/tcp # Local Stripchart 3
star-chart4 6204/tcp # Local Stripchart 4
star-chart5 6205/tcp # Local Stripchart 5
star-chart6 6206/tcp # Local Stripchart 6

where port numbers 6200 — 6206 can be replaced by any set of port numbers.
Step 1
If using the STAR GUIde environment to run a numerical analysis interactively,
StarWatch will start automatically as soon as STAR itself begins execution and will
open a monitoring window like the one shown above (see Chapter 2, “Running a
STAR-CD Analysis”, Step 6). If you are not using STAR GUIde, or if you want to
monitor the progress of another currently active job, you may open the StarWatch
window explicitly by following the steps below:
• Open a new window on your computer or go to an existing one
• Type starwatch, then send this application to the background also. The
StarWatch application panel should appear on your screen.
• Start your STAR job in the same window using the -watch option. Note that
when running a parallel job, the -watch option must precede any other
options used.

Note that you may also start STAR first and then StarWatch.
Step 2
Go to the StarWatch panel and select option Host from the Connect menu. Choose
the name of the machine running your job in the Select Host dialog shown below
and click OK.

Note that:
• Only STAR jobs owned by you and only those that have registered with the
StarWatch daemon can be selected
• Registration usually takes place roughly at the end of the first iteration
• If STAR cannot find the daemon, it will keep trying for a small amount of
time and then continue without trying further contact.

17-16 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
The StarWatch Utility

Now choose the PID of the STAR job you wish to monitor from the list displayed
in the Select STAR Job dialog and click OK.

StarWatch should now start displaying the monitored flow variables against
iteration number or time step.
Choosing the monitored values
The following choices are available:
1. Material (stream) number
In multi-stream applications, select the stream you wish to monitor using the
Material Number slider control.
2. Field or residual values
Select the type of variable to be displayed by clicking the toggle button at the
bottom of the Legend section. The button label changes from Plot Field
Values to Plot Residual Values and vice versa, depending on your choice.
The labelling and scale of the adjacent graph also changes accordingly.
3. Monitored variable
Choose the flow variables to be monitored, in terms of either field or residual
values, by clicking the option buttons next to the variable names. The latter
appear in the Legend section under the Property column and comprise the
three velocity components, turbulence kinetic energy and dissipation rate,
pressure and temperature. The colour used to display each variable is shown
next to the name.
It is also possible to monitor changes in scalar variables, if present in your
model, by selecting View > Selected Data > Scalar Variables from the
menu bar. The contents of the Legend section and the graph labelling will
change accordingly. The method of selecting scalars is the same as for the
main (global) variables. Note that since only seven quantities can be
monitored, option View > Select Scalars lets you decide which scalars you
want to look at; by opening a secondary (Select Scalars) dialog in which the
required scalars and the order in which they appear in the StarWatch display
may be determined.
Controlling STAR
At the beginning of a numerical analysis, STAR reads all files prepared for it by
pro-STAR. Many of the parameters set in pro-STAR can be viewed and altered
dynamically while the solution is in progress by selecting Settings > STAR
Control Variables from the StarWatch menu bar. This brings up the Star Control
Variables dialog shown below:

Version 4.02 17-17


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
The StarWatch Utility

The dialog’s purpose is to allow the user to interactively change the values of
several STAR solution and output control parameters. These are grouped into six
tabs according to function, as shown above, and all act in the same way. The
meaning of the available parameters is listed in the table below:

Parameter Meaning
General Settings
DT Time step size
MAXCOR Maximum number of correctors for the PISO algorithm
RESOC Residual tolerance for the PISO algorithm
SORMAX Overall convergence criterion
IJKMON Monitoring cell number for fluid domains
File Output
ECHO =.T. Control information will be written to file .info
BOECHO =.T. Boundary data will be written to file .info
ITEST =.T. Write all conservation balance information to file .info
IRESI =.T. Write all solver convergence information to file .info
NDUMP Frequency of writing data to file .ccm
NFSAVE Backup frequency (frequency of saving file .pst_iternum)
NCRPR Number of cell Courant numbers (starting from the largest) to
be printed out
NFRRE Iteration frequency for dumping residuals to file .ccm

17-18 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
The StarWatch Utility

Parameter Meaning
Under-Relaxation Factors
FPCR Under-relaxation factor for pressure correction (PISO)
FUVW Under-relaxation factors for velocities
FP Under-relaxation factor for pressure
FTE Under-relaxation factors for k and ε
FT Under-relaxation factor for temperature
FTVS Under-relaxation factor for turbulent viscosity
FDEN Under-relaxation factor for density
FLVS Under-relaxation factor for laminar viscosity
FCON Under-relaxation factor for heat conductivity
FR Under-relaxation factor for radiation
Blending Factors
GGUVW Blending factor for velocities
GGKE Blending factor for k and ε
GGT Blending factor for temperature
GGDEN Blending factor for density
GGSCA Blending factor for scalars
Residual Tolerances
SORU Solver residual for U velocity
SORV Solver residual for V velocity
SORW Solver residual for W velocity
SORP Solver residual for pressure
SORK Solver residual for k
SORE Solver residual for ε
SORT Solver residual for temperature
Number of Sweeps
NSWPU Total number of solver sweeps for U in one run
NSWPV Total number of solver sweeps for V in one run
NSWPW Total number of solver sweeps for W in one run
NSWPP Total number of solver sweeps for P in one run
NSWPK Total number of solver sweeps for k in one run
NSWPE Total number of solver sweeps for ε in one run
NSWPT Total number of solver sweeps for T in one run

Solution control can then be exercised as follows:


1. During execution, monitor the behaviour of normalised residual sums for each
variable being solved for, by looking at the displayed values at the end of each

Version 4.02 17-19


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
The StarWatch Utility

iteration or time step. In addition, look at the flow variable values at the
monitoring location, as specified in the “Monitoring and Reference Data”
STAR GUIde panel.
2. While monitoring this display, you may decide to alter the course of the
calculations by altering a model parameter, e.g. by
(a) re-specifying an under-relaxation factor in order to speed up solution
convergence
(b) increasing the value of parameter SORMAX to stop the run at an earlier
stage

The values currently in use are shown on the dialog. If you want to change
one or more of them, enter the new value in the appropriate box(es) and click
Apply. This change is treated as pending. You can now either Cancel the
change and then make others, or click Send to confirm it.
3. In the latter case, the parameter(s) will change inside STAR from the
beginning of the next iteration (or time step) following the Send operation
and a marker will be placed on the graph indicating the point at which
something was changed.

Note also the following points:


• The colour of marker matches the colour of the tab in which the alteration was
made and STAR itself will print a message indicating the change
• If you make multiple changes, you can highlight any one line and use the
dialog’s Edit menu to copy/paste that line into other boxes and then edit any
of the numbers. If you do not copy a line in, the code assumes that you are
making changes to the last line.
• StarWatch also keeps a control history file called casename.ctrl.hist
recording the changes made during a run. If you re-run a job without
removing the control history file, STAR will make the same changes to the job
that you made during the original run (so you can duplicate and repeat your
changes to, say, under-relaxation factors).
• You do not have to have StarWatch running for the above changes to take
place at various iterations. STAR will read the casename.ctrl file and
make the changes to the run at the appropriate iteration. If you do not want the
run changed the same way, delete casename.ctrl.hist before
re-running a job.
Manipulating the StarWatch display
The monitored variables chosen in the previous section are continuously displayed
in the StarWatch panel as the calculation progresses, in two ways:
• As numerical values in the Iteration / Time Step Data section. The maximum
and minimum values reached so far and the change since the previous
iteration are also shown.
• As a graph of variable value versus iteration number/time step.

The detailed appearance of this graph may be adjusted as follows:

17-20 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
Hard Copy Production

1. Horizontal scale
Use the H: slider to achieve a reasonable scale, depending on the number of
iterations
2. Vertical scale
Use the V: slider to achieve a reasonable scale, depending on the variable
being monitored. Note that this scale changes automatically as you switch
from residual to field values.
3. Horizontal range
Use the Iteration Number / Time Step slider to move the graph window to the
required iteration range, after the job has finished executing.
4. Vertical range
Use the vertical slider to move the graph window to the desired variable value
range. Whether you need to do this or not depends on the vertical scale
chosen.
5. Display size
Select View > Partial View from the menu bar to reduce the extent of the
StarWatch display, which now only shows the graph and associated legend.
Selecting View > Full View restores the original display.

Monitoring another job


If you have several STAR jobs running simultaneously and you want to switch your
monitoring to a different job, follow the procedure below:
Step 1
Select Connect > Disconnect from the menu bar to terminate monitoring of the
current job.
Step 2
Select Connect > Host, enter the name of the machine running the job you wish to
monitor in the Select Host dialog and click OK.
Step 3
Choose the job’s PID from the list displayed in the Select STAR Job dialog and click
OK. StarWatch should now start displaying the monitored variables for the new
job.

Alternatively, you may simply open another window and load another StarWatch
panel, as described in “Running StarWatch”. Note that the number of panels that
may be open simultaneously will depend on the setting specified in file
/etc/services.

Hard Copy Production


Neutral plot file production and use
To obtain hard copy of a screen plot, switch the graphical output temporarily to the
neutral plot file (see “Plot file (.plot)” on page 17-4). Once the required plot is
on-screen, type

TERMINAL,,FILE,RAST
(switches to the neutral plot file in raster, i.e. colour-fill, mode)

Version 4.02 17-21


OTHER STAR-CD FEATURES AND CONTROLS Chapter 17
Hard Copy Production

or

TERMINAL,,FILE,VECT
(switches to the neutral plot file in vector, i.e. line-contour mode)

followed by

REPLOT
(sends the picture to this file)
TERMINAL,,
(switch output back to the screen)

The above process can be repeated as often as is necessary to write all required plot
data to file case.plot.
It is recommended that colour plots destined for a black-and-white printer should
be converted to the grey-scale shading scheme (see “Colour settings” on page 4-10)
before sending them to the neutral plot file. This can be done either by selecting the
Post - Gray option in the Color Tool or by typing command

CLRTABLE,GRAY

To produce the hard copy, process the pictures stored in the neutral plot file outside
the pro-STAR environment using one of the supplied programs in the PLOT suite.
The latter are special graphics post-processors that either
• generate files suitable for plotting on a given type of hard-copy device, or
• display the contents of the neutral plot file on your screen (see Appendix B in
the Post-Processing User Guide for more details).

The PLOT programs available on your particular installation are normally accessed
by opening a window and typing

plot

This produces a response of the form:

Please enter the required plot driver:


Available drivers are:
ai fr gif hp ps pst su x xm [xm]

where

ai — Adobe Illustrator file output


fr — Adobe Freehand file output
gif — GIF file output
hp — HP Graphics Language file output
ps — PostScript file output
pst — utility for adding an extra title to an existing PostScript file
su — utility for reducing the size of an existing neutral plot file by removing
hidden polygons
x — X-windows terminal display
xm — X Motif graphics display

17-22 Version 4.02


Chapter 17 OTHER STAR-CD FEATURES AND CONTROLS
Hard Copy Production

Type the desired option and then follow the instructions on your screen, supplying
additional information as required. Note that options such as xm are suitable for
screen displays while options such as ps are for hard copy production.
Note also that extended mode features such as translucency, layers, and
smooth-shaded contour plots cannot be represented in the neutral plot format. To
produce high-resolution hard copies in extended mode, use the high-resolution
screen capture technique described in Chapter 2, “Screen capture”.
Scene file production and use
STAR-CD scene files provide a convenient way to store a fully post-processed
model in a format that can be subsequently viewed with the lightweight and quick
STAR-View viewer program. A STAR-CD scene file (extension .scn) stores the
current state of the extended-mode graphics window, including the current plot and
any labels, legends, and other screen information. However, unlike conventional
hard copies produced using pro-STAR’s neutral plot facilities, STAR-CD scene
files store a full 3-D representation of the current model so the view can be rotated,
translated, and zoomed interactively in the STAR-View program.
To produce such a file, first generate the desired plot in extended (OpenGL)
mode (see Chapter 2, “Plotting Functions”). This can include any effects available
in extended mode, including multiple layers, translucency, and smooth-shaded
contours. Once the desired plot is achieved, select Utility > Write STAR-CD
Scene File from the main pro-STAR menu. Select or type the desired scene file
name into the File Selection dialog box which appears and press OK to write the
file. Alternatively, pro-STAR command SCENE can be used to record the file.
Once this file is written, simply run the STAR-View program by typing

starview filename.scn
in an X-window, where filename.scn is the file name containing the desired
scene. Once the latter is loaded, the view in the model can be manipulated via the
mouse in exactly the same way as in pro-STAR.

Version 4.02 17-23


APPENDICES

CCM USER GUIDE

STAR-CD VERSION 4.02

CONFIDENTIAL — FOR AUTHORISED USERS ONLY

© 2006 CD-adapco
Appendix A pro-STAR CONVENTIONS
Command Input Conventions

Appendix A pro-STAR CONVENTIONS


Command Input Conventions
1. A single command line may not be longer than 320 characters

2. Input is mostly case-insensitive; both capital and small letters are accepted
(arguments such as file names, titles and screen labels are case-sensitive)

3. Command names may be abbreviated by the first four letters (with one
exception: *ENDIF). Argument keywords may also be abbreviated by the
first four letters (with one exception: parameter arguments for the MEMORY
command)

4. Fields in a command string must be separated by a comma or by any number


of spaces.

5. Multiple commands may be stacked on a single line, separated by a dollar


sign ($).
6. Any command string with an exclamation mark (!) in column 1 is interpreted
as a comment (and therefore not executed).

7. Double plus signs (++) at the end of a line indicates that the next line is a
continuation of the current line. Individual arguments are not continued on a
new line; the new line will begin a new argument. Any number of lines may
be continued in this manner to form a single command line; however, the total
number of characters in a command line formed in this manner may still not
exceed 320 characters.

8. Any command may be entered from any module

9. In NOVICE mode (see command EXPERT), the program will prompt for
arguments needed to execute the command. Command ABORT may be used at
this prompt to abort the current command without performing any action.

10. Basic arithmetic is allowed on all command lines. Each operator must be
separated by blanks or a comma from the numbers or parameters on either
side. For example, the following command

VLIST 10 * 10, A + 7 1000 / B

is interpreted as VLIST 100 to (A+7) by (1000/B), where A and B are


numeric parameters defined by the *ASK, *SET or *GET commands. All
terms are evaluated strictly from left to right.

11. The keyword ‘ALL’ may be used in lieu of any vertex, cell, boundary, etc.
range to denote that all items are to be used for the range. (Examples:
CLIST,ALL and CTMOD,ALL,,,FLUID)

12. The appropriate item set keyword may be used in lieu of most item ranges to
denote that all items in the set are to be used for the range. (Examples:
CPDEL,CPSET and VLIS,VSET,,,1)
Version 4.02 A-1
pro-STAR CONVENTIONS Appendix A
Command Input Conventions

Keyword Item Set

VSET Current vertex set


CSET Current cell set
BSET Current boundary set
SPLSET Current spline set
BLKSET Current block set
CPSET Current couple set
DSET Current droplet set

13. The following keywords may be used in lieu of many item ranges to display
the crosshair cursor in the plot window so the user may select a set to be used
as the range. (Example: CLIST,CCRS).

Keyword Select

VCRS Vertex set


CCRS Cell set
BCRS Boundary set
SCRS Spline set
BLKCRS Block set
DCRS Droplet set

14. The following keywords may be used in lieu of entity numbers. (Example:
V,MXV,1.0,2.0,3.0)

Keyword Interpreted As

MXV Highest numbered vertex + 1


MXC Highest numbered cell + 1
MXB Highest numbered boundary + 1
MXS Highest numbered spline + 1
MXK Highest numbered block + 1
ICUR Currently active coordinate system

15. Certain keywords (which may also be used in lieu of entity numbers) will
cause pro-STAR to display the crosshair cursor in the plot window and expect

A-2 Version 4.02


Appendix A pro-STAR CONVENTIONS
Help Text / Prompt Conventions

the user to select an item, as specified by the following description: (Example:


STLIST,SXT)

Keyword Select Interpreted As

BLKX Block Block number


BX Boundary Boundary number
BXP Boundary Boundary patch number
BXR Boundary Boundary region number
CX Cell Cell number
CXC Cell Cell colour index
CXG Cell Cell group number
CXM Cell Cell material number
CXP Cell Cell porous number
CXS Cell Cell spin index
CXT Cell Cell type number
DRX Droplet Droplet number
DRXT Droplet Droplet type number
SX Spline Spline number
SXC Spline Spline colour index
SXG Spline Spline group number
SXT Spline Spline type number
VX Vertex Vertex number

Help Text / Prompt Conventions


1. Words between slashes (e.g. /ANY/ALL/) represent legal alternatives for the
field.

2. Numbers in parentheses represent defaults for the immediately preceding


variable.

3. Variables beginning with ‘NV’ refer to vertices


Variables beginning with ‘NC’ refer to cells
Variables beginning with ‘NB’ refer to boundaries
Variables beginning with ‘NSPL’ refer to splines
Variables beginning with ‘NBLK’ refer to blocks
Variables beginning with ‘NCP’ refer to couples
Variables beginning with ‘NDR’ refer to droplets
Version 4.02 A-3
pro-STAR CONVENTIONS Appendix A
Control and Function Key Conventions

Control and Function Key Conventions


1. The following short-cuts using the Ctrl key are available:

Control Key Command

Ctrl-a CSET,ALL
Ctrl-e ZOOM,OFF $REPLOT
Ctrl-h Query for help
Ctrl-o ZOOM,OFF $REPLOT
Ctrl-q QUIT
Ctrl-r REPLOT
Ctrl-s SAVE,,
Ctrl-w Zoom out (by a factor of 2)
Ctrl-z Zoom in (by a factor of 2)

2. Function key short-cuts can be defined or changed using the Function Keys
option in the Utility menu. The default function key short-cuts are:

Function Key Default Command

F5 Repeat last command


F6 REPLOT
F7 CPLOT
F8 ZOOM,OFF $REPLOT

File Name Conventions


The default name for any file read or written by the program is casename.ext,
where casename is defined by the user and ext is the file name extension. If you
enclose the file name in quotes, the extension default will be overridden and the
exact name within the quotes will be used.

A-4 Version 4.02


Appendix B FILE TYPES AND THEIR USAGE

Appendix B FILE TYPES AND THEIR USAGE

File Usage
Extension
.ani Default input/output for recording animation commands
.anim Default save file for animation options
Backup (i.e. previous version) of the current pro-STAR model file
.bak
(binary)
.bnd Default input/output for boundary definitions
.btr STAr file used for storing beam tracking data
.ccd STAR file used for storing coal combustion data
.ccm STAR-CD data repository (restart) file (binary/direct access)
.cel Default input/output for cell definitions
.cel Default output for surface cell definitions
.cgns Default input/output for CGNS data files
.cgrd Default input file containing grid change commands
.chm Default output file for chemical scheme definitions (coded)
Default temporary storage of ‘frozen’ vertex data used with the
.cpfz
SAVE and MAP options of command CPFREEZE
.cpl Default input/output for coupled cell definitions
.ctrl Editable file for interactive solution control
.dat Tecplot™ post data output file
.div Post data file created when the solution diverges
.domain ICEM CFD™ post data output file
Default output for droplet definitions (written with command
.drp
PROBLEMWRITE)
.drpc Default input/output for droplet data (coded)
.ecd File for storing engineering data for cell monitoring
File for storing dispersed-phase engineering data for cell moni-
.ecd2
toring
.echo Echo of all input typed by the user
.elem Default input/output for ANSYS™ element definitions
.erd File for storing engineering data for boundary region monitoring
File for storing dispersed-phase engineering data for boundary
.erd2
region monitoring
.evn Default transient event save file (binary/direct access)
.evnc Default input/output for ASCII event data files
.fac File containing cell face definitions

Version 4.02 B-1


FILE TYPES AND THEIR USAGE Appendix B

File Usage
Extension
.fvbnd Default input for GRID3D boundary data files
.g3d Default input for GRID3D cell and vertex data files
.gen Default output for GENERIC data
.grf Default graph register data save file
.grf Default graph register ‘GET’ file
Default combustion data file (binary), used in advanced IC engine
.ics
models
.info Run-time optional output file
.inp Any file containing pro-STAR commands
.inp Default input/output for miscellaneous problem data definitions
.lfb File containing group and colour information for particles
.loop Default save file of current loop information
.mdl Default pro-STAR model file (binary)
.mdl Default input for SMAP-type data
.msh TGRID™ data output file
.nas Default input/output for NASTRAN™ data files
.neu Gambit™ data output file
.node Default input/output for ANSYS™ node definitions
.out Default output file
.pat Default input/output for PATRAN™ data files
Look-up table file created when using PPDF chemical reaction
.pdft
models
.pgr File containing participating media radiation data (binary)
.plot Neutral plot file
.prob Default output for STAR-CD problem data file (coded)
File containing cell-to-processor mapping information used in
.proc
STAR-HPC runs
.pstt Default transient solution file (binary/direct access)
Refinement data file used by the adaptive refinement commands
.refi
(CMREFINE / CMUNREFINE)
Residual history file for phase no. 2 (used in Eulerian two-phase
.reu
problems)
.rsi Default residual history file (binary/direct access)
Default input/output of residual histories for
.rsic
BINARY-CODED-BINARY file conversions
.run Standard run-time output file
.scl Default output for scalar variable definitions (coded)

B-2 Version 4.02


Appendix B FILE TYPES AND THEIR USAGE

File Usage
Extension
Default output for set definitions (written with the SETWRITE
.set
command)
.spd File for storing engine data (coded)
.spl Default input/output for spline definitions
Default output for plotting-surface database (used to skip surface
.srf
creation step in future plots)
.stl Default input for STL data files
.tabl Default input file for droplet spray tables
.tbl Default file for storing general table data
.trk Default input/output for particle/droplet tracks
Default input/output of particle/droplet track data for
.trkc
BINARY-CODED-BINARY file conversions
.trnc Default input for transient load data (coded)
.trns Default transient history save file (binary/direct access
.unv Default input/output for IDEAS™ (SDRC) universal file
.uns Fieldview™ data output file
.usr Default input/output for ASCII post data
.vfs STAR file used for storing view factors
.vrml Virtual reality data output file
.vrt Default input/output for vertex definitions
.vrt Default output for surface vertex definitions

The format for vertex definitions is: (file case.vrt)


Vertex number, X, Y, Z (global coordinates) (I9, 3(1X,G21.14))

The format for boundary definitions is: (file case.bnd)


Boundary number, cell number, face number, region number, patch number,
region type (characters) (5(I9,1X),A)

The format for spline definitions is: (file case.spl)


Spline number, number of vertices, spline type (3I9)
Up to 100 vertex numbers defining the spline (8I9)

The format for couple definitions is: (file case.cpl)


Couple number, number of cells (I8, 1X, I5)
Up to MAXNCP cell number/face number combinations 7(I9,I2)

The format for ASCII input to be used as post-processing data is: (file case.usr)
Vertex and/or cell number (as appropriate), scalar value (I9, 6X, 6G16.9).

Version 4.02 B-3


Appendix C PROGRAM UNITS

Appendix C PROGRAM UNITS


Property Units (SI) Units (English)
AREA m2 ft2
CONDUCTIVITY W/mK Btu/(hr × ft× F)
DENSITY kg/m3 lbm/ft3
DIFFUSIVITY m2/s ft2/s
DYNAMIC VISCOSITY Pa × s psi × s
FORCE N lb
HEAT FLUX W/m2 Btu/(hr × ft2)
HEAT OF FORMATION J/kg Btu/lbm
HEAT OF VAPOURIZATION J/kg Btu/lbm
LENGTH m ft
MASS kg lbm
MASS FLOW RATE kg/s lbm/hour
MOLECULAR WEIGHT kg/kmol lbm/kmol
PRESSURE Pa (N/m2) psi
SPECIFIC HEAT J/(kg × K) Btu/(lbm × F)
SPEED OF SOUND m/s ft/s
SURFACE TENSION COEFFICIENT N/m lb/ft
TEMPERATURE K (° Kelvin) R (° Rankine)
TIME s s
TURBULENCE KINETIC ENERGY k m2/s2 ft2/s2
TURBULENCE DISSIPATION RATE ε m2/s3 ft2/s3
VELOCITY m/s ft/s
VOLUME m3 ft3
VOLUMETRIC EXPANSION COEFF. 1/K 1/R

Version 4.02 C-1


Appendix D pro-STAR X-RESOURCES

Appendix D pro-STAR X-RESOURCES


The Motif version of pro-STAR utilises standard X resources for defining the layout
and look of its windows. While default values for these resources are built into the
program, you can override the defaults in two different ways:
1. The easiest method is to put resource definitions in your .Xdefaults file.
This file is read by the Motif window manager when you log in or restart the
window manager. Any changes made to this file do not take effect until either
you log in again or you issue an xrdb command to re-read the X resource
data base. Typically, you will issue the command as follows:

xrdb -merge .Xdefaults include the full path to the


.Xdefaults file if you are not in your
home directory

2. Any file can be used to set X resources. The only significance of the
.Xdefaults file is that it is read automatically on start-up. You could, for
example, create a file called PROSTAR.resources and put the resource
definitions in that file. In this case, you would have to issue the command:

xrdb -merge PROSTAR.resources

before running pro-STAR in order to activate those definitions

The following describes some useful resource definition commands:

Prostar*background: The default background colour for all


pro-STAR applications

Prostar*foreground: The default foreground colour for all


pro-STAR applications

Prostar.geometry: The size and position of the pro-STAR


graphics window

Prostar.defaultFontList: The font used for the pro-STAR graph-


ics window menus

Prostar.OutputWindow.geometry: The size and position of the pro-STAR


output window

Prostar*cmdForm1Widget.height: The height of the output history por-


tion of the pro-STAR output window

Prostar*cmdForm2Widget.height: The height of the input portion of the


pro-STAR output text window

Prostar*Prostar_Output_Text.fontList: The font used in the pro-STAR output


window

Version 4.02 D-1


pro-STAR X-RESOURCES Appendix D

Prostar*Prostar_Output_Text.foreground: The foreground colour used in the


pro-STAR output window

Prostar*Prostar_Output_Text.background: The background colour used in the


pro-STAR output window

Prostar*panel_name_B1.background: The background colour of button 1 in


the user panel named panel_name.
Buttons in panels are numbered start-
ing from zero and are incremented by
1 from top to bottom and from left to
right. Any panel button can be defined
using the proper panel name and but-
ton number.

Prostar*panel_name_B1.foreground: The foreground colour of button 1 in


the user panel named panel_name.

Prostar*panel_name_B1.fontList: The font used for button 1 in the user


panel named panel_name.

Prostar*macro_editor_text.fontList: The font used for the text section of the


macro edit dialog

Prostar*macro_editor_text.foreground: The text foreground colour used in the


macro edit dialog

Prostar*macro_editor_text.background: The text background colour used in the


macro edit dialog

Prostar*GUIde_INDEXCARD.background: The default background colour for all


index cards (tabs) inside a STAR
GUIde panel

Prostar*GUIde_TABS.background: The default background colour for all


sub-index cards (sub-tabs) inside a
STAR GUIde panel

X colour names are usually (but not always) defined in the file:

/usr/lib/X11/rgb.txt

Geometry definitions are in the form of W × H + X + Y where W is the width (in


pixels), H the height, X the distance (in pixels) from the left of the screen to the left
side of the window, and Y the distance from the top of the screen to the top of the
window. Heights are also defined in pixels.
Available font list names can usually be found by issuing the command:

xlsfonts

D-2 Version 4.02


Appendix D pro-STAR X-RESOURCES

The following shows a sample of resource definitions that could be used with
pro-STAR:

Prostar*background: paleturquoise3
Prostar*foreground: black

Prostar.geometry: 800x800+480+0
Prostar.defaultFontList:
-adobe-helvetica-bold-r-normal--14-140-75-75-p-82-
iso8859-1

Prostar.OutputWindow.geometry: 1000x870+0+0
Prostar*cmdForm1Widget.height: 700
Prostar*cmdForm2Widget.height: 70

Prostar*Prostar_Output_Text.fontList:
-adobe-courier-bold-r-normal--18-180-75-75-m-110-
iso8859-1
Prostar*Prostar_Output_Text.foreground: blue
Prostar*Prostar_Output_Text.background: gray85

Prostar*new_panel_B1.background: Red
Prostar*new_panel_B1.fontList:
-adobe-courier-medium-r-normal--12-120-75-75-m-70-
iso8859-1
Prostar*new_panel_B2.background: Green
Prostar*new_panel_B2.fontList:
-b&h-lucida-medium-r-normal-sans-24-*-*-*-*-*-
iso8859-1

Prostar*macro_editor_text.fontList:
-adobe-courier-bold-r-normal--18-180-75-75-m-110-
iso8859-1
Prostar*macro_editor_text.foreground: blue
Prostar*macro_editor_text.background: skyblue

To customise the opening locations of Tools, Lists, etc. in pro-STAR


If you run the XMotif version of pro-STAR, it is possible to arrange for tools to
open in repeatable locations. This is especially useful if you have a number of
favourite tools that you open each time and can make pro-STAR open them every
time via the PROINIT file.
There are two steps in doing this. The first is finding out where you want the tool
to be. To this end, run pro-STAR and then place (and optionally resize) the tool to
get the desired effect. Follow this by issuing the xwininfo command from an
X-window to get the necessary numbers. For example:

ibm3<68>xwininfo

Version 4.02 D-3


pro-STAR X-RESOURCES Appendix D

xwininfo: Please select the window about which you


would like information by clicking the
mouse in that window.

xwininfo: Window id: 0x54007c2 "Check Tool"

Absolute upper-left X: 587


Absolute upper-left Y: 374
Relative upper-left X: 0
Relative upper-left Y: 0
Width: 630
Height: 590
Depth: 8
Visual Class: PseudoColor
Border width: 0
Class: InputOutput
Colormap: 0x3d (installed)
Bit Gravity State: ForgetGravity
Window Gravity State: NorthWestGravity
Backing Store State: NotUseful
Save Under State: no
Map State: IsViewable
Override Redirect State: no
Corners: +587+374 -63+374 -63-60 +587-60
-geometry 630x590-55-52

This gives us two pieces of information, the name and the location. The name is
enclosed in quotes in the first line of output, for this case it is Check Tool. The
location is given in the last line, -geometry 630x590-55-52. This gives the
width and height as well as the location.
The second step is to feed this information to pro-STAR via Xresources. The
usual way is to edit file .Xdefaults in your home directory. In this case, add the
following line:

Prostar*CheckTool*Geometry: 630x590-55-52

This line is made up as follows:

Prostar*NAME*Geometry: GEOMETRY

where:

NAME is the name of the window stripped of all spaces; capitalisation must
be kept.
GEOMETRY is the location of the window as found from the previous
command.

Once this line has been added to the file, pro-STAR should respond correctly. On
some systems, restarting pro-STAR will suffice. Others may require you to log out
and log in again or issue some variant of the xrdb command.
The above has been tested and works so far on SGI and IBM machines. Other
D-4 Version 4.02
Appendix D pro-STAR X-RESOURCES

machines may work with minor variations.


A suitable PROINIT file will be:

opanel tool$check

Make sure that the PROINIT file is in your current directory or that it is pointed to
by the STARUSR environment variable.

Version 4.02 D-5


Appendix E USER INTERFACE TO MESSAGE PASSING ROUTINES

Appendix E USER INTERFACE TO MESSAGE PASSING


ROUTINES
Some user coding might need access to message passing routines when used in a
parallel run. This appendix lists the parallel message passing calls that may be used
within the user coding.

IGSUM — Global Integer Summation


Synopsis
INTEGER FUNCTION IGSUM (LOCSUM)
Parameters
INTEGER LOCSUM — local value
Returns integer sum of LOCSUM

GSUM — Global Floating Point Summation


Synopsis
REAL1 FUNCTION GSUM (LOCSUM)
Parameters
REAL1 LOCSUM — local value
Returns floating point sum of LOCSUM

DGSUM — Global Double Precision Summation


Synopsis
DOUBLE PRECISION FUNCTION DGSUM (LOCSUM)
Parameters
DOUBLE PRECISION LOCSUM — local value
Returns double precision sum of LOCSUM

LGLOR — Global OR operation


Synopsis
SUBROUTINE LGLOR (LOC,GLO)
Parameters
LOGICAL LOC — local value (input parameter)
LOGICAL GLO — global value (output parameter)

LGLAND — Global AND operation


Synopsis
SUBROUTINE LGLAND (LOC,GLO)
Parameters
LOGICAL LOC — local value (input parameter)
LOGICAL GLO — global value (output parameter)

1. Type REAL becomes DOUBLE PRECISION in double precision runs.


Version 4.02 E-1
USER INTERFACE TO MESSAGE PASSING ROUTINES Appendix E

GMAX — Global MAX operation


Synopsis
REAL1 FUNCTION GMAX (LMAX)
Parameters
REAL1 LMAX — local value
Returns global MAX of LMAX

GMIN — Global MIN operation


Synopsis
REAL1 FUNCTION GMIN (LMIN)
Parameters
REAL1 LMIN — local value
Returns global MIN of LMIN

IGMAX — Global MAX operation


Synopsis
INTEGER IGMAX (ILMAX)
Parameters
INTEGER FUNCTION IGMAX — local value
Returns global MAX of ILMAX

IGMIN — Global MIN operation


Synopsis
INTEGER FUNCTION IGMIN (ILMIN)
Parameters
INTEGER IGMIN — local value
Returns global MIN of ILMIN

1. Type REAL becomes DOUBLE PRECISION in double precision runs.


E-2 Version 4.02
Appendix F STAR RUN OPTIONS
Usage

Appendix F STAR RUN OPTIONS


Usage
star [options] [node1 [node2 [node3 [...]]]]

Options
-version Show STAR version information, which includes patch
number.
-abort Send SIGABRT to stop STAR after the current iteration
or time step.
-batch Generate script for running batch job. Useful if run is to
be submitted via a batch-queuing system like IBM
Loadleveler, LSF, OpenPBS, PBSPro, Sun Grid Engine
or Torque. This requires STAR-NET 3.0.3 or later to be
installed.
-case=casename Select the case name manually. This option is not
needed in general.
-chktime=minutes Enable STAR controlled check-pointing at a regular
interval in minutes for fault tolerance. The default is off.
-chkdir=directory Select directory for storing the check-pointed data. The
default is to use a ‘CHECK’ sub-directory.
-chkpnt Perform manual check-pointing of STAR results now.
This option may be useful for visualising fields while
STAR is still executing in parallel, since it will merge
the case’s results.
-collect Collect and save data from previous crashed run only.
-dp Make STAR-CD run in double precision arithmetic.
Current default is single precision, with the exception of
combustion problems which use either STAR/KINetics
or the Complex Chemistry model, in which case
STAR-CD will execute in double precision.
-devtool="program" Attach a development tool like a debugger to a
STAR-CD run. The use of this option is advised on
sequential runs only. For parallel runs only LAM MPI
and MPICH are fully supported with Totalview.
-fork Enable the use of fork() for starting local external
moving mesh codes and NFS-based communications.
-g Compile ufile source code, so that the user may employ
a debugger to perform a step-by-step analysis of the
coding in the user subroutines. See also option
"-devtool".
-kill Send SIGKILL to terminate STAR immediately.

Version 4.02 F-1


STAR RUN OPTIONS Appendix F
Options

-nolookahead Disable look ahead for socket-based external moving


mesh code communications.
-noramfiles Disable memory based scratch files.
-norecalc Disable the recalculation of radiation view factors.
-norestart Disable restart (if selected in the problem file) by
resetting the restart flag.
-nosave Disable saving of results by using an empty save list.
The "-save=" option can be used to make a new save list.
-noskip Forces geometry decomposition (if applicable), events
preparation (if applicable), user coding compilation and
copying of input files (if applicable) before STAR-CD
starts to execute.
-noufile Ignore user coding in the "ufile" directory, i.e. the
run’s results will not be influenced by the actual user
coding.
-restart Continue the run from an existing restart file by
resetting the restart flag in the problem file.
-save="filelist" Specify additional output files for treatment as results.
On a parallel run, these files will be merged into a single
file. Ideally, these files should be formatted into two
columns: the first column containing an index numeral
that can be ordered (i.e., pro-STAR cell number), and
the second column containing the physical quantity of
interest. Files that should not be merged should be left
out from this option. Wildcards “*” and “?” are
accepted.

Example:
-save="file1.dat file2.dat" or
-save="file1.dat" -save="file2.dat"
-set variable="value" Set environmental variable to a value, especially on a
parallel run, where the variable will be set on all
processes.

Example:
star -set MYVAR="on"
-timer Enable printing of detailed timing data. Use this option
to extract execution time information from the run.
Please note that the use of this option entails a
performance penalty.
-toolchest Build new STAR toolchest from plug-in tools.
-ufile Compile user coding and build new plugable object
only. Useful to verify if user coding compiles, i.e., if it
contains any syntax mistakes.

F-2 Version 4.02


Appendix F STAR RUN OPTIONS
Parallel Options

-uflags="flags" Select additional flags for compiling user coding. This


option gives the user added flexibility in using other
compiler options that may not be listed in installation
scripts.
-ulib="librarylist" Specify precompiled user coding libraries and/or some
additional dynamic shared objects required by user
coding.
-watch Enable connection to the StarWatch daemon. The
daemon itself and the StarWatch GUI still need to be run
separately

Parallel Options
-copy="filelist" Specify additional input files for copying to domains on
a parallel run.

Example:
-copy="file1.dat file2.dat" or
-copy="file1.dat" -copy="file2.dat"
-decomp Run geometry decomposer only. Useful to check the
outcome of the decomposition if it has to satisfy certain
criteria.
-decomphost=hostlist Selects host(s) for running the decomposer (i.e.
host1:host2:…). In particular, for the Parmetis
decomposition option more than two cpu’s (whether or
not on the same machine) should be used. The number
of cpu’s to decompose the mesh can be smaller than the
number of requested partitions.

Example:
-decomphost="host1,2 host2" 5
In the above, STAR will decompose the mesh in 5 parts
using 2 cpu’s on machine "host1" and 1 cpu on machine
"host2".

Version 4.02 F-3


STAR RUN OPTIONS Appendix F
Parallel Options

-decompmeth=method Select one of the decomposition methods listed below.


The abbreviations shown in parentheses can be used
instead. Their individual meanings are:

optimised (o): The decomposition will be read from file


<casename>.proc, composed of two columns:
first column contains cell numbers, second column
contains the process number to which the cells are
going to be assigned.
automatic (a): The decomposition will uniformly
divide the number of cells between the intended
number of processes, based purely on pro-STAR
cell numbering.
manual (m): The decomposition is done according to
cell types, as they have been defined in pro-STAR
sets (s): The decomposition is read from a .set file, as
it has been defined in pro-STAR.
metis (x): The mesh will be partitioned with Metis, a
built-in graph-handling library. By default, each
material domain will be decomposed in turn.
ometis (y): Same as above, but with a lower memory
footprint and higher execution time.
geometric (g): The entire mesh (i.e., the mesh is treated
as if it was just one single material domain) is
decomposed in a single (Cartesian) direction in
which the model is largest.
parmetis (p): Parallel version of the Metis family of
algorithms. Parmetis executes the domain decom-
position step in parallel and requires less memory
than the Metis algorithm. Parmetis calculates
decompositions of similar quality to sequential
Metis. By default, each material domain will be
decomposed in turn. This option is to be used in
conjunction with option ‘-decomphost’, above.

The default is ‘metis’ decomposition, except when the


model contains events, in which case the default
becomes ‘sets’.

Example:
-decompmeth=g

F-4 Version 4.02


Appendix F STAR RUN OPTIONS
Parallel Options

-decompflags="flags" Special options for the domain decomposition step:

vcom: Compress vertex indices on each geometry file; if


compressed, the vertices on each geometry file will be
numbered from 1 to the local maximum number; if not
compressed, the vertices will retain their original num-
bering from the un-decomposed mesh. The vertex num-
bering may be important for the mesh motion operation
(e.g. the vertex movement may be specified relative to a
fixed vertex). The default action is to compress vertices,
except for moving mesh and liquid film cases where the
default is not to compress.

novc: Disable vertex compression.

outproc: If chosen, this option will trigger the creation


of a cell assignment file in the case’s directory; this file
(<casename>.proc), can be loaded into pro-STAR
for the user to visualise the decomposition (with com-
mand RDPROC) or it can be used to repeat the same
decomposition with, for example, a different version of
STAR in conjunction with the ‘-decompmeth=o’
decomposition option.

outsets: this option will trigger the creation of a sets file


in the case’s directory. In this file
(<casename>.sets), each set will contain the cells
that belong to a certain subdomain; this file can be
manipulated from within pro-STAR in the usual manner
or it can be used to repeat the same decomposition with,
for example, a different version of STAR in conjunction
with the ‘-decompmeth=s’ decomposition option.

Example:
-decompflags=”outproc”
-distribute Select distributed data parallel runs using local scratch
disks, as set up at the time when STAR-CD was
initialised. Please see your Systems Administrator for
details.
-loadbalance Select load balancing taking into account the relative
speeds of the hosts, as set up at the time when
STAR-CD was initialised. Please see your Systems
Administrator for details.
-mergehost=hostlist Selects host for merging results (i.e. host1:host2:…).
-mpi=auto Automatic selection of the MPI implementation using
the vendor order shown below. This is the default
behaviour which can be changed by supplying one of
the flags below:

Version 4.02 F-5


STAR RUN OPTIONS Appendix F
Resource Allocation

-mpi=os Select Operating System Vendor’s MPI


-mpi=gm Select MPICH-GM (Myricom GM MPI)
-mpi=hp Select HP MPI
-mpi=intel Select Intel MPI
-mpi=ra Select RA-MPICH (Rapid Array MPI)
-mpi=scampi Select ScaMPI (Scali MPI)
-mpi=score Select SCore MPI
-mpi=sgi Select SGI Itanium MPI
-mpi=lam Select LAM MPI
-mpi=mpich Select MPICH (ANL/MSU MPI)
-mppflags="flags" Select additional flags for message passing protocol.
Use this option to supply additional flags as expected by
the MPI implementation. In general, the user should not
need to use it.
-mpphosts Select non-default network for message passing
protocol using alternative host names, as set up at the
time when STAR-CD was initialised. Please see your
Systems Administrator for details.
-nocollect Disable data collection at the end of a distributed data
parallel run. This also disables saving of results. It is
possible to restart using the data already distributed to
the local scratch disks. Please note that any updates to
these files must be performed manually and the data can
be manually collected using the "-collect" option at the
end of the runs.
-nocopy Disable copying of input files by using an empty copy
list. The "-copy=" option can be used to make a new
copy list.
-nodecomp Do not decompose the computation mesh on a parallel
run and use the last decomposition instead. The user
should not need to use this option in general.
-noshmem Disable shared memory communications for parallel
runs on a single node.
-scratch=directory Select the scratch directory path to use on all nodes for
distributed data runs. This over rides HPC_SCRATCH
settings and must be unique for each running case.

Resource Allocation
The user does not select sequential or parallel STAR runs directly. Instead this is
automatically determined from the resources assigned by the user or the resource
manager. If STAR options are required they need to be specified before the nodes
list.

F-6 Version 4.02


Appendix F STAR RUN OPTIONS
Default Options

-mvmeshhost=host Select additional resource for running external moving


mesh code. The default is to overload the STAR master
CPU with the external moving mesh code, when one is
being used.
-nodefile=file Select nodes to use for running STAR in a file. This can
be specified on a single line or multiple lines.
-nooverload Disable overloading of the STAR master processor with
the external moving mesh code. The number of STAR
domains plus one extra process is needed in the resource
line.
node1 node2 node3 The nodes to use for running STAR. The node is
specified in the format “hostname,np”, where “np” is the
number of processes to use. The local host will be
assumed if the “hostname” is not specified and a single
process will be used if the “,np” parameter is not
supplied.

Default Options
The environment variable STARFLAGS can be set to include some default STAR
options that will be processed before any command line options. Its value is
normally set in the software initialization file (software.ini) to cater for
site-specific STAR solver options that are always used. Examples are:

STARFLAGS=-dp
STARFLAGS=-set VARIABLE="Some Value"
STARFLAGS=-mpi=mpich -noshmem -distribute -timer

The user can reset STARFLAGS manually or use a different .ini file to change
its value.
The options defined in STARFLAGS are always processed first and can be
over-written by additional command-line options, but only if an alternative option
exists. Thus, if

STARFLAGS=-mpi=mpich

the user can still use LAM MPI as follows:

star -mpi=lam

However, if

STARFLAGS=-dp

this setting cannot be modified because a single-precision option is not available at


the command line. Another example is:
STARFLAGS=-set GTIHOME=/users/netapps/gt GTISOFT_LICENSE_FILE=27005@heraclitus

Using STARFLAGS, the software administrator can set things up so that ordinary
Version 4.02 F-7
STAR RUN OPTIONS Appendix F
Cluster Computing

users need to do less work. Other examples are to make everybody run in double
precision, to always use the -distribute option, etc.

Cluster Computing
Cluster computing is widely adopted by STAR-CD users and typically consists of
computing nodes connected by network interconnect devices such as Gigabit
Ethernet, Myrinet and InfiniBand. CD-adapco have been actively working with
computer hardware and software vendors to ensure that STAR-CD takes full
advantage of progress in cluster technology.
An important aspect of this work is STAR-CD’s integration with MPIs that
support various network interconnect devices. Users can type

star -h

to check the available MPI options for the port they are using prior to issuing one of
the "star -mpi=" commands in their session window. By checking which
network interconnect devices are supported by each MPI, users can determine
whether STAR-CD works with a particular MPI/interconnect combination.
STAR-CD’s performance on a cluster is influenced by numerous factors, such as
MPI performance, interconnect latency, interconnect bandwidth and file system
performance. We have been working with our hardware and software partners to
provide benchmark data on various clusters and such data are available on
CD-adapco’s web site.
Due to the extensive range of cluster configurations and the rapid developments
in cluster technology, it is not possible to test all MPI/interconnect combinations
and to measure their performance. Users are advised to contact the relevant system
vendors to check whether a particular combination of MPI implementation and
network interconnects works with STAR-CD.

Batch Runs Using STAR-NET


STAR-NET 3.x is a new, lightweight tool for running applications in sequential and
parallel modes under a batch environment using a resource manager. It is a
completely new design, not compatible with the previous STAR-NET 2.0.x
versions (which only work with STAR-CD in parallel mode). Currently, the IBM
Loadleveler, LSF, OpenPBS, PBSPro, Sun Grid Engine and Torque resource
managers are supported through STAR-NET 3.x compliant plug-ins. Therefore,
you must install STAR-NET 3.x in order to run in batch mode or to use any of the
above resource mangers. Note also that the PBSPro and Torque are only supported
in OpenPBS compatibility mode.
Concise guidelines for running under each system are given below, assuming
prior configuration as detailed in the Installation and Systems Guide.
Running under IBM Loadleveler using STAR-NET
To run STAR-PNP under Loadleveler:
1. Create a batch.sh script by specifying the -batch option:

star -batch <options>

F-8 Version 4.02


Appendix F STAR RUN OPTIONS
Batch Runs Using STAR-NET

where <options> represents all the normal STAR-PNP flags for your job,
as described in the sections above. Note that you cannot assign a node list for
resource allocation in batch mode as this will be performed automatically by
Loadleveler.
2. Submit your job using the llsubmit command. For example:

star -batch <options> -chktime=60


llsubmit batch.sh

The llsubmit command does not allow any resource selection and so this
must be specified correctly in the batch.sh script. The following shows the
most useful settings:

# @ node_usage = shared
# @ class =
# @ node = 3
# @ total_tasks = 8

The above requests 3 nodes and a total of 8 CPUs for running the batch job.
3. The llsubmit command does not support automatic restarts and check-
pointing, so you will need to enable application-level check-pointing by
STAR-PNP as follows:

star -batch <options> -chktime=60


llsubmit batch.sh

Other useful Loadleveler commands:


• Show all my Loadleveler jobs

llq -u username

• Continuously monitor the output of job number 123

tail -f batch.o123

• Terminate job number 123 under Loadleveler

llcancel 123

• Use the built-in GUI interface for submitting and monitoring jobs

xloadl

Running under LSF using STAR-NET


To run STAR-PNP under LSF:
1. Create a batch.sh script by specifying the -batch option:

star -batch <options>

where <options> represents all the normal STAR-PNP flags for your job,
as described in the sections above. Note that you cannot assign a node list for

Version 4.02 F-9


STAR RUN OPTIONS Appendix F
Batch Runs Using STAR-NET

resource allocation in batch mode as this will be performed automatically by


LSF.
2. Submit your job to the queue using the bsub command. For example:
(a) To submit to queue starnet requesting 2 to 4 processors:

bsub -q starnet -n 2,4 batch.sh

(b) To submit to queue starnet requesting 2 to 4 processors with


LSF-controlled automatic restarts and enabling check-pointing every 60
minutes:

bsub -q starnet -n 2,4 -r -k "CHECK 60" batch.sh

It is recommended that you always enable check-pointing and automatic


restarts to allow time-windowing/high-load-enforced job migration to
work.
(c) To submit to a subset of hosts:
bsub -q starnet -m "host1 host2 host3" -n 2,4 -r -k "CHECK 60" batch.sh

Other useful LSF commands:


• Show all my LSF jobs

bjobs

• Continuously monitor the output of job number 123

peek -f 123

• Terminate job number 123 under LSF

bkill 123

• Use the built-in GUI interface for submitting and monitoring jobs

xlsbatch

Alternatively, command starnet can be used to display a brief summary of


the current LSF status.
Running under OpenPBS using STAR-NET
To run STAR-PNP under OpenPBS:
1. Create a batch.sh script by specifying the -batch option:

star -batch <options>

where <options> represents all the normal STAR-PNP flags for your job,
as described in the sections above. Note that you cannot assign a node list for
resource allocation in batch mode as this will be performed automatically by
OpenPBS.
2. Submit your job to the queue using the qsub command. For example, to

F-10 Version 4.02


Appendix F STAR RUN OPTIONS
Batch Runs Using STAR-NET

submit to queue starnet requesting 3 nodes with 2 processors each:

qsub -q starnet -l nodes=3:ppn=2 batch.sh

3. OpenPBS does not support automatic restarts and check-pointing, so you will
need to enable application-level check-pointing by STAR-PNP as follows:

star -batch <options> -chktime=60


qsub -q starnet -l nodes=3:ppn=2 batch.sh

Other useful OpenPBS commands:


• Show all my OpenPBS jobs

qstat -u username

• Continuously monitor the output of job number 123

tail -f batch.sh.o123

• Terminate job number 123 under OpenPBS

qdel 123

• Use the built-in GUI interface for submitting and monitoring jobs

xpbs

Please note that only the OpenPBS features of PBSPro and Torque are supported.
Running under PBSPro using STAR-NET
PBSPro is supported in OpenPBS compatibility mode. This means that only
OpenPBS features are supported (see the description above).

Running under SGE using STAR-NET


To run STAR-PNP under Sun Grid Engine:
1. Create a batch.sh script by specifying the -batch option:

star -batch <options>

where <options> represents all the normal STAR-PNP flags for your job,
as described in the sections above. Note that you cannot assign a node list for
resource allocation in batch mode as this will be performed automatically by
Sun Grid Engine.
2. Submit your job to a queue using the qsub command. For example:
(a) To submit to parallel environment starnet requesting 2 to 4
processors:

qsub -pe starnet 2-4 batch.sh

(b) To submit to a subset of queues:

Version 4.02 F-11


STAR RUN OPTIONS Appendix F
Batch Runs Using STAR-NET

qsub -pe starnet 2,4 -q queue1,queue2,queue3 -ckpt starnet batch.sh

3. Sun Grid Engine supports automatic restarts but not check-pointing, so you
will need to enable application-level check-pointing by STAR-PNP as
follows:

star -batch <options> -chktime=60


qsub -pe starnet 2-4 -ckpt starnet batch.sh

Please note that Sun Grid Engine versions earlier than 5.3 do not support
automatic restarts when the master host fails.

Other useful SGE commands:


• Show all my Sun Grid Engine jobs

qstat -u username

• Continuously monitor the output of job number 123

tail -f batch.sh.o123

• Terminate job number 123 under Sun Grid Engine

qdel 123

• Use the built-in GUI interface for submitting and monitoring jobs

qmon

Alternatively, command starnet can be used to display a brief summary of


the current SGE status.

Running under Torque using STAR-NET


Torque is supported in OpenPBS compatibility mode. This means that only
OpenPBS features are supported (see the description above).

F-12 Version 4.02


Appendix G BIBLIOGRAPHY

Appendix G BIBLIOGRAPHY
[1] Kee R.J., Rupley F.M. and Miller J.A. 1990. ‘The Chemkin
Thermodynamic Data Base’, Sandia Report No. SAND87-8215B.
[2] “CET89 — Chemical Equilibrium with Transport Properties”. 1989.
NASA Lewis Research Center.
[3] Liepman H.W. and Roshko A. 1957. “Elements of Gas Dynamics”. John
Wiley & Sons, New York.
[4] Shapiro A.H. 1953. “The Dynamics and Thermodynamics of
Compressible Fluid Flow — Vol. 1 and Vol. 2”. Ronald, New York.
[5] Gordon S. and McBride B. J. 1994. “Computer Program for Calculation
of Complex Chemical Equilibrium Compositions and Applications, Part I.
Analysis”, NASA Ref. Publ. 1311, NASA Lewis Research Center.
[6] McBride B. J. and Gordon S. 1996. “Computer Program for Calculation
of Complex Chemical Equilibrium Compositions and Applications, Part
II. Users Manual and Program Description”, NASA Ref. Publ. 1311,
NASA Lewis Research Center.
[7] Harten, A., Lax, P.D. and Van Leer, B. 1983. ‘On upstream differencing
and Godunov-type schemes for hyperbolic conservation Laws’, SIAM
Rev., 25, pp. 35-61.

Version 4.02 G-1


INDEX

INDEX
Commands are listed in a separate index immediately following this one

A black body 4-20


abbreviations of commands 16-1 blending factor 14-19, 15-2, 17-19
absorption coefficient 7-4 block set 2-21, A-2
absorptivity boiling. See cavitation
of baffle 4-26 boundary
thermal 4-25 cells 1-4
of wall 4-22 defining 4-2
solar 4-21 layer 1-6
thermal 4-20 in compressible flow 3-10
accuracy turbulent 1-7, 3-14
numerical 1-5 list 4-3
temporal 1-14 location 1-11, 4-1 to 4-5
view factor 7-5 maximum number of 2-18
adaptive mesh refinement. See mesh, adaptive refinement modifying 4-2
add set selection option 2-21, 4-3 monitoring behaviour 5-3
aeroacoustics 13-3 monitoring regions 4-40
aerodynamics problems, boundaries for 4-36 notional 1-4
all set selection option 2-21, 4-3 region 4-1
angle changing type 4-2, 5-4, 5-11
internal 1-5 compress 4-7
warp 1-5 default (region no. 0) 4-7
angular velocity 12-1, 12-8 defining 4-5
defined by user subroutines 14-16 set 2-21, A-2
animation 2-32 selection 4-3
anode 8-33 symmetry 1-3
append mode 2-13 types 4-5
aspect ratio visualisation 4-41
cell 1-5 to 1-6 boundary conditions 1-11
patch 7-6 attachment 4-38 to 4-39
atomisation models 9-1 displaying 12-30
attachment boundaries. See boundary conditions, baffle 4-23 to 4-27
attachment for solar radiation 4-26
autoignition (double-delay model) 8-37 to 8-39 cyclic 1-3, 1-12, 4-27 to 4-32
average nuclear radius 11-7 anticyclic 4-29
axis of rotation 12-4, 12-8 partial 4-30
defined by GT-POWER 4-9
axisymmetric flow. See flow, axisymmetric
defined by tables 2-24, 4-7
baffles 4-27
B free-stream 4-33
background inlet 4-10
colour 2-32 non-reflective 4-19
material, in chemical reactions 8-17 outlet 4-12
background fluid 13-3 pressure 4-14
baffle 4-23 to 4-27 Riemann 4-37
conducting 3-18 stagnation 4-16
expanding 3-18 transient wave transmissive 4-35
in restart after mesh changing 5-17 wall 4-22
porous 4-24, 4-25, 6-5 defined by user subroutines 4-7, 14-5
radiation 4-25 defined using load steps 5-6 to 5-14
side numbering 4-24 degassing 4-40, 10-1
thickness 3-18 for cavitating flows 11-5
transparent 4-26 for compressible flow 3-9
batch mode 2-34 for free surface flows 11-1
beams 7-6 for liquid films 13-6
black and white printing 17-22 for rotating frames 12-6
for subsonic compressible flow 3-9

Version 4.02 1
Index

for supersonic flow 3-9 removal 5-17, 12-14


for transonic flow 3-10 list 2-24, 3-4
free-stream 1-12, 4-32 to 4-34, 14-6 maximum number of 2-18
inlet 4-9 to 4-11, 14-5 monitoring behaviour 5-3
non-reflective 1-12, 4-16 to 4-19, 14-6 near-boundary 1-4
outflow 1-11, 4-9, 4-10, 4-12 near-wall 3-12, 3-13, 3-14
outlet 1-11, 4-11 to 4-12, 14-6 number of 15-1
plots of 4-8 plot 2-23
prescribed flow 1-11 properties 3-1
pressure 1-12, 4-12 to 4-14, 14-6 set 2-21, A-2
for cavitating flows 11-9 volume 3-22
for free surface flows 11-4 shape 1-4
radial equilibrium 4-9 shape changing 5-15
on pressure boundaries 4-12, 4-14 size 1-6
radiation 4-39 to 4-40, 7-8 table 3-1 to 3-3
Riemann 1-12, 4-36 to 4-38, 14-6 compress 3-3
stagnation 1-12, 4-14 to 4-16, 14-6 editor 3-1 to 3-3
for rotating reference frames 12-1, 12-4, 12-9 radiation 7-8
symmetry 1-12, 4-27 number 3-2
table input 2-24 to 2-31 porosity 6-1
time-varying 5-4 tool 3-3
transient wave transmissive 1-12, 4-34 to 4-36, 14-6 type 3-3
wall 4-19 to 4-23, 14-6 change fluid type 12-23
for solar radiation 4-20 change grid (CG) operation 12-9
in turbulent flow 3-12 characteristic
See also wall length. See length, characteristic
buoyancy 1-9 velocity. See velocity, characteristic
See also density check tool 3-5
buoyancy driven flow. See flow, buoyancy driven checking
byte ordering 16-1 model and problem data 15-1
results 1-20
chemical reaction. See reaction
C CHEMKIN 8-11
calculations, checking 1-20 clearing entire geometry 17-14
catalytic converters 8-1 coal combustion 8-41 to 8-47
cathode 8-33 default models 8-42
cavitation 11-5 to 11-10 inlet mass fractions 8-44
defined by user subroutines 14-11 coal particle size distribution 8-45
in free surface flows 11-3, 11-6, 11-8 colour
initialisation 11-8 background 2-32, D-1, D-2
solution algorithms 1-13, 11-9 foreground D-1, D-2
steady-state flows 11-6 options 2-4
temperature calculation 11-8 palette 3-2
vapour properties 11-7 table index 3-2
CEA (Chemical Equilibrium with Applications) colour tool 17-22
program 8-3 combustion. See reaction and coal combustion
cell 1-2 command
attachment 12-18, 12-23 to 12-28 abbreviations 16-1
data 17-3 arithmetic in A-1
detachment 12-23 to 12-28 conventions A-1
face help A-3
boundary 4-1 history 2-13, 2-19, 17-2
matching 4-29 input 2-13
index 3-3 echo 2-19
interface 12-18 output 2-13
layer number of lines 2-19
addition 12-14 commands 2-36, 2-41

2 Version 4.02
Index

compressibility 3-9 monitoring 1-19


compressible flow printout 15-3
Courant number 5-9 differencing schemes 15-2
model setup 3-9 to 3-11 for free surface flows 11-2
outlets 1-11 to 1-12 for steady-state runs 5-2
pseudo-transient approach 5-1 for transient runs 5-11
stagnation boundaries 4-14 for use with DES turbulence models 3-15
transient, boundaries for 4-34 for use with LES turbulence models 3-15
compression wave 4-32 diffusion reaction system 8-1
condensation to liquid films 13-6 diffusivity
conduction thickness 3-19 characteristic 1-14
conduction through baffles 4-25 molecular, defined by user subroutines 14-8
conductivity porous 6-4
defined by user subroutines 14-6 directory, working 2-3, 2-9
in chemical reaction problems 8-17 discrete fourier transform (DFT) algorithm 4-18
in multi-component mixing 13-3 discrete transfer/ordinates radiation. See radiation
connectivity 12-18 discretisation
control keys A-4 error 1-6, 1-21
controlling STAR with StarWatch 17-17 schemes 1-21
convergence 1-19 temporal
in steady-state calculations with SIMPLE 1-16 for cavitating flows 11-9
in transient calculations with SIMPLE 1-17 for DES turbulence models 3-15
rate of 4-13 for free surface flows 11-4
coordinate system for LES turbulence models 3-15
in attachment boundaries 12-19, 12-23 with the SIMPLE algorithm 1-18
in porous media 6-2 to 6-3 time 5-11
corrector step tolerance 1-14 volume 1-3
couple set 2-21, A-2 distance, normal dimensionless. See near-wall,
couples across cyclic boundaries 4-29 dimensionless normal distance and y+ values
Courant number 5-9, 14-18 distributed resistance 6-1, 15-2
for Lagrangian multi-phase flow 9-10 user subroutines 14-8
for LES turbulence models 3-15 divergence 2-6, 2-7
for pseudo-transient calculations 5-2 double precision mode 1-18
crank angle 9-9 drag
cursor select 2-19, A-2 coefficient 14-19
customisation of pro-STAR 2-18, 16-1 force 14-14
cyclic droplet
boundary pair. See boundary conditions, cyclic age 9-7
set list 4-31 collision models 9-1
defined by user subroutines 14-12
diameter distribution function 9-1
D defined by tables 2-26
deleting entire model 17-14 gravitational effect 3-21
density information 9-8
at Riemann boundary 4-37 mass transfer 14-13
calculation 1-9, 3-9 momentum transfer 14-13
defined by user subroutines 14-8 number density 14-12
in aeroacoustic analysis 13-4 positions 9-7
in buoyancy driven flow 3-20 reading data 9-5
in free surface flows 11-3 set 2-21, A-2
in PPDF scheme reactions 8-4, 8-19 set selection 9-6
reference 3-21 track list 9-7
in buoyancy driven flow 4-14 transfer to/from liquid films 13-5, 13-6, 13-7
under-relaxation 1-15, 17-19 user subroutines 9-1 to 9-3, 14-12 to 14-13
dependent variable volume 9-10
in tables 2-27
initialisation 1-10, 4-42

Version 4.02 3
Index

E FASTRAC 7-1
boundary conditions 7-2
emissivity
escape boundaries 7-5
at escape boundaries 4-39
patches 7-2
at radiation boundaries 4-39
symmetry and cyclic boundaries 7-5
of baffle 4-26
user subroutines 7-6
solar 4-26
thermal 4-25, 7-1, 7-4 view factors 7-3
of wall 4-22 with moving mesh 7-7
thermal 4-20, 7-1, 7-4 favourites menu 2-17
with FASTRAC 7-2 file menu 2-16
engine data 9-9 case name 17-1
enthalpy edit file 17-10
defined by user subroutines 14-7, 14-10 model title 2-19
in PPDF scheme reactions 8-19 resume 17-10
stagnation 3-11 resume from 17-2
temperature dependence 14-7 resume model 17-2
environment variables 17-12 save as coded 17-3, 17-10
equations of state 1-9 save model 17-2, 17-10
error system command 2-18
messages 2-19 write geometry file 17-5, 17-10
sweep limits 1-14, 1-15 write problem file 17-6, 17-10
numerical discretisation 1-6, 1-21 files 2-36, 17-1 to 17-12, B-1 to B-3
recovery 2-20 boundary 17-3
round-off 1-18 cell 17-3
splitting 1-13 coded 17-3
escape surfaces 7-2, 7-4 command 17-10
Eulerian multi-phase flow data repository 17-5
boundaries 10-1 droplet data 9-5
interphase 10-2 echo 17-2
model setup 10-1 to 10-4 editing 17-11
phase-escape boundary 4-40 engines 9-9
response coefficient 14-14 event steps 12-12, 12-19
user subroutines 14-14 format of B-3
evaporation from liquid films 13-6 geometry 17-5
event steps load step 5-9
cell attachment 12-22 macros 16-6
cell inclusion/exclusion 12-28 manipulating 17-9
cell removal/addition 12-14 mapping 5-15
deleting 12-12 model 2-4, 17-2
listing 12-12 monitoring engineering data 5-3
modifying 12-12 output 2-7
moving mesh 12-9 panels 16-5
moving pistons 12-13 param.prp 17-13
reading 12-12 plot 17-4
regular sliding 12-18 problem 2-6, 17-6
turning off 12-30 PRODEFS 16-1
writing 12-12 PROINIT 16-1
exhaust gas recirculation 8-11, 8-17 reaction mechanism 8-11
exhaust valve 5-11 relationship between 17-7
expansion wave 4-32 residual 17-6
restart 2-7
exposure
scalar properties 13-3
of baffle 4-26
scene 17-23
of wall 4-20, 4-22
set-up 16-1
solution 17-5
F STAR-CD 3.2x equivalents 17-6
facets 1-4 temporary 17-14

4 Version 4.02
Index

transient 5-5, 5-13, 17-4, 17-7 initialisation 11-3


compressing 5-14 pseudo-transient approach 5-1
vertex 17-3 solution algorithms 1-13, 11-4
view factors 7-3, 7-6 steady-state 11-1
film stripping 13-5, 13-7 surface tension 11-2
flame kernel 8-36 temperature calculation 11-3
flamelet library 8-20 free-stream boundary. See boundary conditions, free-
flow stream
axisymmetric 1-12 function keys 16-9 to 16-11, A-4
buoyancy driven 1-13, 1-15, 3-20, 5-4
pressure boundaries 4-14
under-relaxation 3-21 G
unstable 3-21 gas
cavitating. See cavitation ideal 3-9
chaotic 5-4 law. See ideal gas law
compressible. See compressible flow geometry, modifying 5-15
cyclic 5-4 graph menu 2-17
free surface. See free surface flow graphics driver 2-3
impingement 3-15 group number 3-2
inviscid 4-19 GT-POWER 4-9
enthalpy 3-10
non-Newtonian 3-11
periodic 5-4
H
prescribed 1-11 hard copy 17-21
split 4-11 heat
steady 1-10, 1-15 to 1-17 conductivity
analysis controls 5-1 to 5-4 defined by user subroutines 14-6
output controls 5-2 in chemical reaction problems 8-17
solution controls 5-1 in multi-component mixing 13-3
subsonic 3-9 transfer
supersonic 1-11, 3-9 coefficient 3-17, 14-17
mesh at inlet boundaries 4-10 in baffles 3-18
transient 1-13 to 1-15 in porous media 6-4
analysis controls 5-4 to 5-14 solid-fluid 3-16 to 3-20
output controls 5-5, 5-12 solid-solid 3-20
solution controls 5-5, 5-9 help menu 2-17
transonic 1-11, 3-10 on-line help 2-2
residuals 3-10 pro-STAR help 17-7
turbulent 3-12
unsteady 1-10 I
fluid
I/O window 2-13
background 13-3
IC setup panels 8-23 to 8-37
injection 3-21 to 3-22
ideal gas 3-9
defined by user subroutines 14-10
law 3-20, 8-4
mixture 13-1
ignition 8-10, 8-15, 8-21, 14-15
non-Newtonian 3-11
advanced ICE models 8-24, 8-27, 8-29, 8-31
properties 13-3, 15-2
AKTIM 8-33 to 8-36
stream 15-2
for simulations involving cell layer removal 12-18
multiple 3-5
imbalance 7-5
font size 2-33
independent variables in tables 2-27
fonts D-1, D-2
inflow at outlet boundaries 1-11
force, body 1-9
INFO button 2-22
FORTRAN conventions 14-4
initial conditions 1-10, 1-20, 17-6
free surface flow 11-1 to 11-5
defined by tables 2-24
defined by user subroutines 14-11
for liquid films 13-6
density 11-3
for transient analyses 1-18
differencing schemes 11-2

Version 4.02 5
Index

initialisation procedure boundaries 4-3


in Lagrangian flow using user coding 14-13 cells 3-4
in moving meshes 14-17 cyclic sets 4-31
steady-state run 4-42 droplet tracks 9-7
transient run 4-42 tracks 9-7
injection groups 9-2 lists menu 2-16
injection. See fluid, injection load steps 5-6 to 5-14, 14-18
inlet. See boundary conditions, inlet definition 5-8
inner iterations. See iterations, inner identifying number 5-11
input/output window 2-13 in multi-component mixing 13-3
instability
numerical 1-10, 1-14, 4-13
physical 1-9 M
interface macros 16-6 to 16-9
sliding 12-19 creating 16-7
solid-fluid 3-18 menus 16-8
radiative 7-5 modifying 16-7
invert set selection option 2-21, 4-3 mass
iterations conservation 8-3, 8-10
inner 1-13 flow rate
number of 15-1 defined by tables 2-26
outer 1-13, 1-17 defined by user subroutines 14-10
iterative calculation 1-19 flux 3-22
in excluded cells 12-28
transfer
K coefficient 14-17
Kirchoff’s law 4-20, 4-26 droplet 14-13
in porous media 6-4
material
L number 3-2, 3-7
Lagrangian multi-phase flow properties 3-1
atomisation models 9-1 maximum plot screen 2-32
mesh 9-10 memory requirements of pro-STAR 17-13
model setup 9-1 menus 2-16 to 2-17
nozzle models 9-1 mesh
static displays adaptive refinement 5-17 to 5-19
steady-state 9-5 at non-reflective boundaries 4-18
transient 9-8 block. See block set
trajectory displays 9-8 distortion 1-5 to 1-6
user subroutines 9-1 to 9-3, 9-10, 14-12 problems caused by 1-16
with liquid films 13-5 distribution, near walls 1-7
See also droplet mean dimension of 1-14
length, characteristic 15-2 moving 5-14, 12-9 to 12-13
lift coefficient 14-19 defined by user subroutines 14-16
lighting material 3-2 in porous media 6-5
liquid films 13-5 to 13-7 mesh preview mode 12-13, 12-18
boundary conditions 13-6 parameters 12-10
boundary regions 13-5 post-processing 12-29
evaporation/condensation 13-6 pre-processing 12-28
film stripping 13-5, 13-7 restoration to original state 12-29
gravitational effect 3-21 with radiation 7-3, 7-5
initial conditions 13-6 polyhedral
multi-component 13-6 at boundaries
results 13-6 attachment 4-39
velocity 13-7 free-stream 4-33
with Largrangian multi-phase flow 13-5 pressure 4-14
lists Riemann 4-37

6 Version 4.02
Index

stagnation 4-15 cell 3-12, 3-13, 3-14, 5-18


supersonic inlet 4-10 for compressible flow 3-10
transient wave transmissive 4-35 dimensionless normal distance 1-7, 3-13
walls 4-22 See also y+ values
refinement 5-15 to 5-17 layer (NWL) 1-7, 3-12, 3-14
rotating. See rotating reference frames 12-1 mesh distribution 1-7
sliding 12-18 to 12-22 region 3-13
defined by user subroutines 14-16 neutral plot file 2-31, 17-4
mesh preview mode 12-22 new set selection option 2-21, 4-3
parallel processing 12-22 none set selection option 2-21, 4-3
regular interface 12-18 to 12-22 non-Newtonian flow. See flow, non-Newtonian
without shearing 12-21 no-slip condition 4-19
tetrahedral, at boundaries NOVICE mode A-1
free-stream 4-33 NOx modelling 8-39
pressure 4-14 defined by user subroutines 14-15
Riemann 4-37 nozzle models 9-1
stagnation 4-15 nuclei, number of 11-7
supersonic inlet 4-10 numerical discretisation error 1-6
transient wave transmissive 4-35 Nusselt number 14-15
walls 4-22
visualisation
colour setting 3-1 O
lighting effect 3-1 on-line help 2-2
message passing routines E-1 operate utility 13-4
mixing, multi-component 13-1 operating mode 2-18
mixture outer iterations. See iterations, outer
fluid 13-1 outflow. See boundary conditions, outflow
fraction 8-2, 8-3, 8-10, 8-16 outlet. See boundary conditions, outlet
model output
checking 1-20 controls 5-2, 5-5, 5-12
title 2-19
modelling strategy 1-1
modifying cell type 3-3 to 3-4
P
modules menu 2-16 panels 16-2 to 16-6
transient 5-6, 17-4 creating 16-2
monitoring environment 16-7
engineering data 4-40, 5-3, 5-13 files 16-5
boundary behaviour 5-3 manipulating 16-6
cell behaviour 5-3 menus within 16-3
field data 5-12 modifying 16-2
field variables 2-7, 5-13, 14-19, 15-2 to 15-3, 17-18 panels menu 2-17
numerical solution 5-3, 17-8, 17-19 to 17-21 define macros 16-7
scalars 5-12 define panel 16-2
multi-component parallel processing 2-5
liquid films 13-6 for sliding mesh 12-22
mixing 13-1 run options F-3 to F-7
setting up models 13-1 user subroutines 14-22, E-1
multi-phase flow. See Lagrangian multi-phase flow and with cell layer removal/addition 12-18
Eulerian multi-phase flow with moving mesh 12-13
multiple streams 3-5 to 3-9 parameters 2-36
of fluid mixtures 13-1 varying during run 17-18 to 17-20
parcels 9-2, 9-6
particle radiation. See radiation, coal particles
N patch
natural convection. See flow, buoyancy driven number 4-4
NavCenter 2-38 radiation 7-2 to 7-7
near-wall surface 7-7

Version 4.02 7
Index

permeability function 1-10 R


PISO algorithm 1-13 to 1-15
radiation 7-1 to 7-8, 15-2
under-relaxation 17-19
analysis methods
plot menu 2-17 discrete ordinates 7-3 to 7-5, 7-7 to 7-8
alternate plot mode 2-31 discrete transfer 7-1 to 7-7
background 2-32 at walls 4-20
cell display 4-41, 7-2, 9-5 boundaries. See boundary conditions, radiation
maximum plot screen 2-32 cell table editor 7-8
plot to file 17-4 coal particles 7-5, 8-43
standard plot mode 2-31 cpu time 7-6, 7-7
standard plot screen 2-32 escape boundaries 7-5
plotting hard copies 17-21 FASTRAC. See FASTRAC
porous in coal combustion 8-44
baffles 4-23, 4-25, 6-5 in Eulerian multi-phase flow 10-2
media on baffles 4-25
in Eulerian multi-phase flow 10-2 participating media 7-3
in moving mesh 12-13 patch 7-2 to 7-7
in multi-component mixing 13-2 properties, defined by user subroutines 14-11
user subroutines 14-8 solar 7-1
pressure drops 6-5 baffle boundary conditions 4-26
region modelling 6-1 to 6-5 discrete ordinates 7-8
post menu 2-17 in particpating media 7-3
get droplet data 9-3 wall boundary conditions 4-20
post register 13-4 sub-domains 7-8
post-processor 2-1 surface exchanges 7-1
power law of viscosity 3-11 transparent solids 7-3 to 7-8
Prandtl number 8-17 user subroutines 7-6
defined by user subroutines 14-18 with STAR-HPC 7-6
precision. See solver, precision Rayleigh model 11-7
pre-processor 2-1 reactant 8-2
pressure 4-5 leading 8-2, 8-18
boundary. See boundary conditions, pressure reaction
correction 1-13 to 1-15 advanced ICE models 8-22 to 8-39
drop across porous region 6-5 CFM 8-24 to 8-25
prescribed. See boundary conditions, pressure ECFM 8-26 to 8-27, 8-30
saturation 11-7 ECFM-3Z 8-28 to 8-30
product 8-2 compression ignition 8-29 to 8-30
prolinkl script 17-14 spark ignition 8-28 to 8-29
prosize script 17-13 level set 8-31 to 8-32
pro-STAR 1-2, 2-1 saving data 8-32
customisation 16-1 background material 8-17
display D-1 to D-3 complex chemistry models 8-11, 8-21
executables 17-14 coupled 8-16
launching 2-3, 2-10 eddy break-up reaction 8-13
layout D-3 to D-5 Landau-Teller reaction 8-12
memory 17-13 Lindemann fall-off reaction 8-12
on-line help 2-36, 17-7 SRI fall-off reaction 8-13
quitting 2-21 sub-timestep 8-21
resizing 17-13 three-body reaction 8-12
size 2-18 Troe fall-off reaction 8-12
pseudo-transient calculation 5-1 conventions 8-18
for compressible flow 3-11 copying 8-17
EGR systems 8-17
Q heterogeneous 8-1
quitting pro-STAR 2-21 homogeneous 8-1
in Eulerian multi-phase flow 10-2

8 Version 4.02
Index

local source model 8-2, 8-16 non-reflective boundaries 4-18


models 8-1 run options F-2
NOx formation 8-39 steady-state runs 4-42, 5-4
partially premixed 8-1 transient runs 4-43, 5-6, 5-10
PPDF scheme 8-3 to 8-11, 8-18 to 8-19 turbulence models 3-16
multi-fuel 8-9 view factors 7-5
single-fuel 8-3 with INITFI 14-17
equilibrium models 8-3 restoring sets 2-22
with dilutants 8-9 results checking 1-20
premixed 8-1 RESULTS sub-directory 2-7
rate, defined by user subroutines 14-15 resume mode 2-13
regress variable models 8-10, 8-16 rotating reference frames
eddy break-up models 8-10 arbitrary interface 12-4
flame-area models 8-10 coupling 12-8
CFM-ITNFS 8-10 defined by user subroutines 14-16
Weller 8-10 multiple
Weller 3-equation 8-10 explicit method 12-5 to 12-9
schemes 8-2, 8-14 non-reflecting explicit option 12-9
soot modelling. See soot modelling implicit method 12-2 to 12-5
source term 14-11 single 12-1
temperature limit 8-21 rotation 1-9
turning off 8-17 rotational speeds, defined by tables 2-25
types 8-1 rothalpy, in rotating reference frames 12-5
unpremixed/diffusion 8-1 roughness 14-6
user subroutines 14-15 run time controls, defined by tables 2-25
real constants 5-3 running simulations 2-2, 2-4, 2-11
recovery 2-20 in parallel 2-5
re-executing commands 2-20 on other hosts 2-5
reference temperature 15-2
reflectivity
of baffle 4-26 S
solar 4-26, 7-2 Sauter mean diameter 9-9
thermal 4-25, 7-1, 7-4 saving
of wall 4-22 model 2-42
solar 4-21, 7-2 screen 2-33
thermal 4-20, 7-1, 7-4 sets 2-22
regress variable 8-10 scalar
relaxation factors 1-15, 1-16, 15-2 CAV 11-6, 11-8
residuals 1-15, 1-19, 2-6, 2-7, 15-3, 17-19 in fluid mixtures 13-1
for transonic flow 3-10 initialisation 4-42
inner 1-13, 1-15 numbering 13-3
oscillations 1-19, 2-7 printing 13-2
tolerance 8-16, 10-4, 15-1, 17-19 properties, defining 8-15
resistance, distributed 6-1 variable 8-16
user subroutines 14-8 VOF 11-1, 11-6
resizing pro-STAR 17-13 initialisation 11-4, 11-9
resource allocation F-6 scalar solver 1-19
restart 4-42 scattering coefficient 7-4
aeroacoustic analysis 13-4 Schmidt number 6-4, 8-17
after mesh changes 5-15 defined by user subroutines 14-8, 14-18
coal combustion 8-44 screen
data 5-13 capture 2-33
files 2-7 high-resolution 2-33
flamelet calculations 8-20 display control 2-18
Lagrangian multi-phase 9-10 dump 2-33
moving mesh 12-13 size 2-32
multiple runs 5-14 storage 2-32

Version 4.02 9
Index

set active cell type 3-3 in coal combustion 8-44


sets reacting 8-2
restoring 2-22 specific heat 8-17, 8-19, 15-2
saving 2-22 defined by user subroutines 14-9
set-up files 16-1 in multi-component mixing 13-3
shock wave 4-32 spin
short cut keys. See function keys index 12-3, 12-6
short input history 2-15 parameters 12-1, 12-4 to 12-8
SIMPLE algorithm 1-13, 1-16 to 1-18 velocity 12-1, 12-4
single precision mode 1-18 spline set 2-21, A-2
sliding mesh. See mesh, sliding stability
solid regions 15-2 numerical 1-5, 1-10
initialisation 4-42 dependence on time step 5-9
solid-fluid heat transfer 3-8, 3-16 to 3-20, 15-2 stagnation boundary. See boundary conditions, stagnation
hints 3-19 STAR 1-2
in free surface flows 11-2 defaults F-7
radiative 7-3, 7-5 run options F-1 to F-12
solid-solid heat-transfer 3-20 running 2-4, 2-5, 2-11
solution switches 2-5, F-1 to F-12
algorithms 1-13 to 1-18 STAR-GUIde 2-38
for buoyancy driven flow 3-21 check everything panel 4-6
for cavitating flows 11-9 favourites 2-40
for use with DES turbulence models 3-15 STAR-HPC, with radiation problems 7-6, 7-7
for use with LES turbulence models 3-15 STAR-Launch utility 2-8 to 2-12
in free surface flows 11-4 STAR-NET F-8 to F-12
controls 5-1, 5-5, 5-9 STAR-View 17-23
domain 1-2 StarWatch utility 17-15 to 17-21
mapping 5-15, 5-16 states 17-6
procedure 15-1 steady-state calculation 1-15 to 1-17
solver stoichiometry, checking 8-16
conjugate gradient 1-19 strain rate, at inlet 8-20
precision 1-18 subset set selection option 2-22, 4-3
for Eulerian multi-phase flow 10-3 surface
for liquid films 13-6 tension 11-2
tolerances 1-14, 1-15, 1-16, 15-2 coefficient 14-12
soot modelling 8-39 to 8-41 surface set selection option 2-22
flamelet library model 8-39 sweep limits 1-14, 1-15, 1-16, 15-2
PSDF moments model 8-26, 8-29, 8-40 sweeps 1-13, 17-19
sound, speed of 14-12 switches 5-3
source for ‘prostar’ system command 2-4
aeroacoustic 13-4 for ‘star’ system command 2-5, F-1 to F-12
defined by tables 2-24 symmetry plane. See boundary conditions, symmetry
enthalpy 3-9 system commands, entering in pro-STAR 2-18
defined by user subroutines 14-10
in cavitating flows 11-9
in Eulerian multi-phase flow 10-3 T
in free surface flows 11-4 tables
mass 3-8 dependent variables 2-27
defined by user subroutines 14-10 editor 2-26 to 2-31
momentum 3-8 graphs of 2-29
defined by user subroutines 14-10 hints 2-31
scalar, defined by user subroutines 14-11 independent variables 2-27
turbulence 3-9 title 2-27
defined by user subroutines 14-10 usage in
species boundary conditions 2-24, 4-7
mass fraction 8-4 baffles 4-27
defined by user subroutines 14-16 free-stream 4-33

10 Version 4.02
Index

inlet 4-10 users tool 2-35


non-reflective 4-19 transient calculation 1-13 to 1-15, 3-11
outlet 4-12 completion 2-7
pressure 4-14 full 5-6 to 5-14
Riemann 4-37 single 5-4 to 5-6
stagnation 4-16 transient wave boundary. See boundary conditions,
transient wave transmissive 4-35 transient wave transmissive
walls 4-22 transient waves 4-34
initial conditions 2-24 transmissivity
injectors and sprays 2-26 at solid-fluid interface 7-5
rotational speeds 2-25 of baffle 4-26
run-time controls 2-25 solar 4-26, 7-2
source terms 2-24 thermal 4-25, 7-1, 7-4
tcl/tk interpreter 2-35 of wall 4-22
temperature solar 4-21, 7-2
at free-stream boundary 4-33 thermal 4-20, 7-1, 7-4
at Riemann boundary 4-37 turbomachinery, boundaries for 4-16
at transient wave transmissive boundary 4-35 turbulence 3-12 to 3-16
defined by user subroutines 14-7 changing model 3-16
devolatilisation 8-43 DES models 3-15
distribution 3-17 hybrid wall functions 3-12, 3-14
functional dependence 14-7 in aeroacoustic analysis 13-4
in cavitating flows 11-8 in ECFM combustion models 8-30
in free surface flows 11-3 in porous media 6-4
limit on reaction 8-21 in rotating reference frames 12-2, 12-4, 12-9
radiation 7-2, 7-4 initialisation 4-42
reference 15-2 length scale 14-9
in restart runs 5-17 LES models 3-15
stagnation 3-11 low Reynolds number models 3-12, 3-14
under-relaxation 17-19 models 3-12, 15-2
temporal discretisation. See discretisation, temporal Reynolds stress models 3-15
thermal conditions at boundary
resistance 3-17 free-stream 4-33
runaway 1-10 inlet 4-10
thermal radiation. See radiation Riemann 4-37
time stagnation 4-15
characteristic 1-14 transient wave transmissive 4-35
cpu 2-18, 15-3 two-layer models 1-8, 3-12, 3-13 to 3-14
reducing 1-16 wall functions 3-12, 3-13
elapsed 2-7 non-equilibrium 3-13
elapsed computational 15-3 tutorials 2-37
ignition delay 8-37 two-dimensional flow, axisymmetric. See flow,
scale 5-9 axisymmetric
heat/mass transfer 8-44 two-phase flow. See Lagrangian multi-phase flow and
step 1-10, 5-6, 5-9 Eulerian multi-phase flow
adjusting 1-14, 1-18
defined by tables 2-25
number of 15-1 U
specification 5-11 under-relaxation 1-10, 1-15, 3-20
variable 14-18 density 1-15
varying during run 17-18 for compressible flow 3-10, 3-11
See also Courant number for moving mesh 12-13
tools menu 2-16 for steady-state calculations with PISO 1-15
cell tool 3-3 for steady-state calculations with SIMPLE 1-16
check tool 3-5 for transient calculations with SIMPLE 1-17
colour tool 17-22 in buoyancy driven flow 3-21
convert 17-8 in cell layer removal/addition 12-18

Version 4.02 11
Index

in chemical reaction problems 8-16 of walls 4-22


in Eulerian multi-phase flow 10-4 vertex
in multi-component mixing 13-3 coordinate
pressure 1-16 in moving mesh 5-14, 12-9
pressure correction 1-14 data 17-3, 17-9
varying 17-19 maximum number of 2-18
velocity 1-16 set 2-21, A-2
viscosity 1-15 view factor 7-3, 7-5
units 2-36, C-1 viscosity
unselect set selection option 2-21, 4-3 defined by user subroutines 14-9
unsteady calculation. See transient calculation in chemical reaction problems 8-17
user interface 2-35 oscillations 1-15
user subroutines 2-4, 2-18, 2-36, 14-1 to 14-22, 15-2 power law 3-11
activating 14-2 turbulent 14-9
checking 1-20 under-relaxation 1-15
defining material properties 14-6 to 14-9 viscous sublayer 3-13
editing 14-4, 17-10 volume of fluid (VOF) model 11-1
for boundaries 4-7, 14-5 volume, of droplet 9-10
for chemical reactions 14-15
for Eulerian multi-phase flow 14-14
for heat and mass fluxes 14-6 W
for heat and mass transfer coefficients 13-3, 14-17 wall 1-3
for Lagrangian multi-phase flow 9-10, 14-12 boundary layer 1-6
for moving mesh 12-9, 14-16 data 5-3
for porous media 6-4, 14-8 functions 1-7, 3-12
for rotating reference frames 12-1, 12-4, 12-8, 14-16 hybrid 1-8
for solar radiation 7-6 heat flux 5-5
for turbulence 14-9, 14-10 defined by user subroutines 14-17
in droplet injection 9-1 to 9-3 moving 4-19
in parallel processing runs 14-22, E-1 no-slip 4-19
users tool 2-35 patch 7-2
utility menu 2-17 radiation 4-20 to 4-22
calculate volume 3-22 transmissive external 7-5
capture screen 2-33 transparent 4-21
count 4-2 velocity 4-22
function keys 16-10 See also boundary conditions, wall
save screen as 2-33 wave
solution mapping 5-16 compression 4-32
user subroutines 14-2 expansion 4-32
write STAR-CD scene file 17-23 shock 4-32
transient 4-34

V
vaporization rate 14-12 Y
vapour in cavitating flows 11-7 y+ values 1-21, 3-13, 3-15
variables 2-36 See also near-wall, dimensionless normal distance
vector solver 1-19
velocity
angular. See angular velocity
Z
at stagnation boundaries 4-15 Zeldovich mechanism 8-39
boundary values 4-27
characteristic 1-14
in porous media 6-5
injection 3-22
defined by user subroutines 14-10
of baffles 4-27
of liquid films 13-7

12 Version 4.02
INDEX OF COMMANDS

INDEX OF COMMANDS
This User Guide does not contain comprehensive information on all commands used in pro-STAR.
The Meshing and Post-Processing Guides and on-line STAR GUIde Help should also be consulted

A CZONE 3-4
ABBREVIATE 16-1
ABORT A-1 D
DAGE 9-7, 9-8
B DCOLLISION 14-12
DELTIME 14-18
BATCH 2-18 DENSITY 14-8
BCROSS 4-2, 7-6 DIFFUSIVITY 14-8
BDEFINE 4-2, 7-2 DLIST 9-8
BDELETE 4-4 DRAVERAGE 14-12
BDISPLAY 4-41, 7-2 DRCMPONENT 14-13
BGENERATE 4-2 DRHEAT 14-13
BLIST 4-4 DRMASS 14-13
BLKSET 2-24 DRMOMENTUM 14-13
BMODIFY 4-2, 4-4, 7-6 DRPROPERTIES 14-13
BSET 2-24, 4-3, 4-4 DRUSER 14-13
BSHELL 4-2, 7-2 DRWALL 14-13
DSCHEME 14-19
DSET 2-24, 9-6, 9-8
C DTIME 9-5, 9-7
CASENAME 17-1
CAVERAGE 12-29 E
CAVITATION 14-11
CAVNUCLEI 14-11 EACELL 12-16
CAVPROPERTY 14-12 EACOMPRESS 12-20
CBEXTRUDE 3-18, 3-19 EADELETE 12-20
CCLIST 17-6 EAGENERATE 12-20
CCROSS 3-4 EALIST 12-20
CDELETE 12-29 EAMATCH 12-19, 12-21, 12-27, 12-29
CDISPLAY 4-41, 7-2, 9-5 EATTACH 12-20, 12-21, 12-27
CDSAVE 17-3 ECHOINPUT 2-19
CDSCALAR 13-3 ECLIST 12-16
CDTRANS 5-11, 17-4 ECONDITIONAL 12-26
CFIND 3-4 EDATA 5-3
CJOIN 5-19 EDCELL 12-16
CLOSE 17-10 EDCOMPRESS 12-26
CLRMODE 2-32 EDDELETE 12-26
CLRTABLE 17-22 EDDIR 12-16
CMODIFY 3-5, 16-5 EDETACH 12-21, 12-28
COKE 14-18 EDLIST 12-26
CONDUCTIVITY 14-6 EDRAG 14-14
COUNT 4-2 EECELL 12-28
CPLOT 12-12 EFLUID 12-26
CPOST 5-11 EGRID 12-11, 12-13, 12-16, 12-29 to 12-30
CPRANGE 5-11 EHTRANSFER 14-15
CPRINT 5-11 EICOND 14-17
CPSET 2-24 ETURB 14-14
CREFINE 16-4 EVCHECK 12-30
CRMODEL 14-16 EVCND 12-26
CSET 2-24, 12-29 EVCOMPRESS 12-12, 12-17, 12-20
CTABLE 3-2, 12-15, 12-24 EVDELETE 12-12, 12-17, 12-20, 12-26
CTCOMPRESS 3-2, 3-4 EVEXECUTE 12-29 to 12-30
CTDELETE 3-2 EVFILE 12-11, 12-16, 12-25
CTLIST 3-2 EVFLAG 12-30
CTMODIFY 3-2 EVGET 12-12, 12-17, 12-20, 12-26
CTNAME 3-2 EVLIST 12-12, 12-17, 12-20, 12-27
CTYPE 3-4, 16-5 EVLOAD 12-29 to 12-30
CURSORMODE 2-19 EVOFFSET 12-12, 12-17, 12-20, 12-27
CYCLIC 4-29 EVPARM 12-10
CYCOMPRESS 4-31 EVPREP 12-12, 12-17, 12-20, 12-27
CYDELETE 4-31 EVREAD 12-12, 12-17, 12-21, 12-27
CYGENERATE 4-29 EVSAVE 12-11
CYLIST 4-31 EVSLIDE 14-16

Version 4.02 1
Index of Commands

EVSTEP 12-11, 12-16 O


EVUNDELETE 12-12, 12-17, 12-20, 12-27
EVWRITE 12-12, 12-17, 12-21, 12-27 OFILE 17-10
EXPERT A-1 OPANEL 16-1, 16-6
OPEN 16-10

F
P
FSTAT 2-23, 17-10
PAGE 2-19
PATCH 7-7
G PLATTACH 12-30
PLTBACK 2-33
GEOMWRITE 12-12, 12-17, 12-20, 12-27, 17-5 PLTYPE 12-12
GETCELL 12-12 PMATERIAL 12-25
GETD 8-36 POPTION 12-12
POREFF 14-8
POROSITY 14-8
H PORTURBULENCE 14-8
PRESSURE 12-18, 12-25
HCOEF 14-17 PRFIELD 14-19
HISTORY 2-19 to 2-20 PROBLEMWRITE 12-12, 12-17, 12-20, 12-27, 17-7
HRSDUMP 2-33 PROMPT 16-7
PRTEMP 3-18
PTCONV 17-8
I PTPRINT 9-7, 9-8
IFILE 13-3, 17-2, 17-4, 17-10, 17-12 PTREAD 9-7
IGNMODEL 14-15
INITIAL 14-17
ITER 5-14 Q
QUIT 2-21, 2-42
K
KNOCK 14-15 R
RADPROPERTIES 14-11
RCONSTANT 12-18
L RDEFINE 12-19, 12-23, 14-5, 14-6
RECALL 2-20
LFQSOR 14-14 RECOVER 2-20, 2-42
LFSTRIP 14-14 REPEAT 16-10
LQFBC 14-14 REPLOT 16-3
LQFINITIAL 14-14 RESET 2-32
LQFPROPERTY 14-14 RESUME 2-42, 17-3, 17-10
LSCOMPRESS 5-11 REWIND 17-10
LSDELETE 5-11 RGENERATE 4-7
LSGET 5-11 RRATE 14-15
LSLIST 5-10 RSOURCE 14-10
LSRANGE 5-11 RSTATUS 8-17
LSSAVE 5-10
LSTEP 5-10, 14-18
LVISCOSITY 14-9
S
SAFETY 2-20
M SAVE 2-42, 17-2
SC 13-3
MACRO 16-9 SCCONTROL 13-3
MEMORY 17-14 SCDUMP 2-33
MFRAME 14-16 SCENE 17-23
MLIST 3-9 SCPOROUS 14-8
MMPISTON 12-13 SCPROPERTIES 13-3, 14-16, 14-18
MONITOR 12-18, 12-25 SCRDELETE 2-33
MVGRID 5-11, 12-9, 12-10, 12-16, 14-16, 14-17 SCRIN 2-32
SCROUT 2-32
SCSOURCE 14-11
N SCTRANS 5-11, 13-2
NFILE 17-4 SETADD 2-23
NOX 14-15 SETDELETE 2-22

2 Version 4.02
Index of Commands

SETENV 16-6
SETFEATURE 16-1
SETREAD 2-23
SETWRITE 2-22
SIZE 2-18
SMAP 5-15, 5-16
SMCONV 17-8
SOLAR 14-11
SPECIFICHEAT 14-7, 14-9
SPIN 14-16
SPLSET 2-24
STATUS 2-42, 12-25
STENSION 14-12
STORE 12-12
SUCCEED 2-20
SYSTEM 2-18

T
TBCLEAR 2-28
TBDEFINE 2-28
TBGRAPH 2-28, 2-30
TBLIST 2-30
TBMODIFY 2-30
TBREAD 2-30
TBSCAN 2-31
TBWRITE 2-28
TDSCHEME 5-11
TERMINAL 2-31, 17-4, 17-21
TEXT 2-18
TIME 12-10, 12-15, 14-18
TITLE 2-19
TLMODEL 14-9
TPRINT 2-18
TRFILE 2-42, 5-10, 17-4
TRLOAD 12-12
TSCALE 2-33
TSMAP 5-16
TURBULENCE 14-9

U
USER 2-18
USUBROUTINE 14-3

V
VAPORIZATION 14-12
VFILL 12-11
VLIST 16-5
VMOD 12-11
VMODIFY 2-24
VOLUME 3-22
VSET 2-24, 12-11
VSMOOTH 12-30

W
WHOLE 2-32
WIPEOUT 17-14
WPOST 5-11
WPRINT 5-11

Version 4.02 3

You might also like