Star ccm руководство

Автор:

gudstartup · Опубликовано: 2 часа назад

@VldLg Вроде так и здесь все написано вот только мы не знаем какие именно системные файлы храняться на диске а какие в eprom.

README.TXT

The NC software and the English conversational language are stored in
EPROMs. Other conversational languages are stored on the hard disk. If no
current conversational languages are on the hard disk, load the English
language through machine parameter MP 7230.x. If a software exchange
becomes necessary, HEIDENHAIN provides new EPROMs and setup disks,
or a new complete setup for controls with flash EPROMs.

Вот какая у вас система с eprom или flash eprom?

The following controls are equipped with EPROMs (not flash EPROMs):
TNC 426 CB/PB
TNC430 CA/PA
TNC 426 M (324 990-xx, 324 991-xx, 324 994-xx, 324 995-xx) TNC 430 M (324 992-xx, 324 993-xx, 324 996-xx, 324 997-xx)
The following controls are equipped with flash EPROMs (not EPROMs):
TNC 426 M (344 958-xx, 344 959-xx, 344 962-xx, 344 963-xx) TNC 430 M (344 960-xx, 344 961-xx, 344 964-xx, 344 965-xx)

Я думаю правильно

Исходя из описанного у вас версия 280476130

#define HDDVERS 280476130
/* SETUPID = «286197 15» */
#define TIME 976698456

/*                          nur Ziffern verwenden            */    
/*              iiiiiivvt   iiiiii — Identnummer
                            vv     — Software-Version
                            t      — Teststand                         */

/* ***************** IDENTNUMMER / VERSION **************************/

#define SWIDENT  «280476»          /* Grundidentnummer der Software */
#define SWIDENTE «280477»          /* Grundidentnummer der Embargo-Software */

/* !!!! ACHTUNG: Softwareversion SWVERS muss immer 3 Zeichen haben,
        bei Lieferversionen also letztes Zeichen BLANK   */
 
#define SWVERS    «13 »       /* BLANK ??? Versionsindex der Software  */
#define FLSWVERS  «13»            /* Versionsindex der Software  */
#define SETUPVERS «13»            /* Versionsindex des Setup */

/* ******************************************************************/

/* Pfad, von dem aus relativ Setup-Daten geladen werden: */
#define SETUPPATH «SYS:\\» SWIDENT «\\» SETUPVERS «\\»
/* Pfad, von dem aus relativ die Flash-Software geladen wird: */
#define FLASHPATH «SYS:\\» SWIDENT «\\» FLSWVERS «\\» 

Предложите, как улучшить StudyLib

(Для жалоб на нарушения авторских прав, используйте

другую форму
)

• CCM USER GUIDE

  STAR-CD VERSION 4.02

  CONFIDENTIAL FOR AUTHORISED USERS ONLY

  2006 CD-adapco

• Version 4.02 i

  TABLE OF CONTENTS

  OVERVIEW

  1 COMPUTATIONAL ANALYSIS PRINCIPLESIntroduction
  …………………………………………………………………………………………………
  1-1The Basic Modelling Process
  …………………………………………………………………………
  1-1Spatial description and volume discretisation
  ………………………………………………….. 1-2

  Solution domain definition
  ……………………………………………………………………
  1-3Mesh definition
  ……………………………………………………………………………………
  1-4Mesh distortion
  ……………………………………………………………………………………
  1-5Mesh distribution and density
  ……………………………………………………………….
  1-6Mesh distribution near walls
  …………………………………………………………………
  1-7Moving mesh features
  ………………………………………………………………………….
  1-8

  Problem characterisation and material property definition
  ………………………………… 1-8Nature of the flow
  ………………………………………………………………………………..
  1-9Physical properties
  ……………………………………………………………………………….
  1-9Force fields and energy sources
  …………………………………………………………….
  1-9Initial conditions
  ………………………………………………………………………………..
  1-10

  Boundary description
  ………………………………………………………………………………….
  1-10Boundary location
  ……………………………………………………………………………..
  1-11Boundary conditions
  …………………………………………………………………………..
  1-11

  Numerical solution control
  ………………………………………………………………………….
  1-13Selection of solution procedure
  ……………………………………………………………
  1-13Transient flow calculations with PISO
  ………………………………………………….
  1-13Steady-state flow calculations with PISO
  ……………………………………………..
  1-15Steady-state flow calculations with SIMPLE
  ………………………………………… 1-16Transient flow
  calculations with SIMPLE
  ……………………………………………. 1-17Effect of
  round-off errors
  ……………………………………………………………………
  1-18Choice of the linear equation solver
  ……………………………………………………..
  1-19

  Monitoring the calculations
  …………………………………………………………………………
  1-19Model evaluation
  ……………………………………………………………………………………….
  1-20

  2 BASIC STAR-CD FEATURESIntroduction
  …………………………………………………………………………………………………
  2-1Running a STAR-CD Analysis
  ………………………………………………………………………
  2-2

  Using the script-based procedure
  …………………………………………………………..
  2-3Using STAR-Launch
  ……………………………………………………………………………
  2-8

  pro-STAR Initialisation
  ………………………………………………………………………………
  2-12Input/output window
  ………………………………………………………………………….
  2-13Main window
  …………………………………………………………………………………….
  2-15

• ii Version 4.02

  The menu bar
  …………………………………………………………………………………….2-16General
  Housekeeping and Session Control
  …………………………………………………..2-18

  Basic set-up
  ……………………………………………………………………………………….2-18Screen
  display control
  …………………………………………………………………………2-18Error
  messages
  …………………………………………………………………………………..2-19Error
  recovery
  ……………………………………………………………………………………2-20Session
  termination
  …………………………………………………………………………….2-21

  Set Manipulation
  ………………………………………………………………………………………..2-21Table
  Manipulation
  …………………………………………………………………………………….2-24

  Basic functionality
  ……………………………………………………………………………..2-24The
  table editor
  ………………………………………………………………………………….2-26Useful
  points
  ……………………………………………………………………………………..2-31

  Plotting Functions
  ……………………………………………………………………………………….2-31Basic
  set-up
  ……………………………………………………………………………………….2-31Advanced
  screen control
  ……………………………………………………………………..2-32Screen
  capture
  ……………………………………………………………………………………2-33

  The Users Tool
  …………………………………………………………………………………………..2-35Getting
  On-line Help
  …………………………………………………………………………………..2-35The
  STAR GUIde Environment
  ……………………………………………………………………2-38

  Panel navigation system
  ………………………………………………………………………2-40STAR
  GUIde usage
  ……………………………………………………………………………2-41

  General Guidelines
  ……………………………………………………………………………………..2-413
  MATERIAL PROPERTY AND PROBLEM CHARACTERISATION

  Introduction
  …………………………………………………………………………………………………3-1The
  Cell Table
  ……………………………………………………………………………………………..3-1

  Cell indexing
  ……………………………………………………………………………………….3-3Multi-Domain
  Property Setting
  ………………………………………………………………………3-5

  Setting up models
  …………………………………………………………………………………3-6Compressible
  Flow
  ……………………………………………………………………………………….3-9

  Setting up compressible flow models
  ……………………………………………………..3-9Useful
  points on compressible flow
  ………………………………………………………3-10

  Non-Newtonian Flow
  ………………………………………………………………………………….3-11Setting
  up non-Newtonian models
  ………………………………………………………..3-11Useful
  points on non-Newtonian flow
  …………………………………………………..3-11

  Turbulence Modelling
  …………………………………………………………………………………3-12Wall
  functions
  ……………………………………………………………………………………3-13Two-layer
  models
  ………………………………………………………………………………3-13Low
  Re models
  ………………………………………………………………………………….3-14Hybrid
  wall boundary condition
  …………………………………………………………..3-14

• Version 4.02 iii

  Reynolds Stress models
  ………………………………………………………………………
  3-15DES models
  ………………………………………………………………………………………
  3-15LES models
  ………………………………………………………………………………………
  3-15Changing the turbulence model in use
  …………………………………………………. 3-16

  Heat Transfer In Solid-Fluid Systems
  ……………………………………………………………
  3-16Setting up solid-fluid heat transfer models
  ……………………………………………. 3-17Heat
  transfer in baffles
  ……………………………………………………………………….
  3-18Useful points on solid-fluid heat transfer
  ……………………………………………… 3-19

  Buoyancy-driven Flows and Natural Convection
  …………………………………………… 3-20Setting up
  buoyancy-driven models
  ……………………………………………………..
  3-20Useful points on buoyancy-driven flow
  ……………………………………………….. 3-20

  Fluid Injection
  ……………………………………………………………………………………………
  3-21Setting up fluid injection models
  ………………………………………………………….
  3-22

  4 BOUNDARY AND INITIAL CONDITIONSIntroduction
  …………………………………………………………………………………………………
  4-1Boundary Location
  ……………………………………………………………………………………….
  4-1

  Command-driven facilities
  ……………………………………………………………………
  4-2Boundary set selection facilities
  …………………………………………………………….
  4-3Boundary listing
  ………………………………………………………………………………….
  4-3

  Boundary Region Definition
  ………………………………………………………………………….
  4-5Inlet Boundaries
  …………………………………………………………………………………………..
  4-9

  Introduction
  ………………………………………………………………………………………..
  4-9Useful points
  ……………………………………………………………………………………..
  4-10

  Outlet Boundaries
  ………………………………………………………………………………………
  4-11Introduction
  ………………………………………………………………………………………
  4-11Useful points
  ……………………………………………………………………………………..
  4-12

  Pressure Boundaries
  ……………………………………………………………………………………
  4-12Introduction
  ………………………………………………………………………………………
  4-12Useful points
  ……………………………………………………………………………………..
  4-13

  Stagnation Boundaries
  ………………………………………………………………………………..
  4-14Introduction
  ………………………………………………………………………………………
  4-14Useful points
  ……………………………………………………………………………………..
  4-15

  Non-reflective Pressure and Stagnation Boundaries
  ……………………………………….. 4-16Introduction
  ………………………………………………………………………………………
  4-16Useful points
  ……………………………………………………………………………………..
  4-18

  Wall Boundaries
  …………………………………………………………………………………………
  4-19Introduction
  ………………………………………………………………………………………
  4-19Thermal radiation properties
  ……………………………………………………………….
  4-20Solar radiation properties
  ……………………………………………………………………
  4-20

• iv Version 4.02

  Other radiation modelling considerations
  ………………………………………………4-21Useful
  points
  ……………………………………………………………………………………..4-22

  Baffle Boundaries
  ……………………………………………………………………………………….4-23Introduction
  ……………………………………………………………………………………….4-23Setting
  up models
  ……………………………………………………………………………….4-24Thermal
  radiation properties
  ………………………………………………………………..4-25Solar
  radiation properties
  …………………………………………………………………….4-26Other
  radiation modelling considerations
  ………………………………………………4-26Useful
  points
  ……………………………………………………………………………………..4-27

  Symmetry Plane Boundaries
  ………………………………………………………………………..4-27Cyclic
  Boundaries
  ………………………………………………………………………………………4-27

  Introduction
  ……………………………………………………………………………………….4-27Setting
  up models
  ……………………………………………………………………………….4-28Useful
  points
  ……………………………………………………………………………………..4-30Cyclic
  set manipulation
  ……………………………………………………………………….4-31

  Free-stream Transmissive Boundaries
  …………………………………………………………..4-32Introduction
  ……………………………………………………………………………………….4-32Useful
  points
  ……………………………………………………………………………………..4-33

  Transient-wave Transmissive Boundaries
  ………………………………………………………4-34Introduction
  ……………………………………………………………………………………….4-34Useful
  points
  ……………………………………………………………………………………..4-35

  Riemann Boundaries
  …………………………………………………………………………………..4-36Introduction
  ……………………………………………………………………………………….4-36Useful
  points
  ……………………………………………………………………………………..4-37

  Attachment Boundaries
  ……………………………………………………………………………….4-38Useful
  points
  ……………………………………………………………………………………..4-39

  Radiation Boundaries
  ………………………………………………………………………………….4-39Useful
  points
  ……………………………………………………………………………………..4-40

  Phase-Escape (Degassing) Boundaries
  ………………………………………………………….4-40Monitoring
  Regions
  …………………………………………………………………………………….4-40Boundary
  Visualisation
  ……………………………………………………………………………….4-41Solution
  Domain Initialisation
  ……………………………………………………………………..4-42

  Steady-state problems
  …………………………………………………………………………4-42Transient
  problems
  ……………………………………………………………………………..4-42

  5 CONTROL FUNCTIONSIntroduction
  …………………………………………………………………………………………………5-1Analysis
  Controls for Steady-State Problems
  …………………………………………………..5-1Analysis
  Controls for Transient Problems
  ……………………………………………………….5-4

  Default (single-transient) solution mode
  …………………………………………………5-4

• Version 4.02 v

  Load-step based solution mode
  ……………………………………………………………..
  5-6Load step characteristics
  ……………………………………………………………………….
  5-6Load step definition
  ……………………………………………………………………………..
  5-8Solution procedure outline
  ……………………………………………………………………
  5-9Other transient functions
  …………………………………………………………………….
  5-14

  Solution Control with Mesh Changes
  ……………………………………………………………
  5-15Mesh-changing procedure
  …………………………………………………………………..
  5-15

  Solution-Adapted Mesh Changes
  …………………………………………………………………
  5-176 POROUS MEDIA FLOW

  Setting Up Porous Media Models
  …………………………………………………………………..
  6-1Useful Points
  ……………………………………………………………………………………………….
  6-4

  7 THERMAL AND SOLAR RADIATIONRadiation Modelling for Surface
  Exchanges
  ……………………………………………………
  7-1Radiation Modelling for Participating Media
  …………………………………………………..
  7-3Capabilities and Limitations of the DTRM Method
  …………………………………………. 7-5Capabilities
  and Limitations of the DORM Method
  …………………………………………. 7-7Radiation
  Sub-domains
  …………………………………………………………………………………
  7-8

  8 CHEMICAL REACTION AND COMBUSTIONIntroduction
  …………………………………………………………………………………………………
  8-1Local Source Models
  ……………………………………………………………………………………
  8-2Presumed Probability Density Function (PPDF) Models
  ………………………………….. 8-3

  Single-fuel PPDF
  …………………………………………………………………………………
  8-3Multiple-fuel PPDF
  ……………………………………………………………………………..
  8-9

  Regress Variable Models
  …………………………………………………………………………….
  8-10Complex Chemistry Models
  ………………………………………………………………………..
  8-11Setting Up Chemical Reaction Schemes
  ………………………………………………………..
  8-14

  Useful general points for local source and regress variable
  schemes ……….. 8-16Chemical Reaction Conventions
  ………………………………………………………….
  8-18Useful points for PPDF schemes
  ………………………………………………………….
  8-18Useful points for complex chemistry models
  ………………………………………… 8-21Useful points
  for ignition models
  …………………………………………………………
  8-21

  Setting Up Advanced I.C. Engine Models
  ……………………………………………………..
  8-22Coherent Flame model (CFM)
  …………………………………………………………….
  8-24Extended Coherent Flame model (ECFM)
  ……………………………………………. 8-26Extended
  Coherent Flame model 3Z (ECFM-3Z) spark ignition …………
  8-28Extended Coherent Flame model 3Z (ECFM-3Z) compression ignition
  .8-29Useful points for ECFM models
  …………………………………………………………..
  8-30Level Set model
  …………………………………………………………………………………
  8-31Write Data sub-panel
  ………………………………………………………………………….
  8-32

• vi Version 4.02

  The Arc and Kernel Tracking ignition model (AKTIM)
  ………………………….8-33Useful points for the AKTIM
  model
  …………………………………………………….8-35The
  Double-Delay autoignition model
  ………………………………………………….8-37

  NOx Modelling
  …………………………………………………………………………………………..8-39Soot
  Modelling
  …………………………………………………………………………………………..8-39Coal
  Combustion Modelling
  ………………………………………………………………………..8-41

  Stage 1
  ………………………………………………………………………………………………8-41Stage
  2
  ………………………………………………………………………………………………8-42Useful
  notes
  ………………………………………………………………………………………8-44Switches
  and constants for coal modelling
  …………………………………………….8-45Special
  settings for the Mixed-is-Burnt and Eddy Break-Up models
  …………8-46

  9 LAGRANGIAN MULTI-PHASE FLOWSetting Up Lagrangian Multi-Phase
  Models
  …………………………………………………….9-1Data
  Post-Processing
  …………………………………………………………………………………….9-4

  Static displays
  ……………………………………………………………………………………..9-5Trajectory
  displays
  ……………………………………………………………………………….9-8

  Engine Combustion Data Files
  ……………………………………………………………………….9-9Useful
  Points
  ……………………………………………………………………………………………..9-10

  10 EULERIAN MULTI-PHASE FLOWIntroduction
  ……………………………………………………………………………………………….10-1Setting
  up multi-phase models
  ……………………………………………………………………..10-1

  Useful points on Eulerian multi-phase flow
  …………………………………………..10-411 FREE
  SURFACE AND CAVITATION

  Free Surface Flows
  ……………………………………………………………………………………..11-1Setting
  up free surface cases
  ………………………………………………………………..11-1

  Cavitating Flows
  …………………………………………………………………………………………11-5Setting
  up cavitation cases
  …………………………………………………………………..11-5

  12 ROTATING AND MOVING MESHESRotating Reference Frames
  ………………………………………………………………………….12-1

  Models for a single rotating reference frame
  ………………………………………….12-1Useful points
  on single rotating frame problems
  …………………………………….12-1Models for multiple
  rotating reference frames (implicit treatment)
  …………..12-2Useful points on multiple implicit rotating frame
  problems ……………………..12-4Models for multiple rotating
  reference frames (explicit treatment) ……………12-5Useful
  points on multiple explicit rotating frame problems
  ……………………..12-8

  Moving Meshes
  ………………………………………………………………………………………….12-9Basic
  concepts
  ……………………………………………………………………………………12-9Setting
  up models
  ……………………………………………………………………………..12-10Useful
  points
  ……………………………………………………………………………………12-13

• Version 4.02 vii

  Automatic Event Generation for Moving Piston Problems
  ……………………. 12-13Cell-layer Removal/Addition
  ……………………………………………………………………..
  12-14

  Basic concepts
  …………………………………………………………………………………
  12-14Setting up models
  …………………………………………………………………………….
  12-15Useful points
  ……………………………………………………………………………………
  12-18

  Sliding Meshes
  …………………………………………………………………………………………
  12-18Regular sliding interfaces
  ………………………………………………………………….
  12-18

  Cell Attachment and Change of Fluid Type
  …………………………………………………
  12-22Basic concepts
  …………………………………………………………………………………
  12-22Setting up models
  …………………………………………………………………………….
  12-23Useful points
  ……………………………………………………………………………………
  12-27

  Mesh Region Exclusion
  …………………………………………………………………………….
  12-28Basic concepts
  …………………………………………………………………………………
  12-28

  Moving Mesh Pre- and Post-processing
  ………………………………………………………
  12-28Introduction
  …………………………………………………………………………………….
  12-28Action commands
  …………………………………………………………………………….
  12-29Status setting commands
  …………………………………………………………………..
  12-30

  13 OTHER PROBLEM TYPESMulti-component Mixing
  …………………………………………………………………………….
  13-1

  Setting up multi-component models
  ……………………………………………………..
  13-1Useful points on multi-component mixing
  ……………………………………………. 13-3

  Aeroacoustic Analysis
  ………………………………………………………………………………..
  13-3Setting up aeroacoustic models
  ……………………………………………………………
  13-3Useful points on aeroacoustic analyses
  ………………………………………………… 13-4

  Liquid Films
  ………………………………………………………………………………………………
  13-5Setting up liquid film models
  ………………………………………………………………
  13-5Film stripping
  ……………………………………………………………………………………
  13-7

  14 USER PROGRAMMINGIntroduction
  ……………………………………………………………………………………………….
  14-1Subroutine Usage
  ……………………………………………………………………………………….
  14-1

  Useful points
  ……………………………………………………………………………………..
  14-4Description of UFILE Routines
  ……………………………………………………………………
  14-5

  Boundary condition subroutines
  …………………………………………………………..
  14-5Material property subroutines
  ………………………………………………………………
  14-6Turbulence modelling subroutines
  ……………………………………………………….
  14-9Source subroutines
  ……………………………………………………………………………
  14-10Radiation modelling subroutines
  ………………………………………………………..
  14-11Free surface / cavitation subroutines
  …………………………………………………..
  14-11Lagrangian multi-phase subroutines
  ……………………………………………………
  14-12

• viii Version 4.02

  Liquid film subroutines
  ……………………………………………………………………..14-14Eulerian
  multi-phase subroutines
  ………………………………………………………..14-14Chemical
  reaction subroutines
  ……………………………………………………………14-15Rotating
  reference frame subroutines
  ………………………………………………….14-16Moving
  mesh subroutines
  ………………………………………………………………….14-16Miscellaneous
  flow characterisation subroutines
  ………………………………….14-17Solution control
  subroutines
  ………………………………………………………………14-18

  Sample Listing
  ………………………………………………………………………………………….14-19New
  Coding Practices
  ……………………………………………………………………………….14-20User
  Coding in parallel runs
  ……………………………………………………………………….14-22

  15 PROGRAM OUTPUTIntroduction
  ……………………………………………………………………………………………….15-1Permanent
  Output
  ……………………………………………………………………………………….15-1

  Input-data summary
  ……………………………………………………………………………15-1Run-time
  output
  …………………………………………………………………………………15-3

  Printout of Field Values
  ………………………………………………………………………………15-3Optional
  Output
  ………………………………………………………………………………………….15-3Example
  Output
  ………………………………………………………………………………………….15-4

  16 pro-STAR CUSTOMISATIONSet-up Files
  ………………………………………………………………………………………………..16-1Panels
  ………………………………………………………………………………………………………..16-2

  Panel creation
  …………………………………………………………………………………….16-2Panel
  definition files
  …………………………………………………………………………..16-5Panel
  manipulation
  ……………………………………………………………………………..16-6

  Macros
  ………………………………………………………………………………………………………16-6Function
  Keys
  …………………………………………………………………………………………….16-9

  17 OTHER STAR-CD FEATURES AND CONTROLSIntroduction
  ……………………………………………………………………………………………….17-1File
  Handling
  ……………………………………………………………………………………………..17-1

  Naming conventions
  …………………………………………………………………………..17-1Commonly
  used files
  ………………………………………………………………………….17-1File
  relationships
  ………………………………………………………………………………..17-7File
  manipulation
  ……………………………………………………………………………….17-9

  Special pro-STAR Features
  ………………………………………………………………………..17-12pro-STAR
  environment variables
  ……………………………………………………….17-12Resizing
  pro-STAR
  …………………………………………………………………………..17-13Special
  pro-STAR executables
  …………………………………………………………..17-14Use
  of temporary files by pro-STAR
  …………………………………………………..17-14

  The StarWatch Utility
  ……………………………………………………………………………….17-15

• Version 4.02 ix

  Running StarWatch
  ………………………………………………………………………….
  17-15Choosing the monitored values
  ………………………………………………………….
  17-17Controlling STAR
  ……………………………………………………………………………
  17-17Manipulating the StarWatch display
  …………………………………………………..
  17-20Monitoring another job
  ……………………………………………………………………..
  17-21

  Hard Copy Production
  ………………………………………………………………………………
  17-21Neutral plot file production and use
  ……………………………………………………
  17-21Scene file production and use
  …………………………………………………………….
  17-23

  APPENDICESA pro-STAR CONVENTIONS

  Command Input Conventions
  ……………………………………………………………………….
  A-1Help Text / Prompt Conventions
  …………………………………………………………………..
  A-3Control and Function Key Conventions
  …………………………………………………………
  A-4File Name Conventions
  ………………………………………………………………………………..
  A-4

  B FILE TYPES AND THEIR USAGEC PROGRAM UNITSD pro-STAR
  X-RESOURCESE USER INTERFACE TO MESSAGE PASSING ROUTINESF STAR RUN
  OPTIONS

  Usage
  ………………………………………………………………………………………………………….F-1Options
  ……………………………………………………………………………………………………….F-1Parallel
  Options
  ……………………………………………………………………………………………F-3Resource
  Allocation
  ……………………………………………………………………………………..F-6Default
  Options
  ……………………………………………………………………………………………F-7Cluster
  Computing
  ……………………………………………………………………………………….F-8Batch
  Runs Using STAR-NET
  ………………………………………………………………………F-8

  Running under IBM Loadleveler using STAR-NET
  …………………………………F-8Running under LSF using
  STAR-NET
  …………………………………………………..F-9Running
  under OpenPBS using STAR-NET
  …………………………………………F-10Running under
  PBSPro using STAR-NET
  …………………………………………….F-11Running
  under SGE using STAR-NET
  …………………………………………………F-11Running
  under Torque using STAR-NET
  ……………………………………………..F-12

  G BIBLIOGRAPHY

  INDEX

  INDEX OF COMMANDS

• Version 4.02 1

  OVERVIEWPurpose

  The Methodology volume presents the mathematical modelling
  practices embodiedin the STAR-CD system and the numerical solution
  procedures employed. In thisvolume, the focus is on the structure
  of the system itself and how to use it. Thispresentation assumes
  that the reader is familiar with the background informationprovided
  in the Methodology volume.

  ContentsChapter 1 introduces some of the fundamental principles
  of computationalcontinuum mechanics, including an outline of the
  basic steps involved in setting upand using a successful computer
  model. The important factors to consider at eachstep are mostly
  explained independently of the computer system used to perform
  theanalysis. However, reference is also made to the particular
  capabilities of theSTAR-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 detailedinstructions on how to use some of
  the basic code facilities, e.g. boundary conditionspecification,
  material property definition, etc., and an overview of the GUI
  panelsappropriate to each of them. The description covers all
  facilities (other than meshgeneration) that might be employed for
  modelling most common continuummechanics problems. Mesh generation
  itself is covered in a separate volume, theMeshing User Guide.

  Chapters 2 to 5 should be read at least once to gain an
  understanding of thegeneral housekeeping principles of pro-STAR and
  to help with any problemsarising from routine operations. It is
  recommended that users refer to theappropriate chapter repeatedly
  when setting up a model for general guidance and anoverview of the
  relevant GUI panels.

  Chapters 6 to 13 describe additional STAR-CD capabilities
  relevant to modelsof a more specialised nature, i.e. rotating
  systems, combustion processes,buoyancy-driven flows, etc. Users
  interested in a particular topic should consult theappropriate
  section for a summary of commands or options specially designed
  forthat purpose, plus hints and tips on performing a successful
  simulation.

  Chapter 14 outlines the user programmability features available
  and provides anexample FORTRAN subroutine listing implementing
  these features. All suchsubroutines are readily available for use
  and can be easily adapted to suit themodel’s requirements.

  Chapter 15 presents the printable output produced by the code
  which provides,among other things, a summary of the problem
  specification and monitoringinformation generated during the
  calculation.

  Chapter 16 explains how pro-STAR can be customised, in terms of
  user-definedpanels, macros and keyboard function keys, to meet a
  users individualrequirements.

  Finally, Chapter 17 covers some of the less commonly used
  features ofSTAR-CD, including the interaction between STAR and
  pro-STAR and howvarious system files are used.

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Introduction

  Version 4.02 1-1

  Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLESIntroduction

  The aim of this section is to introduce the most important
  issues involved in settingup and solving a continuum mechanics
  problem using a computational continuummechanics code. Although the
  discussion applies in principle to any such code,reference is made
  where appropriate to the particular capabilities of the
  STAR-CDsystem. It is also assumed that the reader is familiar with
  the material presented inthe Methodology volume.

  The process of computational mechanics simulation does not
  usually start withthe direct use of such a code. It is indeed
  important to recognise that STAR-CD, orany other CFD, CAD or CAE
  system, should be treated as a tool to assist theengineer in
  understanding physical phenomena.

  The success or failure of a continuum mechanics simulation
  depends not only onthe 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 bephysically 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 theabove tasks have been completed.

  The Basic Modelling ProcessThe modelling process itself can be
  divided into four major phases, as follows:Phase 1 Working out a
  modelling strategyThis requires a precise definition of the
  physical systems geometry, materialproperties and flow/deformation
  conditions based on the best availableunderstanding of the relevant
  physics. The necessary tasks include:

  Planning the computational mesh (e.g. number of cells, size and
  distributionof 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.).

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Spatial description and volume discretisation

  1-2 Version 4.02

  The user also has to balance the requirement of physical
  fidelity and numericalaccuracy against the simulation cost and
  computational capabilities of his system.His modelling strategy
  will therefore incorporate some trade-off between these
  twofactors.

  This initial phase of modelling is particularly important for
  the smooth andefficient progress of the computational
  simulation.Phase 2 Setting up the model using pro-STARThe main
  tasks involved at this phase are:

  Creating a computational mesh to represent the solution domain
  (i.e. themodel geometry).

  Specifying the physical properties of the fluids and/or solids
  present in thesimulation and, where relevant, the turbulence
  model(s), body forces, etc.

  Setting the solution parameters (e.g. solution variable
  selection, relaxationcoefficients, etc.) and output data
  formats.

  Specifying the location and definition of boundaries and, for
  unsteadyproblems, further definition of transient boundary
  conditions and time steps.

  Writing appropriate data files as input to the analytical run of
  the followingphase.

  Phase 3 Performing the analysis using STARThis phase consists
  of:

  Reading input data created by pro-STAR and, if dealing with a
  restart run, theresults of a previous run.

  Judging the progress of the run by analysing various monitoring
  data andsolution statistics provided by STAR.

  Phase 4 Post-processing the results using pro-STARThis involves
  the display and manipulation of output data created by STAR
  usingthe appropriate pro-STAR facilities.

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

  Spatial description and volume discretisationOne of the basic
  steps in preparing a STAR-CD model is to describe the geometryof
  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.

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Spatial description and volume discretisation

  Version 4.02 1-3

  Figure 1-1 Example of solution domain subdivision into cells

  This process is called volume discretisation and is an essential
  part of solving theabove equations numerically. In STAR-CD both
  components of the spatialdescription are performed as part of the
  same operation, setting up the finite-volumemesh, but separate
  considerations apply to each of them.

  Solution domain definitionThrough its internal design and
  construction, STAR-CD permits a very general andflexible 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 thesame model, as in problems involving
  fluid-solid heat transfer. The users first taskis therefore to
  decide which parts of the physical system being modelled need to
  beincluded in the solution domain and whether each part is occupied
  by a fluid, solidor porous medium.

  Whatever its composition, the fundamental requirement is that
  the solutiondomain is bounded. This means that the user has to
  examine his systems geometrycarefully and decide exactly where the
  enclosing boundaries lie. The boundaries canbe one of four
  kinds:

  1. Physical boundaries walls or solid obstacles of some
  description thatserve to physically confine a fluid flow

  2. Symmetry boundaries axes or planes beyond which the problem
  solutionbecomes a mirror image of itself

  3. Cyclic boundaries surfaces beyond which the problem solution
  repeatsitself, in a cyclic or anticyclic fashion

  The purpose of symmetry and cyclic boundaries is to limit the
  size of thedomain, and hence the computer requirements, by
  excluding regions wherethe solution is essentially known. This in
  turn allows one to model theproblem in greater detail than would
  have been the case otherwise.

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Spatial description and volume discretisation

  1-4 Version 4.02

  4. Notional boundaries these are non-physical surfaces that
  serve toclose-off the solution domain in regions not covered by the
  other two typesof boundary. Their location is entirely up to the
  users discretion but, ingeneral, 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 Boundarydescription on page 1-10.

  Mesh definitionCreation of a lattice of finite-volume cells to
  represent the solution domain isnormally the most time-consuming
  task in setting up a STAR-CD model. Thisprocess is greatly
  facilitated by STAR-CD because of its ability to generate cells
  ofan arbitrary, polyhedral shape.

  In creating a finite-volume mesh, the user should aim to
  represent accurately thefollowing two entities:

  1. The overall external geometry of the solution domain, by
  specifying anappropriate size and shape for near-boundary cells.
  The latters external faces,taken together, should make up a surface
  that adequately represents the shapeof the solution domain
  boundaries. Small inaccuracies may occur because allboundary cell
  faces (including rectangular faces) are composed of
  triangularfacets, as shown in Figure 1-2. These errors diminish as
  the mesh is refined.

  Figure 1-2 Boundary representation by triangular facets

  2. The internal characteristics of the flow/deformation regime.
  This is achievedby careful control of mesh spacing within the
  solution domain interior so that

  triangular facet

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Spatial description and volume discretisation

  Version 4.02 1-5

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

  Mesh spacing considerationsThe 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 timestep.

  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 inSTAR-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 andsolid mechanics
  to produce the right kind of mesh arrangement.

  Mesh distortionMesh distortion is measured in terms of three
  factors aspect ratio, internal angleand warp angle illustrated in
  Figure 1-3.

  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

  a

  b

  b/a = aspect ratio

  = internal angle

  = warp angle

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Spatial description and volume discretisation

  1-6 Version 4.02

  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 thesefactors 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 dependon local flow conditions.
  However, the following values serve as a useful guideline:

  pro-STAR can calculate these quantities and identify cells
  having out-of-boundsvalues, as discussed in Chapter 3, Mesh and
  Geometry Checking of the MeshingUser Guide.

  What is really important in this respect is the combined effect
  of the variouskinds of mesh distortion. If all three are
  simultaneously present in a single cell, thelimits given above
  might not be stringent enough. On the other hand, the effects
  ofdistortion also depend on the nature of the local flow. Thus, the
  above limits maybe 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 andshould be minimised whenever practicable.

  Mesh distribution and densityNumerical 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
  highmesh density implies a large number of mesh storage locations,
  with associated highcomputing cost. It is therefore advisable,
  wherever possible, to

  ensure that the mesh density is high only where needed, i.e. in
  regions of steepgradients of the flow variables, and low
  elsewhere;

  avoid rapid changes in cell dimensions in the direction of steep
  gradients inthe flow variables.

  The flexibility afforded by STAR-CDs unstructured polyhedral
  meshes facilitatessuch selective refinement. An illustration of
  some of the numerous cell shapes thatmay be employed is given in
  Figure 2-43 and Figure 2-44 of the Meshing UserGuide.

  Aspect Ratio 10Internal angle 45Warp angle 45

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Spatial description and volume discretisation

  Version 4.02 1-7

  Of course, it is not always possible to ascertain a priori what
  the flow structurewill be. However, the need for higher mesh
  density can usually be anticipated inregions 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 wallsAs discussed in Chapter 6, Wall
  Boundary Conditions of the Methodologyvolume, wall functions are an
  economic way of representing turbulent boundarylayers (hydrodynamic
  and thermal) in turbulent flow calculations. These
  functionseffectively allow the boundary layer to be bridged by a
  single cell, as shown inFigure 1-4(a).

  Figure 1-4 Near-wall mesh distribution

  The location y of the cell centroids in the near-wall layer, and
  hence the thicknessof that layer, is usually determined by
  reference to the dimensionless normaldistance from the wall. For
  the wall function to be effective, this distance mustbe

  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, should lie in the approximate range 30 to 150. Note
  that the aboveconsiderations apply to Reynolds Stress models as
  well as several classes of eddyviscosity model (see Chapter 3,
  Turbulence Modelling).

  Alternative treatments that do not require the use of wall
  functions are alsoavailable. These are:

  (b) Two-layer or Low Re models

  Outerregion

  Innerregiony

  (a) Wall function model

  y+

  y+

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Problem characterisation and material property definition

  1-8 Version 4.02

  1. Two-layer turbulence models, whereby wall functions are
  replaced by aone-equation k-l model or a zero-equation
  mixing-length model

  2. Low Reynolds number models (including the V2F model), where
  viscouseffects are incorporated in the k and transport
  equations

  For the above two types of model, the solution domain should be
  divided into tworegions 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 Methodologyvolume (Chapter 6, Two-layer models), the inner
  region should contain at least15 mesh layers and encompass that
  part of the boundary layer influenced by viscouseffects.

  A more recent development, called the hybrid wall function is
  also available thatextends the low-Reynolds number formulation of
  most turbulence models. Thismay be used to capture boundary layer
  properties more accurately in cases wherethe near-wall cell size is
  not adapted for the low-Reynolds number treatment andthus achieve
  independent solutions.

  Moving mesh featuresSTAR-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 theobvious one: avoid creating excessive
  distortion when redistributing the mesh. Thiscaution also applies
  to the use of the other two features, but they have additionalrules
  and guidelines attached to them. These are summarised in the
  Methodologyvolume, Chapter 15 (Internal Sliding Mesh on page 15-5
  and Cell LayerRemoval and Addition on page 15-7). Additional
  guidelines also appear in thisvolume, Cell-layer Removal/Addition
  on page 12-14 and Sliding Meshes onpage 12-18; hence they are not
  repeated here.

  Problem characterisation and material property definitionCorrect
  definition of the physical conditions and the properties of the
  materialsinvolved is a prerequisite to obtaining the right solution
  to a problem, or indeed toobtaining any solution at all. It is also
  essential for determining whether the problemcan be modelled with
  STAR-CD. The user must therefore ensure that the problemis 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

  y+

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Problem characterisation and material property definition

  Version 4.02 1-9

  Nature of the flowIt is very important to understand the nature
  of the flow being analysed in order toselect 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
  unsteadyproblem, 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
  behaviouraccurately.

  Physical propertiesThe specification of physical properties,
  such as density, molecular viscosity,thermal conductivity, etc.
  depends on the nature of the fluids or solids involved andthe
  circumstances of use. For example, STAR-CD contains several
  built-inequations of state from which density can be calculated as
  a function of one or moreof the following field variables:

  Pressure Temperature Fluid composition

  In all cases where complex calculations are used to evaluate a
  material property, thefollowing measures are recommended:

  The relevant field variables must be assigned plausible initial
  and boundaryvalues.

  Where necessary, properties should be solved for together with
  the fieldvariables as part of the overall solution.

  In the case of strong dependencies between properties and field
  variables, theuser should consider under-relaxation of the property
  value calculations, inthe manner described in the Methodology
  volume (Chapter 7, Scalartransport equations).

  When required, STAR-CDs facility for alternative,
  user-programmablefunctions may be used.

  Force fields and energy sourcesAs 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 relativeto the viscous (or turbulent) stresses,
  the flow may become physically unstable. Inthese circumstances it
  is advisable to switch to the transient solution mode.

  It is also possible to insert additional, external force fields
  and energy sourcesvia the user programming facilities of STAR-CD.
  In such cases, it is important tounderstand the physical
  implications and avoid specifying conditions that lead to

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Boundary description

  1-10 Version 4.02

  physical or numerical instability. Examples of such conditions
  are:

  Thermal energy sources that increase linearly with temperature.
  These cangive rise to physical instability called thermal
  runaway.

  Setting the coefficient in the permeability function to avery
  small or zero value. If the local fluid velocity also becomes very
  small,the result may be numerical instability whereby small
  pressure perturbationsproduce a large change in velocities.

  Initial conditionsThe term initial conditions refers to values
  assigned to the dependent variables atall mesh points before the
  start of the calculations. Their implication depends on thetype of
  problem being considered:

  In unsteady applications, this information has a clear physical
  significanceand will affect the course of the solution. Due care
  must therefore be taken inproviding it. It sometimes happens that
  the effects of initial conditions areconfined to a start-up phase
  that is not of interest (as in, for example, flowsthat are
  temporally periodic). However, it is still advisable to take
  someprecautions in specifying initial conditions for reasons
  explained below.

  In calculating steady state problems by iterative means, the
  initial conditionswill usually have no influence on the final
  solution (apart from rare occasionswhen the solution is
  multi-valued), but may well determine the success andspeed of
  achieving it.

  Poor initial field specifications or, for transient problems,
  abrupt changes inboundary conditions put severe demands on the
  numerical algorithm whensubstituted into the finite-volume
  equations. As a consequence, the followingspecial 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
  Numericalsolution control on page 1-13. In either case, increased
  computing times can be anundesirable side effect.

  Boundary descriptionAs stated in Spatial description and volume
  discretisation on page 1-2, boundaryidentification and description
  are intimately connected with the generation of thefinite-volume
  mesh, since the boundary topography is defined by the outermost
  cellfaces. Furthermore, correct specification of the boundary
  conditions is often themain area of difficulty in setting up a
  model. Problems often arise in the followingareas:

  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

  i K i i v i+=

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Boundary description

  Version 4.02 1-11

  first instance.

  Boundary locationDifficulties in specifying boundary location
  normally arise where the flowconditions are incompletely known, for
  example at outlets. The recommendedsolutions, 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 onhow the flow is partitioned between the outlets.

  (b) Prescribed flow split outlets coexisting with prescribed
  mass outflowboundaries in the same domain.

  (c) A combination of prescribed pressure and flow-split outlet
  conditions.

  Boundary conditionsAnother source of potential difficulty is in
  boundary value specification whereverknown 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 boundaryvalues, including the turbulence transport equations
  when they are invoked

  Inappropriate setting of boundary values leads to erroneous
  results and, inextreme cases, to numerical instability

  The following recommendations can be given regarding each
  different type ofboundary:

  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
  massimbalances of up to 5%, correcting them by adjusting the
  outflows. Anerror 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

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Boundary description

  1-12 Version 4.02

  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 ateach location.

  (c) The inapplicability of prescribed split outlets to problems
  where theinflows are not fixed, e.g.

  i) in combination with pressure boundary conditions, orii) 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 theturbulence parameters indirectly, via the turbulence
  intensity and lengthscale or by extrapolating them from values in
  the interior of the solutiondomain.

  4. Stagnation conditions It is recommended to use this condition
  forboundaries lying within large reservoirs where properties are
  not significantlyaffected by flow conditions in the solution
  domain.

  5. Non-reflecting pressure and stagnation conditions A
  specialformulation of the standard pressure and stagnation
  conditions, developed tofacilitate analysis of steady-state
  turbomachinery applications

  6. Cyclic boundaries These always occur in pairs. The main
  points of adviceare:

  (a) Impose this condition only in appropriate
  circumstances.Two-dimensional axisymmetric flows with swirl is a
  good example of anappropriate application.

  (b) For axisymmetric flows, make use of the CD/UD blending
  scheme toapply the maximum level of central differencing in the
  tangentialdirection (the default blending factor is 0.95; see also
  on-line Help topicMiscellaneous Controls in STAR GUIde).

  7. Planes of symmetry It is recommended to use this condition
  fortwo-dimensional axisymmetric flows without swirl

  8. Free-stream transmissive boundaries Used only for modelling
  supersonicfree streams

  9. Transient wave transmissive boundaries Used only in problems
  involvingtransient compressible flows

  10. Riemann boundaries This condition is based on the theory of
  Riemanninvariants and its application allows pressure waves to
  leave the solutiondomain without reflection

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Numerical solution control

  Version 4.02 1-13

  Numerical solution controlProper control of the numerical
  solution process applied to the transport equationsis highly
  important, both for acceptable computational efficiency and,
  sometimes,in order to achieve a solution at all. By necessity, the
  means of controlling theprocess depend heavily on the particular
  numerical techniques employed so nouniversal guidelines can be
  given. Thus, the recommended settings vary with theparticular
  algorithm selected and the circumstances of application.

  Selection of solution procedureThe basic selection should be
  based on a correct assessment of the nature of theproblem and will
  be either

  a transient calculation, starting from well-defined initial and
  boundaryconditions and proceeding to a new state in a series of
  discrete time steps; or

  a steady-state calculation, where an unchanging flow/deformation
  patternunder a given set of boundary conditions is arrived at
  through a number ofnumerical iterations.

  PISO and SIMPLE are the two alternative solution procedures
  available inSTAR-CD. PISO is the default for unsteady calculations
  and is sometimes preferredfor steady-state ones, in cases involving
  strong coupling between dependentvariables such as buoyancy driven
  flows. SIMPLE is the default algorithm forsteady-state solutions
  and works well in most cases.

  SIMPLE is also used for transient calculations in the case of
  free surface andcavitating 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 usuallycheaper
  than outer iterations on all variables within a time step of the
  transientSIMPLE algorithm. However, there are situations in which
  PISO would requiremany correctors or even fail to converge unless
  the time step is reduced, whereasSIMPLE may allow larger time steps
  with a moderate number of outer iterations pertime step. This is
  the case when the flow changes very little but certain
  slowtransients are present in the behaviour of scalar variables
  (e.g. slow heating up ofsolid structures in the case of solid-fluid
  heat transfer problems, deposition ofchemical 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 asteady-state solution or when iterative convergence is
  difficult to achieve, it is betterto perform the calculations using
  the transient option.

  Transient flow calculations with PISOAs stated in The PISO
  algorithm on page 7-2 of the Methodology volume, PISOperforms at
  each time (or iteration) step, a predictor, followed by a number
  ofcorrectors, during which linear equation sets are solved
  iteratively for each maindependent variable. The decisions on the
  number of correctors and inner iterations(hereafter referred to as
  sweeps, to avoid confusion with outer iterationsperformed as part
  of the steady-state solution mode) are made internally on the
  basisof the splitting error and inner residual levels,
  respectively, according to prescribedtolerances and upper limits.
  The default values for the solver tolerances and

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Numerical solution control

  1-14 Version 4.02

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

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

  (1-1)

  Here, U and are a characteristic velocity and diffusivity,
  respectively, and isa mean mesh dimension. Typically, it is
  possible to operate with andstill obtain reasonable temporal
  accuracy. Values significantly above this may leadto errors and
  numerical instability, whereas smaller values will lead to
  increasedcomputing 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 occurduring the start-up phase but
  is nevertheless acceptable if, later on, the warningseither cease
  entirely or only appear occasionally, and the predictions
  lookreasonable. 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
  improvesmatters, then the cause may simply be an excessively large
  .

  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, twofoldchanges are appropriate.

  Pressure under-relaxation a value of 0.8 for pressure
  correctionunder-relaxation, using PISO, may be helpful for some
  difficult cases (e.g. forsevere mesh distortion or flows with Mach
  numbers approaching 1).

  Corrector step tolerance this may be set to a lower value but
  consult

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

  ParameterVariable

  Velocity Pressure Turbulence Enthalpy Mass fraction

  Solvertolerance 0.01 0.001 0.01 0.01 0.01

  Sweep limit 100 1000 100 100 100

  Pressure under-relaxation factor = 1.0

  Corrector limit = 20

  Corrector step tolerance = 0.25

  t

  tc

  tc minLU——

  L2————,

  =

  Lt 50 tc

  t

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Numerical solution control

  Version 4.02 1-15

  CD adapco first.

  Steady-state flow calculations with PISOWhen PISO operates in
  this mode, the inner residual tolerances are decreased
  andunder-relaxation is introduced on all variables, apart from
  pressure, temperature andmass fraction. However, the last two
  variables may need to be under-relaxed forbuoyancy driven problems.
  The standard, default values for these parameters andthe sweep
  limits, which are unchanged from the transient mode, are given in
  Table1-2.

  .

  These settings should, all being well, result in near-monotonic
  decrease in theglobal residuals during the course of the
  calculations, depending on mesh densityand other factors. If,
  thereafter, one or more of the global residuals do not fall,then
  remedial measures will be necessary. In some instances, the
  offendingvariable(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 ofbetween 0.05 and 0.10 and should be applied to the
  velocities if themomentum and/or mass residuals are at fault.

  Decrease in solver tolerances as in the transient case, this may
  provebeneficial, especially in respect of the pressure tolerance
  and its importance tothe flow solution. A twofold reduction should
  indicate whether this measurewill work.

  Increase in sweep limits if warning messages about the limits
  beingreached appear and are not suppressed by measures 1 and 2,
  then it may beworthwhile increasing the limit(s) on the offending
  variables.

  Under-relaxation of density and effective viscosity use of this
  method fordensity 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-Newtonianfluid flow and can be similarly damped by this
  device.

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

  ParameterVariable

  Velocity Pressure Turbulence Enthalpy Mass fraction

  Solvertolerance 0.1 0.05 0.1 0.1 0.1

  Sweep limit 100 1000 100 100 100

  Relaxationfactor 0.7 1.0 0.7 0.95 1.0

  Corrector limit = 20

  Corrector step tolerance = 0.25

  R

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Numerical solution control

  1-16 Version 4.02

  Steady-state flow calculations with SIMPLEAs noted previously,
  the control parameters available for SIMPLE are similar tothose for
  PISO, except that, in the case of the former, a single corrector
  stage isalways used and pressure is under-relaxed. The standard
  (default) settings are givenin Table 1-3.

  .

  In the event of failure to obtain solutions with the standard
  values, then the measuresto be taken are essentially the same as
  those for iterative PISO, given in the previoussection. However,
  here, reduction of the pressure relaxation factor is an
  additionaldevice for overcoming convergence problems. The problems
  usually arise eitherfrom a highly distorted mesh, or from highly
  complex physics (many variablesaffecting each other). If the grid
  is distorted, one should reduce the relaxation factorfor pressure
  from the beginning of the run (e.g. to 0.1). If convergence
  problems arestill encountered, a substantial reduction of the
  under-relaxation factor for velocitiesand turbulence model
  variables should be tried (e.g. to 0.5). If this does not help,
  theproblem may lie in severe mesh defects or errors in the set-up.
  Further reduction ofunder-relaxation factors may be tried if the
  grid is severely distorted and cannot beimproved; otherwise,
  improving the mesh quality can be of much greater help.

  Note that the pressure under-relaxation factor needs to be
  adjusted within thelimits of some range to make the iteration
  process converge, where the number ofiterations required to reach
  such convergence is mainly dictated by thecorresponding factors for
  velocities (and for scalar variables when strongly coupledto the
  flow). In the case of well-behaved flows and reasonable meshes,
  therelaxation factor for pressure can be selected as (1.0 —
  relaxation factor forvelocities), e.g. 0.2 for pressure and 0.8 for
  velocities. Usually, for a given velocityrelaxation factor, the one
  for pressure can be varied within some range withoutaffecting the
  total number of iterations and computing time, but outside this
  rangethe 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 0.05and 0.8 if the velocity
  factor is low, say around 0.5). On the other hand, this
  rangebecomes narrower when the mesh is distorted.

  The limit to which the velocity relaxation factor can be
  increased is bothproblem- and mesh-dependent. When many similar
  problems need to be solved, itis worth trying to work near the
  optimum as this may save a lot of computing time.

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

  ParameterVariable

  Velocity Pressure Turbulence Enthalpy Mass fraction

  Solvertolerance 0.1 0.05 0.1 0.1 0.1

  Sweep limit 100 1000 100 100 100

  Relaxationfactor 0.7 0.3 0.7 0.95 1.0

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Numerical solution control

  Version 4.02 1-17

  On the other hand, for an one-off analysis, it may be more
  efficient to use aconservative setting.

  Note that under some conditions, such as those in Tutorial 13.1,
  a steady-statesolution cannot be achieved due to the inherent
  unsteady character of the flow. Thisis often the case when the
  problem geometry possesses some form of symmetry butthe Reynolds
  (or another equivalent) number is high and recirculation zones
  arepresent. In this case the residuals stop falling at some level
  and then continue tooscillate. The solution at that stage may be
  far from a valid solution of thegoverning equations and should not
  be interpreted as such unless the residual levelis 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 asteady-state solution, while a less diffusive
  turbulence model (such as ReynoldsStress and non-linear
  eddy-viscosity models) combined with a higher-orderdifferencing
  scheme (such as central differencing) may not. In such cases,
  atransient simulation should be performed; the unsteady solution
  may oscillatearound a mean steady state, in which case the
  quantities of interest (drag, lift, heattransfer coefficient,
  pressure drop, etc.) can be averaged over several
  oscillationperiods.

  Transient flow calculations with SIMPLEThe use of this algorithm
  in transient calculations essentially consists of repeatingthe
  steady-state SIMPLE calculations for each prescribed time step. The
  defaultcontrol parameter settings are therefore as summarised in
  Table 1-4.

  .

  The main difference compared to the PISO algorithm lies in the
  fact that alllinearizations and deferred correctors are updated
  within the outer iterations, byrecalculating the coefficient matrix
  and source term. For this reason, solvertolerances do not need to
  be as tight as for PISO; they are actually identical to thoseused
  for steady-state computations. However, since the discretization of
  thetransient term enlarges the central coefficient of the matrix in
  the same way asunder-relaxation does, the relaxation factors for
  velocities and scalar variables canbe increased (the smaller the
  time step, the larger the values that can be used forrelaxation
  factors 0.95 or even more).

  The convergence criterion for outer iterations within each time
  step is by default

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

  ParameterVariable

  Velocity Pressure Turbulence Enthalpy Mass fraction

  Solvertolerance 0.1 0.05 0.1 0.1 0.1

  Sweep limit 100 1000 100 100 100

  Relaxationfactor 0.9 0.3 0.7 1.0 1.0

  Outer iteration limit = 5

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Numerical solution control

  1-18 Version 4.02

  the same as for steady-state flows. However, the number of outer
  iterations is alsoset to a default limit of 10; if substantially
  more iterations are needed to satisfy theconvergence criterion,
  this is a sign that the time step is too large. In such a case,
  itis better to reduce the time step rather than allow more outer
  iterations for a largertime 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, onemay
  increase the time step; experience shows that optimum efficiency
  and accuracyare achieved if 5 to 10 outer iterations per time step
  are performed.

  Note also that the reported mass residuals are computed before
  solving thepressure-correction equation; after this equation is
  solved and mass fluxes arecorrected, the mass residuals are more
  than an order of magnitude lower. For thisreason, one can accept
  mass residuals being somewhat higher than the convergencecriterion
  when the limiting number of outer iterations is reached, provided
  that theresiduals of all other equations have satisfied the
  criterion. In some cases, anincrease in the under-relaxation factor
  for pressure (up to 0.8) can lead to a fasterreduction of mass
  residuals. All these considerations are of course problem-dependent
  and if several simulations over a longer period need to be
  performed, itmay 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 atransient simulation than those at later stages.
  This is the case, for example, whenstarting with a fluid at rest
  and imposing a full-flow rate at the inlet, or full speed
  ofrotation (in the absence of a better initial condition). This is
  equivalent to a suddenchange of boundary conditions at a later
  time, which would also require that thetime step be reduced.
  Another possibility of avoiding problems with abrupt startsfrom
  rest is to ramp the boundary conditions (e.g. a linear increase of
  velocity fromzero to full speed over some period of time).

  The transient SIMPLE algorithm allows you to select either the
  defaultfully-implicit Euler scheme or the three-time-level scheme
  for temporaldiscretisation, described in Chapter 4, Temporal
  Discretisation of theMethodology volume. The latter scheme is
  second-order accurate but is currentlyapplied only to the momentum
  and continuity equations. It should be chosen whentemporal
  variation of the velocity field is essential, e.g. in the case of a
  DES/LEStype 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 qualityetc.), transient SIMPLE may
  allow the use of larger time steps than PISO withoutloss of
  accuracy.

  Effect of round-off errorsEfforts have been made to minimise the
  susceptibility of STAR-CD to the effectsof machine round-off
  errors, but problems can sometimes arise when operating insingle
  precision on 32-bit machines. They usually manifest themselves as
  failure ofthe iterative solvers to converge or, in extreme cases,
  in divergence leading tomachine overflow.

  If difficulties are encountered with problems of this kind, then
  it is clearlyadvisable to switch to double precision calculations.
  Instructions on how to do thisare provided in the Installation
  Manual. As a general rule, however, you should tryto avoid
  generating very small values for cell volumes and cell face areas
  byworking with sensible length units. Alternatively, you could
  re-specify your

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Monitoring the calculations

  Version 4.02 1-19

  problem geometry units while preserving relevant non-dimensional
  quantities suchas Re and Gr.

  Choice of the linear equation solverSTAR-CD offers two types of
  preconditioning of its conjugate gradient linearequations solvers:
  one which vectorises fully, and the other, which is
  numericallysuperior to the first one but vectorises only partially.
  Therefore, the first one (calledvector solver) is recommended when
  the code is run on vector machines (such asFujitsu and Hitachi
  computers), and the second one (called scalar solver) isrecommended
  if the code is run on scalar machines (such as workstations).

  Monitoring the calculationsChapter 5 and the section on
  Permanent Output on page 15-1 give details of theinformation
  extracted from the calculations at each iteration or time step and
  usedfor 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-statecalculations. Ideally, it should be in a sensitive
  region of the flow where theapproach 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 andother
  criteria for choice of location may apply. For example, a location
  maybe chosen so as to confirm an expected periodic behaviour in the
  flowvariables.

  The normalised global residuals for all equations solved. Apart
  fromturbulence dissipation rate residuals (see Chapter 7,
  Completion tests in theMethodology volume), these are used to judge
  the progress and completion ofiterative calculations for steady and
  pseudo-transient solutions. In the earlystages of a calculation,
  the non-linearities and interdependencies of theequations may
  result in non-monotonic decrease of the residuals. If
  theseoscillations persist after, say, 50 iterations, this may be
  indicative of problems.

  Remember that reduction of the normalised residuals to the
  prescribed tolerance ()is a necessary but not sufficient condition
  for convergence, for two reasons:

  1. The normalisation practices used (see Chapter 7, Completion
  tests in theMethodology volume) may not be appropriate for the
  application.

  2. It is also necessary that the features of interest in the
  solution should havestabilised to an acceptable degree.

  If doubts exist in either respect, it is advisable to reduce the
  tolerance and continuethe calculations.

  It follows from the above discussion that strong reliance is
  placed on the globalresiduals to judge the progress and completion
  of iterative calculations of steadyflows. These quantities provide
  a direct measure of the degree of convergence of theindividual
  equation sets and are therefore useful both for termination tests
  and foridentifying problem areas when convergence is not being
  achieved.

  R

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Model evaluation

  1-20 Version 4.02

  Model evaluationChecking the modelSTAR-CD offers a variety of
  tools to help assess the accuracy and effectiveness ofall aspects
  of the model building process. In performing the modelling
  stagesdiscussed previously, the user should therefore take
  advantage of these facilities andcheck that:

  1. The mesh geometry agrees with what it is

• STAR-CCM+ 6.04 : 02.08.2011

  : (Multi-Component Flow Tutorials)

  , :

  , . ( ) . , . :

• STAR-CCM+ 6.04 : 07.08.2011

  2

  (1 , 293 ) . ( 16 /) 323 . , ( ) . . : — , — .

• STAR-CCM+ 6.04 : 07.08.2011

  3

  : ( ) (Dilution Pipe (Steady Flow) Tutorial)

  . 5 / 10 / .

  (Importing the Mesh and Naming the Simulation).

  STAR-CCM+ , , (New Simulation). . . .ccm.

  > > (File > Import > Import Volume Mesh). (Open)
  //multiphase (doc/tutorials/multiphase), , STAR-CCM+,
  dilPipe.ccm.

  (Open), .

• STAR-CCM+ 6.04 : 07.08.2011

  4

  STAR-CCM+ , (Output window). (Graphics) , .

  dilPipeSteady.sim.

  (Visualizing the Geometry) , , . , :

  > 1 > > 1 > (Scenes > Geometry Scene 1 >
  Displayers > Geometry 1 > Parts).

• STAR-CCM+ 6.04 : 07.08.2011

  5

  (Properties) (Parts). (List View), Default_Fluid: symmetry
  plane.

  (Apply) (Close).

  1 (Geometry Scene 1) 180 , , .

  , . , , (baffles), , .

  (Renaming Regions and Boundaries)

• STAR-CCM+ 6.04 : 07.08.2011

  6

  . .

  > Default_Fluid (Regoins > Default_Fluid) Fluid.

  . , , Default_Boundary_Region.

  > > Fluid >

  Default_Boundary_Region (Regions > Boundaries > Fluid >
  Default_Boundary_Region) (Wall).

  (Scaling the Mesh)

  , .ccm . > (Mesh > Diagnostics…), (Output), .

• STAR-CCM+ 6.04 : 07.08.2011

  7

  , 0 5.0 -, 0 5.5 Y-, 0 1.5 Z-. , STAR-CCM+ . 0.0254, .

  > (Mesh > Scale Mesh).

  (Scale Mesh) ,

  Fluid 0.0254 .

• STAR-CCM+ 6.04 : 07.08.2011

  8

  (Apply), , (Close).

  (Apply) .

  _ (Reset_View) , .

  , , (Output).

  (Setting up the Models) , , , , , . . (Segregated Flow)
  K-Epsilon. , , 1 (Physics 1). :

  > 1 (Continua > Physics 1) Dilution Pipe.

• STAR-CCM+ 6.04 : 07.08.2011

  9

  :

  Dilution Pipe (Select models).

  (Physics Model Selection) , , .

  , (Three Dimensional) ( (Space)).

  (Multi-Component) (Material).

  (Non-Reacting) (Reaction Regime).

  (Segregated Flow) (Flow)

  (Ideal Gas) (Equation of State) (Segregated Fluid Temperature) (
  , (Auto-select recommended Physics models) ).

  (Steady) (Time). (Turbulent)

  (Viscous Regime) K-Epsilon (K-Epsilon Turbulence)

  , (Reynolds-Averaged Turbulence).

  (Physics Model Selection) :

• STAR-CCM+ 6.04 : 07.08.2011

  10

  (Close). , Dilution Pipe dilPipeSteady , , .

  Dilution Pipe > (Dilution Pipe > Models). .

• STAR-CCM+ 6.04 : 07.08.2011

  11

  , (Save).

  (Setting Material Properties) . , STAR-CCM+ . : 16 /, 2230 /(*),
  0.038 W/m-K 1.175E-5 Pas. , (Setting up the Models).

  > > (Models > Multi-Component Gas > Gas
  Mixture).

  (Gas Components) .

  (Gas Components) (Select Mixture Components).

• STAR-CCM+ 6.04 : 07.08.2011

  12

  (Select Mixture Components) (Material Databases), (Standard)
  (Gases), .

  (Air) CH4 (),

  . ( ).

  (Apply), (Close). (Gas Components) , .

• STAR-CCM+ 6.04 : 07.08.2011

  13

  CH4 (Methane). , :

  (Gas Components) (Reorder Mixture Components). (Reorder Mixture
  Components) (Methane) .

  (Component

  Properties). (Constant)

• STAR-CCM+ 6.04 : 07.08.2011

  14

  :

  (Dynamic Viscosity) = 1.175E-5 Pa-s (Molecular Weight) = 16
  kg/kg.mol (Specific Heat) = 2230 J/kg-K (Thermal Conductivity) =
  0.038 W/m-K

  .

  (Setting Initial Conditions)

  293 .

  > > > (Dilution Pipe > Initial Conditions >
  Species Mass Fraction > Constant)

  (Properties) (Value).

  (Constant Value)

  1.0 (Air).

• STAR-CCM+ 6.04 : 07.08.2011

  15

  .

  > (Static Temperature > Constant).

  (Properties) 293 .

  .

• STAR-CCM+ 6.04 : 07.08.2011

  16

  (Setting Boundary Conditions and Values)

  , . , .

  > Fluid > > _ > > > (Regions > Fluid >
  Boundaries > Inner_inlet > Physics Values > Mass Fraction
  > Constant).

  (Value) 0.0, 1.0

  > (Static Temperature > Constant).

  (Value) 323 .

  >

• STAR-CCM+ 6.04 : 07.08.2011

  17

  (Turbulence Intensity > Constant).

  (Properties) (Value) 0.1.

  (Turbulent Viscosity Ratio) 100 (Velocity Magnitude) 10 /.

  .

  _ (Outer_inlet) (Physics Values) :

  (Species Mass Fraction) = [1.0, 0.0] (Static Temperature) = 293
  K (Turbulence Intensity) = 0.1 (Turbulent Viscosity Ratio) =

  40 .

  pressure Physics Values. :

  Mass Fraction = [1.0, 0.0] Static Temperature = 308 K Turbulence
  Intensity = 0.1 Turbulent Viscosity Ratio = 40

  , , . , .

  .

  (Setting Solver Parameters and Stopping Criteria)

  , (under-relaxation factors)

• STAR-CCM+ 6.04 : 07.08.2011

  18

  . (under-relaxation factors) . , . 300 , .

  , . (Stopping Criteria) (Maximum Steps).

  (Properties) (Maximum Steps), 300.

  300 , .

  .

  (Visualizing the Solution)

  . .

  (Scenes) > (New Scene > Scalar).

• STAR-CCM+ 6.04 : 07.08.2011

  19

  1 (Scalar Scene 1)

  > -Z > +Y (Look Down > -Z > Up +Y).

  > 1 > > 1 (Scenes > Scalar Scene 1 > Displayers
  > Scalar 1).

• STAR-CCM+ 6.04 : 07.08.2011

  20

  (Properties), (Contour Style) (Smooth Filled).

  1 (Scalar 1) (Parts).

  (Properties) (Parts). (Select All), (Regions) Fluid , Fluid ,
  .

• STAR-CCM+ 6.04 : 07.08.2011

  21

  .

  (Scalar Field).

  (Function)

  (Temperature).

  .

  (Running the Simulation)

  (Run) .

  , > (Solution > Run). (Solution), > > (Tools >
  Toolbars > Solution), . (Residuals), . , (Residuals), . . , .
  (Graphics) .

  1 (Scalar Scene 1) .

• STAR-CCM+ 6.04 : 07.08.2011

  22

  ,

  (Stop) . , (Run). , , 300 .

  , .

  (Visualizing the Results)

  1 (Scalar Scene 1), .

• STAR-CCM+ 6.04 : 07.08.2011

  23

  , .

  > (Mass Fraction > Methane).

  ()

• STAR-CCM+ 6.04 : 07.08.2011

  24

  .

  (Pressure). , , . .

  1 (Scalar 1). (Properties)

  (Transform) 1 (symmetry plane 1).

• STAR-CCM+ 6.04 : 07.08.2011

  25

  .

  (Properties) , (Smooth Shade) .

  , , , . , (Transform) (Outline).

  Fluid: wall Fluid: symmetry plane (Outline), , .

• STAR-CCM+ 6.04 : 07.08.2011

  26

  .

  (Adding Streamlines)

  .

  1 (Geometry Scene 1) ,

  .

  dilutionPipeSteady (Derived Parts) > (New Part >
  Streamline…).

  , .

• STAR-CCM+ 6.04 : 07.08.2011

  27

  , (Input Parts) Fluid.

  , (Vector Field) (Velocity) .

  , (Part Seed)

  (Seed Mode).

  Fluid: Inner_inlet Fluid: Outer_inlet (Seed Parts).

• STAR-CCM+ 6.04 : 07.08.2011

  28

  . , (Display)

  (New Streamline Displayer). :

• STAR-CCM+ 6.04 : 07.08.2011

  29

  (Create), (Close). , , .

• STAR-CCM+ 6.04 : 07.08.2011

  30

  .

  : (Velocity: Magnitude).

  :

  , , , .

  , ,

• STAR-CCM+ 6.04 : 07.08.2011

  31

  , (Derived Parts) streamline, (Source Seed).

  (Properties) (Randomize). :

  .

• STAR-CCM+ 6.04 : 07.08.2011

  32

  (Summary) STAR-CCM+:

  .

  .

  .

  .

  .

  .

  , .

  .

  .

  .

  .

  .

  .

  .

• STAR-CCM+ 6.04 : 07.08.2011

  33

  : ( )

  (Dilution Pipe (Unsteady Flow) Tutorial)

  , ( ) (Dilution Pipe (Steady Flow) Tutorial). , , . —
  (velocity-time profiles) :

  (Loading an Existing Simulation)

  STAR-CCM+ , .

  > (File > Load Simulation…). (Load Simulation).

• STAR-CCM+ 6.04 : 07.08.2011

  34

  (Browse…) dilPipeSteady.sim, ( ) (Dilution Pipe (Steady Flow)
  Tutorial).

  dilPipeSteady.sim (Open)

  .

  (Load Simulation).

  STAR-CCM+ (Explorer) , ( ) (Dilution Pipe (Steady Flow)
  Tutorial), dilPipeSteady.

• STAR-CCM+ 6.04 : 07.08.2011

  35

  (Renaming the Simulation)

  .

  > (File > Save As…), . (Save) , .

  dilPipeUnsteady (File Name).

  (Save).

• STAR-CCM+ 6.04 : 07.08.2011

  36

  (Explorer), (Properties) (Output) .

  (Setting up the Unsteady Flow Model)

  ( ) (Dilution Pipe (Steady Flow) Tutorial), . , .

  (Continua), (Dilution Pipe) (Select models…).

  (Phusics Model Selection) :

  (Steady) (Enabled Physics models) .

  (Time).

  ( ) (Implicit Unsteady). :

• STAR-CCM+ 6.04 : 07.08.2011

  37

  (Close).

  , (Save).

  (Creating Field Functions)

  . STAR-CCM+ . :

  v=10-4t

  v=5+2t :

  (Tools), (Field Functions) (New).

• STAR-CCM+ 6.04 : 07.08.2011

  38

  1 (User Field Function 1).

  (Rename…).

  (Inner Inlet Velocity).

  (Properties)

  (Definition) 1 (User Field Function 1 Definition) ,

• STAR-CCM+ 6.04 : 07.08.2011

  39

  (Outer Inlet Velocity) 5 + 2 * $Time

  .

  (Setting Boundary Conditions and Values)

  , , (Creating Field Functions). :

  > Fluid > > _ > > (Regions > Fluid >
  Boundaries > Inner_inlet > Physics Values > Velocity
  Magnitude).

  (Properties) (Method) (Field Function).

• STAR-CCM+ 6.04 : 07.08.2011

  40

  > (Velocity Magnitude > Field Function).

  (Properties)

  (Inner Inlet Velocity) (Scalar Function).

  > _ (Boundaries > Outer_inlet) , , , . (Outer Inlet
  Velocity).

• STAR-CCM+ 6.04 : 07.08.2011

  41

  .

  (Setting Solver Parameters)

  , ( ) (Dilution Pipe (Steady Flow)) 2 . 0.025, , 80 . :

  > ( ) (Solvers > Implicit Unsteady).

  (Properties) (Time-Step)

  0.025.

  :

  > (Stopping Criteria > Maximum Physical Time).

  (Properties)

  (Max Physical Time) 2.0

  .

• STAR-CCM+ 6.04 : 07.08.2011

  42

  (Autosaving the Simulation)

  0.5 . , . :

  > (File > Auto Save…).

  (Auto Save) , :

  (Close).

  (Running the Simulation)

  , , . 1 (Scalar Scene 1):

  (Scenes), 1 (Scalar Scene 1) (Open).

• STAR-CCM+ 6.04 : 07.08.2011

  43

  (Graphics) .

  (Mass Fraction of Methane), (Graphics).

  , 1 (Scalar

  1) (Transform) (Identity).

  1 (Outline 1) (Transform) (Identity).

• STAR-CCM+ 6.04 : 07.08.2011

  44

  , > -Z > +Y (Look Down > -Z > Up +Y).

  (Run) . , > (Solution > Run). (Solution), > > (Tools
  > Toolbars > Solution), . (Residuals), . , (Residuals), .
  (Graphics) .

  ,

  (Stop) . , (Run). , , 2 .

  1 (Scalar Scene 1), .

• STAR-CCM+ 6.04 : 07.08.2011

  45

  (Visualizing the Results)

  :

  2- :

  (Scenes) > (New Scene > Vector).

  , > -Z > +Y

  (Look Down > -Z > Up +Y).

  > 1 > (Displayers > Vector 1 > Parts).

  (Properties)

  (Parts). (Regions), Fluid .

• STAR-CCM+ 6.04 : 07.08.2011

  46

  , :

  0.5 :

  (File) (Load Simulation…) [email protected]

• STAR-CCM+ 6.04 : 07.08.2011

  47

  :

  (Scenes) 1 (Scalar Scene 1), (Graphics) .

  0.5 , . , :

• STAR-CCM+ 6.04 : 07.08.2011

  48

  [email protected]+00

  1 (Vector Scene 1).

  (opy).

  [email protected] (Scenes).

  (Paste).

• STAR-CCM+ 6.04 : 07.08.2011

  49

  1 (Vector Scene 1). , (Graphics).

  0.5 2 .

  .

• STAR-CCM+ 6.04 : 07.08.2011

  50

  (Summary)

  STAR-CCM+:

  .

  .

  .

  .

  .

  .

  .

  .

  .

Description

From the STAR-CCM+ Home Page: Much more than just a CFD solver, STAR-CCM+ is an entire engineering process for solving problems involving flow (of fluids or solids), heat transfer and stress.

Version

• 15.06.007

• Authorized students in Mechanical Engineering
• Members of the Society of Automotive Engineers at USF

Platforms

• CIRCE cluster
• RRA cluster
• SC cluster

Modules

STAR-CCM+ requires the following module file to run:

• apps/star-ccm/15.06.007

• See Modules for more information.

Running STAR-CCM+ on CIRCE/SC

The STAR-CCM+ user guide is essential to understanding the application and making the most of it. The guide and this page should help you to get started with your simulations. Please refer to the Documentation section for a link to the guide.

• Note on CIRCE: Make sure to run your jobs from your $WORK directory!
• Note: Scripts are provided as examples only. Your SLURM executables, tools, and options may vary from the example below. For help on submitting jobs to the queue, see our SLURM User’s Guide.

If you need more control over your workflow, keep reading below.

If you will be using a Power-on-Demand key, please contact Research Computer for additional instructions.

Interactive Execution via CIRCE/SC Desktop Environment

Establishing a GUI connection to CIRCE/SC

To use STAR-CCM+, you will need to connect to CIRCE/SC with GUI redirection, either using:

• CIRCE/SC Desktop Environment
• SSH with X11 redirection
  • If connecting from OSX or Linux via SSH, please ensure that you use one of the following commands to properly redirect X11:

    • [user@localhost ~]$ ssh -X circe.rc.usf.edu

      or

    • [user@localhost ~]$ ssh -X sc.rc.usf.edu

Once connected to CIRCE/SC, you can open STAR-CCM+ using the steps below:

[user@login0 ~]$ module add apps/star-ccm/15.06.007
[user@login0 ~]$ starccm+

How to Submit Jobs

Provided are batch scripts for running STAR-CCM+ as a single processor and distributed parallel job. Existing STAR-CCM+ SIM files will work in parallel mode with no modification, but only larger models and geometries will see any performance benefit. These scripts can be copied into your work directory (the folder with your input files and database files) so that you can submit batch processes to the queue.

If, for example, you have STAR-CCM+ simulation file called “test.sim”, you would set up your serial/distributed parallel submit scripts like this:

• The scripts below (for testing, name it “starccm-serial-test.sh” or name it “starccm-parallel-test.sh”, respectively) can be copied into your job directory (the folder with your input files) and modified so that you can submit batch processes to the queue.

Serial Submit Script

#!/bin/bash
#
#SBATCH --comment=starccm-serial-test
#SBATCH --ntasks=1
#SBATCH --job-name=starccm-serial-test
#SBATCH --output=output.%j.starccm-serial-test
#SBATCH --time=08:00:00

#### SLURM 1 processor STAR-CCM+ test to run for 8 hours.

module purge
module add apps/star-ccm/15.06.007

export PATH=$TMPDIR:$PATH

starccm+ -pio -batch test.sim

Distributed Parallel Submit script

#!/bin/bash
#
#SBATCH --comment=starccm-parallel-test
#SBATCH --ntasks=32
#SBATCH --job-name=starccm-parallel-test
#SBATCH --output=output.%j.starccm-parallel-test
#SBATCH --time=08:00:00

#### SLURM 32 processor STAR-CCM+ test to run for 8 hours.

module purge
module add apps/star-ccm/15.06.007

export PATH=$TMPDIR:$PATH

# Create our hosts file ala slurm
NODEFILE="$(pwd)/slurmhosts.$SLURM_JOB_ID.txt"
srun hostname -s &> $NODEFILE

starccm+ -np $SLURM_NTASKS -machinefile $NODEFILE -mpi intel -rsh ssh -pio -batch "test.sim"

 
Next, you can change to your job’s directory, and run the sbatch command to submit the job:

[user@login0 ~]$ cd my/jobdir
[user@login0 jobdir]$ sbatch ./starccm-serial-test.sh

• You can view the status of your job with the “squeue -u <username>” command

Home Page, User Guides, and Manuals

• STAR-CCM+ Home Page
  • http://www.cd-adapco.com/products/star-ccm
• Local Documentation
  • /apps/star-ccm/15.06.007/STAR-CCM+15.06.007/doc/

More Job Information

See the following for more detailed job submission information:

• SLURM User’s Guide
• Scheduling and Dispatch Policies
• Advanced Submit Techniques

Reporting Bugs

Report bugs with STAR-CCM+ to the IT Help Desk: rc-help@usf.edu

• CCM USER GUIDE

  STAR-CD VERSION 4.02

  CONFIDENTIAL FOR AUTHORISED USERS ONLY

  2006 CD-adapco

• Version 4.02 i

  TABLE OF CONTENTS

  OVERVIEW

  1 COMPUTATIONAL ANALYSIS PRINCIPLESIntroduction
  …………………………………………………………………………………………………
  1-1The Basic Modelling Process
  …………………………………………………………………………
  1-1Spatial description and volume discretisation
  ………………………………………………….. 1-2

  Solution domain definition
  ……………………………………………………………………
  1-3Mesh definition
  ……………………………………………………………………………………
  1-4Mesh distortion
  ……………………………………………………………………………………
  1-5Mesh distribution and density
  ……………………………………………………………….
  1-6Mesh distribution near walls
  …………………………………………………………………
  1-7Moving mesh features
  ………………………………………………………………………….
  1-8

  Problem characterisation and material property definition
  ………………………………… 1-8Nature of the flow
  ………………………………………………………………………………..
  1-9Physical properties
  ……………………………………………………………………………….
  1-9Force fields and energy sources
  …………………………………………………………….
  1-9Initial conditions
  ………………………………………………………………………………..
  1-10

  Boundary description
  ………………………………………………………………………………….
  1-10Boundary location
  ……………………………………………………………………………..
  1-11Boundary conditions
  …………………………………………………………………………..
  1-11

  Numerical solution control
  ………………………………………………………………………….
  1-13Selection of solution procedure
  ……………………………………………………………
  1-13Transient flow calculations with PISO
  ………………………………………………….
  1-13Steady-state flow calculations with PISO
  ……………………………………………..
  1-15Steady-state flow calculations with SIMPLE
  ………………………………………… 1-16Transient flow
  calculations with SIMPLE
  ……………………………………………. 1-17Effect of
  round-off errors
  ……………………………………………………………………
  1-18Choice of the linear equation solver
  ……………………………………………………..
  1-19

  Monitoring the calculations
  …………………………………………………………………………
  1-19Model evaluation
  ……………………………………………………………………………………….
  1-20

  2 BASIC STAR-CD FEATURESIntroduction
  …………………………………………………………………………………………………
  2-1Running a STAR-CD Analysis
  ………………………………………………………………………
  2-2

  Using the script-based procedure
  …………………………………………………………..
  2-3Using STAR-Launch
  ……………………………………………………………………………
  2-8

  pro-STAR Initialisation
  ………………………………………………………………………………
  2-12Input/output window
  ………………………………………………………………………….
  2-13Main window
  …………………………………………………………………………………….
  2-15

• ii Version 4.02

  The menu bar
  …………………………………………………………………………………….2-16General
  Housekeeping and Session Control
  …………………………………………………..2-18

  Basic set-up
  ……………………………………………………………………………………….2-18Screen
  display control
  …………………………………………………………………………2-18Error
  messages
  …………………………………………………………………………………..2-19Error
  recovery
  ……………………………………………………………………………………2-20Session
  termination
  …………………………………………………………………………….2-21

  Set Manipulation
  ………………………………………………………………………………………..2-21Table
  Manipulation
  …………………………………………………………………………………….2-24

  Basic functionality
  ……………………………………………………………………………..2-24The
  table editor
  ………………………………………………………………………………….2-26Useful
  points
  ……………………………………………………………………………………..2-31

  Plotting Functions
  ……………………………………………………………………………………….2-31Basic
  set-up
  ……………………………………………………………………………………….2-31Advanced
  screen control
  ……………………………………………………………………..2-32Screen
  capture
  ……………………………………………………………………………………2-33

  The Users Tool
  …………………………………………………………………………………………..2-35Getting
  On-line Help
  …………………………………………………………………………………..2-35The
  STAR GUIde Environment
  ……………………………………………………………………2-38

  Panel navigation system
  ………………………………………………………………………2-40STAR
  GUIde usage
  ……………………………………………………………………………2-41

  General Guidelines
  ……………………………………………………………………………………..2-413
  MATERIAL PROPERTY AND PROBLEM CHARACTERISATION

  Introduction
  …………………………………………………………………………………………………3-1The
  Cell Table
  ……………………………………………………………………………………………..3-1

  Cell indexing
  ……………………………………………………………………………………….3-3Multi-Domain
  Property Setting
  ………………………………………………………………………3-5

  Setting up models
  …………………………………………………………………………………3-6Compressible
  Flow
  ……………………………………………………………………………………….3-9

  Setting up compressible flow models
  ……………………………………………………..3-9Useful
  points on compressible flow
  ………………………………………………………3-10

  Non-Newtonian Flow
  ………………………………………………………………………………….3-11Setting
  up non-Newtonian models
  ………………………………………………………..3-11Useful
  points on non-Newtonian flow
  …………………………………………………..3-11

  Turbulence Modelling
  …………………………………………………………………………………3-12Wall
  functions
  ……………………………………………………………………………………3-13Two-layer
  models
  ………………………………………………………………………………3-13Low
  Re models
  ………………………………………………………………………………….3-14Hybrid
  wall boundary condition
  …………………………………………………………..3-14

• Version 4.02 iii

  Reynolds Stress models
  ………………………………………………………………………
  3-15DES models
  ………………………………………………………………………………………
  3-15LES models
  ………………………………………………………………………………………
  3-15Changing the turbulence model in use
  …………………………………………………. 3-16

  Heat Transfer In Solid-Fluid Systems
  ……………………………………………………………
  3-16Setting up solid-fluid heat transfer models
  ……………………………………………. 3-17Heat
  transfer in baffles
  ……………………………………………………………………….
  3-18Useful points on solid-fluid heat transfer
  ……………………………………………… 3-19

  Buoyancy-driven Flows and Natural Convection
  …………………………………………… 3-20Setting up
  buoyancy-driven models
  ……………………………………………………..
  3-20Useful points on buoyancy-driven flow
  ……………………………………………….. 3-20

  Fluid Injection
  ……………………………………………………………………………………………
  3-21Setting up fluid injection models
  ………………………………………………………….
  3-22

  4 BOUNDARY AND INITIAL CONDITIONSIntroduction
  …………………………………………………………………………………………………
  4-1Boundary Location
  ……………………………………………………………………………………….
  4-1

  Command-driven facilities
  ……………………………………………………………………
  4-2Boundary set selection facilities
  …………………………………………………………….
  4-3Boundary listing
  ………………………………………………………………………………….
  4-3

  Boundary Region Definition
  ………………………………………………………………………….
  4-5Inlet Boundaries
  …………………………………………………………………………………………..
  4-9

  Introduction
  ………………………………………………………………………………………..
  4-9Useful points
  ……………………………………………………………………………………..
  4-10

  Outlet Boundaries
  ………………………………………………………………………………………
  4-11Introduction
  ………………………………………………………………………………………
  4-11Useful points
  ……………………………………………………………………………………..
  4-12

  Pressure Boundaries
  ……………………………………………………………………………………
  4-12Introduction
  ………………………………………………………………………………………
  4-12Useful points
  ……………………………………………………………………………………..
  4-13

  Stagnation Boundaries
  ………………………………………………………………………………..
  4-14Introduction
  ………………………………………………………………………………………
  4-14Useful points
  ……………………………………………………………………………………..
  4-15

  Non-reflective Pressure and Stagnation Boundaries
  ……………………………………….. 4-16Introduction
  ………………………………………………………………………………………
  4-16Useful points
  ……………………………………………………………………………………..
  4-18

  Wall Boundaries
  …………………………………………………………………………………………
  4-19Introduction
  ………………………………………………………………………………………
  4-19Thermal radiation properties
  ……………………………………………………………….
  4-20Solar radiation properties
  ……………………………………………………………………
  4-20

• iv Version 4.02

  Other radiation modelling considerations
  ………………………………………………4-21Useful
  points
  ……………………………………………………………………………………..4-22

  Baffle Boundaries
  ……………………………………………………………………………………….4-23Introduction
  ……………………………………………………………………………………….4-23Setting
  up models
  ……………………………………………………………………………….4-24Thermal
  radiation properties
  ………………………………………………………………..4-25Solar
  radiation properties
  …………………………………………………………………….4-26Other
  radiation modelling considerations
  ………………………………………………4-26Useful
  points
  ……………………………………………………………………………………..4-27

  Symmetry Plane Boundaries
  ………………………………………………………………………..4-27Cyclic
  Boundaries
  ………………………………………………………………………………………4-27

  Introduction
  ……………………………………………………………………………………….4-27Setting
  up models
  ……………………………………………………………………………….4-28Useful
  points
  ……………………………………………………………………………………..4-30Cyclic
  set manipulation
  ……………………………………………………………………….4-31

  Free-stream Transmissive Boundaries
  …………………………………………………………..4-32Introduction
  ……………………………………………………………………………………….4-32Useful
  points
  ……………………………………………………………………………………..4-33

  Transient-wave Transmissive Boundaries
  ………………………………………………………4-34Introduction
  ……………………………………………………………………………………….4-34Useful
  points
  ……………………………………………………………………………………..4-35

  Riemann Boundaries
  …………………………………………………………………………………..4-36Introduction
  ……………………………………………………………………………………….4-36Useful
  points
  ……………………………………………………………………………………..4-37

  Attachment Boundaries
  ……………………………………………………………………………….4-38Useful
  points
  ……………………………………………………………………………………..4-39

  Radiation Boundaries
  ………………………………………………………………………………….4-39Useful
  points
  ……………………………………………………………………………………..4-40

  Phase-Escape (Degassing) Boundaries
  ………………………………………………………….4-40Monitoring
  Regions
  …………………………………………………………………………………….4-40Boundary
  Visualisation
  ……………………………………………………………………………….4-41Solution
  Domain Initialisation
  ……………………………………………………………………..4-42

  Steady-state problems
  …………………………………………………………………………4-42Transient
  problems
  ……………………………………………………………………………..4-42

  5 CONTROL FUNCTIONSIntroduction
  …………………………………………………………………………………………………5-1Analysis
  Controls for Steady-State Problems
  …………………………………………………..5-1Analysis
  Controls for Transient Problems
  ……………………………………………………….5-4

  Default (single-transient) solution mode
  …………………………………………………5-4

• Version 4.02 v

  Load-step based solution mode
  ……………………………………………………………..
  5-6Load step characteristics
  ……………………………………………………………………….
  5-6Load step definition
  ……………………………………………………………………………..
  5-8Solution procedure outline
  ……………………………………………………………………
  5-9Other transient functions
  …………………………………………………………………….
  5-14

  Solution Control with Mesh Changes
  ……………………………………………………………
  5-15Mesh-changing procedure
  …………………………………………………………………..
  5-15

  Solution-Adapted Mesh Changes
  …………………………………………………………………
  5-176 POROUS MEDIA FLOW

  Setting Up Porous Media Models
  …………………………………………………………………..
  6-1Useful Points
  ……………………………………………………………………………………………….
  6-4

  7 THERMAL AND SOLAR RADIATIONRadiation Modelling for Surface
  Exchanges
  ……………………………………………………
  7-1Radiation Modelling for Participating Media
  …………………………………………………..
  7-3Capabilities and Limitations of the DTRM Method
  …………………………………………. 7-5Capabilities
  and Limitations of the DORM Method
  …………………………………………. 7-7Radiation
  Sub-domains
  …………………………………………………………………………………
  7-8

  8 CHEMICAL REACTION AND COMBUSTIONIntroduction
  …………………………………………………………………………………………………
  8-1Local Source Models
  ……………………………………………………………………………………
  8-2Presumed Probability Density Function (PPDF) Models
  ………………………………….. 8-3

  Single-fuel PPDF
  …………………………………………………………………………………
  8-3Multiple-fuel PPDF
  ……………………………………………………………………………..
  8-9

  Regress Variable Models
  …………………………………………………………………………….
  8-10Complex Chemistry Models
  ………………………………………………………………………..
  8-11Setting Up Chemical Reaction Schemes
  ………………………………………………………..
  8-14

  Useful general points for local source and regress variable
  schemes ……….. 8-16Chemical Reaction Conventions
  ………………………………………………………….
  8-18Useful points for PPDF schemes
  ………………………………………………………….
  8-18Useful points for complex chemistry models
  ………………………………………… 8-21Useful points
  for ignition models
  …………………………………………………………
  8-21

  Setting Up Advanced I.C. Engine Models
  ……………………………………………………..
  8-22Coherent Flame model (CFM)
  …………………………………………………………….
  8-24Extended Coherent Flame model (ECFM)
  ……………………………………………. 8-26Extended
  Coherent Flame model 3Z (ECFM-3Z) spark ignition …………
  8-28Extended Coherent Flame model 3Z (ECFM-3Z) compression ignition
  .8-29Useful points for ECFM models
  …………………………………………………………..
  8-30Level Set model
  …………………………………………………………………………………
  8-31Write Data sub-panel
  ………………………………………………………………………….
  8-32

• vi Version 4.02

  The Arc and Kernel Tracking ignition model (AKTIM)
  ………………………….8-33Useful points for the AKTIM
  model
  …………………………………………………….8-35The
  Double-Delay autoignition model
  ………………………………………………….8-37

  NOx Modelling
  …………………………………………………………………………………………..8-39Soot
  Modelling
  …………………………………………………………………………………………..8-39Coal
  Combustion Modelling
  ………………………………………………………………………..8-41

  Stage 1
  ………………………………………………………………………………………………8-41Stage
  2
  ………………………………………………………………………………………………8-42Useful
  notes
  ………………………………………………………………………………………8-44Switches
  and constants for coal modelling
  …………………………………………….8-45Special
  settings for the Mixed-is-Burnt and Eddy Break-Up models
  …………8-46

  9 LAGRANGIAN MULTI-PHASE FLOWSetting Up Lagrangian Multi-Phase
  Models
  …………………………………………………….9-1Data
  Post-Processing
  …………………………………………………………………………………….9-4

  Static displays
  ……………………………………………………………………………………..9-5Trajectory
  displays
  ……………………………………………………………………………….9-8

  Engine Combustion Data Files
  ……………………………………………………………………….9-9Useful
  Points
  ……………………………………………………………………………………………..9-10

  10 EULERIAN MULTI-PHASE FLOWIntroduction
  ……………………………………………………………………………………………….10-1Setting
  up multi-phase models
  ……………………………………………………………………..10-1

  Useful points on Eulerian multi-phase flow
  …………………………………………..10-411 FREE
  SURFACE AND CAVITATION

  Free Surface Flows
  ……………………………………………………………………………………..11-1Setting
  up free surface cases
  ………………………………………………………………..11-1

  Cavitating Flows
  …………………………………………………………………………………………11-5Setting
  up cavitation cases
  …………………………………………………………………..11-5

  12 ROTATING AND MOVING MESHESRotating Reference Frames
  ………………………………………………………………………….12-1

  Models for a single rotating reference frame
  ………………………………………….12-1Useful points
  on single rotating frame problems
  …………………………………….12-1Models for multiple
  rotating reference frames (implicit treatment)
  …………..12-2Useful points on multiple implicit rotating frame
  problems ……………………..12-4Models for multiple rotating
  reference frames (explicit treatment) ……………12-5Useful
  points on multiple explicit rotating frame problems
  ……………………..12-8

  Moving Meshes
  ………………………………………………………………………………………….12-9Basic
  concepts
  ……………………………………………………………………………………12-9Setting
  up models
  ……………………………………………………………………………..12-10Useful
  points
  ……………………………………………………………………………………12-13

• Version 4.02 vii

  Automatic Event Generation for Moving Piston Problems
  ……………………. 12-13Cell-layer Removal/Addition
  ……………………………………………………………………..
  12-14

  Basic concepts
  …………………………………………………………………………………
  12-14Setting up models
  …………………………………………………………………………….
  12-15Useful points
  ……………………………………………………………………………………
  12-18

  Sliding Meshes
  …………………………………………………………………………………………
  12-18Regular sliding interfaces
  ………………………………………………………………….
  12-18

  Cell Attachment and Change of Fluid Type
  …………………………………………………
  12-22Basic concepts
  …………………………………………………………………………………
  12-22Setting up models
  …………………………………………………………………………….
  12-23Useful points
  ……………………………………………………………………………………
  12-27

  Mesh Region Exclusion
  …………………………………………………………………………….
  12-28Basic concepts
  …………………………………………………………………………………
  12-28

  Moving Mesh Pre- and Post-processing
  ………………………………………………………
  12-28Introduction
  …………………………………………………………………………………….
  12-28Action commands
  …………………………………………………………………………….
  12-29Status setting commands
  …………………………………………………………………..
  12-30

  13 OTHER PROBLEM TYPESMulti-component Mixing
  …………………………………………………………………………….
  13-1

  Setting up multi-component models
  ……………………………………………………..
  13-1Useful points on multi-component mixing
  ……………………………………………. 13-3

  Aeroacoustic Analysis
  ………………………………………………………………………………..
  13-3Setting up aeroacoustic models
  ……………………………………………………………
  13-3Useful points on aeroacoustic analyses
  ………………………………………………… 13-4

  Liquid Films
  ………………………………………………………………………………………………
  13-5Setting up liquid film models
  ………………………………………………………………
  13-5Film stripping
  ……………………………………………………………………………………
  13-7

  14 USER PROGRAMMINGIntroduction
  ……………………………………………………………………………………………….
  14-1Subroutine Usage
  ……………………………………………………………………………………….
  14-1

  Useful points
  ……………………………………………………………………………………..
  14-4Description of UFILE Routines
  ……………………………………………………………………
  14-5

  Boundary condition subroutines
  …………………………………………………………..
  14-5Material property subroutines
  ………………………………………………………………
  14-6Turbulence modelling subroutines
  ……………………………………………………….
  14-9Source subroutines
  ……………………………………………………………………………
  14-10Radiation modelling subroutines
  ………………………………………………………..
  14-11Free surface / cavitation subroutines
  …………………………………………………..
  14-11Lagrangian multi-phase subroutines
  ……………………………………………………
  14-12

• viii Version 4.02

  Liquid film subroutines
  ……………………………………………………………………..14-14Eulerian
  multi-phase subroutines
  ………………………………………………………..14-14Chemical
  reaction subroutines
  ……………………………………………………………14-15Rotating
  reference frame subroutines
  ………………………………………………….14-16Moving
  mesh subroutines
  ………………………………………………………………….14-16Miscellaneous
  flow characterisation subroutines
  ………………………………….14-17Solution control
  subroutines
  ………………………………………………………………14-18

  Sample Listing
  ………………………………………………………………………………………….14-19New
  Coding Practices
  ……………………………………………………………………………….14-20User
  Coding in parallel runs
  ……………………………………………………………………….14-22

  15 PROGRAM OUTPUTIntroduction
  ……………………………………………………………………………………………….15-1Permanent
  Output
  ……………………………………………………………………………………….15-1

  Input-data summary
  ……………………………………………………………………………15-1Run-time
  output
  …………………………………………………………………………………15-3

  Printout of Field Values
  ………………………………………………………………………………15-3Optional
  Output
  ………………………………………………………………………………………….15-3Example
  Output
  ………………………………………………………………………………………….15-4

  16 pro-STAR CUSTOMISATIONSet-up Files
  ………………………………………………………………………………………………..16-1Panels
  ………………………………………………………………………………………………………..16-2

  Panel creation
  …………………………………………………………………………………….16-2Panel
  definition files
  …………………………………………………………………………..16-5Panel
  manipulation
  ……………………………………………………………………………..16-6

  Macros
  ………………………………………………………………………………………………………16-6Function
  Keys
  …………………………………………………………………………………………….16-9

  17 OTHER STAR-CD FEATURES AND CONTROLSIntroduction
  ……………………………………………………………………………………………….17-1File
  Handling
  ……………………………………………………………………………………………..17-1

  Naming conventions
  …………………………………………………………………………..17-1Commonly
  used files
  ………………………………………………………………………….17-1File
  relationships
  ………………………………………………………………………………..17-7File
  manipulation
  ……………………………………………………………………………….17-9

  Special pro-STAR Features
  ………………………………………………………………………..17-12pro-STAR
  environment variables
  ……………………………………………………….17-12Resizing
  pro-STAR
  …………………………………………………………………………..17-13Special
  pro-STAR executables
  …………………………………………………………..17-14Use
  of temporary files by pro-STAR
  …………………………………………………..17-14

  The StarWatch Utility
  ……………………………………………………………………………….17-15

• Version 4.02 ix

  Running StarWatch
  ………………………………………………………………………….
  17-15Choosing the monitored values
  ………………………………………………………….
  17-17Controlling STAR
  ……………………………………………………………………………
  17-17Manipulating the StarWatch display
  …………………………………………………..
  17-20Monitoring another job
  ……………………………………………………………………..
  17-21

  Hard Copy Production
  ………………………………………………………………………………
  17-21Neutral plot file production and use
  ……………………………………………………
  17-21Scene file production and use
  …………………………………………………………….
  17-23

  APPENDICESA pro-STAR CONVENTIONS

  Command Input Conventions
  ……………………………………………………………………….
  A-1Help Text / Prompt Conventions
  …………………………………………………………………..
  A-3Control and Function Key Conventions
  …………………………………………………………
  A-4File Name Conventions
  ………………………………………………………………………………..
  A-4

  B FILE TYPES AND THEIR USAGEC PROGRAM UNITSD pro-STAR
  X-RESOURCESE USER INTERFACE TO MESSAGE PASSING ROUTINESF STAR RUN
  OPTIONS

  Usage
  ………………………………………………………………………………………………………….F-1Options
  ……………………………………………………………………………………………………….F-1Parallel
  Options
  ……………………………………………………………………………………………F-3Resource
  Allocation
  ……………………………………………………………………………………..F-6Default
  Options
  ……………………………………………………………………………………………F-7Cluster
  Computing
  ……………………………………………………………………………………….F-8Batch
  Runs Using STAR-NET
  ………………………………………………………………………F-8

  Running under IBM Loadleveler using STAR-NET
  …………………………………F-8Running under LSF using
  STAR-NET
  …………………………………………………..F-9Running
  under OpenPBS using STAR-NET
  …………………………………………F-10Running under
  PBSPro using STAR-NET
  …………………………………………….F-11Running
  under SGE using STAR-NET
  …………………………………………………F-11Running
  under Torque using STAR-NET
  ……………………………………………..F-12

  G BIBLIOGRAPHY

  INDEX

  INDEX OF COMMANDS

• Version 4.02 1

  OVERVIEWPurpose

  The Methodology volume presents the mathematical modelling
  practices embodiedin the STAR-CD system and the numerical solution
  procedures employed. In thisvolume, the focus is on the structure
  of the system itself and how to use it. Thispresentation assumes
  that the reader is familiar with the background informationprovided
  in the Methodology volume.

  ContentsChapter 1 introduces some of the fundamental principles
  of computationalcontinuum mechanics, including an outline of the
  basic steps involved in setting upand using a successful computer
  model. The important factors to consider at eachstep are mostly
  explained independently of the computer system used to perform
  theanalysis. However, reference is also made to the particular
  capabilities of theSTAR-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 detailedinstructions on how to use some of
  the basic code facilities, e.g. boundary conditionspecification,
  material property definition, etc., and an overview of the GUI
  panelsappropriate to each of them. The description covers all
  facilities (other than meshgeneration) that might be employed for
  modelling most common continuummechanics problems. Mesh generation
  itself is covered in a separate volume, theMeshing User Guide.

  Chapters 2 to 5 should be read at least once to gain an
  understanding of thegeneral housekeeping principles of pro-STAR and
  to help with any problemsarising from routine operations. It is
  recommended that users refer to theappropriate chapter repeatedly
  when setting up a model for general guidance and anoverview of the
  relevant GUI panels.

  Chapters 6 to 13 describe additional STAR-CD capabilities
  relevant to modelsof a more specialised nature, i.e. rotating
  systems, combustion processes,buoyancy-driven flows, etc. Users
  interested in a particular topic should consult theappropriate
  section for a summary of commands or options specially designed
  forthat purpose, plus hints and tips on performing a successful
  simulation.

  Chapter 14 outlines the user programmability features available
  and provides anexample FORTRAN subroutine listing implementing
  these features. All suchsubroutines are readily available for use
  and can be easily adapted to suit themodel’s requirements.

  Chapter 15 presents the printable output produced by the code
  which provides,among other things, a summary of the problem
  specification and monitoringinformation generated during the
  calculation.

  Chapter 16 explains how pro-STAR can be customised, in terms of
  user-definedpanels, macros and keyboard function keys, to meet a
  users individualrequirements.

  Finally, Chapter 17 covers some of the less commonly used
  features ofSTAR-CD, including the interaction between STAR and
  pro-STAR and howvarious system files are used.

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Introduction

  Version 4.02 1-1

  Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLESIntroduction

  The aim of this section is to introduce the most important
  issues involved in settingup and solving a continuum mechanics
  problem using a computational continuummechanics code. Although the
  discussion applies in principle to any such code,reference is made
  where appropriate to the particular capabilities of the
  STAR-CDsystem. It is also assumed that the reader is familiar with
  the material presented inthe Methodology volume.

  The process of computational mechanics simulation does not
  usually start withthe direct use of such a code. It is indeed
  important to recognise that STAR-CD, orany other CFD, CAD or CAE
  system, should be treated as a tool to assist theengineer in
  understanding physical phenomena.

  The success or failure of a continuum mechanics simulation
  depends not only onthe 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 bephysically 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 theabove tasks have been completed.

  The Basic Modelling ProcessThe modelling process itself can be
  divided into four major phases, as follows:Phase 1 Working out a
  modelling strategyThis requires a precise definition of the
  physical systems geometry, materialproperties and flow/deformation
  conditions based on the best availableunderstanding of the relevant
  physics. The necessary tasks include:

  Planning the computational mesh (e.g. number of cells, size and
  distributionof 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.).

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Spatial description and volume discretisation

  1-2 Version 4.02

  The user also has to balance the requirement of physical
  fidelity and numericalaccuracy against the simulation cost and
  computational capabilities of his system.His modelling strategy
  will therefore incorporate some trade-off between these
  twofactors.

  This initial phase of modelling is particularly important for
  the smooth andefficient progress of the computational
  simulation.Phase 2 Setting up the model using pro-STARThe main
  tasks involved at this phase are:

  Creating a computational mesh to represent the solution domain
  (i.e. themodel geometry).

  Specifying the physical properties of the fluids and/or solids
  present in thesimulation and, where relevant, the turbulence
  model(s), body forces, etc.

  Setting the solution parameters (e.g. solution variable
  selection, relaxationcoefficients, etc.) and output data
  formats.

  Specifying the location and definition of boundaries and, for
  unsteadyproblems, further definition of transient boundary
  conditions and time steps.

  Writing appropriate data files as input to the analytical run of
  the followingphase.

  Phase 3 Performing the analysis using STARThis phase consists
  of:

  Reading input data created by pro-STAR and, if dealing with a
  restart run, theresults of a previous run.

  Judging the progress of the run by analysing various monitoring
  data andsolution statistics provided by STAR.

  Phase 4 Post-processing the results using pro-STARThis involves
  the display and manipulation of output data created by STAR
  usingthe appropriate pro-STAR facilities.

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

  Spatial description and volume discretisationOne of the basic
  steps in preparing a STAR-CD model is to describe the geometryof
  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.

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Spatial description and volume discretisation

  Version 4.02 1-3

  Figure 1-1 Example of solution domain subdivision into cells

  This process is called volume discretisation and is an essential
  part of solving theabove equations numerically. In STAR-CD both
  components of the spatialdescription are performed as part of the
  same operation, setting up the finite-volumemesh, but separate
  considerations apply to each of them.

  Solution domain definitionThrough its internal design and
  construction, STAR-CD permits a very general andflexible 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 thesame model, as in problems involving
  fluid-solid heat transfer. The users first taskis therefore to
  decide which parts of the physical system being modelled need to
  beincluded in the solution domain and whether each part is occupied
  by a fluid, solidor porous medium.

  Whatever its composition, the fundamental requirement is that
  the solutiondomain is bounded. This means that the user has to
  examine his systems geometrycarefully and decide exactly where the
  enclosing boundaries lie. The boundaries canbe one of four
  kinds:

  1. Physical boundaries walls or solid obstacles of some
  description thatserve to physically confine a fluid flow

  2. Symmetry boundaries axes or planes beyond which the problem
  solutionbecomes a mirror image of itself

  3. Cyclic boundaries surfaces beyond which the problem solution
  repeatsitself, in a cyclic or anticyclic fashion

  The purpose of symmetry and cyclic boundaries is to limit the
  size of thedomain, and hence the computer requirements, by
  excluding regions wherethe solution is essentially known. This in
  turn allows one to model theproblem in greater detail than would
  have been the case otherwise.

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Spatial description and volume discretisation

  1-4 Version 4.02

  4. Notional boundaries these are non-physical surfaces that
  serve toclose-off the solution domain in regions not covered by the
  other two typesof boundary. Their location is entirely up to the
  users discretion but, ingeneral, 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 Boundarydescription on page 1-10.

  Mesh definitionCreation of a lattice of finite-volume cells to
  represent the solution domain isnormally the most time-consuming
  task in setting up a STAR-CD model. Thisprocess is greatly
  facilitated by STAR-CD because of its ability to generate cells
  ofan arbitrary, polyhedral shape.

  In creating a finite-volume mesh, the user should aim to
  represent accurately thefollowing two entities:

  1. The overall external geometry of the solution domain, by
  specifying anappropriate size and shape for near-boundary cells.
  The latters external faces,taken together, should make up a surface
  that adequately represents the shapeof the solution domain
  boundaries. Small inaccuracies may occur because allboundary cell
  faces (including rectangular faces) are composed of
  triangularfacets, as shown in Figure 1-2. These errors diminish as
  the mesh is refined.

  Figure 1-2 Boundary representation by triangular facets

  2. The internal characteristics of the flow/deformation regime.
  This is achievedby careful control of mesh spacing within the
  solution domain interior so that

  triangular facet

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Spatial description and volume discretisation

  Version 4.02 1-5

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

  Mesh spacing considerationsThe 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 timestep.

  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 inSTAR-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 andsolid mechanics
  to produce the right kind of mesh arrangement.

  Mesh distortionMesh distortion is measured in terms of three
  factors aspect ratio, internal angleand warp angle illustrated in
  Figure 1-3.

  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

  a

  b

  b/a = aspect ratio

  = internal angle

  = warp angle

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Spatial description and volume discretisation

  1-6 Version 4.02

  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 thesefactors 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 dependon local flow conditions.
  However, the following values serve as a useful guideline:

  pro-STAR can calculate these quantities and identify cells
  having out-of-boundsvalues, as discussed in Chapter 3, Mesh and
  Geometry Checking of the MeshingUser Guide.

  What is really important in this respect is the combined effect
  of the variouskinds of mesh distortion. If all three are
  simultaneously present in a single cell, thelimits given above
  might not be stringent enough. On the other hand, the effects
  ofdistortion also depend on the nature of the local flow. Thus, the
  above limits maybe 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 andshould be minimised whenever practicable.

  Mesh distribution and densityNumerical 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
  highmesh density implies a large number of mesh storage locations,
  with associated highcomputing cost. It is therefore advisable,
  wherever possible, to

  ensure that the mesh density is high only where needed, i.e. in
  regions of steepgradients of the flow variables, and low
  elsewhere;

  avoid rapid changes in cell dimensions in the direction of steep
  gradients inthe flow variables.

  The flexibility afforded by STAR-CDs unstructured polyhedral
  meshes facilitatessuch selective refinement. An illustration of
  some of the numerous cell shapes thatmay be employed is given in
  Figure 2-43 and Figure 2-44 of the Meshing UserGuide.

  Aspect Ratio 10Internal angle 45Warp angle 45

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Spatial description and volume discretisation

  Version 4.02 1-7

  Of course, it is not always possible to ascertain a priori what
  the flow structurewill be. However, the need for higher mesh
  density can usually be anticipated inregions 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 wallsAs discussed in Chapter 6, Wall
  Boundary Conditions of the Methodologyvolume, wall functions are an
  economic way of representing turbulent boundarylayers (hydrodynamic
  and thermal) in turbulent flow calculations. These
  functionseffectively allow the boundary layer to be bridged by a
  single cell, as shown inFigure 1-4(a).

  Figure 1-4 Near-wall mesh distribution

  The location y of the cell centroids in the near-wall layer, and
  hence the thicknessof that layer, is usually determined by
  reference to the dimensionless normaldistance from the wall. For
  the wall function to be effective, this distance mustbe

  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, should lie in the approximate range 30 to 150. Note
  that the aboveconsiderations apply to Reynolds Stress models as
  well as several classes of eddyviscosity model (see Chapter 3,
  Turbulence Modelling).

  Alternative treatments that do not require the use of wall
  functions are alsoavailable. These are:

  (b) Two-layer or Low Re models

  Outerregion

  Innerregiony

  (a) Wall function model

  y+

  y+

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Problem characterisation and material property definition

  1-8 Version 4.02

  1. Two-layer turbulence models, whereby wall functions are
  replaced by aone-equation k-l model or a zero-equation
  mixing-length model

  2. Low Reynolds number models (including the V2F model), where
  viscouseffects are incorporated in the k and transport
  equations

  For the above two types of model, the solution domain should be
  divided into tworegions 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 Methodologyvolume (Chapter 6, Two-layer models), the inner
  region should contain at least15 mesh layers and encompass that
  part of the boundary layer influenced by viscouseffects.

  A more recent development, called the hybrid wall function is
  also available thatextends the low-Reynolds number formulation of
  most turbulence models. Thismay be used to capture boundary layer
  properties more accurately in cases wherethe near-wall cell size is
  not adapted for the low-Reynolds number treatment andthus achieve
  independent solutions.

  Moving mesh featuresSTAR-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 theobvious one: avoid creating excessive
  distortion when redistributing the mesh. Thiscaution also applies
  to the use of the other two features, but they have additionalrules
  and guidelines attached to them. These are summarised in the
  Methodologyvolume, Chapter 15 (Internal Sliding Mesh on page 15-5
  and Cell LayerRemoval and Addition on page 15-7). Additional
  guidelines also appear in thisvolume, Cell-layer Removal/Addition
  on page 12-14 and Sliding Meshes onpage 12-18; hence they are not
  repeated here.

  Problem characterisation and material property definitionCorrect
  definition of the physical conditions and the properties of the
  materialsinvolved is a prerequisite to obtaining the right solution
  to a problem, or indeed toobtaining any solution at all. It is also
  essential for determining whether the problemcan be modelled with
  STAR-CD. The user must therefore ensure that the problemis 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

  y+

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Problem characterisation and material property definition

  Version 4.02 1-9

  Nature of the flowIt is very important to understand the nature
  of the flow being analysed in order toselect 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
  unsteadyproblem, 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
  behaviouraccurately.

  Physical propertiesThe specification of physical properties,
  such as density, molecular viscosity,thermal conductivity, etc.
  depends on the nature of the fluids or solids involved andthe
  circumstances of use. For example, STAR-CD contains several
  built-inequations of state from which density can be calculated as
  a function of one or moreof the following field variables:

  Pressure Temperature Fluid composition

  In all cases where complex calculations are used to evaluate a
  material property, thefollowing measures are recommended:

  The relevant field variables must be assigned plausible initial
  and boundaryvalues.

  Where necessary, properties should be solved for together with
  the fieldvariables as part of the overall solution.

  In the case of strong dependencies between properties and field
  variables, theuser should consider under-relaxation of the property
  value calculations, inthe manner described in the Methodology
  volume (Chapter 7, Scalartransport equations).

  When required, STAR-CDs facility for alternative,
  user-programmablefunctions may be used.

  Force fields and energy sourcesAs 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 relativeto the viscous (or turbulent) stresses,
  the flow may become physically unstable. Inthese circumstances it
  is advisable to switch to the transient solution mode.

  It is also possible to insert additional, external force fields
  and energy sourcesvia the user programming facilities of STAR-CD.
  In such cases, it is important tounderstand the physical
  implications and avoid specifying conditions that lead to

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Boundary description

  1-10 Version 4.02

  physical or numerical instability. Examples of such conditions
  are:

  Thermal energy sources that increase linearly with temperature.
  These cangive rise to physical instability called thermal
  runaway.

  Setting the coefficient in the permeability function to avery
  small or zero value. If the local fluid velocity also becomes very
  small,the result may be numerical instability whereby small
  pressure perturbationsproduce a large change in velocities.

  Initial conditionsThe term initial conditions refers to values
  assigned to the dependent variables atall mesh points before the
  start of the calculations. Their implication depends on thetype of
  problem being considered:

  In unsteady applications, this information has a clear physical
  significanceand will affect the course of the solution. Due care
  must therefore be taken inproviding it. It sometimes happens that
  the effects of initial conditions areconfined to a start-up phase
  that is not of interest (as in, for example, flowsthat are
  temporally periodic). However, it is still advisable to take
  someprecautions in specifying initial conditions for reasons
  explained below.

  In calculating steady state problems by iterative means, the
  initial conditionswill usually have no influence on the final
  solution (apart from rare occasionswhen the solution is
  multi-valued), but may well determine the success andspeed of
  achieving it.

  Poor initial field specifications or, for transient problems,
  abrupt changes inboundary conditions put severe demands on the
  numerical algorithm whensubstituted into the finite-volume
  equations. As a consequence, the followingspecial 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
  Numericalsolution control on page 1-13. In either case, increased
  computing times can be anundesirable side effect.

  Boundary descriptionAs stated in Spatial description and volume
  discretisation on page 1-2, boundaryidentification and description
  are intimately connected with the generation of thefinite-volume
  mesh, since the boundary topography is defined by the outermost
  cellfaces. Furthermore, correct specification of the boundary
  conditions is often themain area of difficulty in setting up a
  model. Problems often arise in the followingareas:

  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

  i K i i v i+=

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Boundary description

  Version 4.02 1-11

  first instance.

  Boundary locationDifficulties in specifying boundary location
  normally arise where the flowconditions are incompletely known, for
  example at outlets. The recommendedsolutions, 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 onhow the flow is partitioned between the outlets.

  (b) Prescribed flow split outlets coexisting with prescribed
  mass outflowboundaries in the same domain.

  (c) A combination of prescribed pressure and flow-split outlet
  conditions.

  Boundary conditionsAnother source of potential difficulty is in
  boundary value specification whereverknown 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 boundaryvalues, including the turbulence transport equations
  when they are invoked

  Inappropriate setting of boundary values leads to erroneous
  results and, inextreme cases, to numerical instability

  The following recommendations can be given regarding each
  different type ofboundary:

  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
  massimbalances of up to 5%, correcting them by adjusting the
  outflows. Anerror 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

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Boundary description

  1-12 Version 4.02

  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 ateach location.

  (c) The inapplicability of prescribed split outlets to problems
  where theinflows are not fixed, e.g.

  i) in combination with pressure boundary conditions, orii) 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 theturbulence parameters indirectly, via the turbulence
  intensity and lengthscale or by extrapolating them from values in
  the interior of the solutiondomain.

  4. Stagnation conditions It is recommended to use this condition
  forboundaries lying within large reservoirs where properties are
  not significantlyaffected by flow conditions in the solution
  domain.

  5. Non-reflecting pressure and stagnation conditions A
  specialformulation of the standard pressure and stagnation
  conditions, developed tofacilitate analysis of steady-state
  turbomachinery applications

  6. Cyclic boundaries These always occur in pairs. The main
  points of adviceare:

  (a) Impose this condition only in appropriate
  circumstances.Two-dimensional axisymmetric flows with swirl is a
  good example of anappropriate application.

  (b) For axisymmetric flows, make use of the CD/UD blending
  scheme toapply the maximum level of central differencing in the
  tangentialdirection (the default blending factor is 0.95; see also
  on-line Help topicMiscellaneous Controls in STAR GUIde).

  7. Planes of symmetry It is recommended to use this condition
  fortwo-dimensional axisymmetric flows without swirl

  8. Free-stream transmissive boundaries Used only for modelling
  supersonicfree streams

  9. Transient wave transmissive boundaries Used only in problems
  involvingtransient compressible flows

  10. Riemann boundaries This condition is based on the theory of
  Riemanninvariants and its application allows pressure waves to
  leave the solutiondomain without reflection

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Numerical solution control

  Version 4.02 1-13

  Numerical solution controlProper control of the numerical
  solution process applied to the transport equationsis highly
  important, both for acceptable computational efficiency and,
  sometimes,in order to achieve a solution at all. By necessity, the
  means of controlling theprocess depend heavily on the particular
  numerical techniques employed so nouniversal guidelines can be
  given. Thus, the recommended settings vary with theparticular
  algorithm selected and the circumstances of application.

  Selection of solution procedureThe basic selection should be
  based on a correct assessment of the nature of theproblem and will
  be either

  a transient calculation, starting from well-defined initial and
  boundaryconditions and proceeding to a new state in a series of
  discrete time steps; or

  a steady-state calculation, where an unchanging flow/deformation
  patternunder a given set of boundary conditions is arrived at
  through a number ofnumerical iterations.

  PISO and SIMPLE are the two alternative solution procedures
  available inSTAR-CD. PISO is the default for unsteady calculations
  and is sometimes preferredfor steady-state ones, in cases involving
  strong coupling between dependentvariables such as buoyancy driven
  flows. SIMPLE is the default algorithm forsteady-state solutions
  and works well in most cases.

  SIMPLE is also used for transient calculations in the case of
  free surface andcavitating 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 usuallycheaper
  than outer iterations on all variables within a time step of the
  transientSIMPLE algorithm. However, there are situations in which
  PISO would requiremany correctors or even fail to converge unless
  the time step is reduced, whereasSIMPLE may allow larger time steps
  with a moderate number of outer iterations pertime step. This is
  the case when the flow changes very little but certain
  slowtransients are present in the behaviour of scalar variables
  (e.g. slow heating up ofsolid structures in the case of solid-fluid
  heat transfer problems, deposition ofchemical 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 asteady-state solution or when iterative convergence is
  difficult to achieve, it is betterto perform the calculations using
  the transient option.

  Transient flow calculations with PISOAs stated in The PISO
  algorithm on page 7-2 of the Methodology volume, PISOperforms at
  each time (or iteration) step, a predictor, followed by a number
  ofcorrectors, during which linear equation sets are solved
  iteratively for each maindependent variable. The decisions on the
  number of correctors and inner iterations(hereafter referred to as
  sweeps, to avoid confusion with outer iterationsperformed as part
  of the steady-state solution mode) are made internally on the
  basisof the splitting error and inner residual levels,
  respectively, according to prescribedtolerances and upper limits.
  The default values for the solver tolerances and

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Numerical solution control

  1-14 Version 4.02

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

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

  (1-1)

  Here, U and are a characteristic velocity and diffusivity,
  respectively, and isa mean mesh dimension. Typically, it is
  possible to operate with andstill obtain reasonable temporal
  accuracy. Values significantly above this may leadto errors and
  numerical instability, whereas smaller values will lead to
  increasedcomputing 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 occurduring the start-up phase but
  is nevertheless acceptable if, later on, the warningseither cease
  entirely or only appear occasionally, and the predictions
  lookreasonable. 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
  improvesmatters, then the cause may simply be an excessively large
  .

  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, twofoldchanges are appropriate.

  Pressure under-relaxation a value of 0.8 for pressure
  correctionunder-relaxation, using PISO, may be helpful for some
  difficult cases (e.g. forsevere mesh distortion or flows with Mach
  numbers approaching 1).

  Corrector step tolerance this may be set to a lower value but
  consult

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

  ParameterVariable

  Velocity Pressure Turbulence Enthalpy Mass fraction

  Solvertolerance 0.01 0.001 0.01 0.01 0.01

  Sweep limit 100 1000 100 100 100

  Pressure under-relaxation factor = 1.0

  Corrector limit = 20

  Corrector step tolerance = 0.25

  t

  tc

  tc minLU——

  L2————,

  =

  Lt 50 tc

  t

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Numerical solution control

  Version 4.02 1-15

  CD adapco first.

  Steady-state flow calculations with PISOWhen PISO operates in
  this mode, the inner residual tolerances are decreased
  andunder-relaxation is introduced on all variables, apart from
  pressure, temperature andmass fraction. However, the last two
  variables may need to be under-relaxed forbuoyancy driven problems.
  The standard, default values for these parameters andthe sweep
  limits, which are unchanged from the transient mode, are given in
  Table1-2.

  .

  These settings should, all being well, result in near-monotonic
  decrease in theglobal residuals during the course of the
  calculations, depending on mesh densityand other factors. If,
  thereafter, one or more of the global residuals do not fall,then
  remedial measures will be necessary. In some instances, the
  offendingvariable(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 ofbetween 0.05 and 0.10 and should be applied to the
  velocities if themomentum and/or mass residuals are at fault.

  Decrease in solver tolerances as in the transient case, this may
  provebeneficial, especially in respect of the pressure tolerance
  and its importance tothe flow solution. A twofold reduction should
  indicate whether this measurewill work.

  Increase in sweep limits if warning messages about the limits
  beingreached appear and are not suppressed by measures 1 and 2,
  then it may beworthwhile increasing the limit(s) on the offending
  variables.

  Under-relaxation of density and effective viscosity use of this
  method fordensity 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-Newtonianfluid flow and can be similarly damped by this
  device.

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

  ParameterVariable

  Velocity Pressure Turbulence Enthalpy Mass fraction

  Solvertolerance 0.1 0.05 0.1 0.1 0.1

  Sweep limit 100 1000 100 100 100

  Relaxationfactor 0.7 1.0 0.7 0.95 1.0

  Corrector limit = 20

  Corrector step tolerance = 0.25

  R

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Numerical solution control

  1-16 Version 4.02

  Steady-state flow calculations with SIMPLEAs noted previously,
  the control parameters available for SIMPLE are similar tothose for
  PISO, except that, in the case of the former, a single corrector
  stage isalways used and pressure is under-relaxed. The standard
  (default) settings are givenin Table 1-3.

  .

  In the event of failure to obtain solutions with the standard
  values, then the measuresto be taken are essentially the same as
  those for iterative PISO, given in the previoussection. However,
  here, reduction of the pressure relaxation factor is an
  additionaldevice for overcoming convergence problems. The problems
  usually arise eitherfrom a highly distorted mesh, or from highly
  complex physics (many variablesaffecting each other). If the grid
  is distorted, one should reduce the relaxation factorfor pressure
  from the beginning of the run (e.g. to 0.1). If convergence
  problems arestill encountered, a substantial reduction of the
  under-relaxation factor for velocitiesand turbulence model
  variables should be tried (e.g. to 0.5). If this does not help,
  theproblem may lie in severe mesh defects or errors in the set-up.
  Further reduction ofunder-relaxation factors may be tried if the
  grid is severely distorted and cannot beimproved; otherwise,
  improving the mesh quality can be of much greater help.

  Note that the pressure under-relaxation factor needs to be
  adjusted within thelimits of some range to make the iteration
  process converge, where the number ofiterations required to reach
  such convergence is mainly dictated by thecorresponding factors for
  velocities (and for scalar variables when strongly coupledto the
  flow). In the case of well-behaved flows and reasonable meshes,
  therelaxation factor for pressure can be selected as (1.0 —
  relaxation factor forvelocities), e.g. 0.2 for pressure and 0.8 for
  velocities. Usually, for a given velocityrelaxation factor, the one
  for pressure can be varied within some range withoutaffecting the
  total number of iterations and computing time, but outside this
  rangethe 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 0.05and 0.8 if the velocity
  factor is low, say around 0.5). On the other hand, this
  rangebecomes narrower when the mesh is distorted.

  The limit to which the velocity relaxation factor can be
  increased is bothproblem- and mesh-dependent. When many similar
  problems need to be solved, itis worth trying to work near the
  optimum as this may save a lot of computing time.

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

  ParameterVariable

  Velocity Pressure Turbulence Enthalpy Mass fraction

  Solvertolerance 0.1 0.05 0.1 0.1 0.1

  Sweep limit 100 1000 100 100 100

  Relaxationfactor 0.7 0.3 0.7 0.95 1.0

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Numerical solution control

  Version 4.02 1-17

  On the other hand, for an one-off analysis, it may be more
  efficient to use aconservative setting.

  Note that under some conditions, such as those in Tutorial 13.1,
  a steady-statesolution cannot be achieved due to the inherent
  unsteady character of the flow. Thisis often the case when the
  problem geometry possesses some form of symmetry butthe Reynolds
  (or another equivalent) number is high and recirculation zones
  arepresent. In this case the residuals stop falling at some level
  and then continue tooscillate. The solution at that stage may be
  far from a valid solution of thegoverning equations and should not
  be interpreted as such unless the residual levelis 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 asteady-state solution, while a less diffusive
  turbulence model (such as ReynoldsStress and non-linear
  eddy-viscosity models) combined with a higher-orderdifferencing
  scheme (such as central differencing) may not. In such cases,
  atransient simulation should be performed; the unsteady solution
  may oscillatearound a mean steady state, in which case the
  quantities of interest (drag, lift, heattransfer coefficient,
  pressure drop, etc.) can be averaged over several
  oscillationperiods.

  Transient flow calculations with SIMPLEThe use of this algorithm
  in transient calculations essentially consists of repeatingthe
  steady-state SIMPLE calculations for each prescribed time step. The
  defaultcontrol parameter settings are therefore as summarised in
  Table 1-4.

  .

  The main difference compared to the PISO algorithm lies in the
  fact that alllinearizations and deferred correctors are updated
  within the outer iterations, byrecalculating the coefficient matrix
  and source term. For this reason, solvertolerances do not need to
  be as tight as for PISO; they are actually identical to thoseused
  for steady-state computations. However, since the discretization of
  thetransient term enlarges the central coefficient of the matrix in
  the same way asunder-relaxation does, the relaxation factors for
  velocities and scalar variables canbe increased (the smaller the
  time step, the larger the values that can be used forrelaxation
  factors 0.95 or even more).

  The convergence criterion for outer iterations within each time
  step is by default

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

  ParameterVariable

  Velocity Pressure Turbulence Enthalpy Mass fraction

  Solvertolerance 0.1 0.05 0.1 0.1 0.1

  Sweep limit 100 1000 100 100 100

  Relaxationfactor 0.9 0.3 0.7 1.0 1.0

  Outer iteration limit = 5

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Numerical solution control

  1-18 Version 4.02

  the same as for steady-state flows. However, the number of outer
  iterations is alsoset to a default limit of 10; if substantially
  more iterations are needed to satisfy theconvergence criterion,
  this is a sign that the time step is too large. In such a case,
  itis better to reduce the time step rather than allow more outer
  iterations for a largertime 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, onemay
  increase the time step; experience shows that optimum efficiency
  and accuracyare achieved if 5 to 10 outer iterations per time step
  are performed.

  Note also that the reported mass residuals are computed before
  solving thepressure-correction equation; after this equation is
  solved and mass fluxes arecorrected, the mass residuals are more
  than an order of magnitude lower. For thisreason, one can accept
  mass residuals being somewhat higher than the convergencecriterion
  when the limiting number of outer iterations is reached, provided
  that theresiduals of all other equations have satisfied the
  criterion. In some cases, anincrease in the under-relaxation factor
  for pressure (up to 0.8) can lead to a fasterreduction of mass
  residuals. All these considerations are of course problem-dependent
  and if several simulations over a longer period need to be
  performed, itmay 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 atransient simulation than those at later stages.
  This is the case, for example, whenstarting with a fluid at rest
  and imposing a full-flow rate at the inlet, or full speed
  ofrotation (in the absence of a better initial condition). This is
  equivalent to a suddenchange of boundary conditions at a later
  time, which would also require that thetime step be reduced.
  Another possibility of avoiding problems with abrupt startsfrom
  rest is to ramp the boundary conditions (e.g. a linear increase of
  velocity fromzero to full speed over some period of time).

  The transient SIMPLE algorithm allows you to select either the
  defaultfully-implicit Euler scheme or the three-time-level scheme
  for temporaldiscretisation, described in Chapter 4, Temporal
  Discretisation of theMethodology volume. The latter scheme is
  second-order accurate but is currentlyapplied only to the momentum
  and continuity equations. It should be chosen whentemporal
  variation of the velocity field is essential, e.g. in the case of a
  DES/LEStype 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 qualityetc.), transient SIMPLE may
  allow the use of larger time steps than PISO withoutloss of
  accuracy.

  Effect of round-off errorsEfforts have been made to minimise the
  susceptibility of STAR-CD to the effectsof machine round-off
  errors, but problems can sometimes arise when operating insingle
  precision on 32-bit machines. They usually manifest themselves as
  failure ofthe iterative solvers to converge or, in extreme cases,
  in divergence leading tomachine overflow.

  If difficulties are encountered with problems of this kind, then
  it is clearlyadvisable to switch to double precision calculations.
  Instructions on how to do thisare provided in the Installation
  Manual. As a general rule, however, you should tryto avoid
  generating very small values for cell volumes and cell face areas
  byworking with sensible length units. Alternatively, you could
  re-specify your

• Chapter 1 COMPUTATIONAL ANALYSIS PRINCIPLES

  Monitoring the calculations

  Version 4.02 1-19

  problem geometry units while preserving relevant non-dimensional
  quantities suchas Re and Gr.

  Choice of the linear equation solverSTAR-CD offers two types of
  preconditioning of its conjugate gradient linearequations solvers:
  one which vectorises fully, and the other, which is
  numericallysuperior to the first one but vectorises only partially.
  Therefore, the first one (calledvector solver) is recommended when
  the code is run on vector machines (such asFujitsu and Hitachi
  computers), and the second one (called scalar solver) isrecommended
  if the code is run on scalar machines (such as workstations).

  Monitoring the calculationsChapter 5 and the section on
  Permanent Output on page 15-1 give details of theinformation
  extracted from the calculations at each iteration or time step and
  usedfor 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-statecalculations. Ideally, it should be in a sensitive
  region of the flow where theapproach 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 andother
  criteria for choice of location may apply. For example, a location
  maybe chosen so as to confirm an expected periodic behaviour in the
  flowvariables.

  The normalised global residuals for all equations solved. Apart
  fromturbulence dissipation rate residuals (see Chapter 7,
  Completion tests in theMethodology volume), these are used to judge
  the progress and completion ofiterative calculations for steady and
  pseudo-transient solutions. In the earlystages of a calculation,
  the non-linearities and interdependencies of theequations may
  result in non-monotonic decrease of the residuals. If
  theseoscillations persist after, say, 50 iterations, this may be
  indicative of problems.

  Remember that reduction of the normalised residuals to the
  prescribed tolerance ()is a necessary but not sufficient condition
  for convergence, for two reasons:

  1. The normalisation practices used (see Chapter 7, Completion
  tests in theMethodology volume) may not be appropriate for the
  application.

  2. It is also necessary that the features of interest in the
  solution should havestabilised to an acceptable degree.

  If doubts exist in either respect, it is advisable to reduce the
  tolerance and continuethe calculations.

  It follows from the above discussion that strong reliance is
  placed on the globalresiduals to judge the progress and completion
  of iterative calculations of steadyflows. These quantities provide
  a direct measure of the degree of convergence of theindividual
  equation sets and are therefore useful both for termination tests
  and foridentifying problem areas when convergence is not being
  achieved.

  R

• COMPUTATIONAL ANALYSIS PRINCIPLES Chapter 1

  Model evaluation

  1-20 Version 4.02

  Model evaluationChecking the modelSTAR-CD offers a variety of
  tools to help assess the accuracy and effectiveness ofall aspects
  of the model building process. In performing the modelling
  stagesdiscussed previously, the user should therefore take
  advantage of these facilities andcheck that:

  1. The mesh geometry agrees with what it is

Description

From the STAR-CCM+ Home Page: Much more than just a CFD solver, STAR-CCM+ is an entire engineering process for solving problems involving flow (of fluids or solids), heat transfer and stress.

Version

• 15.06.007

• Authorized students in Mechanical Engineering
• Members of the Society of Automotive Engineers at USF

Platforms

• CIRCE cluster
• RRA cluster
• SC cluster

Modules

STAR-CCM+ requires the following module file to run:

• apps/star-ccm/15.06.007

• See Modules for more information.

Running STAR-CCM+ on CIRCE/SC

The STAR-CCM+ user guide is essential to understanding the application and making the most of it. The guide and this page should help you to get started with your simulations. Please refer to the Documentation section for a link to the guide.

• Note on CIRCE: Make sure to run your jobs from your $WORK directory!
• Note: Scripts are provided as examples only. Your SLURM executables, tools, and options may vary from the example below. For help on submitting jobs to the queue, see our SLURM User’s Guide.

If you need more control over your workflow, keep reading below.

If you will be using a Power-on-Demand key, please contact Research Computer for additional instructions.

Interactive Execution via CIRCE/SC Desktop Environment

Establishing a GUI connection to CIRCE/SC

To use STAR-CCM+, you will need to connect to CIRCE/SC with GUI redirection, either using:

• CIRCE/SC Desktop Environment
• SSH with X11 redirection
  • If connecting from OSX or Linux via SSH, please ensure that you use one of the following commands to properly redirect X11:

    • [user@localhost ~]$ ssh -X circe.rc.usf.edu

      or

    • [user@localhost ~]$ ssh -X sc.rc.usf.edu

Once connected to CIRCE/SC, you can open STAR-CCM+ using the steps below:

[user@login0 ~]$ module add apps/star-ccm/15.06.007
[user@login0 ~]$ starccm+

How to Submit Jobs

Provided are batch scripts for running STAR-CCM+ as a single processor and distributed parallel job. Existing STAR-CCM+ SIM files will work in parallel mode with no modification, but only larger models and geometries will see any performance benefit. These scripts can be copied into your work directory (the folder with your input files and database files) so that you can submit batch processes to the queue.

If, for example, you have STAR-CCM+ simulation file called “test.sim”, you would set up your serial/distributed parallel submit scripts like this:

• The scripts below (for testing, name it “starccm-serial-test.sh” or name it “starccm-parallel-test.sh”, respectively) can be copied into your job directory (the folder with your input files) and modified so that you can submit batch processes to the queue.

Serial Submit Script

#!/bin/bash
#
#SBATCH --comment=starccm-serial-test
#SBATCH --ntasks=1
#SBATCH --job-name=starccm-serial-test
#SBATCH --output=output.%j.starccm-serial-test
#SBATCH --time=08:00:00

#### SLURM 1 processor STAR-CCM+ test to run for 8 hours.

module purge
module add apps/star-ccm/15.06.007

export PATH=$TMPDIR:$PATH

starccm+ -pio -batch test.sim

Distributed Parallel Submit script

#!/bin/bash
#
#SBATCH --comment=starccm-parallel-test
#SBATCH --ntasks=32
#SBATCH --job-name=starccm-parallel-test
#SBATCH --output=output.%j.starccm-parallel-test
#SBATCH --time=08:00:00

#### SLURM 32 processor STAR-CCM+ test to run for 8 hours.

module purge
module add apps/star-ccm/15.06.007

export PATH=$TMPDIR:$PATH

# Create our hosts file ala slurm
NODEFILE="$(pwd)/slurmhosts.$SLURM_JOB_ID.txt"
srun hostname -s &> $NODEFILE

starccm+ -np $SLURM_NTASKS -machinefile $NODEFILE -mpi intel -rsh ssh -pio -batch "test.sim"

 
Next, you can change to your job’s directory, and run the sbatch command to submit the job:

[user@login0 ~]$ cd my/jobdir
[user@login0 jobdir]$ sbatch ./starccm-serial-test.sh

• You can view the status of your job with the “squeue -u <username>” command

Home Page, User Guides, and Manuals

• STAR-CCM+ Home Page
  • http://www.cd-adapco.com/products/star-ccm
• Local Documentation
  • /apps/star-ccm/15.06.007/STAR-CCM+15.06.007/doc/

More Job Information

See the following for more detailed job submission information:

• SLURM User’s Guide
• Scheduling and Dispatch Policies
• Advanced Submit Techniques

Reporting Bugs

Report bugs with STAR-CCM+ to the IT Help Desk: rc-help@usf.edu