Impulse 4.0 Users Guide SI PDF

TM
AFT Impulse
User’s Guide
AFT Impulse version 4.0

Waterhammer Modeling in Piping Systems
Applied Flow Technology
8/10/2007 4:10:28 PM Page 1 c:\aft manuals\i4\impulse4 manual

(si) g.doc
CAUTION!
AFT Impulse is a sophisticated waterhammer and surge transient
modeling program designed for qualified engineers with experience in
waterhammer analysis and should not be used by untrained individuals.
AFT Impulse is intended solely as an aide for pipe flow analysis
engineers and not as a replacement for other design and analysis
methods, including hand calculations and sound engineering judgment.
All data generated by AFT Impulse should be independently verified
with other engineering methods.
AFT Impulse is designed to be used only by persons who possess a level
of knowledge consistent with that obtained in an undergraduate
engineering course in the analysis of pipe system fluid mechanics and is
familiar with standard industry practice in waterhammer analysis.
AFT Impulse is intended to be used only within the boundaries of its
engineering assumptions. The user should consult the manual for a
discussion of all engineering assumptions made by AFT Impulse.
Information in this document is subject to change without notice. No part of this User’s
Guide may be reproduced or transmitted in any form or by any means, electronic or
mechanical, for any purpose, without the express written permission of Applied Flow
Technology.
© 2007 Applied Flow Technology Corporation. All rights reserved.

Printed in the United States of America.
"AFT Impulse", "AFT Fathom", "AFT Mercury", "Applied Flow Technology", and the
AFT logo are trademarks of Applied Flow Technology Corporation
Chempak is a trademark of Madison Technical Software, Inc.
Microsoft, "Visual Basic", Excel and Windows are trademarks or registered trademarks
of Microsoft Corporation. “CAESAR II” is a registered trademark of COADE, Inc.

(si) g.doc
Contents
Summary
Summary..................................................................................................iii
Detailed Contents ..................................................................................... v
Nomenclature....................................................................................... xxix
1. Introduction .................................................................... 1
2. A Walk Through AFT Impulse...................................... 13
3. Fundamental Concepts................................................ 59
4. The Five Primary Windows.......................................... 71
5. Building and Running Models................................... 125
6. Pipe and Junction Specifications Windows ............ 193
7. Customizing AFT Impulse ......................................... 287
8. Steady-State Hydraulic Theory and Loss Models.... 325
9. Theory of Waterhammer & Solution Methodology .. 367
10. Modeling Time and Event Based Transients ......... 429
8/10/2007 4:10:28 PM Page iii c:\aft manuals\i4\impulse4 manual

(si) g.doc
iv AFT Impulse 4.0 User’s Guide
11. Reducing Surge Pressures ..................................... 441
12. Special Topics: Troubleshooting Models .............. 445
Appendix A. Keyboard Modifiers and Shortcuts ......... 477
Appendix B. Limitations ................................................ 483
Appendix C. Installation Issues .................................... 485
Appendix D. Chempak Technical Information ............. 491
Appendix E. Obtaining Technical Support................... 493
References ...................................................................... 495
Glossary .......................................................................... 497
8/10/2007 4:10:28 PM Page iv c:\aft manuals\i4\impulse4 manual

(si) g.doc
Table of Contents v
Detailed Contents
Summary..................................................................................................iii
Detailed Contents ..................................................................................... v
Nomenclature....................................................................................... xxix
1. Introduction .................................................................... 1
Terminology ............................................................................................. 1
Modeling capabilities ............................................................................... 2
Interface features ...................................................................................... 3
Thermophysical property data .................................................................. 4
Who can use AFT Impulse ....................................................................... 4
Engineering assumptions in AFT Impulse ............................................... 4
Installing AFT Impulse............................................................................. 5
Check hardware and system requirements......................................... 5
Read the README.TXT file............................................................. 5
Run the setup program ....................................................................... 6
Getting started with AFT Impulse ............................................................ 6
Example models ................................................................................. 7
Using online help...................................................................................... 7
Verification models .................................................................................. 7
AFT Impulse overview ............................................................................. 7
Input windows.................................................................................... 8
Output windows ................................................................................. 8
Converting models and databases from previous versions....................... 9
Pipe friction........................................................................................ 9
Infinite pipe junctions ........................................................................ 9
Transient junction data....................................................................... 9
Junction loss models .......................................................................... 9
Reservoir transients............................................................................ 9
Databases ......................................................................................... 10
8/10/2007 4:10:28 PM Page v

c:\aft manuals\i4\impulse4 manual (si) g.doc
vi AFT Impulse 4.0 User’s Guide
Converting models and databases from AFT Fathom or Mercury ......... 10

Combining pipes .............................................................................. 10
Junctions with polynomial loss data or interpolated x-y data.......... 10
Databases ......................................................................................... 10
What’s new in version 4.0 ...................................................................... 11
General ............................................................................................. 11
Pipes ................................................................................................. 11
Junctions........................................................................................... 11
Transient Control ............................................................................. 11
Solver ............................................................................................... 11
Workspace........................................................................................ 12
Model Data....................................................................................... 12
Output............................................................................................... 12
Graph Results................................................................................... 12
Cavitation pressure changes ...................................................... 12
2. A Walk Through AFT Impulse...................................... 13

Step 1. Start AFT Impulse ...................................................................... 13
The Workspace window ............................................................ 14
Step 2. Lay out the model ....................................................................... 15
A. Place the first reservoir ............................................................... 15
Objects and ID numbers ............................................................ 15
Editing on the Workspace.......................................................... 16
B. Place the first branch ................................................................... 16
C. Place surge tank........................................................................... 18
D. Place the second reservoir........................................................... 18
E. Add the remaining junctions........................................................ 19
F. Draw a pipe between J1 and J2.................................................... 20
Reference positive flow direction.............................................. 21
G. Add the remaining pipes ............................................................. 22
Step 3. Set the unit preferences .............................................................. 23
Step 4. Complete the first three checklist requirements ......................... 24
A. Specify Steady Solution Control ................................................. 25
8/10/2007 4:10:28 PM Page vi c:\aft manuals\i4\impulse4 manual

(si) g.doc
Table of Contents vii
B. Specify output control ................................................................. 25

C. Specify system properties............................................................ 26
Step 5. Define the model components (checklist item #4) ..................... 28
Object status .............................................................................. 28
Undefined Objects window ....................................................... 29
A. Define Reservoir J1..................................................................... 29
The Inspection feature ............................................................... 31
B. Define Branch J2 ......................................................................... 32
C. Define Surge Tank J3 .................................................................. 33
D. Define Reservoir J4 and Branches J5 and J6 .............................. 34
E. Define Valve J7 ........................................................................... 34
F. Define Valve J8 ........................................................................... 35
G. Define Pipe P1............................................................................. 37
The Pipe Specifications window ............................................... 39
H. Define Pipes P2 through P7 ........................................................ 40
Reviewing input in Model Data window................................... 40
Step 6. Complete the last two checklist requirements............................ 41
A. Specify pipe sectioning ............................................................... 41
B. Specify transient control.............................................................. 44
Step 7. Run the Solver ............................................................................ 46
The two solvers ................................................................................ 46
The transient output file ................................................................... 47
Step 8. Review the output....................................................................... 48
A. Modify the output format ............................................................ 51
B. Graph the results.......................................................................... 52
C. Animate the Results..................................................................... 53
D. View the Visual Report............................................................... 55
Conclusion .............................................................................................. 57
3. Fundamental Concepts................................................ 59
Pipes and junctions ................................................................................. 59
Convention for flow entering and exiting............................................... 60
Features for modeling irrecoverable losses ............................................ 60
8/10/2007 4:10:28 PM Page vii

viii AFT Impulse 4.0 User’s Guide
Incorporating equivalent length data ......................................... 61

Convention for specifying junction base area.................................. 61
Specifying losses.............................................................................. 62
Modeling valve closures with K factors .......................................... 62
Specifying frictional losses in pipes ................................................ 62
Comment on importance of hydraulic losses during transients ....... 63
Static vs. dynamic losses ........................................................... 63
Variable resistance .................................................................... 63
Introduction to wavespeed and waterhammer phenomenon................... 64
Description of waterhammer............................................................ 64
Instantaneous waterhammer............................................................. 65
Wavespeed ....................................................................................... 66
Communication time in pipes .......................................................... 66
Conceptual example ......................................................................... 66
Phase A: 0 < t < L/a................................................................... 67
Phase B: L/a < t < 2L/a.............................................................. 67
Phase C: 2L/a < t < 3L/a............................................................ 69
Phase D: 3L/a < t < 4L/a............................................................ 69
4. The Five Primary Windows.......................................... 71

Overview................................................................................................. 71
The Workspace window ......................................................................... 72
The Shortcut Button ......................................................................... 73
The Selection Drawing Tool ............................................................ 73
Selecting objects completely or partially inside the box........... 73
The Pipe Drawing Tool.................................................................... 74
Pipe handles and segmenting a pipe .......................................... 76
The Zoom Select Tool...................................................................... 77
The Annotation Tool........................................................................ 78
Junction icons................................................................................... 79
Editing features ................................................................................ 80
Selecting groups of objects........................................................ 80
Selecting objects in a flow path................................................. 81
8/10/2007 4:10:28 PM Page viii c:\aft manuals\i4\impulse4 manual

(si) g.doc
Table of Contents ix
Aligning objects......................................................................... 81
Creating and Managing Saved Groups ............................................ 81
Special group usage.......................................................................... 82
Bookmarks on the Workspace.......................................................... 82
Last View ......................................................................................... 82
Panning the Workspace.................................................................... 82
Customizing features........................................................................ 83
Workspace Preferences ............................................................. 83
Adjusting the Workspace size ................................................... 83
Selective display of pipes and junctions.................................... 83
Scale/flip workspace.................................................................. 84
Renumbering the Workspace objects ........................................ 85
Using the Renumber Wizard ..................................................... 86
Incrementing Object Numbers................................................... 86
Background Graphic ........................................................................ 86
Specifications windows.................................................................... 87
Printing and exporting the Workspace............................................. 88
The Model Data window ........................................................................ 88
Display and printing features ........................................................... 90
The Model Data Control window .................................................... 90
Pipe data display........................................................................ 90
Junction data display ................................................................. 90
Show selected pipes and junctions ............................................ 91
Scenario Format......................................................................... 91
Database connections ................................................................ 92
The Output window ................................................................................ 93
Transient output files ....................................................................... 94
The Output Control window ............................................................ 95
Pipe and Junction parameters .................................................... 95
Pipe Transient parameters ......................................................... 97
General output ........................................................................... 97
Summaries ................................................................................. 97
Format and action ...................................................................... 98
8/10/2007 4:10:28 PM Page ix

x AFT Impulse 4.0 User’s Guide
Show selected pipes/junctions................................................... 99

Database connections .............................................................. 100
Command buttons.................................................................... 100
Output parameter descriptions................................................. 101
Steady flow results ......................................................................... 101
Transient flow results for pipes ..................................................... 101
Transient Max/Min results for pipes ....................................... 102
Summary Max/Min results ...................................................... 103
Transient event messages ........................................................ 104
Output window updates ................................................................. 104
The Graph Results window .................................................................. 106
The Select Graph Data window ..................................................... 107
Transient pipe results............................................................... 107
Transient junction results ........................................................ 108
Profile Along a Flow Path and EGL, HGL and Elevation
Profile… .................................................................................. 108
Animating profile results......................................................... 109
Transient force results ............................................................. 112
Other graph controls ................................................................ 112
The Customize Graph window....................................................... 112
The Auxiliary Graph Formatting window...................................... 114
Printing and exporting the Graph Results image and data............. 114
The Visual Report window................................................................... 114
The Visual Report Control window............................................... 116
Display parameters .................................................................. 117
General display........................................................................ 117
Show selected pipes and junctions .......................................... 118
Color map ................................................................................ 119
Command buttons.................................................................... 121
Visual Report annotations.............................................................. 122
Annotations for the Workspace ............................................... 122
Creating Visual Report annotations ........................................ 122
The Toolbars......................................................................................... 122
8/10/2007 4:10:28 PM Page x c:\aft manuals\i4\impulse4 manual

(si) g.doc
Table of Contents xi
5. Building and Running Models................................... 125

Creating objects .................................................................................... 125
Pipes ............................................................................................... 125
Junctions......................................................................................... 128
Morphing junctions ................................................................. 128
Splitting pipes.......................................................................... 128
Annotations .................................................................................... 129
Moving objects ..................................................................................... 129
Keyboard modifiers........................................................................ 130
Editing objects ...................................................................................... 130
Connecting objects ............................................................................... 130
Defining objects.................................................................................... 132
Specifying required property data.................................................. 132
Highlighting required information ................................................. 133
Use status feature ........................................................................... 133
Undefined objects window............................................................. 133
Satisfying connectivity requirements............................................. 134
Inspecting objects ................................................................................. 134
Using the Checklist............................................................................... 136
Analysis Type................................................................................. 138
Specify Steady Solution Control.................................................... 138
Specify Output Control .................................................................. 138
Specify System Properties.............................................................. 138
Define All Pipes and Junctions ...................................................... 139
Section Pipes .................................................................................. 139
Transient Control ........................................................................... 139
Steady Solution Control........................................................................ 139
Output Control...................................................................................... 140
System Properties ................................................................................. 142
System fluid property variation...................................................... 143
Density and dynamic viscosity....................................................... 144
Bulk modulus of elasticity ............................................................. 144
Vapor pressure and cavitation........................................................ 144
8/10/2007 4:10:28 PM Page xi

xii AFT Impulse 4.0 User’s Guide
The fluid databases ........................................................................ 145

The AFT Standard Database ................................................... 145
The ASME Steam Tables Database ........................................ 145
The Chempak database............................................................ 146
Accuracy option....................................................................... 147
The ASME Steam Tables Database ........................................ 147
Viscosity models ............................................................................ 148
Non-Newtonian flow in non-pipe elements............................. 148
Atmospheric pressure..................................................................... 149
Gravitational acceleration .............................................................. 149
Transition Reynolds Numbers ....................................................... 150
Editing the AFT Standard fluid database....................................... 150
Finding object definition status ............................................................ 151
List Undefined Objects .................................................................. 151
Section Pipes......................................................................................... 152
Numbering convention................................................................... 154
Transient Control.................................................................................. 157
Start and stop time.......................................................................... 157
Save output to file… ...................................................................... 158
Use variable pipe resistance........................................................... 158
Model transient cavitation.............................................................. 159
Stop run if artificial transient detected........................................... 159
Including data in transient output file ............................................ 160
Saving pipe output ................................................................... 160
Disadvantages to saving all stations ........................................ 160
Advantages to saving all stations............................................. 161
Saving junction output ................................................................... 161
Defining force sets ......................................................................... 161
Estimated file size and run time..................................................... 162
Calculating and outputting unbalanced forces ..................................... 163
Including friction and momentum.................................................. 163
Area changes............................................................................ 165
Viewing force data ......................................................................... 165
8/10/2007 4:10:28 PM Page xii c:\aft manuals\i4\impulse4 manual

(si) g.doc
Table of Contents xiii
Exporting transient force results .................................................... 165

Defining pressures for the system ........................................................ 166
Using scenario manager........................................................................ 166
The Scenario Manager window ..................................................... 167
Creating, organizing and editing scenarios .................................... 167
Viewing scenario differences......................................................... 169
Modifying individual scenarios ..................................................... 170
Passing changes to child scenarios................................................. 171
Scenario logic examples .......................................................... 172
Re-establishing broken links ................................................... 176
Fast Scenario Changes ................................................................... 176
Special modeling features..................................................................... 176
Waterhammer Assistant ................................................................. 176
Workspace Find ............................................................................. 177
Reverse Direction........................................................................... 178
Select Special ................................................................................. 178
Special Conditions ......................................................................... 180
Special conditions with no transient data ................................ 181
Special conditions with transient data ..................................... 181
Pump special conditions .......................................................... 182
Special conditions only applied to junctions........................... 183
Transient special conditions........................................................... 183
No reflections – infinite pipe................................................... 183
Special Condition Ignore................................................................ 184
Merging models ............................................................................. 184
Merging models with multiple scenarios................................. 184
Print Preview/Special..................................................................... 185
Transfer Results to Initial Guesses ................................................ 186
Batch runs ...................................................................................... 187
Print Content .................................................................................. 188
Extended Model Check.................................................................. 189
Math calculator .............................................................................. 189
Working with the steady and transient solvers..................................... 189
8/10/2007 4:10:28 PM Page xiii

xiv AFT Impulse 4.0 User’s Guide
Steady-state solver ......................................................................... 189

Recommendations on Steady Solver ....................................... 190
Transient solver.............................................................................. 190
Artificial transients .................................................................. 191
6. Pipe and Junction Specifications Windows ............ 193

Highlight feature ............................................................................ 194
Jump feature................................................................................... 194
Pipe Specifications window ................................................................. 194
Common input parameters ............................................................. 197
Pipe number ............................................................................. 197
Pipe name ................................................................................ 197
Copy Data From Pipe .............................................................. 198
Scenario same as parent........................................................... 198
Connected junctions ................................................................ 198
Pipe Model ..................................................................................... 198
Pipe length ............................................................................... 199
Pipe material ............................................................................ 199
Pipe diameter ........................................................................... 200
Pipe wall thickness .................................................................. 200
Pipe modulus of elasticity ....................................................... 201
Pipe Poisson ratio .................................................................... 201
Friction Model ......................................................................... 201
Pipe Support ............................................................................ 203
Pipe Wavespeed....................................................................... 203
Fittings & Losses ........................................................................... 203
Design Alerts.................................................................................. 205
Optional input ................................................................................ 205
Initial Flow Rate Guess ........................................................... 205
Workspace display................................................................... 205
Design Factor........................................................................... 206
Pipe Line Size and Color ......................................................... 206
Parallel Pipes ........................................................................... 206
8/10/2007 4:10:28 PM Page xiv c:\aft manuals\i4\impulse4 manual

(si) g.doc
Table of Contents xv
Support for Partially Full Pipe................................................. 206

Intermediate Pipe Elevations ................................................... 207
Fluid Properties .............................................................................. 208
Notes .............................................................................................. 208
Status.............................................................................................. 208
Junction Specifications windows ......................................................... 208
Format #1: Junctions with one or two connecting pipes............... 210
Format #2: Junctions with more than two connecting pipes.......... 211
Parameters common to all junction Specifications windows ............... 212
Junction Number ............................................................................ 212
Junction Name................................................................................ 212
Junction Elevation.......................................................................... 212
Database List.................................................................................. 214
Copy Data From Jct list ................................................................. 214
Scenario same as parent........................................................... 214
Pipe connectivity............................................................................ 214
Jumping to another junction........................................................... 214
Transient Data ................................................................................ 215
Repeat Transient...................................................................... 216
Transient Special Conditions................................................... 216
Time based and event based transients.................................... 217
Initial Pressure/Head Guess ........................................................... 217
Displaying junction names and numbers ....................................... 219
Design Factor ................................................................................. 219
Changing the icon graphic ............................................................. 220
Changing the icon size ................................................................... 220
Specifying Base Area..................................................................... 220
Notes .............................................................................................. 220
Status.............................................................................................. 220
Area Change Specifications window.................................................... 221
Assigned Flow Specifications window................................................. 222
Transient Data ................................................................................ 224
8/10/2007 4:10:28 PM Page xv

xvi AFT Impulse 4.0 User’s Guide

Special features for steady flow..................................................... 224
Assigned Pressure Specifications window ........................................... 225
Transient Data ................................................................................ 227
Branch Specifications window ............................................................. 227
Transient Data ................................................................................ 229
Check Valve Specifications window.................................................... 229
Specifying losses............................................................................ 230
Forward Velocity to Close Valve................................................... 231
Delta Pressure to Re-Open ............................................................. 231
Transient Data ................................................................................ 231
Control Valve Specifications window.................................................. 231
Control Valve types ....................................................................... 232
PRV/PSV static vs. stagnation pressure .................................. 233
Action if setpoint not achievable ................................................... 233
Open Percentage Table .................................................................. 235
Transient Data ................................................................................ 235
Failure Transient...................................................................... 235
Control Transient..................................................................... 236
Dead End Specifications window......................................................... 236
Gas Accumulator Specifications window............................................. 237
Initial conditions and polytropic process ....................................... 237
Maximum and minimum volumes.................................................. 239
Interface with pipe system ............................................................. 239
Graphing Gas Accumulator data.................................................... 241
General Component Specifications window ........................................ 241
Liquid Accumulator Specifications window ........................................ 243
Initial conditions and elastic process ............................................. 244
Pump Specifications window ............................................................... 244
Pump Model ................................................................................... 244
Submerged pump ..................................................................... 246
8/10/2007 4:10:28 PM Page xvi c:\aft manuals\i4\impulse4 manual

(si) g.doc
Table of Contents xvii
Do not allow backwards rotation............................................. 246

Check valve at pump discharge ............................................... 246
Pump configurations ...................................................................... 246
Controlled pressure or flow (variable speed)................................. 248
NPSHR and NPSPR (optional in pump configuration) ................. 249
Efficiency and power usage (optional in pump configuration)...... 249
Transient Data ................................................................................ 249
Pumps with curves................................................................... 250
No transient from pump, no back flow.................................... 250
No transient from pump, but back flow possible..................... 250
Known speed transients, no back flow .................................... 250
Pump startup transient with possible reverse flow.................. 251
Pump transients with inertia .................................................... 251
Trip with inertia and no back flow or reverse speed ............... 252
Startup with inertia and no back flow or reverse speed .......... 252
Trip with inertia – four quadrant ............................................. 253
Startup with inertia - four quadrant, known motor
torque/speed…......................................................................... 254
Estimating the pump inertia..................................................... 254
Event modeling – including trips followed by restarts or vice
versa......................................................................................... 254
Controller transient.................................................................. 255
Pumps as fixed flow ................................................................ 255
Representing Multiple Pumps With One Junction......................... 258
Viscosity corrections...................................................................... 258
Graphing pump data....................................................................... 258
Relief Valve Specifications window .................................................... 259
Three relief valve configurations: internal, exit and inline............ 259
Specifying losses............................................................................ 260
Transient valve cracking ................................................................ 260
Transients with variable Cv ........................................................... 260
Reservoir Specifications window ......................................................... 261
Transient Data ................................................................................ 263
8/10/2007 4:10:28 PM Page xvii

xviii AFT Impulse 4.0 User’s Guide
Spray Discharge Specifications window .............................................. 263

Loss Model..................................................................................... 265
Transient Data ................................................................................ 265
Graphing Spray Discharge data ..................................................... 265
Surge Tank Specifications window ...................................................... 265
Tank geometry................................................................................ 266
Interface with pipe system ............................................................. 267
One-Way Surge Tank..................................................................... 268
Transient Data ................................................................................ 268
Surge tank as a branch during steady-state.............................. 269
Surge tank as a pressurizer during steady-state ....................... 270
Surge Tank as a finite size reservoir during steady-state ........ 270
Difference between a Reservoir and Surge Tank .................... 270
Model as dipping tube vessel ......................................................... 270
Graphing Surge Tank data ............................................................. 271
Tee/Wye Specifications window.......................................................... 271
Loss factors .................................................................................... 272
Valve Specifications window ............................................................... 273
Transient Data ................................................................................ 274
Vacuum Breaker Valve Specifications window................................... 274
Three-Stage Valve Modeling ......................................................... 276
Graphing Vacuum Breaker Valve data .......................................... 276
Volume Balance Specifications window.............................................. 276
Global Edit windows ............................................................................ 277
Global Pipe editing......................................................................... 278
Global Junction editing .................................................................. 279
Editing common junction data................................................. 280
Editing specific junction data .................................................. 282
7. Customizing AFT Impulse ......................................... 287

Parameter and Unit Preferences ........................................................... 287
8/10/2007 4:10:28 PM Page xviii c:\aft manuals\i4\impulse4

manual (si) g.doc
Table of Contents xix
Parameter Preferences.................................................................... 288

Unit Preferences............................................................................. 289
Setting preferred units ............................................................. 290
Description of selected unit ........................................................... 291
Database connections..................................................................... 291
Command buttons .......................................................................... 291
Workspace Preferences......................................................................... 292
Pipes and Junctions ........................................................................ 292
Pipe Line Options .................................................................... 292
Special Conditions options ...................................................... 293
Action When Drawing Selection Right To Left...................... 294
Pipe endpoint adjustments....................................................... 294
Modifying icon size ................................................................. 294
Action when dragging junctions.............................................. 294
Auto increment labels.............................................................. 294
Icon source............................................................................... 295
Display Options.............................................................................. 295
Workspace grid........................................................................ 296
Workspace symbols ................................................................. 296
Pipe direction arrows............................................................... 297
Displaying name and/or ID numbers ....................................... 297
Setting the default junction label position............................... 297
Allowable Workspace label movements ................................. 297
Popup menu ............................................................................. 297
Special Conditions Graphics ................................................... 298
Background Picture Scaling .................................................... 298
Colors and Fonts ............................................................................ 298
Sample Workspace......................................................................... 298
Command buttons .......................................................................... 299
Toolbox Preferences............................................................................. 300
Customizing Toolbox contents ...................................................... 300
Toolbox shortcuts .................................................................... 301
8/10/2007 4:10:28 PM Page xix

xx AFT Impulse 4.0 User’s Guide

General Preferences.............................................................................. 302
Customizing the Output........................................................................ 303
Specifying graph style preferences....................................................... 304
Auxiliary Graph Formatting window ...................................... 304
Specifying Visual Report preferences .................................................. 304
Customizing Pipe Fittings & Losses..................................................... 305
Building custom databases ................................................................... 305
Adding custom fluid properties to the Fluid Database .................. 306
Adding custom pipe materials to the Pipe Material Database ....... 309
Friction Data Sets .................................................................... 310
Pipe Physical Properties .......................................................... 311
Adding custom junction data to the Component Database ............ 311
Editing the Component Database................................................... 311
Database Manager ................................................................................ 312
Setting available databases............................................................. 312
Connecting and disconnecting to a database.................................. 313
Editing databases............................................................................ 314
Benefits of shared databases .......................................................... 316
Creating an enterprise-wide, network database system........................ 318
Overview ........................................................................................ 318
Creating database files ................................................................... 319
Sharing database files using DATABASE.LIB ............................. 321
Connecting to the external shared databases ................................. 323
8. Steady-State Hydraulic Theory and Loss Models.... 325

The Steady-State Solver ....................................................................... 325
Stagnation vs. static pressure boundaries....................................... 329
When to use static pressure ..................................................... 331
Static pressure at pressure control valve ................................. 332
Open vs. closed systems ....................................................................... 332
Verifying network solutions ................................................................. 335
Pressure drop calculation methods ....................................................... 335
8/10/2007 4:10:28 PM Page xx c:\aft manuals\i4\impulse4 manual

(si) g.doc
Table of Contents xxi
Pressure drop calculation methods for Newtonian fluids .............. 336

Roughness-based methods ............................................................. 337
Laminar flow ........................................................................... 337
Turbulent flow ......................................................................... 337
Transition flow ........................................................................ 337
Hydraulically smooth ..................................................................... 338
Hazen-Williams method................................................................. 338
Resistance....................................................................................... 338
MIT Equation for crude oil ............................................................ 338
Miller Turbulent method................................................................ 339
Frictionless pipes ........................................................................... 340
Pressure drop calculation methods for non-Newtonian fluids ............. 340
Duffy method for pulp and paper stock ......................................... 340
Brecht & Heller method for pulp and paper stock ......................... 341
Power Law non-Newtonian............................................................ 342
Bingham Plastic non-Newtonian.................................................... 343
Non-Newtonian flow through non-pipes ................................. 344
Design factors ....................................................................................... 344
Steady Solution Control parameters ..................................................... 344
Solution tolerance specification..................................................... 345
How tolerances relate to solution accuracy............................. 348
Relaxation ...................................................................................... 348
Flow Rate Relaxation .............................................................. 349
Pressure Relaxation ................................................................. 350
Avoiding false convergence .................................................... 350
Maximum Iterations ....................................................................... 351
Matrix Method ............................................................................... 351
Solution Progress window and iteration history................................... 351
Warnings in the solution....................................................................... 352
Modeling irrecoverable losses.............................................................. 353
Loss models and reference material ..................................................... 354
Design factors ................................................................................ 355
Area change.................................................................................... 355
8/10/2007 4:10:28 PM Page xxi

xxii AFT Impulse 4.0 User’s Guide
Bend ............................................................................................... 356

90 degree bends ....................................................................... 356
Non-90 degree bends ............................................................... 357
Valve .............................................................................................. 358
Orifice ............................................................................................ 358
Tee/Wye ......................................................................................... 359
Diverging Case ........................................................................ 361
Converging Case...................................................................... 362
K and CV (valve coefficient) .......................................................... 363
K for fire sprinklers........................................................................ 364
Pumps ................................................................................................... 365
Chempak thermophysical property database ........................................ 366
Accuracy options ..................................................................... 366
ASME steam tables database................................................................ 366
AFT Standard fluids ............................................................................. 366
Viscosity......................................................................................... 366
9. Theory of Waterhammer & Solution Methodology .. 367

Description of waterhammer ................................................................ 367
Instantaneous waterhammer ................................................................. 367
Wavespeed............................................................................................ 368
Communication time in pipes............................................................... 370
Method of Characteristics..................................................................... 370
Momentum equation ...................................................................... 371
Mass continuity equation ............................................................... 371
Transient vapor cavitation .................................................................... 376
Pipe sectioning...................................................................................... 377
Effects of flow resistance ..................................................................... 379
Assigned Pressure theory...................................................................... 379
Assigned Pressure vapor cavitation theory .................................... 380
Reservoir theory ................................................................................... 380
Reservoir vapor cavitation theory .................................................. 381
Assigned Flow theory........................................................................... 381
8/10/2007 4:10:28 PM Page xxii c:\aft manuals\i4\impulse4

manual (si) g.doc
Table of Contents xxiii
Assigned Flow vapor cavitation theory.......................................... 381

Dead End theory ................................................................................... 382
Dead End vapor cavitation theory.................................................. 382
Branch theory ....................................................................................... 382
Branch vapor cavitation theory ...................................................... 384
Tee/Wye theory .................................................................................... 384
Valve theory ......................................................................................... 384
Exit valves...................................................................................... 386
Valve vapor cavitation theory ........................................................ 386
Infinite Pipe theory ............................................................................... 387
Infinite Pipe vapor cavitation theory.............................................. 388
Spray Discharge theory......................................................................... 388
Spray Discharge vapor cavitation theory ....................................... 390
Pump theory.......................................................................................... 390
Modeling pump as a known speed ................................................. 391
Pumps with higher order curve fits.......................................... 392
Backflow through the pump – one quadrant ........................... 393
Backflow through the pump – four quadrants ......................... 393
Flow rates greater the pump maximum – one quadrant .......... 393
Flow rates greater the pump maximum – four quadrants ........ 393
Submerged pumps ................................................................... 394
Pump with known speed vapor cavitation theory .......................... 394
Modeling pump with unknown speed ............................................ 395
Trip With Inertia and No Back Flow or Reverse Speed ................ 395
Startup With Inertia and No Back Flow or Reverse Speed............ 397
Trip With Inertia – Four Quadrant................................................. 399
Startup With Inertia – Four Quadrant, w/Motor Torque/Speed..... 404
Estimating the pump inertia ........................................................... 404
Liquid Accumulator theory .................................................................. 405
Liquid Accumulator vapor cavitation theory ................................. 406
Relief Valve theory............................................................................... 406
Internal relief valve ........................................................................ 406
Exit relief valve.............................................................................. 407
8/10/2007 4:10:28 PM Page xxiii

xxiv AFT Impulse 4.0 User’s Guide
Inline Relief Valve ......................................................................... 407

Variable Cv model ......................................................................... 408
Relief Valve vapor cavitation theory ............................................. 409
Internal relief valves ................................................................ 409
Exit relief valves...................................................................... 409
Inline relief valves ................................................................... 409
Check Valve theory .............................................................................. 410
Check Valve vapor cavitation theory............................................. 410
Control Valve theory ............................................................................ 411
Flow control valves ........................................................................ 411
Pressure control valves................................................................... 411
Pressure drop control valves .......................................................... 411
Surge Tank theory................................................................................. 412
Surge Tank vapor cavitation theory ............................................... 413
Gas Accumulator theory....................................................................... 414
Gas Accumulator vapor cavitation theory...................................... 416
Vacuum Breaker Valve theory ............................................................. 416
Sonic inflow ................................................................................... 417
Subsonic inflow.............................................................................. 417
Sonic outflow ................................................................................. 417
Subsonic outflow............................................................................ 417
Equation of state............................................................................. 418
Sonic Inflow ............................................................................ 419
Subsonic Inflow....................................................................... 419
Subsonic Outflow .................................................................... 420
Sonic Outflow.......................................................................... 422
Lumped inertia pipe with orifice theory............................................... 422
Interpreting cavitation results ............................................................... 424
Instantaneous waterhammer equation exceptions ................................ 425
Cavitation ....................................................................................... 426
Pipe size changes............................................................................ 426
Multiple pressure surge sources..................................................... 427
Line pack due to friction ................................................................ 427
8/10/2007 4:10:28 PM Page xxiv c:\aft manuals\i4\impulse4

manual (si) g.doc
Table of Contents xxv
Elevation changes........................................................................... 427
10. Modeling Time and Event Based Transients ......... 429

Time-based transients ........................................................................... 429
Event-based transients .......................................................................... 430
Single event transients ................................................................... 430
Dual event transients: cyclic and sequential .................................. 431
Cyclic events............................................................................ 432
Sequential events ..................................................................... 433
Junctions with inherent event logic ...................................................... 434
Check valve .................................................................................... 434
Relief valve .................................................................................... 434
Vacuum breaker valve.................................................................... 434
Thought experiment to further clarify event transients ........................ 435
Other transient features......................................................................... 435
Absolute vs. relative transient data ................................................ 435
Repeat transient.............................................................................. 436
Add time offset............................................................................... 436
Graphing the transient data ............................................................ 436
Event messages..................................................................................... 436
Transient indicators on the Workspace ................................................ 437
Transient data in Model Data ............................................................... 438
11. Reducing Surge Pressures ..................................... 441

General considerations ......................................................................... 441
Larger pipe sizes................................................................................... 442
Flexible hose......................................................................................... 442
Slower or modified valve closure profiles............................................ 442
Parallel valves....................................................................................... 442
Relief valves ......................................................................................... 442
Vacuum breaker valves for low pressures............................................ 443
Surge tank ............................................................................................. 443
Gas accumulator ................................................................................... 443
8/10/2007 4:10:28 PM Page xxv

xxvi AFT Impulse 4.0 User’s Guide
12. Special Topics: Troubleshooting Models .............. 445

The philosophy of computer modeling................................................. 445
Troubleshooting steady-state models ................................................... 446
Use the Model Data window.......................................................... 446
Poor pump curve fits ...................................................................... 447
Use the Output window Sort feature.............................................. 448
Tee/Wye junction complexity........................................................ 450
Lower the flow rate relaxation....................................................... 450
Try absolute tolerance.................................................................... 451
Make initial flow rate guesses for pipes ........................................ 451
Review the Iteration History .......................................................... 451
Change boundary conditions.......................................................... 451
Turn off parts of the model ............................................................ 452
Break a large model into submodels .............................................. 452
Control valve failure issues in steady-state .......................................... 452
Infinite pipe junctions........................................................................... 453
Application example ...................................................................... 453
How pressure junctions work in steady-state ....................................... 455
The role of pressure junctions in steady-state ...................................... 458
Examples ........................................................................................ 459
Sizing pumps with flow control valves.......................................... 464
Pressure control valves................................................................... 470
Closing parts of a system ............................................................... 472
Effect of pipe sectioning on transient run times................................... 473
Artificial transients ............................................................................... 474
Manually detecting artificial transients.......................................... 476
Appendix A. Keyboard Modifiers and Shortcuts ......... 477

Selection Drawing Tool........................................................................ 477
Double-clicking tool ................................................................ 477
Control key .............................................................................. 477
Shift key................................................................................... 477
Alt key ..................................................................................... 478
8/10/2007 4:10:28 PM Page xxvi c:\aft manuals\i4\impulse4

manual (si) g.doc
Table of Contents xxvii
Drag left to right ...................................................................... 478

Drag right to left ...................................................................... 478
Pipe Drawing Tool................................................................................ 478
Double-clicking tool ................................................................ 478
Control key .............................................................................. 478
Shift key................................................................................... 478
Zoom Select Tool ................................................................................. 479
Shift key................................................................................... 479
Panning the Workspace ........................................................................ 479
Dragging junctions from Toolbox ........................................................ 479
Control key .............................................................................. 479
Shift key................................................................................... 479
Inspecting pipes and junctions.............................................................. 480
Shift key................................................................................... 480
Control key .............................................................................. 480
Moving junctions with connected pipes ............................................... 480
Control key .............................................................................. 480
Pipe and junction specifications window ............................................. 480
F2 function key........................................................................ 480
F5 function key........................................................................ 480
Appendix B. Limitations ................................................ 483
Appendix C. Installation Issues .................................... 485

Customization files ............................................................................... 485
Special file control features.................................................................. 485
What is the Control.aft File?.......................................................... 485
Specifying where Chempak is located ........................................... 486
Network installations............................................................................ 486
Problems loading the Graphics Server ................................................. 487
Appendix D. Chempak Technical Information ............. 491
8/10/2007 4:10:28 PM Page xxvii

xxviii AFT Impulse 4.0 User’s Guide
Appendix E. Obtaining Technical Support................... 493

AFT's three levels of support................................................................ 493
1. Free Software Support for One Year ......................................... 493
2. Annual Support, Upgrade and Maintenance (SUM).................. 493
3. Waterhammer Modeling Consulting Support ............................ 493
Contacting AFT .................................................................................... 494
Telephone support.......................................................................... 494
Web site.......................................................................................... 494
E-Mail support ............................................................................... 494
Mail support ................................................................................... 494
References ...................................................................... 495
Glossary .......................................................................... 497
8/10/2007 4:10:28 PM Page xxviii c:\aft manuals\i4\impulse4

manual (si) g.doc
Nomenclature
a wavespeed
A cross-sectional flow area of a pipe
BM parameter as defined after Equation 9.20
BP parameter as defined after Equation 9.20
CD discharge coefficient
CM parameter as defined after Equation 9.20
CP parameter as defined after Equation 9.20
CV valve coefficient
CHW Hazen-Williams factor
D diameter of a pipe
E modulus of elasticity
e pipe wall thickness
f friction factor
fT turbulent friction factor (used by Crane, 1988)
F error value in mass balance
FB Dimensionless pump torque parameter (see Chapter 9)
FW Dimensionless pump head parameter (see Chapter 9)
g gravitational constant
h head ratio for pump
H head
HGL hydraulic gradeline (piezometric head)
I moment of inertia of pump and entrained liquid
J Jacobian matrix
K loss factor
K constant for Power Law fluid
KB liquid bulk modulus of elasticity
Ksprinkler fire sprinkler loss factor
L length of a pipe
8/10/2007 4:10:28 PM Page xxix c:\aft manuals\i4\impulse4

manual (si) g.doc
xxx AFT Impulse 4.0 User’s Guide
m& mass flow rate

n general constant
n constant for Power Law fluid
N pump synchronous speed (rpm)
P pressure, static
Po pressure, stagnation
Pw wetted perimeter
P power
Q volumetric flow rate
r radius
r relaxation
R, R’ resistance
Rc coefficient of rigidity
Re Reynolds Number
s pump speed
SB parameter as defined in Equation 9.27
SC parameter as defined in Equation 9.28
Sy yield stress
t time
T torque on pump
v pump flow ratio to reference
V velocity
x distance along pipe centerline
z elevation
α, μ, θ angle
α pump speed ratio
β diameter ratio
β torque ratio
λ method of characteristics multiplier
ν kinematic viscosity
ε roughness
8/10/2007 4:10:28 PM Page xxx c:\aft manuals\i4\impulse4

manual (si) g.doc
Nomenclature xxxi
ω pump speed
ρ density
μ dynamic viscosity
Subscripts
f fluid
i junction or pipe station at which solution is sought
j junctions with pipes connecting to junction I
m motor
R rated conditions for a pump, usually BEP
o stagnation
∞ infinity, far away, ambient
new current time for computation
old most recent time for computation
8/10/2007 4:10:28 PM Page xxxi

xxxii AFT Impulse 4.0 User’s Guide
8/10/2007 4:10:28 PM Page xxxii c:\aft manuals\i4\impulse4

manual (si) g.doc
CHAPTER 1
Introduction
Welcome to AFT Impulse™ 4.0 for Windows. AFT Impulse is a

graphical platform for modeling waterhammer and surge transients in
pipe networks.
AFT Impulse’s advanced Windows graphical interface simplifies the
complex process of building waterhammer models. The transient
solution engine employs the proven and well understood Method of
Characteristics to solve the fundamental equations of waterhammer.
An integral part of the interface is AFT Impulse's Workspace window.
Here you place icons and draw lines to represent the components of a
pipe system. You then enter data for the components in the associated
Specifications windows.
AFT Impulse shows you both input data and analysis results in graphical
form, allowing rapid analysis of the model's validity. Identifying poor
assumptions, catching typographical errors, and rerunning models are all
accelerated because of AFT Impulse's graphical environment. This
reduces the possibility of modeling errors.
Whether your pipe system model will be used to evaluate and improve
an existing system or to design a new one, AFT Impulse increases your
productivity in the modeling process.
Terminology
The term “waterhammer” has been used for more than one-hundred
years, and describes a transient phenomenon in piping systems caused by

(si) g.doc
2 AFT Impulse 4.0 User’s Guide
changing pressure or flow boundary conditions, or changing operating
conditions of pipe system equipment. The result is often a pressure surge
that builds much higher than the steady state pressure that preceded it.
There have been and are many terms employed to refer to the
waterhammer phenomenon. The oldest and probably most recognized is
the term waterhammer itself. One drawback of this term is that it
suggests a phenomenon associated only with water. This is misleading,
because the surge pressures resulting from waterhammer occur in all
liquid piping systems.
To help avoid this confusion, the term “fluidhammer” has been
advocated and/or used in some of the literature. Fluidhammer does better
convey the breadth of fluids that can experience waterhammer, but has
not gained wide acceptance.
Another alternate term for waterhammer is “hydraulic transients”.
Again, a suggestion of water-related events are present in this term.
Another term, “fluid transients” was therefore put forth. This term also
has problems. Specifically, a fluid system can have many kinds of
transient behavior, of which waterhammer is only one.
Finally, another commonly heard term is “surge”. This term is more
specific than fluid transients, but is still somewhat ambiguous.
The unfortunate situation, therefore, is that there is no universally
accepted or recognized term to refer to this important phenomenon.
Arguably, the term waterhammer remains the most recognized term,
even with its deficiencies. Therefore, with the qualification that it is not
limited solely to water systems, waterhammer is the preferred term and
the one that will be used most commonly when referring to AFT
Impulse.
Modeling capabilities
AFT Impulse can be used to model a wide variety of pipe system
transients, including:
• Transients in open and closed (recirculating) systems
• Network systems that branch or loop, with no limit on the number of
loops
• Systems with valve transients

(si) g.doc
Chapter 1 Introduction 3
• Systems with pump transients
• Systems with pressure or flow control valve transients
• Systems with transient cavitation and liquid column separation
• Systems with surge suppression devices such as accumulators, surge
tanks and vacuum breaker valves
• Systems with variable density and viscosity
• Multiple design cases in a single model file
• Non-Newtonian fluid behavior
AFT Impulse's steady-state solution engine is based on standard
techniques used for many years in industry. The Newton-Raphson
method is used to solve the fundamental equations of pipe flow that
govern mass and momentum balance. Solutions are obtained by iteration,
and matrix methods optimized for speed are employed to obtain
convergence.
Once a steady-state solution is obtained, AFT Impulse solves the
equations of waterhammer using the Method of Characteristics (MOC).
The MOC is the most widely used method for solving waterhammer
problems.
Interface features
AFT Impulse's graphical interface is based on drag-and-drop operations,
which make it simple to build a model of a generalized pipe system. You
control the arrangement, and you benefit from the direct visual feedback
regarding the layout of your model.
Data is entered for the components in Specifications windows, which are
opened by double-clicking the component of interest. Additional global
editing features simplify making large-scale changes to the model.
AFT Impulse handles both traditional English and SI systems of units.
You assign units to all input parameters by choosing from lists. This
highly flexible approach removes the burden of hand calculated unit
conversions.
8/10/2007 4:10:28 PM Page 3

Numerous output reports can be generated, all of which are
customizable. All printed output is of report quality. Effectively
organized output data also aids in verifying modeling accuracy.
Thermophysical property data

AFT Impulse derives physical properties from one of three sources. The
first is the standard AFT Impulse set of incompressible fluids which
contains data for about 10 common fluids (called AFT Standard). These
are the same fluids as were available in previous versions of AFT
Impulse.
The second are water properties from the ASME Steam Tables database.
The third source is Chempak™. Chempak has a database of
approximately 700 fluids, and supports user specified fluid mixtures.
AFT Impulse is restricted to non-reacting mixture calculations.
The user does have the option of manually entering physical properties.
This is called an Unspecified Fluid.
In the remainder of this User’s Guide it will be assumed in general
discussion that the user has access to the Chempak Database. If the user
does not, then the portions describing Chempak will not be applicable.
Who can use AFT Impulse

AFT Impulse assumes that the user possesses a good general knowledge
of engineering pipe system hydraulics and has at least a basic
understanding of the waterhammer and surge phenomenon. See the
copyright page in this manual for cautionary information.
Engineering assumptions in AFT Impulse

AFT Impulse is based on the following fundamental fluid mechanics
assumptions:
• Liquid flow
• One-dimensional flow

(si) g.doc
• No chemical reactions
• Wavespeed remains constant during transients
• Non-condensable gas release is negligible
• Bubbles that form during transient cavitation do not move
AFT Impulse allows you to change the gravitational acceleration of the
system, a feature useful for aerospace vehicle applications or pipe
systems that are designed for extraterrestrial environments.
Installing AFT Impulse

You install AFT Impulse on your computer by running the SETUP.EXE
program located on the installation CD. The setup program will install
AFT Impulse, the Help system, example models, verification models,
and AFT Impulse support files to your hard disk.
Note: You must use the setup program to install AFT Impulse. You
cannot simply copy the files from the installation disks to your hard
drive.
Before you run the setup program, make sure your computer platform
meets the minimum requirements.
Check hardware and system requirements

To install and run AFT Impulse, your computer must have the features
listed in Table 1.1.
Read the README.TXT file

The README.TXT file is included on the installation CD.
README.TXT contains information about installation and support
material that is more up to date than this User’s Guide.
8/10/2007 4:10:28 PM Page 5

Table 1.1 Computer hardware and software requirements to use
AFT Impulse
Feature Required
Processor Pentium II or higher
Hard disk space 40 MB
CD-ROM drive Yes
Monitor resolution SVGA (800x600)
RAM 128 MB
MS-Windows Windows 98, NT 4.0,
2000, ME, XP or later
versions
Run the setup program

When you run the setup program, you will be prompted for the directory
where you want AFT Impulse installed. To run the setup program:
1. Insert the CD into your CD-ROM drive.
2. If Autorun is disabled, from the Start button choose Run. Assuming
your CD-ROM is drive “d”, type d:\ setup, or click the Browse
button and search for the setup program on your CD-ROM drive in
the Impulse 4.0 directory.
3. Select Impulse 4.0 on the displayed menu.
4. Follow the instructions on the screen.
Getting started with AFT Impulse

If you are not familiar with basic techniques used in Windows
applications (working with the mouse, selecting menus, using dialog
boxes, and so on), first consult your Windows documentation.
To gain an understanding of the powerful features available in AFT
Impulse, read Chapters 3-12 of this User’s Guide. For an overview of the
model building process and AFT Impulse’s capabilities, read Chapter 2
and follow the instructions for building a simple model.

(si) g.doc
Example models
An auxiliary help file (called ImpulseExamples.hlp) is installed with
AFT Impulse and leads the user through modeling a number of real
world systems. This Help file can be accessed by choosing "Show
Examples" from the help menu. The example models discussed in
ImpulseExamples.hlp are installed in the EXAMPLES folder.
Using online help

To access AFT Impulse's online help, press the F1 function key or select
Help from the menu bar. For convenient access, much of the content of
this User’s Guide is included in the Help system. The Help button in
each dialog window provides context-sensitive help on the features of
that window.
In the Help system you have the option of searching for information on
specific topics or searching through the hierarchical layout to find more
general information.
Verification models
A large number of verification models have been built and compared to
published results from the open literature. These are included in the
Verification sub-folder below AFT Impulse. Along with the models you
will find documented comparisons in the Verification help file also
installed in the Verification folder.
AFT Impulse overview

The AFT Impulse window has five subordinate windows that work in an
integrated fashion. You work exclusively from one of these windows at
all times. For this reason they are referred to as primary windows.
Of the five primary windows, two are input windows, two are output
windows, and one displays output and input information. Figure 1.1
shows the relationship between the primary windows.
8/10/2007 4:10:28 PM Page 7

Model Data
Visual Report
Workspace Output
Graph Results
Figure 1.1 Primary window workflow in AFT Impulse
Input windows
The two windows that function exclusively as input windows are the
Workspace window and the Model Data window. These two windows,
one graphical and the other text-based, work together to process model
input data with immense flexibility. The tools provided in these two
windows allow you to model a large variety of pipe networks.
The Visual Report window can function in support of both input and
output data. As an input window, it allows you to see the input data
superimposed on the pipe system schematic created on the Workspace.
Output windows
The two windows that function exclusively as output windows are the
Output window and the Graph Results window. The Output window is
text-based, while the Graph Results window is graphical. These two
windows offer a powerful and diverse range of features for reviewing
analysis results for modeling errors, gaining a deeper understanding of
the pipe system's flow behavior, and preparing the results for
documentation.
As an output window, Visual Report allows you to see the output results
superimposed on the pipe system schematic created on the Workspace.
The five primary windows form a tightly integrated, highly efficient
system for entering, processing, analyzing, and documenting
waterhammer analyses of pipe networks.

(si) g.doc
Converting models and databases from previous versions

AFT Impulse 1.0, 2.0 and 3.0 models can be opened directly by choosing
Open from the File menu. Once saved, they are saved in the version 4.0
format. Importing issues are discussed in the following.
Pipe friction
AFT Impulse 1.0 only worked with constant friction factor models.
When these models are imported into version 4.0, the friction model for
the pipe is set to Explicit Friction Factor or Frictionless, depending on
whether the version 1.0 pipe had a zero or non-zero friction factor.
Infinite pipe junctions

AFT Impulse 1.0 supported a junction called an Infinite Pipe junction.
This feature is still supported in version 4.0 but in a different way. When
importing into version 4.0, Infinite Pipe junctions are converted into
Assigned Pressure junctions, with the Transient Special Condition set to
Ignore Reflections – Infinite Pipe. This was necessary because of the
need to first perform a steady-state solution.
Transient junction data

AFT Impulse 1.0 only supported time-based transients, and thus all
transient initiation data is specified as time-based.
Junction loss models

A loss model found in several junction types in AFT Impulse 1.0 was
K/A^2. This is no longer supported in version 4.0. Such data, whether
steady or transient, is converted to a loss model that is supported such as
K factor or Cv for valves.
Reservoir transients
AFT Impulse 1.0 allowed reservoir surface level transients as a delta
from the steady-state. These will all get converted automatically to the
version 4.0 convention of specifying the actual liquid surface level.
8/10/2007 4:10:28 PM Page 9

Databases
AFT Impulse 1.0 databases can be converted using the AFT Convert
utility provided with the software.
Converting models and databases from AFT Fathom or Mercury

AFT Fathom™ 4.0, 5.0 and 6.0 models, and AFT Mercury™ 5.0 models
can be read directly by AFT Impulse 4.0 by choosing Open from the File
menu and opening the file. Only the base scenario can be imported. If
you want to import scenarios that are not the base, you will need to first
use the Scenario Manager in AFT Fathom or AFT Mercury to save the
scenario into a separate model file.
When imported, all pipe specification data should be maintained,
including friction data.
Combining pipes
AFT Impulse 4.0 will offer to combine pipes for you when you import
AFT Fathom models. This will allow models to run much more
efficiently. When combined, AFT Impulse will create intermediate pipe
elevations at the points where junctions are deleted. All pipe resistance
and junction losses will be maintained in the model.
Junctions with polynomial loss data or interpolated x-y

data
Except for General Components, AFT Impulse 4.0 does not support
polynomial or interpolated x-y loss data. When AFT Fathom or AFT
Mercury models are imported, all such data is automatically converted to
K factors.
Databases
AFT Fathom and AFT Mercury databases can be converted using the
AFT Convert utility.

(si) g.doc
What’s new in version 4.0
General
• Users can create transient force files for use in CAESAR II pipe
stress software
• Select Special on pipe material
• Significantly enhanced global editing of pipes and junctions (data
now shown in categories)
• Batch runs for scenarios
• All printing can go directly to a PDF file
• In grids, cells that have drop-down capability have a persistent drop
down symbol
• Initialization and data files now saved to user folders
Pipes
• Enhanced fittings & losses entry for pipes
• New pipe friction model uses resistance in dH/Q2
Junctions
• Valves now support Cv vs. percent open (like Control Valves)
• More advanced control valve logic
Transient Control
• Users can define “force sets” for purposes of plotting force data
• Runtime estimates based on users previous
Solver
• Power Law and Bingham Plastic now cover both laminar and
turbulent flow
8/10/2007 4:10:28 PM Page 11

Workspace
• Renumber Increment Added
• Shortcut button provides fast access to explanations of shortcut
functions
• Selections can be rotated right or left
• New window added to show all undefined pipes and junctions and
missing properties
Model Data
• Can see data for all scenario direct ancestors on Model Data
Output
• Users can print Model Data with output
Graph Results
• Can graph forces on pipes
Cavitation pressure changes

In AFT Impulse 2.0 and 3.0 there was an attempt at increased numerical
precision by calculating cavitation when the static pressure reached
vapor pressure. Hence the stagnation pressure was always greater than
vapor pressure.
Further review showed that using static pressure as the cavitation criteria
could result in increased numerical noise. A change was therefore made
in AFT Impulse 4.0 to use stagnation pressure as the cavitation criteria.
This will result in static pressures below vapor pressure in certain
situations.

(si) g.doc
CHAPTER 2
A Walk Through AFT Impulse
This chapter is designed to give you the big picture of AFT Impulse's
layout and structure. Some of the more basic concepts will be used to
build an eight-pipe, eight-junction model to solve a waterhammer
problem published in the literature (Karney, et al., 1992). This system
has reservoirs, a surge tank, a relief valve and an exit valve.
This chapter is not intended to replace the more in-depth discussions
given in later chapters. To acquire a more detailed understanding of AFT
Impulse's menus and functionality before creating a model, skip over this
chapter.
A number of other example model discussions are included in a Help file
distributed with AFT Impulse called ImpulseExamples.hlp. It can be
opened from the Help menu by choosing "Show Examples".
Step 1. Start AFT Impulse
¾ To start AFT Impulse, click Start on the Windows taskbar, choose

Programs, then AFT Products then AFT Impulse. (This refers to the
standard menu items created by setup. You may have chosen to specify a
different menu item).
When you start AFT Impulse, the Workspace window is always the
active (large) window. The Workspace window is one of five primary
windows.

(si) g.doc
After AFT Impulse loads, you will notice four windows in the lower part
of the AFT Impulse window; these represent four of the five primary
windows that are currently minimized (see Figure 2.1). The AFT
Impulse window acts as a container for the five primary windows.
Toolbars
Workspace
Toolbox
Minimized
primary
windows
Status Bar
Figure 2.1 The Workspace window is where the model is built
The Workspace window

The Workspace window is the primary vehicle for building your model.
This window has two main areas: the Toolbox and the Workspace itself.
The Toolbox is the bundle of tools on the far left. The Workspace takes
up the rest of the window. See Chapter 4 for more detailed information
on the Workspace.
You will build your pipe flow model on the Workspace using the
Toolbox tools. At the top of the Toolbox are a Shortcut button and four
drawing tools. The Selection Drawing tool, on the upper left below the
Shortcut button, is useful for selecting groups of objects on the
Workspace for editing or moving. The Pipe Drawing tool, on the upper

(si) g.doc
Chapter 2 A Walk Through AFT Impulse 15
right below the Shortcut button, is used to draw new pipes on the
Workspace. Below these two tools are the Zoom Select tool and the
Annotation tool. The Zoom Select tool allows you to draw a box on the
Workspace after which AFT Impulse will zoom into that area. The
Annotation tool allows you to create annotations and auxiliary graphics.
Below the four drawing tools are nineteen icons that represent the
different types of junctions available in AFT Impulse. Junctions are
components that connect pipes and also influence the pressure or flow
behavior of the pipe system. The nineteen junction icons can be dragged
from the Toolbox and dropped onto the Workspace.
When you pass your mouse pointer over any of the Toolbox tools, a
Tool Tip identifies the tool's function.
Step 2. Lay out the model

To lay out the walk-through model, you will place the two reservoir
junctions, three branch junctions, a surge tank junction, a relief valve
junction, and a valve junction on the Workspace. Then you will connect
the junctions with pipes.
A. Place the first reservoir
¾ To start, drag a reservoir junction from the Toolbox and drop it on

the Workspace. Figure 2.2a shows the Workspace with one reservoir.
Objects and ID numbers

Items placed on the Workspace are called objects. All objects are
derived directly or indirectly from the Toolbox. AFT Impulse uses three
types of objects: pipes, junctions and annotations.
All pipe and junction objects on the Workspace have an associated ID
number. For junctions, this number is, by default, placed directly above
the junction and prefixed with the letter “J”. Pipe ID numbers are
prefixed with the letter “P”. You can optionally choose to display either
or both the ID number and the name of a pipe or junction. You also can
drag the ID number/name text to a different location to improve
visibility.
8/10/2007 4:10:28 PM Page 15

The Reservoir you have created on the Workspace will take on the
default ID number of 1. You can change this to any desired number
greater than zero and up to 30,000.
Figure 2.2a Walk Through Model with one reservoir placed
Editing on the Workspace

Once on the Workspace, junction objects can be moved to new locations
and edited with the features on the Edit menu. Cutting, copying, and
pasting are all supported. A single level of undo is available for all
editing operations.
Note: The relative location of objects in AFT Impulse is not important.
Distances and heights are defined through dialog boxes. The relative
locations on the Workspace establish the connectivity of the objects, but
have no bearing on the actual length or elevation relationships.
B. Place the first branch
¾ Next, drag a branch junction from the Toolbox and drop it to the
lower right of the reservoir J1. This becomes junction J2 (Figure 2.2b).

(si) g.doc
Figure 2.2b Walk Through Model with one reservoir and one
branch placed
¾ Now, drag a surge tank junction from the Toolbox and drop it to the
right of the reservoir J1. This junction takes on the number J3 (Figure
2.2c).
8/10/2007 4:10:28 PM Page 17

Figure 2.2c Walk Through Model after placing the Surge Tank
C. Place surge tank
D. Place the second reservoir

The second reservoir can be created the same way as the first one or it
can be derived from the existing reservoir.
¾ To create a second reservoir from the existing one, select junction J1

by clicking it with the mouse. A red outline will surround the junction.
Open the Edit menu and choose Duplicate. If you like, you can Undo the
Duplicate operation and then Redo it to see how these editing features
work.
Drag the new reservoir junction icon to the upper right of the
Workspace. This is junction J4 (Figure 2.2d).

(si) g.doc
Figure 2.2d Walk Through Model after placing the second

Reservoir
E. Add the remaining junctions

Next add the four remaining junctions - two branches, and two regular
valves (Figure 2.2e).
¾ To add a branch junction, select a Branch from the Toolbox and place
it on the Workspace as shown in Figure 2.2e. The Branch will take on
the default number J5. Similarly, add another branch at junction J6.
¾ Add valve junctions by selecting the junction icons from the Toolbox
and placing them on the Workspace as shown in Figure 2.2e. These
junctions will become J7 and J8, respectively.
¾ Before continuing, save the work you have done so far. Choose Save
As from the File menu and enter a file name ("WalkThru", perhaps) and
AFT Impulse will append the “.IMP” extension to the file name.
8/10/2007 4:10:28 PM Page 19

Figure 2.2e Walk Through Model after placing all junctions
F. Draw a pipe between J1 and J2

Now that you have the eight junctions, you need to connect them with
pipes.
¾ To create a pipe, click on the Pipe Drawing tool icon. The pointer will
change to a cross-hair when you move it over the Workspace. Draw a
pipe below the junctions, similar to that shown in Figure 2.2f.
The pipe object on the Workspace has an ID number (P1) shown near
the center of the pipe.

(si) g.doc
Figure 2.2f Walk Through Model after drawing first pipe
¾ To place the pipe between J1 and J2, grab the pipe in the center with
the mouse and drag it so that its left endpoint falls within the J1
Reservoir icon and drop it there. Next, grab the right endpoint of the
pipe and stretch the pipe, dragging it until the endpoint terminates within
the J2 Branch icon (see Figure 2.2g).
Reference positive flow direction

Located on the pipe is an arrow that indicates the reference positive flow
direction for the pipe. AFT Impulse assigns a flow direction
corresponding to the direction in which the pipe is drawn. You can
reverse the reference positive flow direction by choosing Reverse
Direction from the Arrange menu or selecting the reverse direction
button on the Toolbar.
8/10/2007 4:10:28 PM Page 21

Figure 2.2g Walk Through Model after placing first pipe
In general, the reference positive flow direction indicates which

direction is considered positive. However, when used with pumps and
certain other junction types the pipes must be in the correct flow
direction because that is how AFT Impulse determines which side is
suction and which is discharge. If the reference positive direction is the
opposite of that obtained by the Solver, the output will show the flow
rate as a negative number.
G. Add the remaining pipes

A quicker way to add a pipe is to draw it directly between the desired
junctions.
¾ Activate the pipe drawing tool again and press the mouse button down
on the J2 Branch. Stretch the pipe up to the J3 Surge Tank and release
the mouse button. Draw a third pipe from the J3 Surge Tank to the J4
Reservoir. Draw a fourth pipe from the J3 Surge Tank to the J5 Branch.
Draw a fifth pipe from the J6 Branch to the J5 Branch. Draw a sixth pipe
from the J2 Branch to the J6 Branch. Draw a seventh pipe from the J6
Branch to the J7 Valve.. Finally, draw an eight pipe from the J6 Branch
to the J8 Valve. Your model should now look similar to Figure 2.2h.

(si) g.doc
At this point all the objects in the model are graphically connected. Save
the model by selecting Save in the File menu or by clicking on the
diskette button on the Toolbar.
Note: It is generally desirable to lock your objects to the Workspace

once they have been placed. This prevents accidental movement and
disruption of the connections. You can lock all the objects by choosing
Select All from the Edit menu, then selecting Lock Object from the
Arrange menu. The lock button on the Toolbar will appear depressed
indicating it is in an enabled state, and will remain so as long as any
selected object is locked. Alternatively, you can use the grid feature
enabled on the Workspace Preferences window and specify that the
pipes and junctions snap to grid.
Figure 2.2h Walk Through Model with all pipes and junctions
placed
Step 3. Set the unit preferences
¾ Open the Parameter and Units Preferences window from the Options
menu. This Walk Through model is in metric units. The default in AFT
8/10/2007 4:10:28 PM Page 23

Impulse is to offer both metric and traditional English units to the user,
with English as the default. For consistency, we want to use metric as
the default for this model, and we can specify that here.
¾ Select the Unit Preferences tab and then select the Default Unit System
as SI. Click the Apply Default Units button. See Figure 2.3. Finally click
OK to close the window.
Figure 2.3 The Parameter and Unit Preferences window allows

you to specify the default unit system
Step 4. Complete the first three checklist requirements
¾ Next, click the checkmark on the Toolbar that runs across the top of
the AFT Impulse window. This opens the Checklist window (see Figure
2.4). The checklist contains six items. Each item needs to be completed
before AFT Impulse allows you to run the Solver.

(si) g.doc
The Status Bar at the bottom of the AFT Impulse window also reflects
the state of each checklist item (see Figure 2.1). Once the checklist is
complete, the Model Status light in the lower left corner turns from red
to green.
Figure 2.4 The Checklist tracks the model’s status
A. Specify Steady Solution Control

The first item, Specify Steady Solution Control, is always checked when
you start AFT Impulse because AFT Impulse assigns default solution
control parameters for the steady-state part of the analysis. In general,
you do not need to adjust Steady Solution Control values. If necessary,
you can make adjustments by opening the Steady Solution Control
window from the Analysis menu.
B. Specify output control

The second item on the checklist is Specify Output Control. Like
Solution Control, this item is always checked when you start AFT
Impulse. Default Output Control parameters and a default title are
assigned.
You may want to add a descriptive title for the model. To do this, open
the Output Control window (see Figure 2.5) and enter a title on the
General tab. In addition, this window allows you to select the specific
output parameters you want in your output. You also can choose the
units for the output.
8/10/2007 4:10:28 PM Page 25

¾ Close the checklist and select Output Control from the Analysis menu.
(Figure 2.5 shows the Output Control window). Click the General
Output tab, enter a new title (if you like you can title this “Transient
With Surge Tank”), then click OK to accept the title and other default
data.
Figure 2.5 The Output Control window lets you customize the
output
C. Specify system properties

The third item on the checklist is Specify System Properties. To
complete this item, you must open the System Properties window (see
Figure 2.6). This window allows you to specify your fluid properties
(density, dynamic viscosity, bulk modulus and optional vapor pressure),
viscosity model, gravitational acceleration and atmospheric pressure.
For models with variable fluid properties, the values for density,
viscosity and bulk modulus are default fluid properties. You can then
enter different property values, if desired, for any pipe in the Pipe
Specifications window.

(si) g.doc
Figure 2.6 The System Properties window lets you enter physical
properties of the fluid
You can model the fluid properties in one of five ways.

1. Unspecified fluid – This fluid model allows you to directly type in
the density, viscosity, bulk modulus and vapor pressure.
2. AFT Standard fluid – This fluid model accesses fluid data from the
AFT Standard database. These fluid properties are either
temperature dependent or dependent on the solids concentration.
You type in the desired condition (e.g., temperature), click the
Calculate Properties button and the required properties are
calculated. Users can add their own fluids to this database. Custom
fluids are created by opening the Fluid Database window from the
AFT Impulse Database menu or by clicking the Edit Fluid List
button in the System Properties window.
3. Chempak Fluid – This fluid model allows you to select a single fluid
from the Chempak database list. These fluid properties are pressure
and temperature dependent, although some are temperature
dependent only. Chempak is an optional add-on to AFT Impulse.
8/10/2007 4:10:28 PM Page 27

4. Chempak Mixture – This fluid model allows you to create a liquid
mixture from among the Chempak database fluids. These fluid
properties are pressure and temperature dependent. Upon entering a
pressure, AFT Impulse will display a temperature range applicable
to the liquid. Chempak mixture data can be exported to a data file,
and then be shared with other AFT applications or users. Chempak is
an optional add-on to AFT Impulse.
5. Water Data from ASME Steam Tables – As its name implies, this
fluid model obtains water properties from ASME steam tables. This
model is pressure and temperature dependent.
¾ Select System Properties from the Analysis menu to open the System
Properties window. For this example, select the AFT Standard fluid
option, then choose “Water at 1 atm” from the list and click the Add to
Model button. The properties for AFT Standard water are given only as
a function of temperature. Enter 20° C in the temperature box, click the
Calculate Properties button and click OK.
¾ Open the checklist once more or observe the Status Bar and you should
now see the third item checked off.
Step 5. Define the model components (checklist item #4)

The fourth item on the checklist, Define All Pipes and Junctions, is not
as straightforward to satisfy as the first three. This item encompasses the
proper input data and connectivity for all pipes and junctions.
Object status
Every pipe and junction has an object status. The object status tells you
whether the object is defined according to AFT Impulse's requirements.
To see the status of the objects in your model, click the floodlight on the
Toolbar (alternatively, you could choose Show Object Status from the
View menu). Each time you click the floodlight, Show Object Status is
toggled on or off.
When Show Object Status is on, the ID numbers for all undefined pipes
and junctions are displayed in red on the Workspace. Objects that are
completely defined have their ID numbers displayed in black. (These

(si) g.doc
colors are configurable through Workspace Preferences from the
Options menu.)
Because you have not yet defined the pipes and junctions in this sample
problem, all the objects' ID numbers will change to red when you turn on
Show Object Status.
Undefined Objects window

The Undefined Objects window lists all undefined pipes and junctions
and further displays the items that are not yet defined.
A. Define Reservoir J1
¾ To define the first reservoir, open the J1 Reservoir Specifications

window by double-clicking on the J1 icon. Enter in a reservoir surface
elevation of 200 meters. You can assign any unit of length found in the
adjacent drop-down list box of units.
Note: You can also open an object's Specifications window by selecting

the object (clicking on it) and then either pressing the Enter key or by
clicking the Open Pipe/Jct Window icon on the Toolbar.
¾ Enter surface pressure of 1 atmosphere (atm) and a reservoir depth of

50 meters in the table on the Pipe Depth and Loss Coefficient tab.
Note: You can specify preferred units for many parameters (such as
meters for length) in the Parameter & Unit Preferences window.
You can give the component a name, if desired, by entering it in the
Name field at the top of the window. In Figure 2.7, the name of this
reservoir is Supply Tank A. By default the junction’s name is the
junction type. The name can be displayed on the Workspace, Visual
Report or in the Output.
Most junction types can be entered into a custom database allowing the
junction to be used multiple times or shared between users. To select a
junction from the custom database, choose the desired junction from the
Database list. The current junction will get the properties from the
database component.
8/10/2007 4:10:28 PM Page 29

The Copy Data From Jct list will show all the junctions of the same type
in the model. This will copy selected parameters from an existing
junction in the model to the current junction.
The pipe table on the Pipe Depth and Loss Coefficients tab allows you to
specify entrance and exit loss factors for each pipe connected to the
reservoir (in this case there is one). You can enter standard losses by
selecting the option buttons at the right. The default selection is the
Custom option with loss factors specified as zero. To later change the
loss factors, click within the pipe table and enter the loss. You can also
specify a depth for the pipe.
The Optional tab allows you to enter different types of optional data.
You can select whether the junction number, name, or both are displayed
on the Workspace. Some junction types also allow you to specify an
initial pressure as well as other junction specific data. The junction icon
graphic can be changed, as can the size of the icon. Design factors can
be entered for most junctions, which are applied to the pressure loss
calculations for the junction in order to give additional safety margin to
the model.
Each junction has a tab for notes, allowing you to enter text describing
the junction or documenting any assumptions.
The highlight feature displays all the required information in the
Specifications window in light blue. The highlight is on by default. You
can toggle the highlight off and on by double-clicking anywhere in the
window or by pressing the F2 key. The highlight feature can also be
turned on or off by selecting it on the Options menu.
¾ Click OK. If Show Object Status is turned on, you should see the J1 ID
number turn black again, telling you that J1 is now completely defined.

(si) g.doc
Figure 2.7 The Reservoir Specifications window for J1
The Inspection feature

You can check the input parameters for J1 quickly, in read-only fashion,
by using the Inspection feature. Position the mouse pointer on J1 and
hold down the right mouse button. An information box appears, as
shown in Figure 2.8.
Inspecting is a faster way of examining the input in an object than
opening the Specifications window.
8/10/2007 4:10:28 PM Page 31

Inspection
window
Figure 2.8 Inspecting from the Workspace with right mouse

button
B. Define Branch J2
¾ Open the J2 Branch Specifications window. In this window, three

connecting pipes will be displayed in the pipe table area. Branches are
connector points for up to twenty-five pipes.
¾ Enter an elevation of 100 meters for the J2 Branch junction (an

elevation must be defined for all junctions). You can also specify a
transient flow source or sink at the junction.
¾ Click the Optional tab and enter an imposed flow rate of -2 m3/sec (the
negative sign means that the flow is out of the junction – a flow sink).
¾ Click OK to accept the input and exit the window.

(si) g.doc
C. Define Surge Tank J3
¾ Open the J3 Surge Tank Specifications window and enter an elevation

of 150 meters. Enter a tank area of (a constant) 5 square meters. This is
the cross-sectional area of the surge tank, which allows AFT Impulse to
determine how the surface elevation changes with a given amount of
flow into the tank. Leave the tank height blank. This is the physical
height of the tank. If a height is entered, AFT Impulse will not allow the
liquid height to exceed the tank height (which results in spillage over the
tank edge). Figure 2.9 shows the Surge Tank window.
¾ Click the check box to Model Short Connector Pipe. This allows you
to model a short pipe which is lumped together with the surge tank for
each solution. The data for the connector pipe is as follows: Friction
factor is 0.02, Pipe Diameter is 0.5 meters, Pipe Area is 0.1963 square
meters (it is cylindrical), Pipe Length is 30 meters, and Elevation
Change is 30 meters (i.e., the pipe is vertical because the elevation
change is the same as the length).
You also can model a flow restrictor (e.g., orifice) if one exists.
Figure 2.9 The Surge Tank specifications window for J3
8/10/2007 4:10:28 PM Page 33

D. Define Reservoir J4 and Branches J5 and J6
¾ Open the J4 Reservoir Specifications window. Enter a surface

elevation of 175 meters and a depth of 25 meters (on the Pipe Depth…
tab). Leave surface pressure as atmospheric.
¾ Open the J5 Branch Specifications window. Enter an elevation of 100

meters and an imposed flow of -1 m3/sec (on the Optional tab). This is
an outflow type of branch similar to J2.
¾ Open the J6 Branch Specifications window. Enter an elevation of 50

meters. There is no imposed flow at this junction.
E. Define Valve J7
¾ Open the J7 Valve Specifications window. Valve junctions connect

with two pipes if they are internal and one pipe if they are exit valves.
This will be an exit valve. Enter an elevation of 25 meters and select the
Exit Valve option at the lower left. Choose Head and enter an exit
pressure of 25 meters, which is equivalent to 1 atm (because exit
atmospheric head is equal to elevation).
¾ Choose the loss model as Cv. This valve will initiate the transient for
this system. Until the valve changes position, the entire pipe system is in
a steady-state condition. Enter a Cv of 2392.6. This is the steady-state
value.
¾ Select the Transient Data tab. In this transient the valve partially
closes over a period of 10 seconds, then holds steady at the new position
for the remainder of the simulation. Enter the following Cv vs. time data
in the table (see Figure 2.10):

(si) g.doc
Figure 2.10 Valve Specifications for J7 initiates the transient for

the system
Time Cv
0 2392.6
10 797.4
100 797.4
F. Define Valve J8
¾ Open the J8 Valve Specifications window. For this example, a regular

valve junction will be used to model the behavior of a relief valve. This
relief valve will vent to an external source, so it should be modeled as an
Exit Valve, with an exit pressure of 1 atm. Enter an elevation of 50
meters.
¾ Choose the loss model as Cv and enter a Cv value of 651.2. This

represents the valve's Cv after it fully opens. The valve remains closed
8/10/2007 4:10:28 PM Page 35

until the cracking pressure is exceeded (defined on the Transient Data
tab), at which point the valve opens to relieve pressure.
¾ Click the Transient Data tab and select the Single Event option for the
Initiation of Transient. Select Pressure Stagnation at Pipe as the Event
Type, Greater Than or Equal To as the Condition, and a Value of 1.688
MPa at the outlet of Pipe 8. This Event represents the cracking pressure
required initiate the opening transient of the relief valve.
Enter the valve transient Cv data below in the Transient Data table (see
Figure 2.11a):
Time Cv
0 0
3 651.2
63 0
100 0
The data represents the valve initially closed at time zero. Time zero is
the moment of cracking. The valve then opens and gradually closes
again over a period of 60 seconds.
Because the valve is closed initially, set the valve Special Condition on
the Optional Data tab to Closed (see Figure 2.11b).

(si) g.doc
Figure 2.11a Valve Specifications for J8 initiates the transient for

the valve
G. Define Pipe P1
The next step is to specify all the pipes. To open the Pipe Specifications
window, double-click the pipe object on the Workspace.
¾ First open the Pipe Specifications window for Pipe P1 (Figure 2.12).
Choose the Pipe Material as Unspecified, choose the User Specified
Wavespeed option, and select the Friction Model Data Set as
Unspecified with the Friction Model as "Explicit Friction Factor". Enter
a length of 1001.2 meters, a diameter of 1.5 meters, and a friction factor
of 0.012.
8/10/2007 4:10:28 PM Page 37

Figure 2.11b Valve Specifications for J8 has a Special Condition of

Closed.
Note: This example from Karney assumes all friction factors are
known ahead of time. Normally this will not be the case. In most cases
you will access the roughness for the pipe from the pipe material
database supplied with AFT Impulse, or input your own roughness
value. Then AFT Impulse will calculate the friction factor using standard
methods (see Chapter 8).
The wavespeed is a very important parameter in waterhammer analysis.

The wavespeed can be calculated with reasonable accuracy from fluid
and pipe data, or it may be available from test data or industry
publications. In this example the wavespeed is known to be 996.3 meters
per second. Therefore, the wavespeed can be entered directly. This is
why we selected the User Specified Wavespeed option. If the wavespeed

(si) g.doc
were not known (which is typical), then the Calculated Wavespeed
option would be preferred. In this case, data would be required for pipe
wall thickness, modulus of elasticity, Poisson Ratio, and pipe support
details. Data for pipe wall thickness, modulus of elasticity, Poisson Ratio
are built into the pipe material databases supplied with AFT Impulse.
The Pipe Specifications window

The Pipe Specifications window offers control over all important flow
system parameters related to pipes.
The Inspect feature also works within the Pipe Specifications window.
To Inspect a connected junction, position the mouse pointer on the
connected junction's ID number and hold down the right mouse button.
This is helpful when you want to quickly check the properties of
connecting objects. (You can also use this feature in junction
Specifications windows for checking connected pipe properties.)
By double-clicking the connected junction number, you can jump
directly to the junction's Specifications window. Or you can click the
Jump button to jump to any other part of your model.
Figure 2.12 Pipe Specifications window for P1
8/10/2007 4:10:28 PM Page 39

H. Define Pipes P2 through P7
¾ Open the Specifications window for each of the other pipes and enter
the following data (pipe 1 data was just entered so ignore it):
Pipe Length Diameter Friction Wavespeed
(meters) (meters) Factor (meters/sec)
1 1001.2 1.5 0.012 996.3
2 2000 1 0.013 995.3
3 2000 0.75 0.014 995
4 502.5 0.5 0.015 1000
5 502.5 0.5 0.015 1000
6 1001.2 1 0.014 996.3
7 2000.2 0.75 0.013 995.1
8 100.12 1 0.014 996.3
After entering the data for all the pipes, the fourth checklist item should
be completed. If it is not, see if the Show Object Status is on. If not,
select Show Object Status from the View menu or toolbar. If the fourth
checklist item is not completed at this point, see if any of the pipes or
junctions have their number displayed in red. If so, you did not enter all
the data for that item.
¾ Before continuing the model, save it to file one more time. It is also a
good idea to review the input using the Model Data window.
Reviewing input in Model Data window

The Model Data window is shown in Figure 2.13. To change to this
window, you can select it from the Window menu, the Toolbar, pressing
Ctrl-M, clicking anywhere in the Model Data window if it has been
restored or, if minimized, clicking on the minimized window at the
bottom of the screen and restoring it. The Model Data window gives you
a text-based perspective of your model. Selections can be copied to the
clipboard and transferred into other Windows programs, saved to a
formatted file, printed to an Adobe PDF, or printed out for review.
Figure 2.14 shows an expanded view of the Transient Data tab from
Figure 2.13. Here all transient input data for the model is shown.

(si) g.doc
The Model Data window allows access to all Specifications windows by
double-clicking the appropriate ID number in the far left column of the
table. You may want to try this right now.
Customize view
General data
Pipe data
Junction data
Figure 2.13 The Model Data window shows the input data in text
form
Step 6. Complete the last two checklist requirements
A. Specify pipe sectioning

The fifth checklist item is Section Pipes. This window cannot be opened
until sufficient data is entered previously. First, the fluid must be
selected (we did this Step 4C when we entered System Properties). And
second, all pipes must have a length and wavespeed entered. After this
data has been entered, the pipes can be sectioned. The Section Pipes
8/10/2007 4:10:28 PM Page 41

window helps divide the pipes into computation sections in a manner
which is consistent with the Method of Characteristics (MOC).
Figure 2.14 The Transient Data tab in the Model Data Junction
data area shows all transient data entered
¾ Select Section Pipes on the Analysis menu to display the Section Pipes
Window (Figure 2.15). For this model the controlling pipe is P8. This is
the pipe with the shortest end-to-end communication time (i.e., L/a – the
length divided by the wavespeed). To satisfy the MOC, the following
equation must be applied:
Li
ni =
ai Δt
where n is the number of sections in pipe i, L is the length, and a is the
wavespeed. The Δt is the time step. Since all pipes in the network must
be solved together, the same time step must be used for each pipe. With
a given length and wavespeed for each pipe, it can be seen from the
above equation that it is unlikely that the number of required sections, n,
for each pipe will be a whole number.

(si) g.doc
To address this situation, it is helpful to recognize that the wavespeed, a,
is the least certain input parameter. It is therefore acceptable to allow up
to a 15% uncertainty in wavespeed. By adjusting the wavespeed for each
pipe within this tolerance the sectioning can be made to come out as
whole numbers for each pipe. The Section Pipes window automates this
process by searching for sectioning which satisfies the required
tolerance.
You can specify the tolerance on wavespeed by entering it into the Max.
Percentage Error. The minimum and maximum allowable sections in the
controlling pipe narrows the search space. Also, the Percentage
Increment directs the routine in how fine to search the search space.
This model has been specified by Karney in such a way that the pipe
sectioning falls closely along whole numbers.
Figure 2.15 The Section Pipes window automates the sectioning

process and calculates the time step.
¾ For Min. and Max. Sections in Controlling Pipe enter 1 section for
each. Limit the Max Error to 1% for the search (usually 5 or 10% is
8/10/2007 4:10:28 PM Page 43

fine). Check the box for Sort Sectioning by Minimum Error. Then click
the Search button.
A list of possible sectioning is displayed. Click the top line with 1.0000
sections in the controlling pipe. The resulting time step will be displayed
as 0.100492 seconds.
The sectioning and resulting errors in the remaining pipes are displayed
in the table near the bottom. Click OK and the fifth checklist item should
be completed.
Note: The error in the Section Pipes window relates only to sectioning
roundoff and not to overall model accuracy.
B. Specify transient control

The sixth and final checklist item is Transient Control. This window
allows you to specify the time at which the transient starts and ends, as
well as how much data to include in the output file.
¾ Select Transient Control on the Analysis menu to display the Transient

Control window (Figure 2.16 top). Enter zero for Start Time and 65 for
Stop Time.
The Transient Control window lets you enable or disable transient
cavitation modeling. It also offers control over how AFT Impulse should
respond to artificial transients. Artificial transients are a problem that
can sometimes occur when steady-state and transient conditions are
inconsistent.
Data can be saved for all pipe stations or only selected stations. The
selected stations are shown on the Pipe Station Output tab (see Figure
2.16 bottom). With the list next to the “Change All Pipes To” button set
to “All Stations”, click the “Change All Pipes To” button. This will save
all pipe station data for all pipes, which will be useful later for animation
purposes.

(si) g.doc
Figure 2.16 The Transient Control window offers features to

specify the time span for the transient and what
output data is written
8/10/2007 4:10:28 PM Page 45

At the bottom of the window the projected output file size is shown. You
should pay attention to this number, as the output file size can grow very
large. In this case the output file will be 1.0MB. If the output file does
become excessively large, you will want to limit the number of time
steps and pipe output written to disk. This is discussed in Chapter 5.
¾ Click OK to accept the current settings. The last checklist item should
be completed. The model is ready to be solved.
Step 7. Run the Solver
¾ Choose Run from the Analysis menu or click the arrow icon on the
toolbar. During execution, the Solution Progress window displays
(Figure 2.17). You can use this window to pause or cancel the Solver's
activity.
The two solvers

AFT Impulse has two Solvers. The first is called the Steady-State Solver,
which as its name suggests obtains a steady-state solution to the pipe
network. The second Solver is called the Transient Solver. This solves
the waterhammer equations.
Before a transient simulation can be initiated, the initial conditions are
required. These initial conditions are the steady-state solution to the
system. After the steady-state solution is obtained by the Steady-State
Solver, AFT Impulse uses the results to automatically initialize the
Transient Solver and then run it.

(si) g.doc
Figure 2.17 The Solution Progress window displays the state of

the simulation
¾ When the solution is complete, click the View Output button to

display the text-based Output window. The information in the Output
window can be reviewed visually on the screen, saved to file, exported
to a spreadsheet-ready format, copied to the clipboard, printed to an
Adobe PDF file, and printed out on the printer.
The transient output file

When the Transient Solver runs, the transient output data is written to a
file. This file is given the same name as the model itself with a number
appended to the name, and with an OUT extension appended to the end.
For all transient data processing, graphing, etc., the data is extracted
from this file. The number is appended because AFT Impulse allows the
user to build different scenarios all within this model. Each scenario will
have its own output file, thus the files need to be distinguishable from
each other.
8/10/2007 4:10:28 PM Page 47

The output file will remain on disk until it is erased by the user or the
input model is modified. This means that if you were to close your
model right now and then reopen it, you could proceed directly to the
Output window for data review without rerunning your model.
Step 8. Review the output

The Output window (Figure 2.18) is similar in structure to the Model
Data window. Three areas are shown, and you can enlarge each area by
selecting the options from the Show list box on the Toolbar or from the
View menu. The items displayed in the tables are those items you chose
in the Output Control window (checklist item # 2).
The Output window allows you to review both the steady-state and
transient results. A summary of the maximum and minimum transient
results for each pipe is given on the Transient Max/Min tab in the pipe
area (Figure 2.19). You can also review the solutions for each time step
(i.e., a time history) for which data was written to file. These two data
sets are located on the Transient Output tab and Transient Max/Min tab
in the pipe area of the Output window (see Figures 2.17 and 2.18).

(si) g.doc
Customize view
General results
Pipe results
Junction results
Figure 2.18 The Output window displays output results in text

form
8/10/2007 4:10:28 PM Page 49

Figure 2.19 The Transient Output and Transient Max/Min results

tabs are in the Output window Pipe section area.
These show transient results for pipes.

(si) g.doc
A. Modify the output format

If you selected the default AFT Impulse Output Control, the Pipes table
of steady-state results (the tab on the far left in the pipe area) will show
volumetric flow rate in the second column with units of m3/sec (cubic
meters per second), as shown in Figure 2.18.
¾ Select Output Control from the Analysis menu or Toolbar. On the

right side of the Pipe tab is the list of selected output parameters. Click
Volumetric Flow Rate and change the units by selecting liter/sec from
the list at the bottom.
¾ Click OK to display changes to the current results. You should see

the volumetric flow rate results, still in the second column, in units of
liter/sec. Notice the Velocity results in the third column.
¾ Select Output Control from the Analysis menu one more time. Select
the Pipe tab. The Reorder scroll bar on the far right allows you to reorder
parameters in the list.
¾ Select the Velocity parameter and use the Reorder scroll bar to move it
up to the top of the parameter list.
¾ Click OK to display the changes to the current results. You will see
in the Pipes table that the first column now contains velocity and the
third column contains the volumetric flow rate. The Output Control
window allows you to obtain the parameters, units and order you prefer
in your output. This flexibility will help you work with AFT Impulse in
the way that is most meaningful to you, reducing the possibility of
errors.
¾ Lastly, double-click the column header Velocity in the Output

window Pipes Table. This will open a window in which you can change
the units once again if you prefer. These changes are extended to the
Output Control parameter data you have previously set.
8/10/2007 4:10:28 PM Page 51

B. Graph the results
¾ Change to the Graph Results window by choosing it from the

Windows menu, Toolbar, pressing Ctrl-G, clicking anywhere on the
window if it has been restored or, if minimized, clicking on the
minimized window at the bottom of the Workspace and restoring it. The
Graph Results window offers full-featured Windows plot preparation.
¾ Choose Select Graph Data from the View menu or the Toolbar to
open the Select Graph Data window (see Figure 2.20). From the All Pipe
Stations… list choose pipe 7. Select the Outlet of pipe 7. This is the
station connected to the exit valve. A graph of this station shows the
pressure and flow histories at the valve.
Figure 2.20 The Select Graph Data window controls the Graph
Results content
¾ Click the Add >> button to add this station to the list on the right.

(si) g.doc
¾ From Graph Parameters select Hydraulic Gradeline (HGL) and
choose meters as the Y-Axis units. Then click the Show button to
display the graph (Figure 2.21).
This pressure/head history can be compared to that published in Karney
(1992). This is done in the Verifications documentation provided with
AFT Impulse.
Figure 2.21 The Graph Results window offers full-featured plot

generation.
C. Animate the Results

The animation features in AFT Impulse can greatly aid the
understanding of complex transient behavior. Open the Select Graph
Data window from the View menu or toolbar again. Select the “Profile
Along a Flow Path” tab. There select pipes 1, 6 and 7 from the Pipes List
(this is the sequence of pipes from the Supply Tank A at J1 to the
discharge valve at J7), select HGL from the Graph Parameters List, and
select Animate From Output File option in the Parameter Values area
(see Figure 2.22). Select the X-Axis Units and HGL Units as meters.
Click the Show button.
8/10/2007 4:10:28 PM Page 53

Figure 2.22 The Select Graph Data window allows selection of

animation parameters.
Additional animation control features will be displayed on the Graph

Results window (see Figure 2.23). Push the Play button and AFT
Impulse will run through all of the results in pipes 1, 6 and 7 for the
entire transient which animates the results. The animation can be paused
or stopped, and the time can be reset to any desired time using the time
slider. If the animation is too fast, slow it down using the animation
speed control at the right.
You can use the other buttons in the Graph Results window to change
the graph appearance and to save and import data for cross-plotting. The
Graph Results window can be printed, saved to file, copied to the
clipboard, or printed to an Adobe PDF file. The graph’s x-y data can be
exported to file or copied to the clipboard.

(si) g.doc
Figure 2.23 The Graph Results window allows animation of output

to be displayed.
D. View the Visual Report
¾ Change to the Visual Report window by choosing it from the Window

menu, Toolbar, pressing Ctrl-I, clicking anywhere on the window if it
has been restored or, if minimized, clicking on the minimized window at
the bottom of the Workspace and restoring it. This window allows you to
integrate your text results with the graphic layout of your pipe network.
8/10/2007 4:10:28 PM Page 55

Figure 2.24 The Visual Report Control window selects content for
the Visual Report window
¾ Click the Visual Report Control button on the Toolbar (or View
menu) and open the Visual Report Control window, shown in Figure
2.24. Default parameters are already selected, but you can modify these
as desired. For now, select Max Pressure Stagnation and Min Pressure
Stagnation in the Pipe Results area. Click the Show button. The Visual
Report window graphic is generated (see Figure 2.25).
It is common for the text in the Visual Report window to overlap when
first generated. You can change this by selecting smaller fonts or by
dragging the text to a new area to increase clarity (this has already been
done in Figure 2.25, as has the selection to show units in a legend). This
window can be printed, copied to the clipboard for import into other
Windows graphics programs, saved to file, or printed to an Adobe PDF
file.

(si) g.doc
Figure 2.25 The Visual Report window integrates results with the
model layout
Conclusion
You have now used AFT Impulse's five primary windows to build a
simple model. Review the rest of this User’s Guide for more detailed
information on each of the windows and functions.
8/10/2007 4:10:28 PM Page 57


(si) g.doc
CHAPTER 3
Fundamental Concepts
This chapter explains the fundamental concepts and conventions AFT

Impulse employs in processing pipe system analysis data, as well as how
to interpret the information AFT Impulse processes.
Pipes and junctions

AFT Impulse employs two fundamental pipe system constructs: pipes
and junctions.
Pipes are conduits for one-dimensional liquid flow. Junctions are
connector points for pipes and are elements at which flow balances are
made and where transients are initiated. Some junction types can only
connect to one pipe; others can connect with up to twenty-five. AFT
Impulse provides a total of nineteen junction types.
In addition to balancing flow, junctions also influence the flow or
pressure behavior of the system. For example, a reservoir junction
applies a constant pressure at a location, and the flow there is free to
adjust in whatever manner is consistent with the governing equations. In
contrast, an assigned flow junction applies a known flow rate at its
location, allowing the pressure to adjust to that level dictated by the
governing equations.
The nineteen junction types allow you to specify special kinds of
transient behavior, irrecoverable pressure losses or fluid behavior.
Among the nineteen junctions there is some overlap in features.

(si) g.doc
Junctions communicate with each other through the pipes connecting
them. Each pipe must be connected to two junctions. There are no
exceptions to this rule.
A pipe differs from a junction in that it has a reference positive flow
direction. To say a pipe has a flow rate of 1 lbm/sec is meaningless
unless the flow direction is specified.
In general, you do not need to specify the actual flow direction in a pipe,
because AFT Impulse sorts out the true physical flow directions of the
system you define. For example, if you specify the reference positive
flow direction from left to right, and the solution is in fact the opposite
direction, then that information will be presented in the output as a
negative flow rate. Inlet and outlet conditions correspond to the
reference positive flow direction, not the actual flow direction.
However, it is important to specify the proper flow direction for pipes
which connect to junctions such as pumps or control valves. AFT
Impulse uses the flow direction of the pipes to determine in which
direction to increase or reduce pressure.
Convention for flow entering and exiting

Many of the junction types permit fluid flow to come into or pass out of
the system in a prescribed manner. The most obvious of these is the
reservoir junction. Reservoirs maintain a constant pressure head at a
node, and the flow adjusts according to the governing equations. While
in principle the fluid is stored or removed from the reservoir, AFT
Impulse considers the fluid as coming into or going out of the system.
The stored reservoir fluid is not part of the system.
The convention for defining inward or outward flow for the system or
any AFT Impulse object is that flow in is positive (that is, the object
gains the flow) and flow out is negative (the object loses the flow).
Features for modeling irrecoverable losses

AFT Impulse works almost exclusively with the concept of the loss
factor, also known as K factor or resistance factor. No provision has
been made for the equivalent length approach to solving pressure drop
problems. In some Specifications windows, other popular methods for

(si) g.doc
Chapter 3 Fundamental Concepts 61
specifying loss information are provided where they are consistent with
the loss factor method. For example, you will find Cv for valves.
Incorporating equivalent length data

The equivalent length method is a popular method for modeling losses in
fittings and equipment. There are two ways you can incorporate
equivalent loss data into AFT Impulse – one way which is acceptable
and one which isn't.
First we will discuss the unacceptable way. You can account for losses
by adding to the actual pipe length, thus creating a fictitious length. For
steady flow calculations this is acceptable. However, in modeling
waterhammer transients the true length is a critical aspect of reflection of
pressure waves. If the length is artificially increased, this will change the
dynamic characteristics of the pipe and yield incorrect results. Any
approximations which result in artificially changing pipe length are
unacceptable.
The second way to incorporate equivalent length data is to apply a
design factor to the pipe. The design factor is specified on the Pipe
Specifications window Optional tab, and acts as a multiplier on the
standard pipe friction. Thus if the pipe is physically 30 meters long, and
has fittings with 3 meters of equivalent length, a design factor of 1.1 can
be used to account for this. This maintains the true physical length and
thus the dynamic behavior.
Convention for specifying junction base area

Loss factors are area dependent. When specifying a loss factor at a
junction, the base area is of critical importance. By default AFT Impulse
adheres to the following convention: whenever it makes sense
physically, the base area is always the upstream pipe flow area.
However, this can be modified by the user. In most junction
Specifications windows, the base area can be specified as the upstream
pipe, downstream pipe, user specified area or diameter.
However, some junction types, especially those that allow three or more
connecting pipes, specify loss factors based on each connecting pipe
flow area, whether upstream or downstream. An example of this is a wye
in which each of the connecting pipes has a different diameter. The idea
of specifying a loss factor based on upstream flow area loses meaning
because there may be more than one upstream pipe. In these cases the
8/10/2007 4:10:28 PM Page 61

loss factor is referenced to the base area of the actual downstream
pipe(s) where it connects to the junction. The types of junctions that do
not always follow the upstream pipe area convention are Tee/Wye,
Branch, Reservoir, Assigned Pressure, and Assigned Flow. Other
exceptions are Valves and Relief Valves specified as exit valves.
Specifying losses
Local losses can be specified in two ways. The most flexible way of
including a loss factor in an AFT Impulse model is to associate it with a
junction specifically defined for that type of loss.
Secondly, you can include loss factors with pipes. When you specify
Fittings & Losses in the Pipe Specifications window, the loss is assumed
to be distributed evenly along the pipe length, much like friction losses.
For this reason a pipe-associated loss factor is referred to as a distributed
loss. The losses at junctions, on the other hand, are point losses.
When modeling for waterhammer, it is preferred that losses for all static
components be included in pipes as Pipe Fittings & Losses. This reduces
the number of pipes and model run time.
Modeling valve closures with K factors

You can close a valve by specifying the transient K factor as a –1
(negative one).
Specifying frictional losses in pipes

AFT Impulse also offers a flexible approach to selecting friction models
for pipes. For each pipe in the model you may specify an absolute
roughness, a roughness that is relative to the pipe diameter, a Hazen-
Williams factor, a hydraulically smooth roughness, an explicit friction
factor, a pipe resistance, or a frictionless pipe. There are some
restrictions to where frictionless pipes can be located in a model. These
are discussed in Chapter 8.
There are also two special loss models. The MIT equation is applicable
for petroleum. And the Miller turbulent is applicable for light
hydrocarbons.

(si) g.doc
Comment on importance of hydraulic losses during

transients
Generally speaking, hydraulic losses (also called system resistance) from
pipes, fittings or equipment, have a second order effect on waterhammer
transients. The system resistance can be important in determining local
steady-state pressures, and thus initial pressures in the system. It is
worthwhile to account for all system resistances in some manner.
However, component resistances are frequently lumped together for
simplicity. Lumping resistances together also results in models that run
faster.
Static vs. dynamic losses

Pipe system fittings and equipment can be static or dynamic. Whenever
possible, static elements should be lumped together with other elements
or, better, into pipes. Examples of static elements that should be lumped
are elbows, orifices, filters, and valves whose position remains constant
over time. To discourage users from using static elements as junctions in
their model, such junctions are not offered in AFT Impulse. There are no
junction types for elbows, orifices or filters. However, valves can be
either static (i.e., no position changes), or dynamic (i.e., they can open or
close during transient modeling). The user can thus model a static valve
if he/she chooses. However, it is better to only use valves for dynamic
cases, as this will improve model run time.
Static losses (such as elbows, orifices, filters, and valves) can be
included as Fittings & Losses in pipes.
Variable resistance
When experiencing waterhammer transients, the flowrate in a pipe will
vary for each pipe computing station at an instant in time, and all pipe
stations will vary over time. As the flowrate varies, so does the Reynolds
number. This means that, strictly speaking, the pipe friction factor will
change as the flowrate changes, and will also change from station to
station.
This change in friction factor is typically neglected for two reasons.
First, it complicates the computation method. And second, it is usually
not important.
AFT Impulse allows the user to model either constant friction factor or
variable. The variable friction factor model results in model runtimes
8/10/2007 4:10:28 PM Page 63

roughly three times as long. The choice is available in the Transient
Control window in the Use Variable Pipe Resistance field.
Introduction to wavespeed and waterhammer phenomenon
Description of waterhammer
Waterhammer is a phenomenon that occurs in all liquid piping systems
whenever some event disturbs the steady state. Rapid disturbances can
cause large transient pressures that in extreme cases can cause a
catastrophic failure of the piping system and/or damage to associated
equipment.
For more slowly occurring disturbances, the transient pressures may be
small or even unnoticeable. In such cases it is sometimes said that
waterhammer doesn’t exist. What is meant by such statements is that
large, potentially destructive waterhammer doesn’t exist. Regardless of
the magnitude, waterhammer always exists when the liquid velocity in a
piping system is changed. In cases of small velocity changes
waterhammer transients are small and are typically and justifiably
ignored by engineers.
The disturbances consist of coupled pressure and velocity waves that
propagate throughout a pipe system. Across these waves a discontinuity
in magnitude exists. An example is given later in this section to clarify.
To picture what happens during a waterhammer event, consider a 1220
meter long pipeline flowing water in a steady state condition. Any
disturbance in the fluid will propagate near the acoustic speed of the
liquid in both the upstream and downstream directions.
It turns out that the propagation speed is influenced by the piping
material, wall thickness, diameter and structural support method. These
other factors combine into a parameter called the wavespeed, which is
traditionally referred to as a. The wavespeed is thus analogous to the
acoustic speed, but combines the effect of pipe interaction. Although the
wavespeed varies from case to case, a nominal value of 1220 m/sec is
reasonable for water.
Now picture a transient event in this pipe, specifically that the valve is
suddenly closed at the downstream end (see Wylie, et al., 1993, pp. 7).

(si) g.doc
Assume the upstream supply is a reservoir, which acts as a fixed
pressure.
When the valve closes this fact is communicated up the pipe to the
supply, telling it to stop supplying fluid because fluid is no longer
needed. This communication takes a finite time, and this time is in fact
related to the wavespeed. In our system it will take one full second for
the disturbance downstream to be communicated to the upstream end
because the wavespeed is 1220 m/sec and the pipe is 1220 meters long.
During this second, the flow continues because it has not yet received
the message to stop.
Once the message reaches the upstream supply, however, it takes another
full second for the supply’s response to reach the valve. There is thus a
full two second period in which the pipe disturbance communicates
upstream and the response comes back.
Instantaneous waterhammer
With few exceptions, it is safe to calculate the maximum possible
waterhammer pressure surge by using the instantaneous waterhammer
equation. The instantaneous waterhammer equation assumes that the
transient event occurs either instantaneously or rapidly enough such that
it is in effect instantaneous.
In such a case, it can be shown (Wylie, et al., 1993, pp. 4) by use of the
mass and momentum equation that the pressure transient is given by the
following equation:
ΔP = − ρ a ΔV (3.1)
where:
ΔP = pressure surge
ρ = density
a = wavespeed
ΔV = velocity change
By adding the pressure surge to the existing static pressure, one can
obtain the maximum theoretical pressure in the pipe. However, in some
cases the pressure can exceed the instantaneous prediction. These are
discussed in Chapter 9.
8/10/2007 4:10:28 PM Page 65

Here’s an example. Assume flow in a pipe filled with water is
instantaneously stopped by a valve closure. Assume the density is 1000
kg/m3, the wavespeed is 1220 meters per second, and the initial velocity
is 3 meters per second. Also assume that before the transient the static
pressure is 350 kPa.
kg m m
Pmax = 1000 3
* 1220 * 3 + 350 kPa
m s s
Pmax = 3660 + 350
Pmax = 4010 kPa
Wavespeed
When a transient event is initiated in a pipe system, the remainder of the
system must adjust to the new conditions. In order to adjust, the
existence of the event must be communicated to the rest of the system.
This communication takes place at the wavespeed of the fluid. The
wavespeed is somewhat analogous to the sonic speed of the liquid.
However, the wavespeed is affected by the pipe structure.
See Chapter 9 for a mathematical description.
Communication time in pipes

Transient events communicate their effect through the pipe at the
wavespeed of the pipe and fluid. When transient events occur, they are
communicated upstream (or downstream) so the system can adjust to the
new conditions. No response to the changing conditions can be obtained
until the communication wave travels to the other end of the pipe and
back. This communication time is thus given by the following:
L
Δt = 2 (3.2)
a
where L is the pipe length and a is the wavespeed.
Conceptual example
Let's consider a simple example of a valve closure. A frictionless pipe is
connected to an upstream reservoir and a downstream valve. Fluid is

(si) g.doc
flowing at a steady state. At time zero the valve is instantaneously
closed. We will look at four phases of the transient response.
Phase A: 0 < t < L/a

The first phase occurs after the valve has closed but before the resulting
wave has traveled the pipe length and reached the upstream reservoir
(see Figure 3.1). The wave, moving at speed a, travels from the valve to
the reservoir (see Figure 3.1, Phase A – the wave initially travels from
right to left). The fluid upstream of the wave front (on the left) does not
yet know of the closed valve, and thus remains at the steady-state
pressure and velocity. Thus before the wave reaches the inlet, fluid
continues to flow into the pipe at velocity Vsteady even though the valve is
closed.
The portion of the pipe to the right, where the wave has already passed,
knows of the closed valve and thus the motion has stopped (the velocity
is zero). The kinetic energy of the fluid has been converted into potential
energy by increasing the fluid pressure and expanding the pipe
circumference. Figure 3.1, Phase A, shows the pipe expanded to the right
(exaggerated for effect), but at the original diameter on the left. It further
shows a plot of the pressure and velocity in the pipe.
Since the valve was closed instantaneously, the magnitude of the
pressure wave can be obtained from the instantaneous waterhammer
equation (Equation 3.1).
Phase B: L/a < t < 2L/a

The second phase occurs after the wave has reflected from the inlet
reservoir but before it has come back to the valve. The wave, still
moving at speed a, travels from the reservoir to the valve (see Figure 3.1,
Phase B). The fluid upstream of the wave front (on the left) now is
aware of the closed valve. Further, the high pressure that built up during
Phase A means there is excess fluid in the pipe. During Phase B this
fluid is expelled to the only place that can receive it – the upstream
reservoir. Thus backflow occurs as the pressure returns to the steady-
state level.
The fluid downstream of the wave front (on the right) does not yet know
the reservoir is available to relieve the pressure, so it remains at the high
pressure and zero velocity.
8/10/2007 4:10:28 PM Page 67

V = Vsteady V = Vsteady
V=0 V=0
a a
P P
Psteady ΔPinstantaneous Psteady ΔPinstantaneous
x x
V V
Vsteady
x x
-Vsteady
Phase A: 0 < t < L / a Phase B: L / a < t < 2L / a
V = Vsteady V = Vsteady
V=0 V=0
a a
P P
Psteady Psteady
ΔPinstantaneous ΔPinstantaneous
x x
V V
Vsteady
x x
-Vsteady
Phase C: 2L / a < t < 3L / a Phase D: 3L / a < t < 4L / a
Figure 3.1 Pressure and velocity in a pipe after instantaneous

valve closure. Pipe diameter changes with pressure
exaggerated for effect

(si) g.doc
Phase C: 2L/a < t < 3L/a
The third phase occurs after the wave has reflected from the inlet
reservoir, traveled back to the valve and reflected from the valve, and is
now moving back towards the reservoir (see Figure 3.1, Phase C). The
wave is still moving at speed a.
Due to the pipe's attempt to establish equilibrium by flowing back
towards the reservoir in Phase B to relieve the high pressure, the system
has overcompensated. During Phase C the pressure drops below the
steady value. The pressure drop is the same magnitude as the pressure
rise in Phase A, but opposite in sign.
As the wave passes each location during Phase C, the fluid motion stops
and the pressure is reduced.
Phase D: 3L/a < t < 4L/a

The fourth phase occurs after the wave has reflected from the inlet
reservoir for the second time and is moving back towards the valve (see
Figure 3.1, Phase D). The fluid upstream of the wave front (on the left)
has returned to the original steady values of pressure and velocity. The
fluid downstream of the wave front (on the right) is still at a reduced
pressure.
The fourth phase occurs after the wave has reflected from the inlet
reservoir for the second time , traveled back to the valve and reflected
from the valve, and is now moving back towards the reservoir (see
Figure 3.1, Phase C). The wave is still moving at speed a.
Once the wave reaches the valve for the second time, the four phases are
repeated. If the pipe is frictionless, the wave motion will continue
indefinitely. If friction effects are included the wave will decay over
time and motion would eventually stop altogether.
8/10/2007 4:10:28 PM Page 69


(si) g.doc
CHAPTER 4
The Five Primary Windows
This chapter explores the features found in AFT Impulse's five primary
windows and explains the role each plays in the analysis process.
Overview
AFT Impulse has five primary windows. The primary windows are
subordinate to the AFT Impulse window and can be maximized or
minimized within the boundaries of the AFT Impulse window. The
primary windows are permanent in that you work in one of these
windows at all times.
The primary windows work together to provide tools for entering model
input, analyzing results for accuracy, and preparing results for
documentation. Figure 4.1 summarizes the workflow using the primary
windows.
• The Workspace window allows you to build the model visually and
see the model layout.
• The Model Data window is the text-based complement to the
Workspace window. The Model Data window shows input data in
text form. This window works hand-in-hand with the Workspace
window to provide exceptional flexibility in manipulating data.
• The Output window displays the results of the analysis in text form
and lets you produce attractive, effectively organized printed output.
• The Visual Report window merges the output data with the pipe
system layout from the Workspace to present a unique perspective of

(si) g.doc
the results. This window also allows customized layout of the
information for documentation and presentation purposes. An
alternate use for the Visual Report window is to show textual input
data combined with the input schematic.
• The Graph Results window provides full-featured Windows plotting
capability. Here results can be viewed in a variety of ways for
evaluating system performance or identifying important trends. The
Graph Results window also allows graph customization.
You can change between the five primary windows by selecting them
from the Window menu, the Toolbar, or by clicking the appropriate
icons within the AFT Impulse window.
Model Data
Visual Report
Workspace Output
Graph Results
Figure 4.1 Primary window workflow in AFT Impulse
The Workspace window

The Workspace window is the sole area where you specify the model
layout and connectivity. The Workspace window has two separate
functional areas: the Toolbox on the far left and the Workspace itself
encompassing the rest of the screen (see Figure 4.2).
At the top of the Toolbox (Figure 4.3), a Shortcut button and four tool
icons are slightly offset from the rest of the junction icons. These are the
Selection Drawing Tool, Pipe Drawing Tool, Zoom Select Tool, and the
Annotation Tool.

(si) g.doc
Chapter 4 The Five Primary Windows 73
The Shortcut Button

The Shortcut button pops up an information window describing different
keyboard shortcuts available on the Workspace and Specifications
windows. This information is also available by clicking Keyboard
Shortcuts on the Help menu, and is described with more detail in
Appendix A. The Shortcut button can be hidden using features on the
Toolbox Preferences window.
The Selection Drawing Tool

The drawing tool in the upper left corner (and below the Shortcut
button) of the Toolbox is the Selection Drawing Tool (see Figure 4.3).
This tool is one means of selecting groups of objects on the Workspace
for group operations such as dragging or duplication.
To select a group of objects, click the Selection Drawing Tool and draw
a box around the desired objects on the Workspace.
If you Double-click the Selection Drawing Tool it remains active until
you click it again a single time. This allows you to make a series of
selections.
Three keyboard modifiers extend the Selection Drawing Tool's
functionality. If you hold down the CTRL key when completing the
selection drawing (just before releasing the mouse button), the Selection
Drawing Tool remains active. The CTRL key thus performs the same
function as double-clicking the tool.
If you first double-click the Selection Drawing Tool or hold down the
CTRL key, you can also hold down the SHIFT key while drawing the
selection box and the enclosed pipes and junctions will be deselected.
Finally, if you first double-click the Selection Drawing Tool or hold
down the CTRL key, you can also hold down the ALT key while
drawing the selection box and the enclosed pipes and junctions will
toggle. That is, any pipes or junctions that were selected will become
deselected, and vice versa.
Selecting objects completely or partially inside the box

If you select the objects by dragging left to right, you will select all
objects completely inside the box.
8/10/2007 4:10:28 PM Page 73

If you select the objects by dragging right to left, you will select all
objects completely or partially inside the box (if this feature is enabled
in the Workspace Preferences window).
Toolbars
Workspace
Toolbox
Minimized
primary
windows
Status Bar
Figure 4.2 The Workspace window is where you assemble your

model
The Pipe Drawing Tool

The tool in the upper right corner (and below the Shortcut button) of the
Toolbox is the Pipe Drawing Tool. With this tool you can draw new
pipes on the Workspace, starting and ending at any desired location.
To draw a pipe, click the Pipe Drawing Tool. The mouse pointer will
change to a crosshair when you move it over the Workspace. Position
the cursor at the pipe’s starting point, hold down the left mouse button
while dragging the cursor to the pipe endpoint, then release the mouse
button.

(si) g.doc
Two keyboard modifiers extend the Pipe Drawing Tool's functionality. If
you hold down the SHIFT key down while drawing on the Workspace,
you can draw pipes that are perfectly horizontal or vertical. This may
improve the aesthetics of your model.
If you hold down the CTRL key when completing the pipe drawing (just
before releasing the mouse button), the Pipe Drawing Tool remains
active, and you can draw a series of pipes without returning to the
Toolbox each time.
Shortcut Button
Selection Drawing Tool Pipe Drawing Tool
Zoom Select Tool Annotation Tool
Branch Reservoir
Assigned Flow Assigned Pressure
Area Change Relief Valve
Tee/Wye Valve
Control Valve Check Valve
Dead End Gas Accumulator
Surge Tank Liquid Accumulator
Vacuum Breaker Valve Spray Discharge
Pump Volume Balance
General Component
Figure 4.3 Tools on the Workspace window Toolbox
If you double-click the Pipe Drawing Tool it remains active until you
click it again a single time. This allows you to draw a series of pipes
without returning to the Toolbox each time.
8/10/2007 4:10:28 PM Page 75

Pipe handles and segmenting a pipe
When a single pipe is selected, two black squares (or more if the pipe is
segmented), called handles, appear. When the cursor is placed over a
pipe handle it changes to a crosshair and the pipe end can be moved or
stretched by dragging with the left mouse button.
A pipe can be segmented or “bent” in up to ten segments (see Figure
4.4). This does not affect the model or the results, only the visual
representation. It does not add an “elbow” to the model or the resulting
pressure drop effects of an elbow.
Figure 4.4 Pipe with five segments
This feature is useful in graphically bending the pipe around another part
of the model or graphically showing that there are embedded elbows as
Fittings & Losses.
You can add a segment to a pipe in two ways. The easiest way to add a
single segment is to select Pipe Segments and Add from the Arrange
menu. If there are no segments, then a single segment can also be added
by selecting the Add/Remove Segment from the Toolbar. If the pipe

(si) g.doc
already has one or more segments, this toolbar selection will remove all
segments.
A more flexible way of adding or removing segments is to use the Pipe
Segment Manager. To open the Pipe Segment Manager, choose Pipe
Segments and Manager from the Arrange menu. A window will display
(Figure 4.5) where you can add, remove and merge segments.
Figure 4.5 The Pipe Segments window allows you to create,

merge and remove pipe segments.
The Zoom Select Tool

The Toolbox tool below the Selection Drawing Tool is the Zoom Select
Tool. With this tool you can draw a box that will cause the Workspace to
zoom to that box. The maximum zoom size is 200%.
8/10/2007 4:10:28 PM Page 77

One keyboard modifier alters the Zoom Select Tool's behavior. If you
hold down the SHIFT key while selecting the zoom area, the zoom state
will always be 100%.
The Annotation Tool

The Toolbox tool below the Pipe Drawing Tool is the Annotation Tool.
With this tool you can draw a box that will contain an Annotation
displayed on the Workspace.
As shown in Figure 4.6, the Annotation Tool allows you to create a text
message and, by clicking on the appropriate tab, show an outline around
the message and draw lines with pointers to certain areas of the model.
Different style, color and font options are provided.
Figure 4.6 The Workspace Annotation tool dialog box.
Annotations can be “attached” to the Workspace, a pipe or a junction.

When attached to the Workspace, the annotation always remains in the
same location until you move it. When attached to a pipe or junction, the
annotation will move when the pipe or junction is moved. The Scenario
Options tab allows you to specify which scenarios in which the
annotation appears.

(si) g.doc
Annotations print out with the model. Figure 4.7 shows the Workspace
with an Annotation displayed.
The Annotation Manager, available on the View menu, allows you to
specify whether an annotation will be shown or hidden, and in which
scenarios it should appear.
Figure 4.7 Annotation added to the workspace
Junction icons
Below the four drawing tools are nineteen junction icons. The junctions
allow you to model a large variety of pipe system components important
in waterhammer. To add a junction to a model, drag the desired icon
from the Toolbox and drop it anywhere on the Workspace.
As you move the mouse pointer over the Toolbox, a Tool Tip identifies
the type of junction under the mouse pointer.
The types and order of junction icons on the Toolbox can be modified
through the Toolbox Preferences window. For more details on how to
customize the Toolbox, refer to Chapter 7.
8/10/2007 4:10:28 PM Page 79

Editing features
Junction and pipe objects on the Workspace can be manipulated as
individual items or as groups. This applies to cutting, copying, pasting,
deleting, duplicating, dragging, locking, and undoing. Pipe objects on the
Workspace can be stretched by selecting the pipe at one end and
dragging it to a new location.
Selecting groups of objects

A group of objects can be selected for manipulation in five ways:
1. By clicking the Selection Drawing Tool and drawing a box around
the desired objects
2. By holding down the SHIFT key while clicking each object in the
group
3. By selecting a pipe or junction, then choosing Select Flow Path from
the Edit menu
4. By choosing Select All from the Edit menu
5. By choosing Select Special from the Edit menu
6. By choosing a group from the Edit menu (Groups -> Select)
When the SHIFT key is depressed, you can click repeatedly on the same
object to select and deselect it. Selected junctions are outlined in red;
selected pipes are also shown in red. The selection color is configurable
in the Workspace Preferences.
The Copy and Paste operations allow you to duplicate a group on the
Workspace. When you copy an item, it is placed on the Windows
clipboard; from there you can paste it into another Windows graphics
program as a picture.
New objects added to the Workspace from the Toolbox are undefined
(refer to Chapter 5 for details on defining objects). Objects derived
through duplication or copying retain all relevant Specifications window
information. If a group of fully defined objects is duplicated or copied,
AFT Impulse retains the object status and connectivity within the new
group.

(si) g.doc
Selecting objects in a flow path
The pipes and/or junctions in a flow path may all be easily selected
using the Select Flow Path item on the Edit menu, Toolbar or the Popup
menu (shown when the right mouse button is pressed while the cursor is
over an empty area on the Workspace). A flow path is a sequence of
pipes and junctions where there is no branching or flow sources or sinks.
The endpoint junctions can be included in the flow path if the All
Objects Including Endpoints option is chosen.
If a junction is selected which has multiple flow paths attached to it
(e.g., a branch with 3 or more pipes) then objects in each of the
connected flow paths will be selected. This feature is very useful in
duplicating or deleting sequences of pipes and junctions. It can also be
used to quickly define groups for graphing or for selected Output or
Model Data listings.
Aligning objects
Tools to align components in your model are located on the Arrange
menu.
Creating and Managing Saved Groups

Groups of pipes and junctions can be named for later recall. This is done
using the Group Tools on the Edit menu. The easiest way to create a
named Group is as follows:
1. First select on the Workspace the pipes and junctions you want in
the group
2. Choose Groups -> Create from the Edit menu or toolbar.
3. You will be prompted for a name for the group. Enter a meaningful
name and click OK.
4. The Group Manager then opens. The selection of pipes and
junctions in the Group will match those currently selected on the
Workspace. If you want to modify the selections once more, you can
do it here.
You can include Subgroups inside of your Groups. Subgroups are
existing Groups. If, for example, you included a Subgroup in your
Group, and later change the contents of the Subgroup, the Group which
contains the Subgroup would also be changed.
8/10/2007 4:10:28 PM Page 81

After you have created a Group, you can then quickly reselect that
Group of pipes or Junctions on the Workspace by choosing Group ->
Select from the Edit menu or Toolbar and choosing the name of the
Group.
The Group Manager can be opened by choosing Group -> Manager from
the Edit menu or Toolbar. The Group Manager allows you to modify,
create, delete or rename Groups.
Another place Groups are available is in the Select Special window.
Special group usage

Besides being useful for selection and general data management, groups
of pipes can be used in specifying pipe sequences that can be used in
creating profile graphs.
Bookmarks on the Workspace

Bookmarks can be added on the Workspace in relation to individual
pipes or junctions. You can quickly navigate around the Workspace by
moving from Bookmark to Bookmark.
Use the Bookmark tools on the View menu to set Bookmarks and to
display different bookmarked pipes or junctions.
Last View
The last view of the Workspace is always saved in memory and can be
displayed by selecting Last View from the View Menu or Toolbar.
Panning the Workspace

If the Workspace has been enlarged beyond the default one screen size
(Workspace scrollbars will be visible), then the Workspace can be
panned. To do this, hold down the CTRL key, press the mouse down on
the Workspace, and then drag in the direction you want to move the
Workspace. The cursor will change to any open hand while you are
panning.

(si) g.doc
Customizing features
Workspace Preferences
The Workspace Preferences on the Options menu lets you customize the
appearance and behavior of the Workspace. New defaults can be saved
so AFT Impulse will always appear and behave according to your
preferences.
See Chapter 7 for more information on Workspace Preferences.
Adjusting the Workspace size

Should the model expand beyond a single screen (as it often will), the
Workspace size can be expanded by choosing Workspace Size from the
View menu. By clicking the boxes in this dialog window you can expand
or contract the Workspace area. Each box represents the area of the
original Workspace screen. When the Workspace is enlarged, scroll bars
automatically appear to simplify navigation of the enlarged Workspace.
The View menu also allows you to zoom in and out at ratios of 25%,
50%, and 75% of the normal size. The Custom Zoom feature allows you
to specify the exact zoom value you desire. There is also a Zoom to Fit
selection, which zooms the model so that one can see the entire model.
Selective display of pipes and junctions

Pipes and junctions can be shown or not shown on the Workspace by
choosing Display Pipes and Junctions on the View Menu (see Figure
4.8).
8/10/2007 4:10:28 PM Page 83

Figure 4.8 Pipes and junctions can be selectively displayed on

the Workspace using the Display Pipes and Junctions
window on the View menu.
Additionally, junctions may be alternatively displayed as an outline or a

solid gray box. By selecting the pipe(s) or junction(s) in the lists and
pressing the << Shift or Shift >> buttons, an “x” will move left or right
indicating how the objects will be displayed on the Workspace. A
junction which is shown as an outline or a solid box may still be opened
for editing.
AFT Impulse does not limit the number of junction icons you can show
at a time.
Scale/flip workspace
The Workspace can be scaled horizontally and/or vertically by choosing
Scale/Flip Workspace on the Arrange menu (see Figure 4.9). Scaling
will move the junctions and will stretch or shrink the pipes so that the
area the model covers is changed by the percentage specified.
Flipping the Workspace horizontally and/or vertically will move the
pipes and junctions to their mirror image locations. This is useful when

(si) g.doc
you are building a system which has symmetrical or mirrored segments.
For example, you could select the segment, duplicate it, and then flip it
horizontally, vertically, or both, with Scale/Flip Selected Objects Only
checked.
Figure 4.9 The Workspace can be scaled or flipped
Renumbering the Workspace objects

Objects on the Workspace can be renumbered in one of three ways:
Renumber Automatic, Renumber Wizard and Renumber Increment. By
choosing Renumber Automatic from the Edit menu (see Figure 4.10).
Selected pipes and/or junctions can be renumbered starting at the lowest
selected number or at a specified number. If a pipe or junction which is
not selected to be renumbered already has the next number in sequence,
it will be changed only if the Force Sequential checkbox is checked.
Impulse will start at the beginning of the flow path containing the lowest
numbered selected pipe and will attempt to march down the pipes in the
direction of flow. It is recommended that renumbering be performed
after the model has been completed so that the junctions and pipes are
completely connected.
If you want certain parts of the model to be numbered in sequence, then
use the Renumber Automatic feature on the selected parts only. It is also
preferable in the case of large or complicated models to renumber small
sections at a time.
8/10/2007 4:10:28 PM Page 85

Figure 4.10 Pipes and junctions can be renumbered using the

Renumber Automatic window opened from the Edit
menu
Using the Renumber Wizard

The Renumber Wizard (see Figure 4.11) allows you to renumber pipes
and/or junctions by simply clicking on them on the Workspace. Normal
Workspace functions are suspended when the Renumber Wizard is
active. The Renumber Wizard is activated on the Edit menu.
Incrementing Object Numbers

The Renumber Increment feature, initiated from the Edit menu, allows
you to add or subtract an increment from select pipes and junctions.
Background Graphic
You can load a graphic into the background of the Workspace and build
your model on top of that graphic. The graphic could be, for instance, a
topographic map or a facility drawing. Such graphics are loaded from the
File menu by choosing Load Background Picture. At your option the
background graphic can be included in Workspace printouts.

(si) g.doc
Figure 4.11 The Renumber Wizard allows you to renumber objects

by simply clicking on them.
Specifications windows
Each object on the Workspace has an associated Specifications window.
The Specifications window lists all the input data for the selected object.
There are three ways to open an object’s Specifications window from the
Workspace:
1. Double-click the object
2. Select the object and click the Open Pipe/Jct Window icon on the
Toolbar
3. Select the object and press ENTER
The Specifications windows can also be accessed through the Model
Data window.
8/10/2007 4:10:28 PM Page 87

Printing and exporting the Workspace

The Workspace image can be:
• Copied to the clipboard
• Saved to a picture file
• Saved directly to an Adobe PDF file
• Sent to the printer
The Print Preview/Special window allows the above outputs to be
applied to all or selected Workspace objects.
The Model Data window

The Model Data window offers a text-based perspective on all
engineering information in the pipe flow model (see Figure 4.12). This
window is immensely useful for obtaining a global view of the model
and rapidly checking the input.
The Model Data window does not offer any tools to build or add to a
model; all model assembly must occur in the Workspace.
The Model Data window allows you to manipulate existing elements
from the Workspace. Therefore, a complete model can be assembled in
the Workspace window without ever opening the Model Data window.
In fact, for small models this may be the preferred approach.
The Pipes table allows display of all input data for all pipes in the model.
Details of fittings and losses are displayed in the Pipe Loss Data table,
accessed with an adjacent tab. The Pipe Loss Data table displays all
fittings with K factors grouped into columns.

(si) g.doc
Open Model
Data Control
General data
section
Customize
View
Pipe data
section
Junction data
section
Figure 4.12 The Model Data window provides a convenient

summary of model element specifications
The other display tab in the pipe area is the Pipe Detail Summary, which
assembles a list of details about each pipe in another format useful to
some engineers.
The junction data area is below the pipe data area. Here the junction data
is separated into tables for each junction type. Click the folder tabs to
see the data for any type of junction.
The transient data for each junction is shown on the Transient Data tab
(see Figure 4.12).
To edit information in the Model Data window, you can open the
Specifications window for the pipe or junction. This can be done by
double-clicking the far left column where the pipe or junction ID
numbers are shown. Alternatively, you can use the Global Edit windows
opened from the Edit menu.
8/10/2007 4:10:28 PM Page 89

Display and printing features

The Show Data options on the View menu (and Toolbar) lets you
specify the combination of pipe, junction, and general information you
want to see in the Model Data window.
Frequently, there are empty columns in the Model Data window tables.
These can be hidden by selecting Hide Empty Columns from the Options
menu.
The Print Content window lets you specify the content to include in the
printed Model Data report and select the font to be used for printing.
This window is opened from the View menu (or Toolbar).
The Model Data window contents can be:
• Saved to a formatted text file
• Exported to a delimited file suitable for direct importing into
spreadsheet software
The Model Data Control window

The Model Data Control allows you to customize what data to show for
the pipes and for each junction type and also which pipes and junctions
to show (see Figure 4.13). It can be opened from the View menu (or
Toolbar).
Pipe data display

The data to be shown for pipes is selected on the left side of the Model
Data Control window. Only the items with check marks will be shown in
the Model Data window. All pipes shown will report the selected items.
Junction data display

The data to be shown for the junctions is selected on the right side of the
Model Data Control window. The data is grouped by junction type.
However, some data is general and applied to each of the junction types.
Select the junction type from the list and then select the data for that

(si) g.doc
junction type. Each type also includes the general data which may be
selected/deselected. You can select/deselect the general data for all
junction types simultaneously by clicking Set Common.
Figure 4.13 The Model Data Control window allows selection of

the parameters to show in the Model Data window
Show selected pipes and junctions

Specific pipes and junctions can be shown in Model Data. Click the
Show Selected Pipes/Junctions tab, then click the items in the lists to
select the pipe(s) and junction(s) to be shown.
Scenario Format
A powerful Model Data feature when using scenarios is to show all
direct ancestor data in the Model Data along with data from the current
scenario. The Scenario Format features allow you to do this, and to
highlight where data changes occur. See Figure 4.14. Note that the
scenario name is appended to the pipe and junction number in the left
columns.
8/10/2007 4:10:28 PM Page 91

Figure 4.14 Model Data display of scenario ancestor data. This

functionality is enabled in Model Data Control.
Database connections
In the lower left of the Model Data Control window there is a Database
checkbox. If this box is checked, your Model Data Control parameters
are set up as determined by the database to which you are connected.
This is referred to as an active database. To make it inactive, uncheck
the box or change one of the Model Data Control settings controlled by
the database.
If the checkbox is unchecked, but enabled, you are connected to a
database but the settings are not being passed to the Model Data Control
window. The database is thus inactive. To make it active, check the box
then click the OK button.
If the checkbox is disabled, there is no connected database.

(si) g.doc
See Chapter 7 for an in-depth discussion of how databases are
configured and administered through your local or wide area network.
The Output window

The Output window is the primary vehicle for communicating the results
of an analysis in text form (see Figure 4.15). The Output window
follows the same general window organization as the Model Data
window.
When a model is submitted to the Solver, the Solution Progress window
appears. After processing is complete, you can view the results in the
Output window by clicking the View Output button.
If the information in the Output window is not in the desired format, you
can choose Output Control from the Analysis menu (or Toolbar) to
modify the format. The Output Control window (see Figure 4.16) lets
you specify the parameters, their units, and the order of their display in
the Output window (Figure 4.15).
Data in the Output window can be:
• Saved to a formatted text file
The printed and Adobe PDF output follows the same flexible, organized,
and accessible format as the Output window itself.
The Show options on the View menu (or Toolbar) lets you enlarge each
of the output sections for greater clarity. The Print Content window,
available on the View menu (or Toolbar), lets you specify the content to
include in the printed Output report and select the font to be used. The
Sort window, opened from the Arrange menu (or Toolbar), offers the
ability to sort the Pipe Results or Junction Results tables according to the
values in any of the columns. The Transfer Results to Initial Guess
feature on the Edit menu (and Toolbar) sets all pipe and junction initial
8/10/2007 4:10:28 PM Page 93

flows and pressures equal to the steady-state results, which causes future
steady-state runs to converge much faster.
You can view the input data for individual pipes or junctions directly
from the Output window in one of two ways. First, if you press down the
right mouse button on the pipe or junction number in the far left column
of the table you will see the Inspection window. Second, if you double-
click the pipe or junction number you will open the Specifications
window in read-only form where you can review input in detail.
Open Output Control
Customize view
General section
Pipe output section
Junction output section
Figure 4.15 The Output window displays output in text form
Transient output files

The transient output is saved to an external file with the extension ".out".
The file name will be the same as the input filename with a number
appended. The filename number allows AFT Impulse to distinguish
among output files for different scenarios. If you want to know which

(si) g.doc
specific output file is used for a scenario, the name is given in the
General section of the Output window.
The content of the Output file is specified in the Transient Control
window. Details are given in Chapter 5, and summarized here. As the
Transient Solver progresses, solutions are obtained for all pipe stations.
Each pipe is divided into multiple computing stations. In order to
perform the simulation, solutions must be obtained for each station for
each time step. When all pipe stations and time steps are considered, the
amount of output data can be enormous. Saving all of this data can thus
create a huge output file. The Transient Control window allows you to
only save selected pipe stations at selected time intervals to the output
file. This allows more reasonable output file sizes, while still retaining
the data of most interest.
The output file saves the fundamental pressure and flow data that allows
all important transient parameters to be determined. However, the user
can only review output data that exists in the output file. If data for a
pipe station is not saved, the user cannot review data in the Output
window or Graph Results window at that location because the data was
not saved.
The Output Control window

The Output Control window (see Figure 4.16) allows you to customize
the parameters shown on the Output window. The control features are
grouped on eight tabs.
Pipe and Junction parameters

The first two tabs control the parameters to be shown for the pipe and
junction steady-state results. The list on the left side indicates the
parameters not shown but available. This list can be viewed in
alphabetized or categorized mode, which makes it easier to find a
parameter. Choose a parameter to be shown by selecting it on the left-
hand list and clicking the Add >> button. The selected item will move to
the list on the right which shows the order of the parameters to be
displayed in the output and the units (if any) in which to display the
results.
You can change the order of the output by selecting a parameter on the
right-hand list, then clicking the Reorder scroll bar on the far right either
8/10/2007 4:10:28 PM Page 95

up or down. The list will be reordered accordingly. The output is
Change output order presented in the order defined in this list.
using this tool
These parameters are

shown in the output
These parameters are not

shown but may be added
Available parameters may

be shown alphabetically
or by category
Description of
output terms
Change units here
Database connection
Figure 4.16 The Output Control window selects parameters for the
Output window. Here the Pipe parameters (for steady-
flow results) are shown.
The units used for a parameter can be changed by selecting the

parameter then choosing the desired units from the dropdown list below
the list of parameters. By selecting the units of interest, you can obtain
your results in whatever units you find most convenient and meaningful.
This means that you can enter all your input parameters in one set of
units (or a variety of units, for that matter), and have all your output
parameters in a completely different set of units. This feature has
obvious benefits for international applications.
To set selected units for both pipes, junctions and summary reports to
the preferred units (set in Parameter and Unit Preferences), click the Use
Preferred Units button.
You will note that some output parameters do not offer the ability to
specify output units. These parameters either do not have units
associated with them (such as Reynolds Number or friction factor), or
are in fact echoes of input parameters (such as Pipe Nominal Size).

(si) g.doc
Pipe Transient parameters
The third and fourth tabs control the parameters to be shown for the pipe
and junction transient results. The list on the left side indicates the
parameters not shown but available. Choose a parameter to be shown by
selecting it on the left-hand list and clicking the Add >> button. The
selected item will move to the list on the right which shows the order of
the parameters to be displayed in the output and the units (if any) in
which to display the results.
Reordering the parameters and changing the units works similarly to the
steady-state methods described previously.
General output
The General section is the text area located at the top of the Output
window. The project title, reference information and other settings
affecting the General Output are defined here.
You can enter a descriptive title which will be used in the Model Data,
Output and Visual Report windows. A title is required and can have up
to 100 characters.
You can keep a lengthy explanation or any other documentation about
your model in the Reference Information section. Names of projects,
individuals, and assumptions can all be kept with the model. This
information can then be included in your reports to improve tracking.
The checkboxes allow you to enable certain special reports to be shown
in the General Section.
Summaries
Several Special Summary Reports can be displayed in the Output
window. These include a Pump Summary and Valve Summary. The
contents of these summary reports and the engineering units can be
specified in the Summaries list at the right.
8/10/2007 4:10:28 PM Page 97

Format and action
The results can be automatically sent to the printer or saved to a file
when the model has reached convergence. Additionally, the results can
be automatically saved as initial conditions by checking the Transfer
Results to Initial When Done option. The model can also be
automatically saved after the initial conditions are transferred. The valve
state (open, closed or failed) can also be saved along with the initial
conditions.
When enabled, the Wrap Table Column headers option makes the output
tables wrap the column headers an extra line if the header is wider than
usual. There are four formatting methods for the pipe and junction
output data:
1. Align on the decimal point
2. Show only the specified number of digits
3. Align only when the differences in magnitudes are less than a user-
set limit
4. Use exponential notation with the specified number of digits when
the differences in the magnitudes are greater than a user-set limit
The first method allows the magnitude of the data to be easily compared
as the data is read down the column. However, a number whose
magnitude is small may force larger numbers to show many digits. For
example, assume that the minimum number of digits is set to four, and
there are two output values: 1234.5 and .09876. These two values will be
displayed in the output as:
1234.50000
0.09876
The second method will force the data to use exactly the number of
digits specified, which may result in the decimal point location changing
from value to value. Using this method with the above example and
specifying four digits, the output will be displayed as:
1235.
0.09876
The third method will align the data on the decimal point only if the
maximum value in the column divided by the minimum value is less than
a limiting order of magnitude that you set. If this ratio is greater than the
limit then each value in the column will be exactly the number of digits

(si) g.doc
specified. Using the above example with a limit set to 4 then
1234.5/.09876 = 12500 or an order of magnitude of 5. This is greater
than the limit. The entire column will not be aligned and will be
displayed as:
1235.
0.09876
The fourth method is similar to the third except that if the ratio is greater
than the limit then the entire column will be shown in exponential
format. Using the above example, the column would be shown as:
1.235E+03
9.876E-02
These formatting options are for the entire Output window; however, the
last two are calculated and applied on a column basis. In other words, if
you choose the third option, one column may be aligned but another may
not be.
The output can be forced to always display in exponential notation by
choosing the checkbox.
For transient results, the maximum and minimum values for all selected
parameters are shown in the Output window. The results can show the
max/min values for all pipe stations, or only the max/min values for
anywhere in the pipe. You can choose which type of report you want by
selecting Detailed or Summary in the Max/Min Transient Table
Contents area. For models with a large number of pipe sections, the
Summary report is generally preferable because of the time required to
display the Detailed table.
Show selected pipes/junctions

You can choose which pipes and junctions are to be displayed in the
Output window. Clicking a selected item in the list on the Show Selected
Pipes/Junctions tab will deselect it. Clicking a deselected item will select
it. This feature is especially useful when you are interested only in a
certain area of the model.
You can quickly define the output you want by doing the following:
1. Select the pipes and junctions on the Workspace
2. Open the Output Control window and select the Show Selected
Pipes/Junctions tab
8/10/2007 4:10:28 PM Page 99

3. Click the Workspace button for both pipes and junctions
4. Click the Output Control window OK button
In the lower left of the Output Control window there is a Database
checkbox. If this box is checked, your Output Control parameters are set
up as determined by the database to which you are connected. This is
referred to as an active database. To make it inactive, uncheck the box or
change one of the Output Control settings controlled by the database.
database but the settings are not being passed to the Output Control
See Chapter 7 for an in-depth discussion of how databases are
Command buttons
There are eight buttons at the bottom of the Output Control window.
Impulse has built-in default parameters, units and settings which you can
choose by clicking the Impulse Default button. You can also develop
your own settings, tailored to your project or industry, and have these
used by default (instead of Impulse’s defaults). To make your own
default, first select the output parameters, units and settings you would
like to use then click the Set As Default button. Your settings will be
saved and will be used each time any new project is initiated. If you
make changes to the settings, and want to get back to your defaults, click
the User Default button. The saved default settings are updated only
when you click Set As Default.
You can save the output settings (except for the selected pipes/junctions
and junction deltas) to a file by pressing the Save Control Format button
and entering a file name. These settings are loaded again by pressing the
Load Control Format button and choosing the file name. For example,
you may have a final report format that is always desired. You can load
this format before generating final results. If you have another format
you use for reviewing model accuracy, you may want a larger number of
parameters in the output.
8/10/2007 4:10:28 PM Page 100 c:\aft manuals\i4\impulse4

manual (si) g.doc
The format files you create can be placed on a network for sharing
among a group or company, or incorporated into a company-wide
database, allowing standardized reporting.
If you have made changes you don’t want to keep, click the Cancel
button. Click OK to use the settings you have specified.
Output parameter descriptions

At the bottom of the Pipe, Junction, Pipe Transient and Junction
Transient tabs a description is given for the selected parameter. There
are three lines shown.
First is the Output Window table header. This is the text that is shown in
the headers of the pipe and junction tables on the Output window.
Second is the Visual Report Abbreviation. This is the text shown on the
Visual Report window when you display an output parameter. This text
is very short so as to save space.
Third is the description of the term.
Steady flow results

Figure 4.17 shows the steady flow results in tables for pipes and
junctions as specified in the Output Control window.
Transient flow results for pipes

You can review the transient results for any pipe station for any time
which was saved to file as specified in the Transient Control window.
These results are shown in the Output window Transient Output table
(Figure 4.18)
Figure 4.18 shows the transient flow results at the initial time step
(which is equal to zero here). The columns of data shown in Figure 4.18
are specified in the Output Control window.
Note that the pipe stations shown are those saved to the output file.
8/10/2007 4:10:28 PM Page 101

Figure 4.17 The Output window shows steady flow results in the
Pipes table and the various junction tables. Transient
data is shown on the transient tabs.
Transient Max/Min results for pipes

You can review the maximum and minimum transient results for all pipe
stations for all times. The data shown is not dependent on which pipe
stations or time steps were saved to file. Figure 4.19 shows the Transient
Max/Min table in the Output window. The parameters that are shown are
the same as those specified in Output Control. However, there are four
columns for each parameter. The four columns show the maximum and
minimum values for the parameter, and the times for each of those.

manual (si) g.doc
Buttons to
increment or
decrement
currently
displayed time
Time step for

currently viewed
results
Time step
selection tool
slides up and
down
Figure 4.18 The Transient Output table in the Output window

shows output results for pipe stations and at specified
times.
Summary Max/Min results

There is a summary version of the Transient Max/Min table. The version
shown in Figure 4.19 is what is called the Detailed version. You can
switch between detailed and summary views by selecting the option on
the Output Control "Format & Action" tab (see Figure 4.20).
The summary version shows only maximum and minimum values for the
entire pipe, as opposed to each pipe station. The summary also shows the
pipe station that is the overall maximum or minimum for the pipe.
Therefore, for each Output Control transient parameter there are six
8/10/2007 4:10:28 PM Page 103

columns in the summary version of the Transient Max/Min table. See
Figure 4.21 for an example.
Figure 4.19 The Transient Max/Min table details the maximum and
minimum values for the entire simulation.
Transient event messages

When events occur at junctions during the transient simulation, a log of
all events is displayed in the two Event Message lists. In the first list
events are sorted by junction number. In the second list they are sorted
by time. The list includes when check valves close (and re-open), relief
valves open (and re-close), control valves lose control (and regain it), as
well as all standard event transients (see Chapter 10).
Output window updates

Whenever you change the input model, all output windows are erased, as
is the output file. This prevents you from changing your input model,

manual (si) g.doc
forgetting to rerun it, and mistaking the previous Output window for the
current model's results.
Figure 4.20 Detailed and Summary versions are available for the
Transient Max/Min table
Figure 4.21 Summary version of the Transient Max/Min table
8/10/2007 4:10:28 PM Page 105

The Graph Results window

The Graph Results window, shown in Figure 4.22, allows you to
generate high quality printed graphs.
The Select Graph Data window, opened from the View menu (and
Toolbar), lets you control the content of the graph and offers several
types of standard graphs. The Customize Graph window, opened from
the Options menu (and Toolbar), offers you control of the graph's
general style. Fonts and colors can be selected.
The graphical data can be exported to a file for later import as a cross
plot against results for a different model. For example, let's say you
would like to see the effect on the system pipe pressure drop when you
change a valve loss factor. You could run the Solver for one value of K
and export the results to a file. Then you could rerun the Solver for a
new value of K and import the results from the previous case for cross
plotting. These features are accessed by choosing Save Graph Data and
Import Graph Data from the File menu.
Figure 4.22 The Graph Results window offers full-featured plot

generation

manual (si) g.doc
Graph “Sets” can be created and saved by the user using the tools in the
Select Graph Data window.
The Graph Results window is always erased when you make changes to
model input. Rerun the model to graph the new results.
The Select Graph Data window

The Select Graph Data window defines what type of graph you want
created and with what parameters and units (see Figure 4.23). You can
create five types of graphs. Three types use time as the independent
variable, and the other two use pipe length.
Figure 4.23 The Select Graph Data window controls the Graph
Results window
Transient pipe results

Transient pipe results can be plotted vs. time for any pipe station that is
saved to the output file. As summarized earlier in this chapter and
discussed in detail in Chapter 5, the Transient Control window allows
the user to specify that data for selected pipe stations be saved. As
shown in Figure 4.23, each pipe is shown in an expandable list. When a
8/10/2007 4:10:28 PM Page 107

pipe is expanded, the pipe stations for which data was saved are shown.
These pipe stations can be added to the list on the right.
The user can plot the data for the entire simulation, or only for some
specified time frame. Further, if Design Alerts are specified for the pipe
(e.g., max/min operating pressures), these can be cross-plotted against
the transient data.
Transient junction results

Similar to transient pipe results, transient junction results can be plotted
vs. time for any junction that is saved to the output file. Junction data
includes items such as pump speed and accumulator volume. Similar to
pipe output, these are specified in the Transient Control window.
Profile Along a Flow Path and EGL, HGL and Elevation Profile
These two types of graphs will plot the selected parameter along a flow
path. The pipes that comprise the path must form a single, continuous
path. The independent variable is the length along the pipe(s).
A path of pipes can be specified in one of two ways. First, the user can
select the pipes in the displayed list. This permits only a single path to be
specified.
Second, one or more flow paths can be specified using groups (see
Figure 4.24). Only groups which consist of a pipe sequence will be
displayed. Groups are set up on the Workspace using the Groups tool on
the Edit menu.
The results can be the values at a particular point in time, or the overall
maximum and/or minimum values. The Profile Along a Flow Path graph
type also allows cross-plotting of Design Alerts such as max/min
allowed pressures.
Both of these graph types support animation, discussed in the next
section.

manual (si) g.doc
Figure 4.24 Multiple paths can be cross-plotted using predefined

groups.
Animating profile results

There are two animation options: Animate Using Output File, and
Animate Using Solver.
The option to Animate Using Output File uses data saved to the transient
output file (discussed earlier in this chapter). This option can only be
used with paths of pipes for which transient data was saved for all
stations. The reason being that to show how a parameter such as pressure
changes along a pipe, the data at all computation stations must be
available. Saving selected pipe stations is specified in the Transient
Control window, and is discussed in Chapter 5.
8/10/2007 4:10:28 PM Page 109

As its name implies, the Animate Using Output File option obtains the
data for animation by reading in data saved to the transient output file
and then displaying it in the graph.
The option to Animate Using Solver does not use the transient output
file at all. Instead, it actually reruns the Transient Solver in the
background to regenerate the transient data. This rerun does not write
any output to file. Because the model is being rerun, all data for all pipe
stations is available for each time step.
Figure 4.25 shows the Graph Results window when using animation. As
shown, additional controls become visible to allow the user to adjust
animation. At the left is the Play button, which starts the animation. Next
to it are the Pause and Stop buttons. If the Pause button is pushed, the
animation stops at that time step. One can print the graph or copy the
data if desired or, if using the Animate Using Output File option, use the
time slider controls (to the right of the Stop button) to set the time
forward or backwards. If using the Animate Using Solver, the solver can
be made to march one step at a time when one clicks the “+” button. The
Stop button stops the transient and resets the time to zero.
The current time (in seconds) and time step are shown to the right of the
time controls. At the far right are the speed controls. Slowing the
animation involves adding time delays to the data display. Increasing the
speed involves skipping time steps. The default speed, right in the
middle, has no delays or skips.
If plotting a single path using the list of pipes in Select Graph Data, one
can cross-plot on the animation design alerts (in Profile Along a Flow
Path) or steady-state results (in EGL, HGL and Elevation Profile).
Each of these options has advantages and disadvantages. The Animate
Using Output File option is usually a little faster. It also allows you to
use the time slider to reset the time to any time step forwards or
backwards. The reason being that the transient output file has all of the
required data over time, and thus obtaining data for any other time is just
a matter of reading the data from the proper place in the output file. The
Animate Using Solver does not have this flexibility, since it can only
obtain the data by marching through all of the time steps.

manual (si) g.doc
Figure 4.25 The animation options display additional controls
Another advantage of the Animate Using Output File option is that it can
be made to run much faster. When the animation speed control is set to
increase speed, intermediate time steps are skipped and not plotted. If
the speed is increased too much, the quality of the animation deteriorates
because too much missing time data makes it difficult to follow what is
happening. When the speed is increased, the Animate Using Output File
option just skips time steps and reads in the data for each required time
step.
On the other hand, the Animate Using Solver option cannot skip
calculation steps because it is running the Transient Solver. To obtain
the results for a future time step, it must calculate all of the intermediate
time steps.
The biggest advantage of the Animate Using Solver option is that it does
not require one to save all of the required pipe station data to disk in
order run an animation. Because it is obtaining data from the Transient
Solver, it can always run any animation desired. If one uses the Animate
Using Output File option and has not saved all of the required data, the
8/10/2007 4:10:28 PM Page 111

data must be specified in Transient Control (as discussed in Chapter 5)
and then the entire transient rerun. The Animate Using Solver option
allows one to run animations while keeping the output file sizes to a
minimum.
Transient force results

If transient force sets have been defined in the Transient Control window
(see Transient Control section in Chapter 5), the resulting force data can
be plotted vs. time. Figure 4.26 shows the Select Graph Data window
when plotting transient force data.
The differential pressure forces for each force set selected from the list
of defined force sets will be plotted. If desired, the user may select to
include momentum and friction effects in the force balance calculation,
as well.
The differential force data can be plotted for the entire transient
simulation, or the user has the option of selecting a specific time frame
to display.
The options to include pipe friction and momentum in the force balance
calculation are discussed in Chapter 5. Force data can be exported to a
file for use with CAESAR II pipe stress software, which is also
discussed in Chapter 5.
Other graph controls

When you change graph types or parameters you can choose to keep the
labels and/or color styles by clicking the appropriate checkboxes on the
lower right.
The Customize Graph window

The Customize Graph window allows you to customize the plot by
giving you control over the many aspects including:

manual (si) g.doc
1. Fonts for titles, numbers, and legend
2. Colors for the graph background, foreground, and data sets
3. Line types and thicknesses
4. Symbol shapes
5. Normal, semi-log, or log-log plots
6. Borders and gridlines
7. Printing preferences (color or monochrome)
You can save the settings you set up by clicking the Save button on the
System tab. If you make changes to the settings, and want to get back to
previous settings, click the Load button on the System tab and load the
file.
Figure 4.26 Transient Force data can be plotted using force sets
defined in the Transient Control window
8/10/2007 4:10:28 PM Page 113

The Auxiliary Graph Formatting window

The Auxiliary Graph Formatting window allows you to more easily
change graph colors and curve thickness. These preferred settings can be
saved as the default for future use. In addition, these preferences are
used in all graphs throughout AFT Impulse. This window is opened from
the Options menu or toolbar.
Printing and exporting the Graph Results image and

data
The Graph Results image can be:
The x-y data in the Graph Results graph can be:
The Visual Report window

The Visual Report will help you visualize the model data and steady-
state and transient solutions in relationship to the model layout. The
Visual Report window displays the pipe system layout from the
Workspace window with information in text form (see Figure 4.27).
Additionally, magnitudes of steady-state and transient parameters may
also be shown with color variations.
The values to be shown may come from the input data (when in input
mode), Steady-State Output results, or Transient Output results. You can
change between these by selecting the option at the top of the Visual
Report Control window.
In the Visual Report window, the positions of the junctions and pipes are
fixed based on their locations on the Workspace. While the pipe system
layout cannot be edited in the Visual Report window, the integrated text

manual (si) g.doc
results for each pipe or junction can be rearranged by dragging and
dropping to improve the final appearance of the window.
The contents of the Visual Report window can be:
The View menu allows you to zoom in and out at ratios of 25%, 50%,
and 75% of the normal size, as well as choose custom sizes. This feature
compresses more results onto a single page. The printed size is
consistent with the zoom state. That is, if you zoom out to 50%, the
Visual Report screen prints at 50% of its normal size. The Custom Zoom
feature allows you to specify the exact zoom value you desire. There
also is a Zoom to Fit selection, which zooms the model so that one can
see the entire model.
The Visual Report window is always erased when you make changes to
model input. If viewing output results, rerun the model to view the new
results.
Figure 4.27 Visual Report window integrates input data and

output results with model layout
8/10/2007 4:10:28 PM Page 115

The Visual Report Control window

The Visual Report Control lets you control the content and appearance
of the Visual Report window (see Figure 4.28). You can also save
custom Visual Report window layouts to disk for later recall, so you
don't have to reformat this window repeatedly.
When the Visual Report window is opened and no output results are
currently displayed, Visual Report Control will be in the Input Display
Mode. If output results are displayed, it will be opened in the Steady
Output or Transient Output Display Mode. Parameters available for
display vary depending upon the mode, i.e. input, steady output or
transient output. If output results exist, the display mode may be changed
at any time.
Figure 4.28 The Visual Report Control window allows you to

configure the Visual Report display

manual (si) g.doc
Display parameters
The parameters to be shown in the Visual Report window are selected
from the Display Parameters tab. The left list selects the pipe parameters
and the right list selects the junction parameters. In these lists you can
show only the parameters you have selected in the Output Control
window, or all output parameters. You can select more than one
parameter; however, the Visual Report window may get crowded if too
many are selected.
General display
The General Display tab allows you to configure several other items.
The title, defined in the Output Control window, and filename can be
shown or not shown depending on your preference. The scenario name
can also be shown.
Figure 4.29 The Visual Report with the units and Color Map in the
Legend
The units for all parameters can be shown in a legend instead of next to
the parameter value (see Figure 4.27 or 4.28). This will reduce the
amount of text and crowding shown in the Visual Report window. The
legend can be moved to any location in the Visual Report window. If this
8/10/2007 4:10:28 PM Page 117

option is chosen while in input-only mode, all values will be converted
to a consistent set of units.
You can change the font for pipe and junction text by clicking the Fonts
button and choosing the font desired or by using the font size change
buttons on the Toolbar.
The number of digits to show in the numerical display can be configured
using the provided drop-down list.
Normally on the Workspace, closed objects are displayed using special
graphics (dashed lines, X’s, etc.). The number of a junction which has
special conditions set, by default, has an X preceding the J (e.g. XJ56).
If you want these symbols to be used in the Visual Report window also,
choose the Show Closed Symbols checkbox. Normally, pipes in closed
sections are shown as dashed lines, and junctions outlined with a dashed
line. If you want these symbols to be used in the Visual Report window
also, choose the Use Closed Object Style checkbox.
If a background picture exists on the Workspace, it can be shown in the
Visual Report by selecting the checkbox provided. Similarly, if a grid is
shown on the Workspace it can be shown on the Visual Report.
Show selected pipes and junctions

You can choose, on an individual pipe and junction basis, what you
would like to show in the Visual Report Control window (see Figure
4.30). This display customization feature is useful for large models
where only some of the data points are of interest. For simplicity, this
discussion will focus on just the pipes; however, the junctions behave in
the same fashion.
Each pipe is listed with a set of x’s to the left of the name. The x’s
indicate what will be shown. To change the settings, first choose the
pipe(s) to change. Then use the checkboxes on top to set or clear the x’s.
If more than one pipe is selected and the current options are different,
the checkbox will look grayed and the first click will clear the checkbox
and the second click will set it. You can have any combination of
settings, including no settings at all. To restore the settings to what is
shown in the Workspace and the data, click the Workspace Settings
button.

manual (si) g.doc
Color map
The Color Map feature allows you to visualize magnitudes of various
parameters by coloring the pipes (see Figures 4.28 and 4.30). There is
one Color Map for the steady output mode, one for transient output
mode, and another for the input-only mode. A Color Map is useful to
show graphically where, for example, high velocities or reverse flows
are located or to visualize the pipe diameters or lengths in a model.
To create a Color Map, first choose a parameter and units for which all
pipes will be categorized. Then choose a color, select the operator (i.e. >,
>=, =, <=, <) and enter a value, then press Add to Map. Values can be
categorized on an absolute value basis by selecting this option.
Figure 4.30 The Visual Report Control window allows you to

individually select the object, name, number and data
to be shown in the Visual Report window.
The order of entry is not important. Impulse will sort the entries first by
value and then by the operator, with the > being above the <. The pipes
8/10/2007 4:10:28 PM Page 119

will be colored using the first match found in the map, starting from the
top of the map. If no match is found, the pipe will be the Workspace pipe
color. Be careful to avoid overlapping operators and values. If, for
example, the map had the following four colors,
1) >=60
2) <60
3) >=20
4) <20
then there is no way colors 3 and 4 will be used because any pipe less
than 20 is also less than 60. Since <60 is before <20 in the map, the pipe
will be colored using 2.
Figure 4.31 The Visual Report Control Color Map allows you to
specify the display colors for pipes according to their
output values.

manual (si) g.doc
To avoid overlapping colors, only use less than (<) as the last entry. A
better map would be to use
1) >=60
2) >=20
3) <20.
You can choose to categorize the pipes based on their absolute values by
choosing the Display Absolute Value checkbox. To delete a map value,
click the color you want to delete then click the Remove Setting button.
Instead of coloring the pipes, you could choose to color just the pipe
numbers. This is done by choosing the Color Pipe Labels and clearing
the Color Pipes checkboxes in the Color Options area. You could also
apply the colorization to both. Choose the Print Black & White
checkbox if you don’t have a color printer. The pipes will be printed in
black.
The color map values can be displayed in a legend by checking the
Display in Legend checkbox (see Figure 4.31). The color map will
appear in the legend after the pipe and junction units, if that option was
selected. The legend may be dragged and dropped to any location in the
Visual Report window.
Command buttons
You can save the output settings (except for the selected pipes/junctions)
to a file by clicking the Save Options button and entering a file name.
These setting are loaded again by pressing the Load Options and
choosing the file name. This file may be shared among engineers or
incorporated into a company-wide database.
Impulse will initially try to place the text for pipes and junctions based
on their positions on the Workspace. You can drag the text to new
locations which will give the best visual presentation. These positions
are saved with the model. You can reset the locations back to where
Impulse initially placed the text by clicking the Reset Locations button
and then confirming the operation.
If you have made changes which you don’t want to keep, click the
Cancel button. Click Show to apply your settings.
8/10/2007 4:10:28 PM Page 121

Visual Report annotations
Annotations for the Workspace

Annotations created on the Workspace can be displayed on the Visual
Report by opening the Workspace Annotation window (double-click the
annotation) and selecting the Also Show on Visual Report on the
Scenario Options tab.
Workspace annotations displayed on the Visual Report can be moved on
the Visual Report but not edited.
Creating Visual Report annotations

An annotation can be created specifically for the Visual Report by
choosing Create Annotation from the Edit menu or Visual Report
Toolbar. When the Create Annotation tool is activated, you can draw an
annotation on the Visual Report just like on the Workspace.
Visual Report annotations have all the properties of a Workspace
annotation, except they can be displayed only on the Visual Report.
The Toolbars
The Toolbars offer quick access to the features used most frequently in
AFT Impulse. Figure 4.32 shows the Toolbars. There are actually six
toolbars in all. One is the Common Toolbar, and it is always shown. Its
features are common to all five Primary windows. In addition, there is
one Toolbar for each Primary window, offering features appropriate for
work in that window. Each of the functions on the Toolbar is also
available through the menu system.

manual (si) g.doc
Figure 4.32 The Toolbars allow you to quickly access important

features. There is one common Toolbar and a special
Toolbar for each of the five Primary windows. You see
only two toolbars at a time in AFT Impulse. The
function of each tool is indicated in the ToolTips that
appear when you hold the cursor over the button.
8/10/2007 4:10:28 PM Page 123


manual (si) g.doc
CHAPTER 5
Building and Running Models
This chapter discusses the basics of building and running a waterhammer

model with AFT Impulse. AFT Impulse assumes that the user has a good
general understanding of engineering pipe hydraulics and a basic idea of
what waterhammer is.
Creating objects
The pipe flow model is assembled on the Workspace by arranging
graphical objects (pipes and junctions) that represent parts of a physical
pipe flow system. In addition to pipe and junction objects, annotation
objects can also be added.
Pipes
To create a pipe, select the Pipe Drawing Tool from the Toolbox and
draw a line on the Workspace to represent the pipe. To draw multiple
pipes more quickly, hold down the CTRL key while drawing each pipe.
This keeps the Pipe Drawing Tool active, so you don't have to select it
for each pipe. Alternatively, you can double-click the Pipe Drawing Tool
to “lock” it down. You can then continue to draw pipes until you click
the Pipe Drawing Tool a second time.
To draw a vertical or horizontal pipe, hold down the SHIFT key while
drawing the pipe.

manual (si) g.doc
The length of the line on the Workspace has no relationship to the
physical length of the pipe. The Workspace pipe is merely an abstract
representation of a pipe.
If a pipe is drawn as an extremely short line, AFT Impulse eliminates the
line immediately after it is drawn.
Another way to create new pipes is to duplicate existing pipes with the
Duplicate or the Copy and Paste features found on the Edit menu. When
a new pipe is derived from a previous pipe, the new pipe retains the
engineering information associated with the original pipe.
When a new pipe is created it is assigned a default ID number. When the
pipe is displayed on the Workspace, the ID number is preceded by the
letter “P” to indicate that it is for a pipe. An arrow on the pipe indicates
the reference positive flow direction for the pipe. From the Pipe
Specifications window you can change the ID number to any value
greater than zero and up to 30,000. The reference positive flow direction
can be reversed by choosing Reverse Direction from the Edit menu or by
selecting the reverse direction button on the Toolbar.
Tip: The pipe or junction ID number text can be dragged to a new

location to improve visibility. The new location is saved with the model.
If you want the text returned to its default location, double-click it. If
you want all pipe and junction text returned to the default locations,
choose Reset Label Locations from the Arrange menu.
Each pipe must be connected to two junctions. There are no exceptions

to this rule.
A pipe can be segmented or “bent” in up to ten segments (see Figure
5.1). This does not affect the model or the results, only the visual
representation. It does not add an “elbow” to the model or the resulting
pressure drop effects of an elbow.

manual (si) g.doc
Chapter 5 Building and Running Models 127
Figure 5.1 Pipe with five segments
This feature is useful in graphically bending the pipe around another part
of the model or graphically showing that there are embedded elbows as
Fittings & Losses.
You can add a segment to a pipe in two ways. The easiest way to add a
single segment is to select Pipe Segments and Add from the Arrange
menu. If there are no segments, then a single segment can also be added
by selecting the Add/Remove Segment from the Toolbar. If the pipe
already has one or more segments, this toolbar selection will remove all
segments.
The other way to segment a pipe is to select the pipe on the Workspace
then choose the Pipe Segments and Manager from the Arrange menu. A
window will display where you can add, remove and merge segments. A
“handle” will appear in the middle of the pipe for each new segment. As
the cursor passes over the handle, the cursor will change to a crosshair.
Press and hold down the left mouse button and drag the middle of the
pipe to the desired location. By holding the SHIFT key the pipe segment
will move only at right angles.
8/10/2007 4:10:28 PM Page 127

Junctions
To place a junction, drag the icon from the Toolbox onto the Workspace.
The nineteen available junctions represent various pipe system
components.
When you duplicate a junction using the Duplicate or Copy and Paste
features on the Edit menu, the new junction retains the engineering
information associated with the original junction.
Junctions are shown on the Workspace with default ID numbers, which
you can change in the Junction Specifications window. The letter “J”
precedes the ID number to signify that it is for a junction. From the
junction’s Specifications window you can change the ID number to any
desired value greater than zero and up to 30,000. Identical ID numbers
can be assigned to both a pipe and a junction because the “P” or “J” will
distinguish the two.
Unlike pipes, junctions do not have reference positive flow directions.
Junctions for which the flow direction is important (such as pumps and
control valves) derive the flow direction from the flow direction of the
connecting pipes.
Morphing junctions
A junction can be “morphed” from one type to another. To morph a
junction, hold down the CTRL key and select a junction from the
Toolbox and drop it onto an existing Workspace junction. The junction
type will change to the new junction type, and data that can be kept will
be copied into the new junction.
Splitting pipes
If you hold down the SHIFT key while selecting a junction from the
Toolbox and then drop the junction onto an existing pipe, the pipe will
split into two pipes. The physical length of the original pipe will be
automatically halved, and the new pipe will be assigned the balance.
Thus the sum of the two pipe lengths will equal the length of the original
pipe. Any fittings/losses in the original pipe will be left in the original
pipe, and the new pipe will have no fittings/losses.

manual (si) g.doc
Annotations
To create an annotation, select the annotation tool on the Toolbox and
draw the outline of the annotation. When you release the mouse button,
the Workspace Annotation window will display. Here you can enter text
for the annotation, draw lines and arrows, change colors, and draw
outlines. You also can change the colors of the annotation.
Annotations can be cut, copied, pasted, duplicated, and moved like pipe
and junction objects. When you select an existing annotation, resizing
handles appear.
The Annotation Manager assists you in specifying which annotations
will display in which scenarios.
Moving objects
Objects on the Workspace can be moved individually or as groups.
Clicking a Workspace object selects it. Pipe objects change color when
selected. Junction and annotation objects are outlined in red when
selected. This color is configurable in the Workspace Preferences.
To move an object, select it, drag it within the Workspace, and drop it in
the desired location. When an object is dragged beyond the existing
Workspace area, the Workspace is expanded accordingly.
A pipe object can be stretched by grabbing the handles (small black
squares) at the pipe endpoints and moving an endpoint to a new location.
To prevent accidental movement, lock the objects on the Workspace.
The Lock feature is accessed from the Edit menu or the lock button on
the Toolbar.
To group multiple objects for movement or other operations, hold down
the SHIFT key when selecting the objects. Objects can also be grouped
by drawing a box around them using the Selection Drawing Tool on the
Toolbox. You can use the Select Special on the Edit menu to group
objects based on specified criteria.
You can also select all objects in a flow path as follows:
• Select one object in the flow path
• Choose Select Flow Path from the Edit menu or Toolbar
8/10/2007 4:10:28 PM Page 129

• Choose what type of objects you want to select
The endpoints of a flow path are the junctions which connect more than
one flow path, like a branch junction connected to three or more pipes.
You can choose to include or exclude these end point junctions.
Tip: You can undo all pipe and junction movements using Undo on the
Edit menu or Toolbar. You also can press the escape key during the
movement to cancel the move.
Keyboard modifiers
When you drag a junction, the endpoints of any connected pipes are
moved with the junction, thus maintaining the graphical connection.
This functionality may be changed through the Workspace Preferences
window.
Editing objects
The objects you place on the Workspace can be edited with the editing
commands from the Edit menu or the Toolbar. Objects can be cut,
copied, pasted, duplicated, and deleted. These operations can be
performed on individual objects or on groups. The Edit menu provides
one level of undo for each editing operation. Alternately, you can use the
Undo button on the Toolbar.
The Copy Graphics feature on the Edit menu will copy the specified
Workspace objects to the Windows clipboard. This image can then be
pasted into other Windows applications.
Connecting objects
During construction of a new model, objects can be placed anywhere on
the Workspace. In order to assemble a model that is ready to submit to
the Solver, you must connect the objects properly.
Remember that connectivity only exists between junctions and pipes.
There are no junctions that connect to junctions, and no pipes that
connect to pipes.

manual (si) g.doc
AFT Impulse will attempt to automatically connect pipes and junctions
when they are moved. This eliminates unnecessary opening of the pipe
and junction Specifications windows. If Impulse cannot make the
connection automatically, then the following three steps may be required
to establish a connection between a junction and a pipe:
1. Graphically connect the objects on the Workspace. The pipe
endpoint must terminate within the boundaries of a junction icon.
Figure 5.2 shows a graphical connection between a pipe and a
junction.
2. After the pipe is graphically connected to two junctions, Double-
click the pipe to open its Specifications window, then click OK to
accept the connected junctions as determined by AFT Impulse. It is
most efficient to do this when you are entering the pipe's data in its
Specifications window.
3. Open the Specifications window of the corresponding junction and
accept the pipe connectivity determination. You can do this as you
enter the junction's data in its Specifications window.
Figure 5.2 A pipe and junction graphical connection
8/10/2007 4:10:28 PM Page 131

If the model seems to not be connected properly, you can also use the
Extended Model Check feature on the View menu.
When you accept AFT Impulse's connectivity interpretation, the
connection advances from a graphical connection to a model connection
(For more information on Specifications windows, refer to Chapter 6.)
The model connectivity you establish on the Workspace is retained only
as long as you maintain the graphical objects in their current visual
relationship to each other. If you move a pipe or a junction and break the
graphical connection, AFT Impulse disconnects the two objects.
However, if you move a group of connected objects, their model
connectivity is maintained.
Defining objects
Each Workspace object in a model must be defined before AFT Impulse
can obtain a solution. AFT Impulse examines your model for proper
definition before it gives you access to the Solver.
To define an object, you must specify all the required property data for
the object and satisfy its connectivity requirements as described in the
following sections.
The Show Object Status feature identifies pipe and junction objects in
the model that are not completely defined.
Specifying required property data

To specify the required property data for an object, open its
Specifications window. This can be done in six ways:
1. Double-click the object on the Workspace
2. Select the object (by clicking on it once, or using the Selection
Drawing Tool) and press ENTER
3. Select the object and click the Open Pipe/Junction Window icon on
the Toolbar
4. Double-click the left column of the appropriate table in the Model
Data window (where the object ID number is shown)
5. Jump from another Specifications window

manual (si) g.doc
6. Double-click the connected pipe or junction number in a
Specifications window
When editing in a Pipe Specifications window, pressing the F5 function
key jumps automatically to the next higher numbered pipe. Pressing F5
while holding the CTRL key jumps to the next lower number. Junction
windows behave similarly.
Parameters that functionally describe the pipe or junction are entered in
the Specifications window. Some parameters in the Specifications
windows are required while some are optional. The optional parameters
offer you added control over your model and the ability to obtain special
information.
Highlighting required information

You can toggle the highlight feature on and off in any of the following
ways:
• Double-click anywhere in the specifications window
• Press the F2 function key when a specifications window is open
• Choose “Highlight in Pipe and Jct Windows” from the Options
menu
The highlighting feature may be especially useful when you are first
learning to use AFT Impulse or when you are having difficulty obtaining
a defined object status.
A more detailed discussion of the contents of pipe and junction
Specifications windows is given in Chapter 6.
Use status feature

Each pipe and junction specifications window has a Status tab. The
Status tab shows what required information for the pipe or junction is
missing.
Undefined objects window

You can open the Undefined Objects window to view all undefined
objects and the undefined properties. This is available on the View
menu.
8/10/2007 4:10:28 PM Page 133

Satisfying connectivity requirements

In addition to entering the required information for each object in the
model, you must satisfy these connectivity requirements:
• You must accept the connectivity information shown in each object's
Specifications window (refer to “Connecting objects” earlier in this
chapter).
• Each junction must have the appropriate number of pipes connected
to it. For example, an area change junction must have exactly two
connecting pipes, whereas a reservoir may have one to twenty-five
connecting pipes. (Details on each of the junctions are given in
Chapter 6 and summarized in Table 6.1.)
• Pipes must always be connected to two junctions, one at each pipe
endpoint.
Should you ever delete any of the required information in the
Specifications window or move a connected object from its connection,
the object's status reverts to undefined.
For example, say you place a pump on the Workspace. A pump requires
inlet and outlet piping (unless it is a submerged pump). AFT Impulse
reflects this by requiring each pump to be assigned an upstream pipe and
a downstream pipe. Entering only the pump constants and setting an
elevation is necessary but not sufficient information to define the pump
according to AFT Impulse's requirements. Once you have entered the
pump data and properly connected it, the pump is defined.
Inspecting objects
The Inspection feature is a time saving feature for reviewing the data
associated with a Workspace pipe or junction object. Inspection displays
the data for an object in read-only format (see Figure 5.3).

manual (si) g.doc
Figure 5.3 Inspecting from the Workspace with right mouse

button
To inspect a pipe or junction object, position the mouse pointer on the

object and hold down the right mouse button. Inspecting is quicker than
opening the Specifications window if you want to look at an object's
input data but not edit it.
If you press the SHIFT key when you perform the inspection, only the
items not yet specified are displayed. If you press the CTRL key when
you perform the inspection, the output data will be displayed (if output
exists).
The Inspection feature is also available in each Specifications window.
This is useful if you want to check the input data of objects connected to
the one on which you are working.
For example, when working in an Valve Specifications window, you
may wish to quickly check the upstream and downstream pipe diameters
previously defined. Holding down the right mouse button on each
connected pipe ID number will show you that pipe's primary input
parameters.
8/10/2007 4:10:28 PM Page 135

The Inspection feature works similarly in Pipe Specifications windows
when you inspect connected junctions. Judicious inspecting can save you
time when assembling or troubleshooting a model.
Using the Checklist

Assembling a model that is ready to submit to the Solver is not, at first
glance, an obvious process. The new user will wonder where to begin,
and even the experienced user will at times lose track of what
information has already been entered and what information remains to
be entered. To help you in this process, the Checklist window tracks the
status of your model and communicates to you where you lack complete
information.
The Checklist shows six items (see Figure 5.4). The first four are always
required, and the last two are required if it is a transient model. When a
requirement has been met, you will see a check mark next to the item in
the Checklist window. You can view the Checklist by selecting
Checklist from the View menu or clicking the check button on the
Toolbar. As a guide to the user, the Checklist items are also shown on
the Status Bar at the bottom of the AFT Impulse window (Figure 5.5).
When you have entered the proper information for all six Checklist
requirements your model is complete and ready to submit to the Solver.
As long as the Checklist remains uncompleted the Model Status Light in
the lower left corner is red. If you try to run a model with an incomplete
Checklist, you are informed what Checklist items are not completed. The
Model Status Light will turn green when the checklist is complete.
Tip: If you double-click the checkbox on the Checklist window or Status
Bar, the appropriate window will be opened. For example, double-
clicking Specify Output Control will open the Output Control window.

manual (si) g.doc
Figure 5.4 The six Checklist items help you track the status of
your model
Status bar
Status light
Figure 5.5 The Status Bar shows the status of each Checklist
item. The Model Status Light turns green when the
model is complete.
8/10/2007 4:10:28 PM Page 137

Analysis Type
The first item on the Analysis menu is Analysis Type. This drops down
to two menu choices: Transient and Steady Only. By default the
Transient option is selected when a new model is opened. A transient
model always performs a steady-state analysis first, and then proceeds to
the transient solution. The Transient menu selection requires all six
Checklist items to be completed. The Steady Only option requires only
the first four.
Specify Steady Solution Control

For a new model, the Checklist shows two items checked and four
unchecked. The first item is Steady Solution Control; this is checked
because the solution control parameters have been set to default values
which do not generally require modification.
AFT Impulse attempts to shield less experienced analysts from the
details of numerical convergence by providing robust defaults. More
experienced users of numerical modeling software, however, will desire
access to the parameters that control the behavior of the Steady-State
Solver. The Steady Solution Control parameters can be modified by
opening the Steady Solution Control window from the Analysis menu.
More information on Steady Solution Control is given later in this
chapter and in Chapter 8.
Specify Output Control

The second Checklist item, Output Control, is also checked initially
because Output Control default parameters have been specified.
However, you will probably want to change the title of your analysis
from the default (on the General tab).
The Output Control window allows you to choose the parameters you
want included in your output and the units for those parameters. You can
also rearrange the output parameters to appear in the order you find most
productive. These features are discussed in more detail later in this
chapter.
Specify System Properties

The third Checklist item can be addressed by changing to the System
Properties window. Here the working fluid(s), gravitational level, and

manual (si) g.doc
atmospheric pressure are entered. Defaults are assigned for gravity and
atmospheric pressure. You can choose to use an Unspecified Fluid, an
AFT Standard fluid, Water Properties from the ASME Steam Tables, or,
if you have licensed the Chempak add-on, the Chempak fluids. Chempak
fluids permit user specified mixtures.
Define All Pipes and Junctions

The fourth Checklist item, Define All Pipes and Junctions, is more
abstract and cannot be addressed like the first three items by simply
opening a window and entering the proper information. This item
requires that the proper specifications have been entered and
connectivity requirements have been met for all objects in the model.
When you have properly defined all Workspace objects, the fourth
Checklist item will be completed.
Section Pipes
The fifth Checklist item, Section Pipes, is only used when the Analysis
Type is Transient. The details of pipe sectioning are related to the
transient solution method, the Method of Characteristics (MOC)
discussed in Chapter 9. In summary, the pipe must be broken down into
computing sections so the MOC can be used. There are strict criteria on
how this is done, and the Section Pipes window automates this process.
The Section Pipes window can only be opened after required pipe input
and fluid properties has been entered, as these data impact how the
sectioning is performed.
Transient Control
The sixth Checklist item, Transient Control, is only used when the
Analysis Type is Transient. Here the model start and stop times are
defined, as well as what data to save to the output file and estimates of
model run time and output file size.
Steady Solution Control

The solution control parameters you can specify are tolerances for
pressure and flow rate balance, two numerical relaxation parameters, and
the number of iterations to perform before stopping (see Figure 5.6).
8/10/2007 4:10:28 PM Page 139

Figure 5.6 The Steady Solution Control window allows control

over parameters that affect the Steady-State Solver
To open the Steady Solution Control window, select Steady Solution

Control from the Analysis menu. The parameters that control the
behavior of the Steady-State Solver are closely linked to the solution
methodology employed by AFT Impulse. Therefore, a detailed
discussion of the features in the Steady Solution Control window is
presented in Chapter 8.
The default Steady Solution Control settings have been chosen to
provide the most robust solution settings applicable to most models. In
general, you do not need to modify these parameters. Change them only
when you are comfortable with the effect the changes will have on your
analysis or if you have a model that does not converge with the default
settings.
Output Control
Most engineering analyses involve processing a very large amount of
information, some of which is critical but most of which is of lesser

manual (si) g.doc
importance. The key to accurately and efficiently analyzing engineering
systems is the proper identification of the critical information that affects
the system; when this is accomplished, resources and attention can be
focused where they will have the most benefit. Too much secondary
information can obscure the critical information and result in lost time
or, worse, erroneous conclusions.
AFT Impulse recognizes this dilemma and provides the Output Control
window for customizing your output. The Output Control window is
available on the Analysis menu and is shown in Figure 5.7. This window
is described in detail in Chapter 4 and will be described only briefly
here. The Output Control window allows you to specify the following:
• Pipe and junction steady and transient output parameters to be
included in the output
• Engineering units in which the output parameters will be expressed
• The order in which the output parameters will appear
• Content and order of summary reports
• Which pipes and junctions to show in the output
• Changes in parameter values between junctions
• Title appearing on the output report
• Model reference information for detailed documentation purposes
• Formatting preference of the numbers which appear in the output
parameters
• Where to direct the output once it has been obtained
8/10/2007 4:10:28 PM Page 141

Change output order

using this tool
These parameters are

shown in the output
These parameters are not

shown but may be added
Available parameters may

be shown alphabetically
or by category
Description of
output terms
Change units here
Database connection
Figure 5.7 The Output Control window selects parameters for

Output. The list of available parameters (on the left)
can be viewed alphabetically or by category.
System Properties
The System Properties window gives you control over the fluid and
environmental properties that influence your pipe system analysis.
Figure 5.8 shows the System Properties window, which is accessed
through the Analysis menu.
The three fluid properties that are required for all analyses are density,
dynamic viscosity and liquid bulk modulus of elasticity. If only a steady-
state analysis is being performed, the liquid bulk modulus of elasticity
will not be used although input is still required. Fluid vapor pressure is
an optional specification.
These properties can be entered by hand when the fluid is selected as
Unspecified, or they can be obtained from the database list of AFT
Standard or Chempak fluids. When selected from the list, density,

manual (si) g.doc
viscosity and bulk modulus cannot be edited as they depend on the
specified temperature (and pressure if Chempak or ASME Steam
Tables).
Figure 5.8 The System Properties window is where you choose

the fluid(s) for your model. The case above uses an
AFT Impulse Standard fluid.
System fluid property variation

The System Properties window allows you to choose from one of two
fluid property variation models. The first fluid property variation model
is for constant fluid properties. When this model is chosen, all pipes in
the system are assigned the fluid properties you specify in the System
Properties window.
The second model is for variable properties. In this model, the fluid
properties assigned in the System Properties window are used as the
default fluid properties for each pipe. You can then assign different fluid
properties within each Pipe Specifications window. This allows you to
model systems that may be subject to large temperature variations or
systems that contain more than one fluid.
8/10/2007 4:10:28 PM Page 143

If you choose a variable fluid property method, you can also constrain
the density to remain constant by selecting the Always Use Constant
Density checkbox. By maintaining a constant density, the hydraulic
grade line concept retains meaning, even though the viscosity may vary.
Density and dynamic viscosity

The fluid density and dynamic viscosity can be obtained from
handbooks, special databases or literature, vendors, or from your own
measurements. AFT Impulse allows these parameters to be entered in
whatever units you find most convenient.
Bulk modulus of elasticity

The fluid bulk modulus of elasticity can be difficult to find, although it
can be calculated with density and pressure data. The definition of bulk
modulus is (Wylie, et al, 1993, p. 5):
ΔP
KB = ρ
Δρ
Vapor pressure and cavitation

The vapor pressure does not affect the steady-state solution, and the
transient solution can be obtained without vapor pressure. However,
certain important phenomena occur when the fluid static pressure drops
to the vapor pressure. These phenomena all result from cavitation.
Cavitation occurs when the fluid static pressure drops to the fluid vapor
pressure. In steady flow this is usually a result of flow through a
restricted area of some kind. In transient flow it frequently occurs
because of reflected pressure spikes that are negative in magnitude.
Small bubbles or “cavities” of vapor tend to occur and persist until the
fluid moves to a location where the static pressure is increased
sufficiently above the vapor pressure to cause them to collapse. In
addition to causing undesired noise, cavities that collapse in proximity to
a structural element can cause severe damage to the structure.
Pumps generally have a required NPSH (Net Positive Suction Head) or
NPSP (Net Positive Suction Pressure) that must be provided by the pipe
system to prevent cavitation in the pump. Pump cavitation may lead to
degraded pump performance, or, in the worst case, it may damage or

manual (si) g.doc
even destroy the pump. In addition to pump cavitation, cavitation may
occur downstream of restrictions such as valves and orifices.
The fluid databases

The AFT Standard Database offers to the user 10 common fluids. The
optional Chempak Database offers approximately 700 fluids, and also
offers liquid and gas mixture capabilities. Water properties from the
built-in ASME Steam Tables are also available.
The AFT Standard Database

The AFT Standard Database offers property data based on temperature
or, if entered by the user, solids concentration.
To add a fluid into your model, select it from the list at the top and
choose the Add to Model button. The fluid will appear in the lower list
of Fluids in Current Model. Enter a temperature and click the Calculate
Properties button to display the density, viscosity, bulk modulus and
vapor pressure.
The AFT Standard Database supports two different independent
variables. Most commonly, the independent variable will be temperature.
All fluids provided by AFT are temperature dependent. User created
fluids can depend on either temperature or solids concentration. Fluid
properties dependent on solids concentration are useful for modeling
sludge in pipe systems. The choice of which independent variable to use
is offered in the Add New Fluid window opened from the Fluid
Database window.
More information on the custom fluids is given in Chapter 7.
The ASME Steam Tables Database

The ASME Steam Tables Database offers users access to water
properties obtained from the IAPWS Industrial Formulation 1997 for the
Thermodynamic Properties of Water and Steam (ISPWS-IF97), (see
ASME Press in References). This database is selected by choosing the
option, at which point “ASME ’97 Water” is automatically displayed in
the Fluids in Current Model area.
8/10/2007 4:10:28 PM Page 145

The Chempak database
The optional Chempak database offers the user approximately 700
fluids. In addition, it offers non-reacting mixture calculations.
A Chempak fluid is added the same way as a Standard fluid, by selecting
it in the list at the top and choosing the Add to Model button (see Figure
5.9). You also have the option of creating a mixture of Chempak fluids
by choosing the Create New Mixture and Add button. This will open the
Create Mixture window shown in Figure 5.10. Here you can specify a
mixture of fluids in any composition ratio you desire.
Figure 5.9 The System Properties window is where you choose

the fluid for your model. The case above uses
methanol, a Chempak fluid.
The basis of all mixtures is either mass or mole, and is specified on the
System Properties window itself. There are no limits to the number of
components in a particular mixture.
The mixtures you create here are referred to as pre-mixtures because you
are specifying the composition before the model is run.

manual (si) g.doc
Once a Chempak mixture has been defined, the mixture data can be
exported to a data file which can be imported by other AFT applications
or users.
Accuracy option
There is one accuracy option for Chempak fluids and mixtures. The
State Property Accuracy option influences the number of elements used
in curve fits for calculating property data. Typically you should use the
high accuracy option.
Figure 5.10 The Create Mixture window (opened from the System
Properties window) allows you to create predefined
mixtures for the model.
The ASME Steam Tables Database

8/10/2007 4:10:28 PM Page 147

ASME Press in References). This database is selected by choosing the
option at the top, at which point “ASME ’97 Water” is automatically
displayed in the Fluids in Current Model area.
Viscosity models
The default viscosity model is Newtonian, which applies to a wide
variety of important liquids including water. Some liquids exhibit a
dependence of viscosity on the fluid dynamics, and these fluids are
called non-Newtonian. When one of these model is selected, additional
input parameters are required (see Figure 5.11).
AFT Impulse offers four non-Newtonian viscosity models. Two of the
models apply to pulp and paper. These are the Duffy model and the
Brecht & Heller model. Also available are models for Power Law fluids
and Bingham Plastics. See Tilton, 1997 for more information on these
viscosity models.
Also available are models for Power Law fluids and Bingham Plastics.
See Darby, 2001 for more information on these viscosity models.
More extensive discussion of these viscosity models is given in Chapter
8.
Non-Newtonian flow in non-pipe elements

The non-Newtonian viscosity models are designed for calculating
pressure loss in pipes. But what about other elements such as valves?
There does not seem to be any standard calculation methods for such
elements. AFT Impulse therefore offers an approximate method. At your
option, standard K values and/or loss data will be modified by ratioing
the provided loss values (assumed to be Newtonian) by the non-
Newtonian and Newtonian friction factors. See Chapter 8 for more
discussion.
If you wish to use these correction factors, select the appropriate option
in the provided fields.

manual (si) g.doc
Figure 5.11 Non-Newtonian fluid data on System Properties.
Atmospheric pressure
AFT Impulse uses a default atmospheric pressure of 1 standard earth
atmosphere, or 101.325 kPa. You are free to change this value. This
pressure is also used to convert absolute pressures to gauge pressures.
Keep in mind that atmospheric pressure varies with altitude. If you are
doing one project at Sea Level and another at 2500 meters you should
adjust the Atmospheric Pressure. See Figure 5.12 for the System Data
tab.
Gravitational acceleration
The default for gravitational acceleration is 1 standard earth
gravitational acceleration, or 9.81 m/s2. You can change this to a
multiple of standard earth accelerations (that is, number of g's) according
to the design environment for your system. In principle, the body force
on a fluid system does not have to be due to gravity. See Figure 5.12 for
the System Data tab.
8/10/2007 4:10:28 PM Page 149

Figure 5.12 System Properties window System Data tab
Transition Reynolds Numbers

AFT Impulse allows you to modify the transition Reynolds Numbers.
The default values are Reynolds Number less than 2300 is laminar and
greater than 4000 is turbulent. In the transition zone a linear
interpolation is assumed. This applies for friction factors. See Figure
5.12 for the System Data tab.
You can modify these transition Reynolds Numbers if you choose.
Editing the AFT Standard fluid database

AFT Impulse provides a fluid database for numerous common fluids. If
the working fluid is not in the database, you may elect to enter the fluid
properties in a custom database for later recall.
To enter or modify fluid properties in the fluid list database, click the
Edit Fluid List button or select Fluid Database from the Database menu.
The fluids you define are incorporated into the fluid list as if they were
native to AFT Impulse. The fluid data is saved to disk and read in when
AFT Impulse is loaded. More information on building custom fluid
databases into AFT Impulse is given in Chapter 7.

manual (si) g.doc
If you want to change the data for one of the AFT Standard fluids you
can select the fluid, click Add To Database, then enter a new name for
the fluid. You can now edit this copy of the original fluid data (the
original will still exist).
Finding object definition status

In order to complete the fourth checklist item you must properly define
all objects on the Workspace.
The Show Object Status feature helps you identify which objects on the
Workspace are defined and which are not. This feature can be toggled on
and off by selecting it from the View menu or by clicking the flood light
button on the Toolbar.
Turning Show Object Status on displays the undefined junction and pipe
ID numbers in the Undefined Color, which is red by default. The color
can be changed through the Workspace Preferences window. Show
Object Status helps you rapidly assess where your model requires
additional information.
Leaving Show Object Status always toggled on is not recommended for
large models, as it will slow the Workspace graphics response. When
this feature is on, every Workspace graphical operation you perform
must be accompanied by a check of all pipe and junction connections.
This check takes time, and could degrade your graphics performance.
The most efficient way to use Show Object Status is to turn it on only
when you have finished adding and arranging the Workspace objects. If
you need to graphically edit your model, it is best to turn Show Object
Status off, perform your graphical editing, then turn it back on when you
need it.
List Undefined Objects

The Undefined Objects window can be opened from the View menu or
Workspace Toolbar. It lists all undefined pipes and junctions and
displays the undefined properties as well. See Figure 5.13.
8/10/2007 4:10:28 PM Page 151

Figure 5.13 The Undefined Objects window displays undefined

pipe and junctions and the undefined properties of
each.
Section Pipes
In a transient simulation, a common time step must be used for all pipes.
The Method of Characteristics (MOC) requires that each pipe section
satisfies the following:
Δx
a= (5.1)
Δt
where a is the wavespeed in the pipe, Δx is the length of the section and
Δt is the time step. The length of a pipe section is just the length of the
pipe divided by the number of sections in that pipe:
L
Δx = (5.2)
n
The time step can be obtained by combining Equations 5.1 and 5.2:

manual (si) g.doc
L
Δt = (5.3)
na
The maximum time is obtained from the pipe with the smallest value of
L/a. This pipe will have one section (n = 1) and thus the time step will
be:
L
Δt max = (5.4)
a
min
In any pipe system there will be one pipe that is the controlling pipe.
The controlling pipe is that pipe which has the least number of sections,
frequently only one. Once the controlling pipe is identified, the time step
is determined by solving Equation 5.4 for Δt. Then the number of
sections in the remaining pipes is obtained from:
Li
ni = (5.5)
ai Δt
The Section Pipes window automatically determines which pipe is the
controlling pipe. It also provides searching tools that allow the user to
identify a sectioning strategy that has acceptable error levels.
Error is introduced at this point because of an approximation that is
usually required in multi-pipe systems. Because the number of sections
in pipes is derived from Equation 5.5, typically the number of sections,
n, will not be a whole number. Since partial sections in a pipe cannot be
modeled, an alternative must be found.
The wavespeed for each pipe is the least certain parameter. Therefore,
the wavespeed in each pipe is allowed to depart slightly from its original
calculated value in order to cause n to be a whole number. Introducing
this error term modifies Equation 5.5 to the following:
Li
ni =
(1 ± ψ )ai Δt
where ψ is the accepted uncertainty in wavespeed up to ±15% (Wylie,
et al., 1993, pp. 54).
8/10/2007 4:10:28 PM Page 153

Figure 5.14 The Section Pipes window automates the process of

obtaining appropriate sectioning criteria
Figure 5.14 shows the Section Pipes window. The controlling pipe (#8)
is shown at the right. The sectioning criteria are shown at the upper left,
where the average and maximum sectioning errors are shown. The table
at the bottom shows the selected sectioning strategy and resulting
modified wavespeed. The resulting time step size is shown at the right.
Numbering convention
Once a pipe is broken into sections, computation takes place where the
sections join together. These are called pipe stations. The total number
of stations in a pipe is equal to the number of sections plus 1. The station
at the beginning of the pipe is numbered zero. See Figure 5.15.

manual (si) g.doc
Computing
stations 0 1 2 3 4 5 6
Pipe sections 1 2 3 4 5 6
Figure 5.15 The computing stations are at the pipe section

boundaries. The number of stations is one more than
the pipe sections.
To obtain the pressure difference across a junction, compare the outlet

computing station of the junction’s upstream pipe to the first computing
station of the junction’s downstream pipe (see Figure 5.16).
J1 J2 J3
P1 P2
P1 J2 P2
1,0 1,1 1,2 1,3 1,4 2,0 2,1 2,2 2,3
Figure 5.16 The pressure drop across the valve at J2 in the AFT
Impulse model at the top is given by the difference
between the solutions at Pipe 1, station 4 and Pipe 2,
station 0.
For junctions with more than one connecting pipe, the principle is the
same (see Figure 5.17).
8/10/2007 4:10:28 PM Page 155

J1 J3
P1 P2
J2
P3
J4
P2 2,0
1,0 P1 2,1
1,1 J2 2,2
1,2 2,3
2,4
1,3
3,0 P3
3,1
3,2
3,3
3,4
Figure 5.17 Pipe station convention for AFT Impulse model. Pipes
1, 2 and 3 have 3, 4 and 4 pipe sections, respectively.
Because there is no loss at the J2 branch, the
pressure solution for Pipe 1, computing station 3 (i.e.,
1,3) is the same as both Pipe 2, computing station 4
(i.e., 2,4) and Pipe 3 computing station 0 (i.e., 3,0). The
flow solutions at these three station will sum to zero
at all times (in the absence of vapor cavitation).

manual (si) g.doc
Transient Control
Figure 5.18 shows the Transient Control window. The first tab is an area
called the Transient Solver Control area. This area offers numerous input
fields detailed in the following descriptions.
Figure 5.18 The Transient Control window configures the

behavior of the transient solver and output file data
Start and stop time

The Start Time is the time at which you want the simulation to begin.
Usually this will be zero, but you can define the starting time as any time
that is convenient for you.
Note the junction transient data that is time-based is referenced to the
start time. That is, if you want the simulation to begin at -10 seconds, the
initial transient data point should be at -10 seconds. The Stop Time is of
course when you want the simulation to end.
Shown right below is the Time Step Size. This is selected in the Section
Pipes window, and is statically shown here for reference. Along with the
8/10/2007 4:10:28 PM Page 157

Start and Stop Times, the Total Time Steps can be determined. This is
shown statically below the Time Step Size.
Save output to file…

When AFT Impulse runs a simulation, the transient output data that is
generated is written to an output file. This is a file with a .out extension.
There are two options in the lower left that allow you to specify how
frequently data is written to the output file. In the Figure 5.18 example,
there are 21468 time steps that will be calculated in the simulation. The
first option, "Save Output to File Every Time Step", will write 21468
time steps of data to file if chosen.
Output files can become extremely large if the data is saved for every
time step, and in such cases the second option (in Figure 5.18) "Save
Intermediate Output to File…" is a better choice. With this option AFT
Impulse will skip over writing output to file at certain time steps. It will
write it on increments of time step as specified in the input box. It is
recommended this value never be set above 25. In Figure 5.18, results
for every tenth time step will be written, for a total of 2146 time steps.
When output data is skipped, it will not appear in the output results or
graphs. The justification for skipping over data is that over small time
increments the transient behavior is sufficiently linear that the graphed
results, for example, will appear the same as when all data is graphed.
It should be noted that AFT Impulse tracks the maximum and minimum
values for pressure and flowrate, and does this for all computations
whether saved to the output file or not. Thus a review of the maximum
and minimum results in the Output window will alert you to whether any
significant data was skipped by using the intermediate output to file
feature.
Use variable pipe resistance

Traditional waterhammer modeling has ignored the effect of varying
pipe resistance (i.e., friction) with flowrate. This is for several reasons.
First, for highly turbulent flow (i.e., large Reynolds numbers) the pipe
resistance does not depend on flowrate at all. On Moody charts this is
sometimes referred to as the "fully rough zone". Second, for turbulent
flows that are not in the fully rough zone, the dependence of pipe
resistance is weak. Third, the waterhammer phenomenon is not affected

manual (si) g.doc
very much by relatively small changes in pipe resistance. And finally,
introducing varying pipe resistance complicates the calculation process.
Therefore, accounting for varying pipe resistance is not important in
many waterhammer applications. One application where it is important
is when the initial steady conditions have no flow in some pipes. For
instance, when the transient being modeled is a pipe with a closed valve
which is being opened. In such cases, it is difficult to obtain the steady-
state pipe resistance because the flow is zero and thus no Reynolds
number exists. Thus it is probably best to use variable resistance.
If you choose to not use variable resistance, the pipe frictional resistance
obtained in the steady-state part of the simulation will be used as a
constant in the transient part of the simulation. On the other hand, if you
allow the resistance to vary, the pipe friction will differ for each pipe
section because the flowrate (and hence Reynolds number) is varying
along the pipe due to the waterhammer transient. Further, the resistance
will vary with time as the flowrate changes.
If variable resistance is modeled, the required run time increases.
Experience suggests the run time increases by about a factor of three.
Model transient cavitation

When the transient pressure drops below the vapor pressure, transient
cavitation will occur. You can model this using the cavitation models
provided. In AFT Impulse 4.0 there is only one model and it is called the
Discrete Vapor Cavity Model. If you do not model cavitation, then the
fluid pressure can go below the vapor pressure and, additionally, below
absolute zero pressure.
If the steady-state results indicate cavitation is occurring, this will cause
a problem for the transient simulation and an artificial transient may be
generated. If so, a warning will be given.
Stop run if artificial transient detected

Artificial transients occur when there is a mismatch between the steady-
state results and the initial transient time step calculation. Many steps
have been taken to avoid artificial transients, but it is always wise to
ensure they do not exist. If you want the transient simulation to
immediately stop if artificial transients are detected, then select the
option to have the simulation stop. If you choose to not stop the
8/10/2007 4:10:28 PM Page 159

simulation, the artificial transients will be displayed in the warnings
section when the simulation is complete.
Including data in transient output file

Earlier in this chapter it was discussed how one can skip saving output
for certain time steps in order to minimize output file size. One can also
choose to skip saving output for certain pipe stations. By default, all
inlet and outlet pipe stations are saved but no junction data is saved.
Here we will discuss how to specify what data is saved and the
advantages and disadvantages associated with these choices.
Saving pipe output

The Pipe Station Output tab allows you to specify which pipe station
data is saved for each pipe. For pipes you have six choices for each pipe:
1. All Stations
2. Inlet and Outlet
3. Inlet Only
4. Outlet Only
5. None
6. User Specified (Up to Five)
As mentioned, the default is to save data for all inlet and outlet pipe
stations. If the user is interested in Animation, the first choice to save All
Stations provides the most animation flexibility. The last choice, User
Specified, allows one to specify up to five stations that will be saved to
the output file.
By selecting in the table your choices, you can save all stations for some
pipes and no stations for the others. If you want to change settings for all
pipes, use the "Change All Pipes To" button.
Disadvantages to saving all stations

When saving all stations, the output file size may be increased
dramatically. In addition, saving output to file takes a lot of time, and
run times can be increased significantly if all stations are saved. In such
cases, a good compromise is to save only the inlet and outlet stations.

manual (si) g.doc
Advantages to saving all stations
If one wants to see the pressure profile along a pipeline at certain times
during the simulation, then one must save all the pipe stations. However,
if one is interested only in the maximum and minimum pressures, these
are always saved to the file regardless of how many stations are saved.
The maximum and minimum profile can be graphed in the Graph Results
window.
Saving junction output

The Junction Output tab allows you to save transient junction data. As
mentioned, by default no junction output is saved. In general, most data
for junctions can be obtained from the pipes. For instance, the flowrate
through a valve over time can be obtained by looking at the upstream
pipe flow rate. However, certain types of junction data is not obtainable
in this way. In such cases, the data can be saved separately to file and
reviewed in the output or graphed.
Here are some examples of junction data that can be saved:
• Surge tank liquid surface height, inflow and total liquid volume
• Gas accumulator volume and pressure
• Spray discharge outflow
• Vacuum breaker valve inflow and total volume
• Pump speed
You can select which junction data you want to save by selecting the
Junction Output tab and then selecting the junctions in the list.
If you want to save junction data by default, select the check box on the
Junction Output tab and junction data will always be saved. The only
exception is branch junctions, which in most cases have no transient
data.
Defining force sets

In order to process force data, the locations where the forces are to be
calculated must be defined. The forces may be defined as differential
forces, or as point forces.
8/10/2007 4:10:28 PM Page 161

Forces are defined on the Force Output tab in the Transient Control
window, as shown in Figure 5.19. New force sets are added by clicking
the New button. Force sets can be duplicated and deleted using the
appropriate buttons on the tab, as well as moved up and down the list of
force sets by using the reorder buttons. By right-clicking on the start pipe
or end pipe in a force set, a popup window displaying input information
about the pipe is displayed.
Force sets are defined by a Start Node and an End Node. The node
locations are selected from the pipe stations calculated during the Pipe
Sectioning process. For each node, the user must select the pipe where
the force is to be calculated, and specify the distance from the start of the
pipe (Station 0) to the location of the node.
In the example shown in Figure 5.19, the first force set (Node 1) is
defined with the Start Node located on pipe P1, at a distance of 0 meters
from the start of the pipe. The End Node is located along pipe P3 at a
distance of 6 meters from the start of the pipe. The closest pipe station to
the specified distance along the pipe will be selected automatically. In
this example, the nearest pipe station to the Start Node is Station 0, and
the nearest pipe station to the End Node is Station 2 (at 6 meters).
Once a force set is defined, AFT Impulse can calculate the differential
force between the two defined points, and the resulting forces can be
displayed in the Graph Results window, or they can be exported to a
Force/Time file which can be imported into the CAESAR II pipe stress
analysis program. A typical application of force set data would be to
select the pipe endpoints between a pair of pipe elbows.
A point force can be defined by selecting the Point option in the Force
Set column. When a user defines a point force, the cells for the end node
are disabled as they have no meaning. For point forces, an ambient
pressure can be specified against which this force will act. The ambient
pressure can be different for each force set if desired.
Estimated file size and run time

At the bottom of the Transient Control window (see Figure 5.18) is an
estimate of the output file size and runtime.

manual (si) g.doc
Figure 5.19 Force set data is defined on the Forces tab in the
Transient Control window
Calculating and outputting unbalanced forces

The preceding section discussed creating force sets in Transient Control.
Here the calculation and output of such forces will be discussed.
Including friction and momentum

When calculating the resulting force unbalance between two pipe
stations, multiple forces come into play. It is traditional to calculate the
force imbalance based solely on the difference in pressure between
points. Since the pressure can differ based on elevation, the
hydropressure is included in the force balance in order to adjust the
pressure difference so this important effect does not bias the pressure
difference.
What is typically not taken into account is the effect of fluid momentum
change and friction forces. Each of these will bias the pressure
difference and hence force calculation. If these effects are included in
8/10/2007 4:10:28 PM Page 163

the force balance, then the pressure difference can be adjusted to account
for them and hence a force calculation based purely on imbalanced
forces can be obtained.
To understand this, consider the steady-state pressure difference between
elbows as shown in Figure 5.20. It is clear there is a pressure difference
between nodes A and B. The difference in this case is due solely to
frictional pressure drop in the pipe and across the valve. However, as the
flow is steady-state, all forces at Node A are balanced by Node B and
hence there is no force unbalance. Thus even though there is a pressure
difference, the pressure difference is caused by the friction which in fact
acts as another force in the opposite direction and of equal magnitude
such that all forces in the x-direction balance to zero.
Node A Node B
Flow
y
PA
Pressure
PB
Distance
Figure 5.20 Steady-state pressure difference between elbows
If the frictional force is not included in the force balance calculation,

then the steady-state situation will show a force unbalance which is
obviously not correct. Similarly, during transient conditions the
frictional force should be included to calculate the true unbalanced
force. The “Include Friction” option in the Force Data graphing and in
the CAESAR II Export windows account for this (see Figure 5.21).

manual (si) g.doc
Another component of the force balance is the unbalanced momentum at
Nodes A and B in Figure 5.20. In steady state these will be in balance,
but during a transient they usually will not. Figure 5.21 shows the option
to include momentum in the force balance calculation.
When including friction in the force calculation, AFT Impulse can only
output force balances for pipe stations within a single pipe object, or two
pipes which are directly connected with each other (with a single
junction in between).
Should one desire to calculate forces for pipes which are not directly
connected, a sequence of force sets can be created between connected
pipes and summed together. AFT Impulse does not perform this
summation, and hence the user must perform it as a side calculation –
typically with spreadsheet software.
Area changes
In cases where force sets span across two pipes, and the pipes have a
different diameter, the area change itself cannot physically create an
unbalanced force even though the static pressure changes. AFT Impulse
will automatically account for this in the force balance calculation.
See the Examples Help for additional explanation of these issues.
Viewing force data

Force data can be graphed in the Graph Results window. See Chapter 4
for more information.
Exporting transient force results

The transient force data can be exported by selecting the Export
CAESAR II Force File from the File menu. This will open the Export
CAESAR II Force File window (see Figure 5.21). This window allows
the user to define the content of a force/time data file and to export it.
This data file may then be imported into the CAESAR II pipe stress
analysis application.
8/10/2007 4:10:28 PM Page 165

Figure 5.21 This window exports data to file for direct use with
CAESAR II. It is accessed from the File menu in any of
the output windows.
Defining pressures for the system

AFT Impulse needs a pressure specified in the system, to be used as the
reference point upon which all other pressures are based. In certain
cases, however, more than one pressure must be specified. These cases
include systems which contain pressure control valves, flow control
valves, internal relief valves, or fixed flow pumps.
Chapter 12 contains a comprehensive discussion on this topic. See the
Role of pressure junctions.
Using scenario manager

Scenario Manager is a powerful tool for managing variations of a model,
referred to as scenarios. Scenario Manager allows you to:
• Create, name and organize scenarios

manual (si) g.doc
• Select the scenario to appear in the Workspace (the ‘current’
scenario)
• Delete, copy and rename scenarios
• Duplicate scenarios and save them as separate models
• Review the source of a scenario’s specifications
• Pass changes from a scenario to its variants
All scenarios are saved within the model file.
The Scenario Manager window

Illustrated in Figure 5.22, the left side of the Scenario Manager window
contains the scenario tree, displaying the names of the currently defined
scenarios, their relationship to other scenarios and which is the current
Workspace scenario (the current scenario is also identified in AFT
Impulse’s Status Bar). On the right there is a place where one may
record notes related to that scenario.
Those attributes whose value comes from a different scenario are linked,
while those values coming from the selected scenario are not linked.
Linked attributes will inherit changes made in the same object and
attribute of the linked ancestor. This inheritance will occur over as many
generations as the link exists.
Creating, organizing and editing scenarios

The “All Scenarios Defined in Model” portion of the Scenario Manager
window shows the name and relationship of all currently defined
scenarios. Scenarios are referred to as either a parent or child, plus the
Base Scenario, which may be thought of as the root scenario. Upon
opening a new model file, only a Base Scenario will be present. To
create a new scenario below the Base Scenario, select the Base Scenario
by clicking on it, then click the Create Child button and enter a name for
the new scenario.
Figure 5.22 illustrates a model with a Base Scenario with three children,
‘3 in PVC Pipe’, ‘3.5 in PVC Pipe’ and ‘4 in PVC Pipe’, representing
three different size pipe runs. Each of these, in turn, has two children,
‘North hydrant open’ and ‘South hydrant open’, for the two operating
cases.
8/10/2007 4:10:28 PM Page 167

Current
Workspace
scenario
Scenario
Tree
Rename,
delete, clone,
promote &
save
scenarios by
clicking here
Create a new
scenario by
clicking here
Figure 5.22 Scenario Manager Window
‘3 in PVC Pipe’, ‘3.5 in PVC Pipe’ and ‘4 in PVC Pipe’ were created by
first selecting the Base Scenario then clicking on Create Child. ‘North
hydrant open’ and ‘South hydrant open’ were, in turn, created by
selecting each of the above children scenarios for the three different pipe
run sizes, then clicking on Create Child (Note: alternatively, one could
have made the Base Scenario current, created the ‘3 in PVC Pipe’ child,
then selected it and created the ‘North hydrant open’ and ‘South hydrant
open’ scenarios below it, then, with ‘3 in PVC Pipe’ still selected, click
on Other Actions / Clone With Children twice, then rename the new
clones ‘3.5 in PVC Pipe’ and ‘4 in PVC Pipe’). The ‘North hydrant’ and
‘South hydrant’ scenarios may be thought of as grand children of the
Base Scenario. This concept of scenario ancestry is helpful in keeping
track of the differences and similarities between scenarios discussed
later.
A box with a plus sign next to a scenario indicates it may be expanded to
reveal children scenarios while a box with a minus sign indicates the
level is fully expanded with all children scenarios shown, if any. No box
indicates there are no children. Levels may be expanded by clicking on
the plus sign and condensed by clicking on the minus sign.

manual (si) g.doc
Clicking on the Other Actions button reveals a menu providing the
following choices:
• Rename – renames the selected scenario
• Delete – deletes the selected scenario
• Clone Without Children – creates a copy of the selected scenario at
the same level (a sibling)
• Clone With Children – creates a copy of the selected scenario along
with its children
• Promote – raises a scenario one level, i.e. from a child to a
sibling…active only for scenarios more than one level below the
base scenario
• Save Scenario to File Without Children – saves the selected scenario
to a new file without its children
• Save Scenario to File With Children – saves the selected scenario
and its children to a new file
It is important to remember that AFT Impulse’s Edit menu Undo does
not apply to Scenario Manager operations. To undo a Rename scenario
operation, you can rename the scenario again to the old name. To undo a
Clone or Save Scenario operation, you can delete the clone or file
created. Once a scenario is deleted or promoted, there is no way to un-
delete or un-promote the scenario.
Viewing scenario differences

A scenario’s data and that for all of its direct ancestors can be viewed in
the Model Data window (Figure 5.23). This functionality can be enabled
in the Model Data Control window. Output from different scenarios can
also be displayed in the Output window using Output Control features.
8/10/2007 4:10:28 PM Page 169

Figure 5.23 Model Data display of scenario ancestor data. This

functionality is enabled in Model Data Control.
Modifying individual scenarios

When first created, a scenario is identical to its parent. To modify or run
a scenario, it must first be made the current Workspace scenario by
selecting it within the Scenario Manager window, then clicking on the
Load As Current Scenario button. (The name of the current scenario is
displayed in both the Scenario Manager window and in AFT Impulse’s
Status Bar.) Once a scenario is the current scenario, changes may be
made to differentiate it from its parent using any of AFT Impulse’s
editing tools and functions. When an attribute is changed within a
scenario it breaks the link for this attribute. Subsequent changes in the
scenario’s ancestors will no longer affect this attribute.
Properties of a scenario that may differ from its ancestors include:

manual (si) g.doc
• Presence and location of objects – one may add or delete pipes and
junctions within a scenario and the workspace locations may vary
from scenario to scenario
• Pipe and junction specifications – all values set in the specification
windows
• System properties – fluid selected, temperature of fluid, viscosity
model, system data
• Solution control – solution method, tolerances, relaxation
• Output Control
• Transient Control
• Database Manager
• Visual Report Control
Workspace Preferences, Parameter and Unit Preferences, General
Preferences, Toolbox Preferences and Model Data Control apply to all
scenarios.
Passing changes to child scenarios

Scenario Manager will pass changes made in a scenario to linked
attributes of its descendants, allowing multiple scenarios to be modified
in one operation and maintain similarity between scenarios where
desired.
Any attribute whose value is the same as that of the scenario’s parent is
linked to the parent. In turn, if the parent’s attribute value is the same as
that of its parent, then the child’s attribute is linked to its grandparent.
This ancestral linking of attributes can exist across any number of
generations. A change made to a scenario will be passed downward to its
children, grandchildren, etc., as far as the link exists. Changes are NOT
passed upward from child to parent.
Once an attribute is changed within a scenario, the link is broken.
Subsequent changes in the scenario’s ancestors, parent, grandparent,
etc., will no longer affect that attribute.
8/10/2007 4:10:28 PM Page 171

Scenario logic examples
For many users, it is easiest to grasp Scenario Manager when it is
explained how the coding logic is actually implemented. Blank fields for
children, grandchildren, etc., mean to look to the parent for the data. The
Base Scenario never has blank fields (Figure 5.24a). Data only passes
downwards, never upwards.
Diameter Length
Base 3 25
Child #1 __ __
GrandChild #1 __ __
Figure 5.24a Scenario Manager logic – Blank fields for Child #1 and
GrandChild #1 mean that the data is to come from the
parent. If the Base scenario data is changed, all
descendants are changed.
If a child scenario does not have a blank field, then data for that property
is initiated at that scenario level (see Figure 5.24b).
Diameter Length
Base 3 25
Child #1 2 __
GrandChild #1 __ __
Figure 5.24b Scenario Manager logic – Child #1 does not have a

blank field, so its Diameter would be 2, not 3, as would
GrandChild #1.
If a scenario is changed, and its child has different data, then the change
will not pass downwards (see Figure 5.24c).

manual (si) g.doc
Diameter Length
Base 3 25
Child #1 2 __
GrandChild #1 __ __
Diameter Length
Base 6 40
Child #1 2 __
GrandChild #1 __ __
Figure 5.24c Scenario Manager logic – Changing the Base Scenario

Diameter from 3 to 6 would not impact Child #1 or any
descendents in that line. Changing the Length from 25
to 40 would also change the length in Child #1,
GrandChild #1, and any descendents of GrandChild
#1.
If a child scenario has data that is different than the parent, its children
cannot relink to the parent (see Figure 5.24d).
If a child scenario’s data, which was previously changed and is thus
different from the parent, is changed back to the same vale as the parent,
the inheritance link is re-established (see Figure 5.24e). Its descendant’s
link is also re-established.
8/10/2007 4:10:28 PM Page 173

Diameter Length
Base 3 25
Child #1 2 __
GrandChild #1 3 __
Figure 5.24d Scenario Manager logic – Even if the GrandChild #1

has the same Diameter as the Base, it is not linked to
the Base because it and its parent are not blank.
Diameter Length
Base 3 25
Child #1 3 __
GrandChild #1 3 __
Diameter Length
Base 3 25
Child #1 __ __
GrandChild #1 __ __
Figure 5.24e Scenario Manager logic – If the Diameter in Child #1 is

changed to be the same as the Base, it will be
“blanked out” the next time the scenario is loaded and
the link re-established. And so will GrandChild #1, if
its Diameter is also the same.

manual (si) g.doc
Figure 5.24f shows data for two pipe properties across three scenarios.
The data that would be used when each scenario is loaded is as follows:
1. Base scenario
a. Diameter = 3
b. Length = 25
c. Changes to Diameter will not pass downwards
d. Changes to Length will affect only Child #1.
2. Child #1 scenario
a. Diameter = 2
b. Length = 25
c. Changes to Base Diameter will not affect Diameter
d. Changes to Base Length will affect Length
3. GrandChild #1 scenario
a. Diameter = 2
b. Length = 15
c. Changes to Base Diameter will not affect Diameter
d. Changes to Child #1 Diameter will affect Diameter
e. Changes to Base Length or Child #1 Length will not affect
Length
Diameter Length
Base 3 25
Child #1 2 __
GrandChild #1 __ 15
Figure 5.24f Scenario Manager logic – Input data for these

properties explained in the text.
8/10/2007 4:10:28 PM Page 175

Re-establishing broken links
A link may be re-established by returning the attribute to the same value
as that of its parent. This can be done manually by entering the value or
selecting the Copy Data From Pipe list and selecting the Parent Pipe
Data option. Within other data windows it is done by selecting the Same
As Parent option in Solution Control, Output Control, System Properties,
Cost Settings, Database Manager or Visual Report Control.
Since links are identified by comparing attribute values of pipes or
junctions with the same Workspace ID number, renumbering a scenario
will break the links of all pipes and junctions renumbered. Since
numbers must be unique, once a link has been broken by renumbering, it
may not be re-established.
Fast Scenario Changes

AFT Impulse remembers the most recent scenario you had open, and
allows you to quickly change back to that scenario by using the Last
Scenario feature. Last Scenario is found on the View menu and Toolbar.
Special modeling features

The modeling utilities described in this section help you build your
model more rapidly and keep track of what you are doing and what
assumptions you have made.
Waterhammer Assistant
The Waterhammer Assistant can be opened from the Help menu or
Toolbar. The Waterhammer Assistant reviews your model and offers
helpful advice (see Figure 5.25).
By default, whenever a model is run AFT Impulse checks with the
Waterhammer Assistant to see if it has any advice. If it does, the
Waterhammer Assistant is displayed. The user then has the option of
canceling the run or ignoring the advice and continuing. If you do not
want the Waterhammer Assistant to be consulted before each run, you
can specify this choice in the General Preferences window opened from
the Options menu.

manual (si) g.doc
Figure 5.25 The Waterhammer Assistant offers advice on

modeling issues
Workspace Find
The Find window utility shown in Figure 5.26 helps you find a particular
pipe or junction by its ID number and name. The Find utility becomes
more useful for models that expand over multiple Workspace pages.
You can access the Find utility by selecting Find on the View menu or
by clicking the button on the Toolbar. The Workspace is moved so that
the pipe or junction is located near the center of the screen.
8/10/2007 4:10:28 PM Page 177

Figure 5.26 The Find feature for the Workspace is helpful for
locating a specific pipe or junction. It is opened from
the View menu or Toolbar.
Reverse Direction
Each pipe you create has a reference positive flow direction as indicated
by the arrow on the pipe. If you want to reverse the reference positive
flow direction, you can do so by manually moving the pipe endpoints.
However, this procedure is tedious. The Reverse Direction utility toggles
the reference positive flow direction for the selected pipe or pipes.
The Reverse Direction utility is accessed from the Arrange menu or from
the reverse direction button on the Toolbar.
Select Special
Select Special is a tool for selecting or deselecting objects based on
certain criteria (see Figure 5.27) and allows you to quickly select a group
of objects.

manual (si) g.doc
Figure 5.27 You can customize your selections using Select

Special. It is accessed from multiple locations within
AFT Impulse to assist with selecting groups of pipes
and/or junctions.
When first opened, Select Special will select all the objects currently
selected on the Workspace. You may alternatively select objects based
on several other criteria available from the Selection Type drop-down
menu. Objects in the Pipe Selections or Junction Selections list may be
selected or deselected based on the criteria by clicking the Select Pipes
or Select Junctions button or the Deselect Pipes or Deselect Junctions
button.
Invert will deselect all selected objects and select all unselected objects.
Clicking on the Select All or None buttons will result in all of the
objects being selected or deselected respectively.
Several select/deselect operations may be performed in succession.
When finished, click on OK button.
8/10/2007 4:10:28 PM Page 179

This tool is available from several windows and will select the items on
the window from which it is opened. For example, it could be used to
display only the closed valves in the Output window when it is accessed
from the Show Selected Pipes/Junctions tab in the Output Control
window.
Tip: To quickly show only a part of the model in the Output window,
first select the objects on the Workspace for which you want to see
output. Then, on the Show Selected Pipes/Junctions tab in the Output
Control window, select the Workspace buttons in the pipe and junction
areas. Finally, click OK in the Output Control window and only the
pipes and junctions you selected on the Workspace will be shown in the
Output.
Special Conditions
Some junctions have Special Conditions that alter the normal state of the
junction. For example, a valve can be closed, a pump turned off, or a
relief valve opened.
The special conditions are set in one of three ways:
1. By selecting the junction(s) on the Workspace and choosing Special
Conditions from the Edit Menu
2. By selecting the junction(s) on the Workspace and clicking the
Special Condition icon on the Toolbar
3. By opening the junction’s Specifications window and clicking the
appropriate condition on the Optional tab
The junctions that have special conditions set are shown using a special
symbol before the ID number (an “X” by default). This symbol can be
customized in the Workspace Preferences window.
As shown in Figure 5.28, when the special condition results in a section
of the model being closed or “turned off”, the pipes which will have zero
steady-state flow are displayed as dotted lines. This indicates visually
that they are cut off from flow; the junctions are outlined with a dotted
line. The default dotted line can be changed in the Workspace
Preferences window.

manual (si) g.doc
An X next to
the junction
number
indicates a
special
condition has
been set
Dashed lines
for pipes and
around
junctions
indicate closed
model sections
Figure 5.28 Workspace display when a valve is closed using the

Special Condition. Notice the “X” next to the valve
number (J14), which is shown in red by default. Also
notice the connecting pipes shown with dashed lines
and junctions with dashed outlines which indicate
nonflowing sections of the model.
Special conditions with no transient data

If a Special Condition is applied to a junction without any transient data,
then the transient solution will retain the Special Condition. For
instance, if a Special Condition is set for a valve junction, the steady-
state solution will solve with the valve closed. In addition, the transient
solution will run the entire simulation with the valve closed as well
because there is no transient data entered which implies the valve state is
not intended to change.
Special conditions with transient data

If a Special Condition is applied to a junction that has transient data,
then the transient solution will ignore the Special Condition. It is then
incumbent upon the engineer to relate the Special Condition to the initial
transient data.
8/10/2007 4:10:28 PM Page 181

For instance, assume the user wants to simulate the transient that occurs
during a valve opening. The steady-state solution will have the valve
closed, and this is modeled by using a Special Condition for the valve
junction. In such a case, the Cv data entered on the valve junction's Loss
Model tab will be ignored.
On the Transient tab, a transient can be entered. The first data in the
table should then be Cv = 0, which corresponds to a closed valve. From
there the Cv can be increased above zero as desired to open the valve.
If the user were to keep the Cv equal to zero, then that would be
equivalent to the previous case where the user set the Special Condition
but did not enter any transient data.
Pump special conditions

Pump Special Conditions are slightly more complicated than for valves.
In fact, pump junctions have two types of Special Conditions. The first is
to turn the pump off and have no flow through it. This is called "Pump
Off No Flow". The second type of Special Condition turns the pump off
but allows flow to go through the pump. This is called "Pump Off With
Flow Through".
Why would one want to use one Special Condition rather than the other?
The simple answer is that the first Special Condition is more appropriate
for positive displacement type pumps and the second for centrifugal
pumps.
Here's why. When a positive displacement pump is turned off and has a
pressure difference across it, it will usually act like a closed valve and
not allow flow to go through it. Thus the first Special Condition would
be most appropriate. For instance, assume one wants to model the
transient that occurs during the startup of a positive displacement pump.
The best way to do this would be to use the first Special Condition with
no flow, and then input a flowrate transient in which the first data point
is zero flow.
On the other hand, when a centrifugal pump is turned off and has a
pressure difference across it, in the absence of other valving which
prevents flow the pump will usually allow flow to go through it. Thus
the Special Condition that allows through flow would be most
appropriate. For example, assume one wants to model the transient
during the startup of a centrifugal pump. One would use the second

manual (si) g.doc
Special Condition, and, in conjunction with a pump curve entered at
100% speed, input a speed transient with the initial speed as zero.
Special conditions only applied to junctions

Special Conditions in AFT Impulse can only be used for junctions.
Unlike AFT Fathom, Special Conditions cannot be applied to pipes. The
reason is fairly simple. In AFT Fathom a pipe Special Condition closes
the pipe. But in reality one does not close a pipe. One instead closes a
valve that is in the pipe. For user convenience AFT Fathom overlooks
this fact.
In contrast, AFT Impulse does not have as much freedom in such cases.
When transients occur, there are reflections off system components such
as valves. If one were to "close a pipe" in AFT Impulse, it would not be
known exactly where the closure exists and thus the transient reflections
could not be properly modeled.
Transient special conditions

Junctions which accept transient data also support Transient Special
Conditions. The default for all junctions is None. The other choice is
Ignore Transient Data. This allows one to have no transient initiation at
that junction without having to delete the transient data.
When a Transient Special Condition is set to Ignore, a # symbol is
displayed adjacent to the junction number on the Workspace. This
symbol can be customized in the Workspace Preferences window.
No reflections – infinite pipe

The Assigned Pressure and Assigned Flow junctions support a third type
of Transient Special Condition. It is called No Reflections (Infinite
Pipe). This condition is appropriate for the entrance to a long pipe line
from which no reflections will occur during the simulation period.
When this Transient Special Condition is specified, a * symbol is
displayed adjacent to the junction number on the Workspace. This
symbol can be customized in the Workspace Preferences window. A
discussion of infinite pipe behavior is given in Chapter 12.
8/10/2007 4:10:28 PM Page 183

Special Condition Ignore

The Vacuum Breaker Valve, Gas Accumulator, Liquid Accumulator,
Surge Tank, and Relief Valve junctions have a unique kind of Special
Condition called Ignore. This makes the steady-state and transient
solvers ignore the presence of the junction. Thus not only is the data
ignored, the junction itself is ignored. This is convenient when locating
one of these devices by trial and error. Rather than having to delete the
junction, it can be ignored.
Here is an example of when this feature might be useful. Assume that
one is trying to find the best location for a gas accumulator. If one does
this with a single gas accumulator junction, it will be required to run a
case, view the results, then relocated the junction by changing the pipe
lengths and rerunning. If one uses the Special Condition Ignore, one can
place multiple gas accumulator junctions in the model and then set all of
them except one to Ignore so that only one of the junctions is used at a
time.
Merging models
Two models can be merged together using the Merge command on the
File menu. Here’s how:
• Open the first model
• Choose Merge on the File menu
• From the file list, select the model to be merged
If there are duplicate numbers, the objects in the second model will be
changed. The second model will be selected (highlighted) so that you
can move it into position (be careful not to click off the selected portions
or the second model may be deselected).
Tip: To move a selected group of objects, use a junction icon, instead of

a pipe, to drag the group. This will avoid missing the pipe and clicking
on an empty area of the Workspace which will deselect your group.
Merging models with multiple scenarios

If the model you are merging from has multiple scenarios, only the base
scenario can be merged. If you want to merge one of the children of that
model, do the following:

manual (si) g.doc
1. Open the model which you want to merge.
2. Open the Scenario Manager.
3. Select the child scenario you want to merge.
4. On the Other Actions button, choose the Save Scenario to File
Without Children.
5. Enter a name and click OK. This process creates a new Impulse
model with the desired scenario as the base scenario.
6. Open the original model and merge in the new model file created in
Step 5.
Print Preview/Special
The Print Preview/Special window shown in Figure 5.29 allows you to
access special print features such as Print Preview and Fit to One Page.
It is accessed from the File menu.
Figure 5.29 The Print Preview/Special window is opened from the

File menu or Toolbar
The Print Preview/Special window also offers you a preview of the print
content and format for all primary windows. By clicking the Print
Preview button you can review your printed material at several zoom
states. Individual pages can be printed.
8/10/2007 4:10:28 PM Page 185

The Fit On One Page feature causes the graphic image generated from
the Workspace or Visual Report windows to be compressed onto a single
page. This is a convenient feature for creating reports of your model.
AFT Impulse supports whatever printer page sizes are supported by your
printer, so use of legal paper or larger may allow you to get more visible
details. The graphics can also be centered vertically and/or horizontally.
You also can fit your content onto multiple pages.
The Print Preview/Special window allows you to set the page
orientation, which can also be changed once you are in the Print Preview
window itself.
Several other special printing features affect the Workspace and Visual
Report. The Only Print Selected Pipes and Junctions works with
Workspace printing by selectively printing the pipes and junctions
currently selected on the Workspace.
The Include Hidden Pipes and Junctions feature includes hidden pipes
and/or junctions on the Workspace or Visual Report in the printed
content. By default this box is not selected, and hidden pipes and
junctions are not included in Workspace and Visual Report printouts.
By default, AFT Impulse prints the junction icons for the Workspace and
Visual Report with the same colors as displayed on the screen. It also
prints the background color as the same color as your Workspace or
Visual Report background. You can also specify that the junction icons
be printed in black and white (still with the color background) or that the
entire printout be black and white.
If you added a Workspace Picture into the background, you can include
this picture in the printout.
Finally, if you have a grid displayed on the Workspace or Visual Report,
you can include the grid in the printout.
Transfer Results to Initial Guesses

AFT Impulse’s iterative Steady-State Solver discussed in Chapter 8
takes initial guesses (also called initial conditions) and applies the
principles of mass and momentum balance to improve the guesses. This
is called iteration. It continues to refine the solutions until agreement
with the governing equations is obtained.
After a converged answer is obtained, you can choose to take these
results and use them as the initial guesses for a subsequent run. This

manual (si) g.doc
causes much faster convergence in subsequent runs because you are
starting the Steady-State Solver at or very close to the actual solution.
This feature is called Transfer Results to Initial Guess, and it is on the
Output window Edit menu and Toolbar.
Batch runs
AFT Impulse scenarios or models can be run in batch mode, a feature
that is especially useful for running a group of scenarios or models
during lunch or overnight.
¾ To run models in batch mode, follow these steps:

1. Click “Start Batch Run” on the File menu.
2. Select the Batch Run Type. This is whether you want to run
scenarios from the current model or different model files. If the
current model does not have any scenarios then only the second
option will be available.
3. If the Batch Run Type is “Scenarios in Current Model”, then choose
the Add Scenarios button and select the scenarios you wish to run.
4. If the Batch Run Type is “Models from Different Files”, then the
models can be selected individually using the “Add Model Files”
button, or loaded from a Batch File (by Choosing the “Load From
Batch File” button). A Batch File is a text file listing all models you
would like to run. At any time a selected model file can be removed
by clicking the “Remove Selections”. The current list of models can
be saved to a Batch File for use in the future by clicking the “Save
List to File” button.
5. Use the Output Options in the lower left to modify style and content
of output, and whether to save the output data to a file and/or send it
to a printer or Adobe PDF file. All reports will use the specified
font.
6. Click the Start Run button to begin the batch run.
AFT Impulse then opens each scenario or model file in sequence,
automatically runs each scenario/model that has a completed Checklist,
and sends results to the specified destination. All pertinent information
that is generated during the batch run (such as error messages) are
automatically displayed when the final scenario/model is finished.
8/10/2007 4:10:28 PM Page 187

The entire batch run can be canceled at any time by clicking the Cancel
button in the Solution Progress window.
Print Content
With either the Model Data or Output window active, you can open Print
Content from the View menu (see Figure 5.30), allowing you to specify
the content and format of Model Data and Output reports.
You can include Model Reference Information (entered in the Output
Control window) to document your model or make important comments
you want kept with the model.
Figure 5.30 Print Content window specifies the print content of

the Model Data or Output window. It is opened from
the View menu or Toolbar.
The various tables in the General section, Pipe section, and Junction
section can be selectively printed. All Pipe and Junction Notes are
shown in the Model Data General section and can be printed at your
option.

manual (si) g.doc
Empty columns in the three Model Data tables can be ignored in the
printed report. You can also have each of the Print Content sections start
on a new page. Both of these are formatting options in the Print Format
window.
The printer font and style of your choice can be selected for the printed
report. Large reports can be condensed by selecting small fonts. A good
choice for large reports is Arial True Type Regular Style at a 6 or
8-point size.
Extended Model Check

The connections and settings of each pipe and junction are exhaustively
checked using the Extended Model Check item on the View menu. This
is also helpful if the model seems to not be connected correctly.
Math calculator
For convenience, AFT Impulse offers you quick access to the Windows
Calculator; simply select Calculator from the View menu or press
CTRL+K in any specifications window.
Working with the steady and transient solvers

The term Solver conveys a mathematical equation solution method
whose purpose is to model some physical phenomenon. The
implementation is in the form of algorithms and subroutines which solve
the equations. AFT Impulse 4.0 has two Solvers because it is solving two
problems.
Steady-state solver
The Steady-State Solver obtains a balanced steady-state solution to the
system being modeled. It employs a Newton-Raphson matrix method
that is iterative nature. An initial guess to the steady-state solution is
made, and the Newton-Raphson method progresses the guess towards a
final solution that agrees with the governing equations within some
tolerance. Details are given in Chapter 8.
8/10/2007 4:10:28 PM Page 189

Control of the Steady-State Solver is given in the Steady Solution
Control window opened from the Analysis menu. Here the user can
specify items such as tolerances, relaxation and matrix methods.
The Steady-State Solver is based on that in AFT Fathom™, a
commercial steady-state modeling software developed by Applied Flow
Technology. AFT Impulse contains most of the capability found in AFT
Fathom for solving systems with known density, and can optionally be
used for steady-state modeling without reference to transient modeling.
If you want to run a model in steady-state mode only, on the Analysis
menu specify the Analysis Type as "Steady Only".
When you specify the model as Steady Only, certain input data normally
required for transient models are no longer required. Some examples are
wavespeed for pipes, liquid bulk modulus, and input for the Section
Pipes window and Transient Control on the Analysis menu. In fact,
when modeling Steady Only the Checklist adjusts to only require four
items to be completed.
Recommendations on Steady Solver

It is recommended that before running a transient simulation, the user
first run the model in Steady Only mode to help ensure the model is
assembled properly. Once confidence is established in the steady-state
results, the Transient feature can be enabled and run.
Transient solver
The Transient Solver models the surge transients in a pipe system using
the Method of Characteristics (MOC). Unlike the iterative Newton-
Raphson matrix method used for steady-state solutions, the MOC is a
direct solution method that marches in time. Details are given in Chapter
9.
To function properly, the MOC requires an accurate steady-state solution
to provide the initial conditions. Thus when one runs a transient
simulation in AFT Impulse 4.0, the Steady-State Solver is always run
first. The steady-state solutions are then automatically passed on to the
Transient Solver, which uses them as the initial conditions for the MOC.
This will reduce artificial transient problems to a bare minimum.

manual (si) g.doc
Artificial transients
The fundamental cause of artificial transients is mismatched steady-state
conditions. One would hope that artificial transients would never occur
in AFT Impulse 4.0, but that is not the case. If they do occur, there is one
of two causes. First, it may be an indication of a problem with the steady
solution. This could be a problem in the Steady-State Solver, or an input
problem that could not be identified when the model was built.
Secondly, a user could potentially specify mismatched steady-state and
transient behavior of a junction. For example, for a valve junction the
user inputs a steady-state valve Cv on the Loss Model tab. In addition,
the user can enter a transient valve Cv on the Transient tab. The initial
transient Cv should match the steady-state Cv on the loss model tab. If
they do not, then an artificial transient would occur.
AFT Impulse attempts to identify such inconsistencies when the data is
first input and then again before the model is run. If for some reason
AFT Impulse does not catch this problem and bring it to the user's
attention, then an artificial transient will occur.
8/10/2007 4:10:28 PM Page 191


manual (si) g.doc
CHAPTER 6
Pipe and Junction Specifications

Windows
This chapter discusses the Pipe and Junction Specifications windows,

the features they offer, and how you use these features to obtain a proper
model of your system.
Each pipe and junction on the Workspace has an associated
Specifications window. In the Specifications window, you enter the
engineering data that represents the pipe or junction you are modeling.
The Specifications window also contains AFT Impulse's interpretation of
the connectivity within your model.
Each of the nineteen junction types has a separate Specifications
window. The Specifications window is the primary vehicle for entering
input data for the components of your model. The Global Edit windows,
discussed at the end of this chapter, can be used to change data for
multiple pipes and junctions all at the same time.
You can access an object's Specifications window in six ways:
1. Double-clicking the object on the Workspace
2. Selecting the object on the Workspace and pressing ENTER
3. Selecting the object on the Workspace and clicking the Open
Pipe/Jct Window icon on the Toolbar
4. Double-clicking the left column of the appropriate table in the
Model Data window (where the object ID number is shown)
5. Clicking the Jump button in another Specifications window

manual (si) g.doc
6. Double-clicking the connected pipe or junction number in another
Highlight feature
The highlight feature identifies required input data in the Specifications
window. You can toggle the highlight feature on and off in any of the
following ways:
• Double-click within the specifications window (outside of the
tabbed area)
• Press the F2 function key
• Choose Highlight in Pipe and junction Windows from the Options
menu
The highlighting feature may be especially useful when you are first
learning to use AFT Impulse or when you are having difficulty obtaining
a defined object status.
Jump feature
By clicking the Jump button in a Specifications window, you can move
directly to the Specifications window of a selected pipe or junction.
Pipe Specifications window

Each pipe on the Workspace has engineering information associated
with it. Each pipe must have an ID number, a length, a diameter, a
friction model specification, and two connecting junctions. For transient
analyses, it must also have a wavespeed. Figures 6.1a-e show the Pipe
Specifications window tabs.

manual (si) g.doc
Chapter 6 Pipe and Junction Specifications Windows 195
Figure 6.1a Pipe Specifications window Pipe Model tab.
Figure 6.1b Pipe Specifications window Fittings & Losses tab.
8/10/2007 4:10:28 PM Page 195

Figure 6.1c Pipe Specifications window Design Alerts tab.
Figure 6.1d Pipe Specifications window Optional tab.

manual (si) g.doc
Common input parameters

The common input parameters are located above the tabbed folders
(Figure 6.1a-e) and are discussed below.
Pipe number
Every pipe must have an ID number. When you create a new pipe, AFT
Impulse assigns a default pipe number. The pipe number can be
changed; however, duplicate pipe ID numbers are not accepted, and ID
numbers must be greater than zero and up to 30,000. The pipe numbers
you choose have no bearing on the model connectivity, direction, or
layout. They are merely convenient identifiers.
Pipe name
A name can be assigned to each pipe for reference purposes. The default
name is simply “Pipe.” Names do not need to be unique. The name can
be shown on the Workspace, in the Visual Report and in the Output
window. By using names, attention can be called to important or critical
pipes.
Figure 6.1e Pipe Specifications window Fluid Properties tab

(available only when Variable Fluid Properties are
selected in System Properties).
8/10/2007 4:10:28 PM Page 197

Copy Data From Pipe
The Copy Data From Pipe dropdown list allows you to conveniently set
the pipe properties to be the same as any existing pipe. You can copy all
or part of the parameters from an existing pipe in the model to the
current pipe by choosing one of the pipes in the list and then specifying
which parameters to copy. Only the parameters you choose will be
copied to the current pipe.
The Copy Previous button functions the same way as the Copy Data
From Pipe except that the last pipe edited is used. This is convenient
when there are several pipes that have the same or similar parameters.
Scenario same as parent

If the current model is a child scenario, in the Copy Data From Pipe list
there is an option to set the pipe to Parent Pipe Data. This will change
the selected pipe to be the same as its parent.
Connected junctions
The connected junctions area shows you AFT Impulse's interpretation of
the junctions that are connected to the particular pipe. Every pipe must
be connected to two junctions in order to be completely defined. When a
junction does not yet exist at one of the pipe endpoints, “None” appears
in the area where the junction ID number would normally be displayed.
All pipes on the Workspace have a reference positive flow direction. The
reference positive direction is indicated on the Workspace by an arrow
on the pipe. Based on the reference direction, AFT Impulse identifies an
upstream junction and a downstream junction.
To determine the properties of the connected junctions, you can use the
Inspection utility. To inspect, position the mouse pointer over the
connected junction ID number and hold down the right mouse button.
Pipe Model
The Pipe Model tab (Figure 6.1a) allows you to input geometric data for
the pipe.

manual (si) g.doc
Pipe length
The Length entry represents the length of pipe between the two
connecting junctions. The length entered in the Pipe Specifications
window has no relationship to the length of pipe you draw on the
Workspace. The Workspace pipe is merely an abstract used for
conceptual modeling purposes. The pipe drawn on the Workspace is
similar to the conceptual pipe drawn on an engineer's calculation pad.
Every pipe must be assigned a length. This length can be any positive
value. To the right of the Length field, a dropdown list allows you to
choose the units of the pipe length.
Pipe material
The Pipe Specifications window allows you to leave the pipe material
unspecified or to choose a pipe material from a list.
AFT Impulse provides default pipe material data for several different
kinds of pipe. In addition to these pipe materials, you may add your own.
Access to the Pipe Material Database is available on the Database menu.
By entering your own custom pipe materials, you can build a pipe
material database that is saved to disk and read in during startup. Once
you enter materials data, AFT Impulse treats the pipe materials as if they
were native to the program. More detail on building a custom pipe
material database is given in Chapter 7.
When you select a pipe material from the materials list, two things
happen:
1. In the Size area, the dropdown lists for nominal pipe Size and Type
are enabled
2. In the Friction Model area, a default friction data set and
accompanying handbook pipe roughness value is entered
The Type entry allows you to select the specific type, class, or schedule
of material for the nominal pipe size. Once the type is selected, the pipe
diameter is shown in the diameter field. This diameter cannot be edited.
In addition, if transient modeling is enabled and the user has selected
Calculated Wavespeed, the pipe wall thickness, modulus of elasticity,
and Poisson ratio are shown. All of these parameters come from the pipe
material database, and cannot be edited. These three parameters are
needed for wavespeed calculations.
8/10/2007 4:10:28 PM Page 199

If you select the Pipe Material as Unspecified, these three fields are
editable and you will need to enter data in each for wavespeed
calculations.
If User Specified Wavespeed is selected, these three fields will be
disabled no matter what Pipe Material is specified, as they are not
needed when the user supplies the wavespeed. Finally, if the model is for
steady-flow only, the three fields will always be disabled as they have no
bearing on steady-state models.
In the Friction Model area, the value shown is not to be construed as the
correct pipe roughness value for your application. The pipe roughness
depends on many factors, most importantly the age of the pipe. AFT
Impulse offers the pipe material roughness as a guide. You will want to
enter the roughness value that is most representative of your application.
Multiple friction data sets can be compiled for each pipe. For example,
friction data for pipes with different ages can be set up. These different
data sets are then made available in the Data Set dropdown list. More
information on creating multiple friction data sets is given in Chapter 7
in the section on custom pipe materials.
Pipe diameter
The inner diameter of the pipe is entered in the Inner Diameter field in
the Pipe Specifications window. For pipes whose material is unspecified,
the diameter and units may be entered directly.
If the material is specified from the Pipe Material list, the diameter
cannot be edited; the value is derived from the database for the selected
material, size, and type.
The reduction in diameter due to scaling can be accounted for by
entering a percent reduction in the ID Reduction - Scaling field. Zero
percent would represent no reduction.
Pipe wall thickness

The wall thickness is entered in the Size area of the Pipe Specifications
window. Wall thickness is required for pipes in which the wavespeed is
calculated.
If the material is specified from the Pipe Material list, the wall thickness

manual (si) g.doc
material, size, and type. This field is disabled if only a steady-state
model is being run or if the User Specified Wavespeed is selected.
Pipe modulus of elasticity

The modulus of elasticity is entered in the Size area of the Pipe
Specifications window. Modulus of elasticity is required for pipes in
which the wavespeed is calculated.
If the material is specified from the Pipe Material list, the modulus of
elasticity cannot be edited; the value is derived from the database for the
selected material. This field is disabled if only a steady-state model is
being run or if the User Specified Wavespeed is selected.
Pipe Poisson ratio

The Poisson Ratio is entered in the Size area of the Pipe Specifications
window. Poisson Ratio is required for pipes in which the wavespeed is
calculated.
If the material is specified from the Pipe Material list, the Poisson ratio
material. This field is disabled if only a steady-state model is being run
or if the User Specified Wavespeed is selected.
Friction Model
A common problem in specifying pipe roughness is that long term
operation of the pipe often results in deposits forming on the inside of
the pipe. This is commonly referred to as fouling. Fouling increases
roughness and, in severe instances, decreases the area available for fluid
transport (also causing greater pressure loss in the pipe). You are
encouraged to seek out applicable pipe design data when specifying pipe
roughness in your model.
The Friction Model area of the Pipe Specifications window offers eight
methods for specifying frictional models. Methods 1, 2, 4 and 5 all rely
on the Darcy-Weisbach friction factor method of calculating pipe
pressure drop.
1. Absolute roughness – AFT Impulse's default method is to specify
the roughness as an absolute average roughness height. Values of
pipe roughness can be found in many pipe handbooks or from
manufacturer's data.
8/10/2007 4:10:28 PM Page 201

2. Relative roughness – Some pipe roughness specifications are given
as a relative roughness. In this case, the roughness height is divided
by the pipe diameter. Relative roughness has no units, so when you
enter a relative roughness, the Units dropdown list is disabled.
3. Hazen-Williams – The Hazen-Williams method uses an empirical
factor (CHW) to relate the flow rate to the pressure drop in the pipe.
This method is still in common use in the field of water distribution.
You can specify the Hazen-Williams factor for any pipe.
4. Explicit Friction factor – If you already know the friction factor for
the pipe, you can enter the value explicitly.
5. Hydraulically smooth – You can also specify a pipe as
hydraulically smooth. Modeling a pipe as hydraulically smooth
implies that its roughness is negligible. However, having a small
roughness is not the same as being frictionless. Rather, the pipe
friction factor follows the hydraulically smooth curve in the
turbulent region of a standard Moody diagram. You are encouraged
to consult a standard fluid mechanics textbook for additional
discussion of this subject.
6. Resistance – You can specify a pipe resistance in terms of head loss
and volumetric flow rate. The pipe head loss will then be as follows:
dH = RQ2.
7. Frictionless – For modeling purposes, it is occasionally useful to
model a pipe as having no friction. There are limitations to where
such a pipe can be located in your model. Basically, a frictionless
pipe cannot be the sole connection between two branching junctions.
8. MIT Equation – The MIT Equation is appropriate for crude oil. See
Chapter 8 for a discussion of how this method is implemented.
9. Miller Turbulent – The Miller Turbulent method is appropriate for
light hydrocarbons. See Chapter 8 for a discussion of how this
method is implemented.
Because you are free to specify the friction model for each pipe
individually, you can mix and match the friction models throughout your
pipe system if you choose.
If you find it more convenient to use a certain type of friction
specification, like the Relative Roughness method for example, you may
want to change the defaults in the Parameter and Unit Preferences
window to reflect this. When you change the default method, the other

manual (si) g.doc
methods are still accessible, but the Pipe Specifications window always
comes up with your preferred method selected.
The preceding friction models all apply to Newtonian viscosity models.
The viscosity model is specified in System Properties window. If the
user chooses one of the non-Newtonian models, the preceding friction
data may not be used in the calculation.
A more detailed theoretical discussion of the different friction models is
given in Chapter 8.
Pipe Support
The Pipe Support selection is made from the provided drop down list to
obtain the constant c1. This constant is required for pipes in which the
wavespeed is calculated. AFT Impulse provides seven common pipe
support types.
The meaning of the c1 constant and its relationship to the seven support
types is given in Chapter 9.
Pipe Wavespeed
The wavespeed is the speed at which transient events propagate through
the pipe. Typical values for the wavespeed are 700-1600 m/sec (2000-
5000 ft/sec).
This parameter depends on the liquid acoustic velocity and the pipe
material and support. AFT Impulse will calculate the wavespeed for you,
or you can provide your own user specified wavespeed.
Fittings & Losses

The Fittings & Losses K factor is one way to enter loss factors in your
model. The Fittings & Losses K factor is added to the friction loss
calculated by the Steady-State Solver. This is useful when a pipe section
contains elbows and other fittings.
You can use any of the losses supplied in the Impulse database or you
can enter a specific loss factor. To enter losses, click the Fittings &
Losses tab (Figure 6.1b) and then click the Specify Fittings & Losses
button to open the Fittings & Losses window (Figure 6.2). This window
allows you to specify losses from a table of standard losses.
8/10/2007 4:10:28 PM Page 203

Because many of the fittings & losses are dependent on pipe diameter,
you must specify a diameter before you can open the Pipe Fittings &
Losses window.
The types of pipe losses are divided into several categories represented
by tabs. To add a loss to the pipe, click the table cell which describes the
fitting or loss and type in the quantity. When you are done entering all
the losses in each of the categories, press OK and you will see a
summary list in the Pipe Specifications window.
Figure 6.2 Fittings & Losses window opened from the Pipe
The Area Changes loss does not change the size of the pipe, it only puts
in a loss factor based on the pipe diameter and the area ratio of the loss.
When this option is selected, notice the illustration of how the loss is
defined, where Apipe is the area of the pipe and As is the area of the loss.
These area changes are useful to model fittings between a pipe and
another junction like a pump or control valve.

manual (si) g.doc
The abbreviations used in the Fittings & Loss window are listed at the
bottom of the window. You can double-click the image to see the
reference information for the loss.
Design Alerts
A design alert (Figure 6.1c) allows you to specify certain maximum
and/or minimum parameters for the pipe. After running the Steady-State
or Transient Solvers, any exceeded value will be identified in the
Warnings section of the Output window and the specific cell containing
the value in the Pipes section will be highlighted. Design Alerts can be
cross-plotted in the Graph Results window vs. the actual results. One
good example where this is useful is for cross-plots of maximum and
minimum allowed operating pressures.
Optional input
There are several optional input parameters (Figure 6.1d). These are
described in the following sections.
Initial Flow Rate Guess

The Initial Flow Rate Guess is the initial guess at the mass flowrate,
volumetric flowrate, or velocity solution through the pipe. Specifying a
good initial guess of the flow solution decreases the number of iterations
the Steady-State Solver must execute in order to arrive at a converged
solution. However, specifying initial guesses takes time and good
judgment.
If you do not specify an initial flow guess, AFT Impulse implements a
proprietary method of generating first guesses. This method is usually
sufficient to get the Steady-State Solver going in the right direction so a
converged solution results. However, the method of generating first
guesses is not 100% effective, and there may be times when you will
have to specify the first guess in order to obtain a solution. More
discussion on this subject is given in Chapter 8.
Workspace display
You have the option of specifying what to display on the Workspace.
You can choose to display the pipe ID number, the pipe name, the pipe
nominal size, and the pipe schedule. This is helpful in large models to
8/10/2007 4:10:28 PM Page 205

reduce the text on the Workspace and to focus on what you think is
important. The default setting for this can be changed in the Workspace
Preferences window.
Design Factor
You can specify two design factors for each pipe – one for the pipe
friction and one for fittings & losses that you add to the pipe.
Design factors allow you to specify multipliers on the pipe friction
calculations and fittings & losses that are applied by the Steady-State
and Transient Solvers during the solution process. Design factors are
helpful for adding safety margins to your design calculations.
Pipe Line Size and Color

The default thickness of the line shown on the Workspace and default
pipe color is defined in the Workspace Preferences window. You can
change the displayed line thickness and color for each pipe individually
if you desire.
Parallel Pipes
Here you can specify that the pipe is one of a number of parallel pipes.
The pipe properties thus represent a single pipe. This is useful, for
example, to represent a bundle of tubes in a heat exchanger.
Support for Partially Full Pipe

There is checkbox which indicates the pipe supports partially full pipe
calculations. This means partially full along the axial direction, as
opposed to partially full along the radial direction. This function only
works for pipes that have a slope, and when the pipe is partially full it
drains from the end with the higher elevation.
If this feature is not selected, the pipe will always be assumed to be
liquid full. Selecting this feature implies that one end of the pipe is
connected to a junction at its higher elevation endpoint which allows gas
to flow in, thus allowing the pipe to drain.
If the option is selected, and the entry is 100%, it means the pipe is
initially 100% full, but can drain if backflow occurs at the high elevation
endpoint. The percent full can also be set to something less than 100%.
In this case, the pipe is assumed to be partially full initially. This is only

manual (si) g.doc
valid when the steady-state condition for the pipe is such that there is no
flow. Because if there were flow in that pipe during steady-state, the
pipe would not be partially full.
This feature only works with pipes which connect to a vacuum breaker
valve, spray discharge, assigned pressure or exit valve junction at the
high elevation end.
When selected, the predicted level of the liquid surface and other related
parameters can be graphed in the Graph Results window. When pipes
can be partially full, more accurate predictions result when the pipe is
modeled with more sections. A single section is not recommended. At
least two sections should be used, and more if possible.
Intermediate Pipe Elevations

Pipe inlet and exit elevations are obtained from the upstream and
downstream junction elevation data. The pipe elevations in between are
assumed to vary linearly.
In general, the pipe elevation profile will have no effect on the hydraulic
results, and the linear assumption is suitable. However, even though the
hydraulics are predicted accurately, the intermediate pipe pressures are
not (unless the elevation is in fact linear). As long as these pressures are
within the allowables for the pipe, and do not drop below vapor pressure,
then no problems arise. However, if the pressures do exceed these
conditions, it is important to find this out.
One way to account for this is to break the pipe up into smaller
individual pipes, with additional junctions (usually branch junctions)
connecting them. These junctions can be set at the proper elevations, and
the actual elevation profile can thus be modeled. However, this too has
ramifications, especially for transient models. With the increased
number of pipes, this can force a shorter time step selection, which can
drastically increase model runtime.
The solution to the just described dilemmas is to use intermediate pipe
elevations. Here an elevation profile can be entered directly for the pipe,
and only a single pipe is required as long as the diameter is constant.
To use intermediate pipe elevations, tick the check box for Use
Intermediate Pipe Elevations, and enter the elevation profile. To do this,
the pipe length entered on the Pipe Model tab (Figure 6.1a) is required.
AFT Impulse will ensure the total length in the intermediate elevation
8/10/2007 4:10:28 PM Page 207

area matches the length on the Pipe Model tab. It does so by fixing the
length of the final section to the difference between the sum of the
lengths in the table and the actual length entered by the user. The end
elevation is also fixed to that of the downstream junction. If the user
later modifies the downstream junction elevation, the pipe’s final
elevation section will be automatically updated.
Fluid Properties
This tab is only visible when variable properties are being modeled
(Figure 6.1e). Variable property modeling is enabled in the System
Properties window.
For models with variable fluid properties, you can assign different fluid
properties to each pipe in the model. This is convenient for modeling
systems that are subject to temperature variations and systems that have
multiple fluids. The pipe fluid properties are displayed in the Pipe
Specifications window, and can be modified by selecting the Fluid
Properties button. Besides varying the fluid properties, the viscosity
model can varied from pipe to pipe. Thus some pipes in the model can
be modeled as Newtonian, while others as non-Newtonian.
Notes
Each pipe can have notes associated with it. This is useful for listing
assumptions, reference documents, drawings, measurements, etc.
Status
The Status tab shows which required input data has not been entered.
Junction Specifications windows

The nineteen different types of junctions on the Toolbox offer a great
degree of freedom in assembling a pipe network model. While there is
some redundancy in capabilities between the junctions, each of the
nineteen types of junctions offers specialized features to allow you to
prepare a good conceptual model of the physical system of interest.
Table 6.1 compares the features of the junction types.

manual (si) g.doc
Table 6.1 Comparison of junction capabilities
Junction Type Connecting Special Custom Window
Pipes Conditions Database Format
Area Change 2 √ 1
Assigned Flow 1 √ √ 1
Assigned Pressure* 1/1-25 √ 2
Branch 2-10 √ 2
Check Valve 2 √ 1
Control Valve 2 √ √ 1
Dead End 1 √ 1
Gas Accumulator 1-25 √ √ 2
General Component 2 √ 1
Liquid Accumulator 1-25 √ √ 2
Pump 1-2 √ √ 1
Relief Valve 1-2 √ √ 1
Reservoir 1-25 √ 2
Spray Discharge 1-4 √ √ 2
Surge Tank 1-25 √ √ 2
Tee/Wye** 3 √ 2
Vacuum Breaker Valve 1-2 √ √ 1
Valve 1-2 √ √ 1
Volume Balance 2 1
*Assigned Pressure junction connects to only 1 pipe for static pressure,
and up to 25 for stagnation
**Tee/Wye junction has a unique format closer to Format #2.
There are nineteen junction Specifications windows, one for each

junction type. Each junction Specifications window falls into one of two
basic window formats, depending on the number of connecting pipes
allowed for that junction type.
8/10/2007 4:10:28 PM Page 209

Note: AFT Impulse does not prevent negative flow through junctions.
Should this occur, AFT Impulse will use the same loss factor or pressure
drop data referenced to the defined upstream pipe, just as in forward
flow. Because it is possible that the loss factor or pressure drop data
would not be the same for both flow directions, the flow solution could
be misleading. For this reason it is important that you either properly
define the pipe flow directions for junctions or pay close attention to the
flow directions that result. Should negative flow occur in any junction,
AFT Impulse will give a warning in the output.
Format #1:
Junctions with one or two connecting pipes
The first basic window format is for junction types that allow only one
or two connecting pipes (see Figure 6.3). While junctions do not have an
explicitly defined flow direction like pipes do, typically those with two
pipes or less adopt a direction from the connecting pipes. For these
junction types, upstream and downstream pipes are recognized by AFT
Impulse based on the reference positive flow direction of the connecting
pipes. The upstream and downstream pipes are displayed separately near
the top of the window.
In this first basic window type, it is generally important to have the
pipe’s reference positive flow directions specified in the physically
correct directions. A good example of this is a Pump junction which
will add pressure to the system in the direction of positive flow through
the connecting pipes. The Pump junction interprets where to add the
pressure based on the directions of the connecting pipes.
To inspect the connected pipe information in this window format,
position the mouse pointer over the connected pipe ID and hold down
the right mouse button.

manual (si) g.doc
Fixed display of
connecting pipes
Figure 6.3 Specifications window format #1 (fixed connecting

pipe display).
Format #2:
Junctions with more than two connecting pipes
The second basic window format for junction Specifications windows is
for junctions that allow more than two connecting pipes (see Figure 6.4).
These junctions typically allow up to twenty-five connecting pipes; the
exception is the Tee/Wye junction, which allows only three.
Because the number of pipes connected to a junction may vary, the
second basic format uses a table sized according to the number of
connecting pipes.
An example of this second window format is the Branch Specifications
window, shown in Figure 6.4, which has three connecting pipes. To
determine AFT Impulse's interpretation of the model connectivity for
this second window format, you can review the contents of the table
areas. There are separate loss factors for flow going into the pipe from
the junction and flow going out of the pipe into the junction. The
8/10/2007 4:10:28 PM Page 211

direction of the pipe in reference to the junction is also shown in the pipe
table.
To inspect connected pipe information in this window format, position
the mouse pointer over the pipe ID number in the left column of the pipe
table and hold down the right mouse button.
Parameters common to all junction Specifications windows

There are several junction parameters common to most junctions
described in this section.
Junction Number
Each junction has an ID number (see Figures 6.3 and 6.4) that you are
free to change. However, no two junction numbers can be the same, and
all junction numbers must be greater than zero and up to 30,000.
Duplicate junction numbers are not accepted. The junction numbers you
choose are arbitrary and have no bearing on model connectivity,
direction, or layout. They are merely convenient identifiers.
Junction Name
Each junction can be assigned a name for reference purposes (see
Figures 6.3 and 6.4). The default name is the junction type. The name
does not need to be unique. The name can be displayed on the
Workspace, Visual Report, Model Data and Output windows.
Junction Elevation
Each junction has an entry for the elevation of the junction (see Figures
6.3 and 6.4). Next to each elevation entry is a set of units. For junctions
that use Format #1, the default inlet elevation is entered by the user, and
the outlet elevation is assumed to the same. A different outlet elevation
can be entered by clearing the “Same as Inlet” checkbox. Different inlet
and outlet elevations will have no effect on the flow solution, but will
offset the local static pressure at the connecting pipe by the hydrostatic
pressure difference.

manual (si) g.doc
Figure 6.4 Specifications window format #2 (pipes displayed in a

table).
Junctions that follow Format #2 have a single elevation and offsets from
that elevation (if they exist) for each individual connecting pipe entered
in the connecting pipe table.
Junction elevations are used to account for pressure changes due to
gravity and thus to allow calculation of absolute pressures in pipes. In a
sense, junction elevations are arbitrarily defined. The magnitude of the
elevation is not important from a solution standpoint, but the relative
elevation difference between junctions is of critical importance because
of its effect on the pressure results.
Elevation is required input for every junction. AFT Impulse uses this
input to determine the relative height changes between one end of a pipe
and the other.
In the Parameter and Unit Preferences window you can set a default
elevation for your system. When a default elevation is specified, each
time you open a new junction Specifications window the default
8/10/2007 4:10:28 PM Page 213

elevation is automatically entered. For systems that reside primarily at
the same elevation, this can save time in entering data.
Database List
The Database List (see Figures 6.3 and 6.4) is a list of all custom
equipment you have entered into the Component database. It displays
custom equipment from both the local and network database sources.
Selection of an item automatically retrieves the data and enters it into the
junction. Changes to the custom input data are not accepted. To stop
using a database selection, choose (None) from the database list.
Copy Data From Jct list

The Copy Data From Jct list (see Figures 6.3 and 6.4) is a list of all the
junctions of the same type in the model. This displays a list of
parameters from which you can choose to apply to the current junction.
Scenario same as parent

If the current model is a child scenario, in the Copy Data From Jct list
there is an option to set the pipe to Parent Junction Data. This will
change the selected junction to be the same as its parent.
Pipe connectivity
In each junction Specifications window (see Figures 6.3 and 6.4), you
can see AFT Impulse's interpretation of your system's connectivity. All
junctions must have the proper number of connecting pipes in order to
be completely defined (refer to Table 6.1).
Jumping to another junction

You can jump to another junction from within the Specifications
window (see Figures 6.3 and 6.4). This will save the current junction
information before jumping to the other junction. This method avoids
having to close the current junction Specifications window before
opening another Specifications window.

manual (si) g.doc
Transient Data
Many junction types have an area where active transient data can be
specified. Table 6.2 lists the junction types.
Table 6.2 Junctions that support transient data

Junction Type Type of Transient
Assigned Flow Flowrate
Assigned Pressure Pressure
Branch Flowrate source or sink
Check Valve Opening/Closing profiles
Control Valve Opening/Closing profiles, pressure or flow
Pump Speed, flowrate, control, trip or startup parameters
Relief Valve Opening/Closing profile
Reservoir Surface level changes
Spray Discharge Discharge opening/closing profiles
Surge Tank Surface Pressure
Valve K or Cv profile for opening and closing in part or full
The active transient data is entered on the Transient Data tab (Figure
6.5). Table 6.2 lists the type of data that can be entered for each junction
type. If a junction does not have any active transient behavior, there is no
need to enter any data on the Transient Data tab.
In most cases, transient data can be entered as absolute values or as a
percentage of steady-state. The distinction is made when choosing
between Absolute Values and Relative to Steady-State Value (Figure
6.5).
The first data point always needs to match the steady-state value which
is usually entered on the tab at the far left. If, for example in Figure 6.5,
the steady-state data for Cv of 1000 is changed to 750, the transient data
will also need to be changed. However, if in Figure 6.5 the Relative
option is specified, the first data point would be 100%, and when the Cv
is steady-state is changed from 1000 to 750 there is no need to change
the transient data since the first data point is still 100%. But it is now
relative to a different steady-state value.
8/10/2007 4:10:28 PM Page 215

Figure 6.5 Example of junction transient data. A here a valve is

being closed over five seconds.
Repeat Transient
If the transient data is periodic, you can enter the data for one cycle of
the period and then tick the check box for Repeat Transient. This will
cause the one cycle of transient data to be repeated once it has reached
the end. The repetition will continue until the end of the simulation.
Transient Special Conditions

There may be occasions where, once having entered the transient data,
you do not wish the transient to activate for a particular case. One option
is to just delete the transient data. A second option is to specify the
Transient Special Condition as Ignore Transient Data. In such a case, the
data can be left in the window for future use, but will not activate during
the run.

manual (si) g.doc
A third type of Transient Special Condition is available for Assigned
Flow and Assigned Pressure junctions. This is the No Reflections
(Infinite Pipe) option. In such a case, the flow or pressure is used only
for steady-state calculations. During the transient the flow or pressure is
not fixed, but varies in accordance with the concept that no transient
reflections occur. All transient data at this junction will be ignored.
The reasons for using this feature are important, but somewhat difficult
to grasp. A lengthy discussion is given in Chapter 12.
Time based and event based transients

In Figure 6.5, there is an area called Initiation of Transient, and the
selection is Time. This means that the specified transient is linked to the
time of the simulation. Specifically, the valve transient data is activated
as soon as the simulation is started. Because the transient data is
constant for one second, the valve does not start to close until the
simulation time reaches one second.
Transients can also be initiated based on events. For instance, we can
change the valve in Figure 6.5 to start closing when the pressure at some
other system location reaches a certain level. Or maybe when a tank has
filled up and thus reached a certain level. In such a case, the time zero
data in Figure 6.5 does not refer to the start of the simulation, but is
relative to when the event occurs. If, for instance, the pressure never
reaches the specified level, the valve never closes.
Chapter 10 discusses in much more detail how event transient modeling
can be used.
Initial Pressure/Head Guess

On the Optional tab of each junction Specifications window (see
example in Figure 6.6), except for Reservoir and Assigned Pressure
junctions, is an area for entering an initial guess at the pressure (or head)
at the junction. This is only the initial guess and the solution will vary
from this value; it does not specify a fixed pressure at the junction. This
guess is used in setting up the steady flow solver.
8/10/2007 4:10:28 PM Page 217

Figure 6.6 The Optional tab offers several optional data inputs
and is very similar for each junction type
Specifying a good initial guess decreases the number of iterations the

Steady-State Solver must execute in order to arrive at a converged
solution. However, similar to specifying Initial Flow guesses in pipes,
specifying initial guesses takes time and good judgment.
The Reservoir and Assigned Pressure junctions do not allow you the
option of specifying an initial guess. Because the pressure is a known
quantity, no guess is needed. The pressure solution is already available.
If you do not specify an initial guess, AFT Impulse implements a
proprietary method of generating first guesses. This method is usually
sufficient to get the Steady-State Solver going in the right direction so a
converged solution results. However, AFT Impulse's method of
generating first guesses is not 100% effective, and there may be times
when you will have to specify first guesses in order to obtain a solution.

manual (si) g.doc
Displaying junction names and numbers

Also on the Optional tab of each junction Specifications window (see
example in Figure 6.6) is the option to display the junction ID number,
the junction name, both, or neither. This is helpful in large models for
reducing the text on the Workspace and focusing on what you think is
important. The default setting for this can be changed in the Workspace
Preferences window.
Special Conditions
Some junctions allow you to set Special Conditions (see example in
Figure 6.6). This feature allows you to override the default behavior of
the junction. For example, the special condition for a Valve junction is
to close the valve. The special condition for a Relief Valve is for it to be
open no matter what the pressure. Not all junction types support special
condition settings.
The junction for which the Special Condition is set will have an ‘X’
placed before the junction ID number. If the special condition is to close
the junction, by default it will be outlined with a dashed line. These
settings may be changed in the Workspace Preferences window.
Chapter 5 discusses how to apply transient data to junctions with steady-
state Special Conditions.
The Vacuum Breaker Valve, Gas Accumulator, Liquid Accumulator,
Surge Tank, and Relief Valve junctions have a unique kind of Special
Condition called Ignore. This makes the steady-state and transient
solvers ignore the presence of the junction. Thus not only is the data
ignored, the junction itself is ignored. This is convenient when locating
one of these devices by trial and error. Rather than having to delete the
junction, it can be ignored.
Design Factor
You can specify a design factor for each junction (see example in Figure
6.6). Design factors allow you to specify multipliers on the junction
pressure loss. These multipliers are applied by the Steady-State and
Transient Solvers during the solution process. In the case of a pump, the
design factor multiplies the pressure rise. Design factors are helpful for
adding safety margins to your design calculations.
8/10/2007 4:10:28 PM Page 219

Note: If you are designing for some minimum condition, you may want
to use the design factors differently for pumps than for other junction
types. If you use 1.1 for all junctions with a pressure loss, you will get an
extra 10% pressure drop. Using the same 1.1 on your pump will give an
extra 10% pressure rise. This will work against the 10% margin on your
pressure losses. If you want to use a pump design factor in this case, it
might be best to use 0.9.
Changing the icon graphic

You may want to change the icon to better represent a specific junction
type. This is accomplished by clicking the Change Icon button on the
Optional tab (see example in Figure 6.6). The Change Icon window
appears, allowing you to choose a new icon from a list of available
icons.
Changing the icon size

You may want to change the size of the junction to distinguish it from
other junctions. This is accomplished by clearing the Use Workspace
Default for Size checkbox and selecting the desired size in the drop
down list (see example in Figure 6.6).
Specifying Base Area

Some junction windows that follow Format #1 allow selection of the
base area for loss models. By default the base area will be the upstream
pipe area. The user has the option of using the downstream pipe area or
directly specifying an area or diameter.
Notes
Each junction can have associated notes. This is useful for listing
assumptions, reference documents, drawings, measurements, etc. Notes
are entered in the provided area on the Notes tab.
Status
The Status tab shows which required input data has not been entered.

manual (si) g.doc
Area Change Specifications window

The Area Change Specifications window is shown in Figure 6.7. The
Area Change junction type must always have two connecting pipes. This
junction type allows you to model the irrecoverable loss that occurs at
the area transition between two pipes, whether expansion or contraction.
The Area Change Specifications window follows the first of the two
basic Specifications window formats, displaying the connecting pipes in
a fixed format. A flow direction through the junction is adopted from the
defined directions of the two connecting pipes. Consistent with AFT
Impulse's convention, the loss factor base area is referenced to the
upstream flow area shown as the Base Area.
Figure 6.7 The Area Change Specifications window
The Area Change Specifications window offers two standard types of

area change losses: the Conical Transition and the Abrupt Transition.
The window shows a schematic of the selected geometry.
The critical parameter that influences the magnitude of the area change
loss is the area ratio between upstream and downstream pipes. Until you
8/10/2007 4:10:28 PM Page 221

have input the information for the upstream and downstream pipes that
allow for flow area determination, the standard Area Change junction
loss factor is not calculated. For conical transition geometries you also
need to specify the conical angle.
If the standard loss factors provided are inadequate for your application,
you may specify your own custom loss factor referenced to the upstream
pipe flow area.
More information on the standard loss factor types for Area Change
junctions is given in Chapter 8.
Assigned Flow Specifications window

The Assigned Flow Specifications window is shown in Figure 6.8. The
Assigned Flow junction type allows you to connect one pipe and to input
a known steady-state and/or transient flowrate entering or leaving at a
particular location in your system. You also can include an irrecoverable
loss with this junction type.
The Assigned Flow Specifications window follows the first of the two
basic Specifications window formats, displaying the connecting pipe
information in a fixed format. Depending on whether you specify the
junction as an inflow or outflow type, the required connecting pipe
(upstream/inlet or downstream/outlet) will be enabled while the other
pipe is disabled. Note that if the flow is specified as Inflow (to the
system) then the reference positive flow direction of the connected pipe
must be away from the junction (i.e. the junction is upstream of the
pipe).

manual (si) g.doc
Figure 6.8 The Assigned Flow Specifications window
Tip: If you see that the required pipe in the Specifications window states
“None” but a pipe is connected to the junction graphically, then check
the Inflow/Outflow setting on the junction or the reference positive flow
direction of the pipe.
The Assigned Flow junction type allows you to specify positive flow
rates as mass flow rates. Negative flow rates are not accepted for steady
flow, although are for transient flow. If, for example, you have an
outflow type junction and your system is physically flowing in, you
cannot assign a negative flow rate to the junction. You must reverse the
connecting pipe flow direction to be consistent with the actual direction
and change from an outflow type to an inflow type.
8/10/2007 4:10:28 PM Page 223

Transient Data
On the transient data tab a flowrate transient can be entered. A special
feature is the ability to model a sinusoidal flowrate transient. Input for
the amplitude and frequency is required. This transient is summed to that
specified in the transient data table. A chopped sine wave takes the
absolute value of the sine function.
More information on transient data, including event transients, is given
earlier in this chapter and in Chapter 10.
Special Conditions
You can set a Special Condition for an Assigned Flow junction, which
will turn the flow off and make it act like a Dead End during steady-
state.
Special features for steady flow

If the steady-state flow is zero at the junction, there may be occasions
where you want the Assigned Flow junction to act like a pressure
junction during the steady-state calculation. This would occur, for
instance, if a known flowrate profile, initially zero, exists at the junction,
and there are no other suitable pressure junctions in the system. In such a
case, the pressure is assigned on the Optional tab (see Figure 6.9).
Usually this feature would be used in conjunction with a Special
Condition on the junction where its flow was turned off during steady-
state.

manual (si) g.doc
Figure 6.9 Optional tab allows junction to act like a fixed

pressure during steady-state if initial flow is zero.
Assigned Pressure Specifications window

The Assigned Pressure Specifications window is shown in Figure 6.10.
The Assigned Pressure junction type allows you to connect up to twenty-
five pipes. The Assigned Pressure junction is convenient for specifying a
known steady-state and/or transient pressure in your system. It also
allows you to specify irrecoverable losses due to the fluid dynamic
effects of the connecting pipe geometry.
8/10/2007 4:10:28 PM Page 225

Figure 6.10 The Assigned Pressure Specifications window
The Assigned Pressure junction type has much in common with the
Reservoir junction. In each case you specify parameters in order to
achieve a known pressure. However, the Assigned Pressure junction
allows you to specify either stagnation or static conditions. This is useful
if, for example, you are modeling a system where the pressure in a pipe
is known and the location is being used as a boundary in the model. If
the measured conditions are for a location with a velocity, then they
represent static conditions and the choice of static properties should be
selected. In this case, the choice of a Reservoir junction would be
inappropriate. Similar to the Reservoir, an Assigned Pressure junction
causes the rest of the system to distribute the flow in a manner consistent
with the defined pressure. You can specify the units for pressure by
selecting from the adjacent drop-down list.
The distance of each connecting pipe (if any difference exists) from the
elevation can be entered in the pipe table.

manual (si) g.doc
If the static pressure option is selected, only one connecting pipe is
accepted. The reason for this restriction is that multiple connecting pipes
could have different flows and, hence, velocities. Different velocities
will change the static pressure, thus a unique static pressure cannot be
specified.
This limitation does not affect specified stagnation pressure because
stagnation pressure does not change with velocity. See Chapter 8 for
more information on this subject.
Transient Data
On the transient data tab a pressure transient can be entered. A special
feature is the ability to model a sinusoidal pressure transient. Input for
the amplitude and frequency is required. This transient is summed to that
specified in the transient data table. A chopped sine wave takes the
absolute value of the sine function.
Branch Specifications window

The Branch Specifications window is shown in Figure 6.11a-b. The
Branch junction allows up to twenty-five connecting pipes. The Branch
junction is the most flexible junction type for building network models
that have flow splits. Other significant features are the ability to impose
a steady-state and/or transient flow source or sink at the junction and the
ability to specify separate loss factors for each connecting pipe. The
Branch junction must have at least two connecting pipes.
The Branch Specifications window follows the second of the two basic
Specifications window formats. A table on the Loss Coefficients tab
displays the connecting pipe information. This table grows in size to
accommodate up to twenty-five connecting pipes. After you add a fifth
pipe a scroll bar appears, allowing you to review and enter loss factors
for all pipes in the table.
8/10/2007 4:10:28 PM Page 227

Figure 6.11a The Branch Specifications window Loss Coefficients

table.
For each connecting pipe, the pipe table lists the reference flow direction
and up to two loss factors. The first loss factor is for physical flow out of
the pipe and into the branch. The second is for physical flow into the
pipe and out of the branch. The loss factors can be specified
independently or left as zero. To edit within the pipe table, simply click
in the appropriate column and row.
The distance of each connecting pipe (if any difference exists) from the
elevation can be entered in the pipe table.
The Branch junction also offers you the option of specifying a flow
source or sink at the junction. According to AFT Impulse's convention, a
source of flow into the system is defined as positive, while a
consumption of flow with flow out of the system (a sink) is defined as
negative.

manual (si) g.doc
Figure 6.11b The Branch Specifications window Optional tab

allows you to specify a steady-flow source or sink.
Transient Data
On the transient data tab a transient source or sink flowrate can be
entered. More information on transient data, including event transients,
is given earlier in this chapter and in Chapter 10.
Check Valve Specifications window

The Check Valve Specifications window is shown in Figure 6.12. The
Check Valve junction type requires two connecting pipes. This junction
type allows you to model the effect of a check valve which closes to
prevent back flow.
The Check Valve Specifications window follows the first of the two
8/10/2007 4:10:28 PM Page 229

a fixed format. The Check Valve junction does not have an explicit flow
direction, but adopts a flow direction from the connecting pipes.
Consistent with AFT Impulse's convention, the loss factor base area is
referenced to the upstream flow area.
Figure 6.12 The Check Valve Specifications window
A check valve is a device that allows flow in only one direction. AFT
Impulse assumes that the Check Valve is initially open. If the flow
solution indicates that forward flow will not occur, AFT Impulse closes
the valve. The valve can close during the steady-state or transient portion
of the simulation.
A check valve Special Condition can be either open or closed.
Specifying losses
You specify the losses for the Check Valve on the Valve Model tab in
the Specifications window. For convenience, you can specify the
constant loss characteristics of a Check Valve as a valve coefficient (CV)

manual (si) g.doc
or as a loss factor (K). Chapter 8 details the relationship between K and
CV.
Forward Velocity to Close Valve

This is the velocity at which the valve starts to close. By default the
velocity is zero. However, some valves are designed to start closing at
some negative velocity. You can specify that here.
Delta Pressure to Re-Open

Frequently after a check valve closes, there is pressure across the valve
at which it will open again. You can specify that delta pressure here.
Transient Data
The closing of a check valve is a complicated transient process because
the mechanical movement is coupled to the hydraulics of the pipe.
Sophisticated models exist in the literature which attempt to model this
complicated process, but these require extensive data about the valve
and typically more simplified models are employed.
AFT Impulse offers a more simplified model, whereby the user can
specify a closing profile transient of Cv or K vs. time. Remember that
when using the K loss model, a value of –1 (negative one) means the
valve is closed. The closing will occur as soon as the closing velocity
criteria is met (discussed above). If no closing transient profile is
entered, the valve will close instantaneously.
You can also specify an opening profile of Cv or K vs. time, which is the
profile the valve will follow if it reopens. If no opening transient profile
is entered, the valve will open instantaneously.
The zero time for the closing transient is relative to when the valve starts
to close because of backflow. For the opening transient, it is relative to
when the valve starts to open.
Control Valve Specifications window

The Control Valve Specifications window is shown in Figure 6.13.
Control Valve junctions are always internal to the system, with two
connecting pipes. This junction type allows modeling of valves that offer
8/10/2007 4:10:28 PM Page 231

special pressure or flow control characteristics at a location in the pipe
system. In addition, the controlled pressure or flow can be changed with
time.
The Control Valve Specifications window follows the first of the two
a fixed format. The Control Valve junction does not have an explicit
flow direction, but adopts a flow direction from the connecting pipes.
Figure 6.13 The Control Valve Specifications window
Control Valve types

You can model four types of control valves:
• Pressure Reducing Valves (PRV’s)
• Pressure Sustaining Valves (PSV’s)
• Flow Control Valves (FCV’s)
• Pressure Drop Control Valves (PDCV’s)

manual (si) g.doc
Loss information for a control valve is not required, because control
valves are dynamic devices that change their geometry in response to the
pipe system behavior. The loss that results is that required to maintain
the control parameter. You can, however, specify the full-open loss. This
is the loss that will occur should the valve fail to a full-open state.
A PRV is a device that controls the pressure in a pipe system. The PRV
maintains a constant control pressure downstream of the junction as long
as the upstream pressure exceeds the control pressure. If the upstream
pressure is lower than the control pressure, the ability to control pressure
is lost.
A PSV is similar to a PRV in that it controls pressure in a pipe system.
While the PRV maintains a constant downstream pressure, the PSV
maintains a constant upstream pressure. If the downstream pressure rises
higher than the control pressure, the ability to control pressure is lost.
An FCV is a device that maintains a constant flow rate in a pipe system.
By setting the junction to an FCV type and entering a flow rate, the
junction will limit the flow through the connecting pipes to be equal to
the control flow rate. The FCV can lose its ability to control flow when
the pressure drop across it becomes zero.
A PDCV is a device that maintains a constant (stagnation) pressure drop.
For this option, valve failure cannot be modeled. An indicator that an
unrealistic pressure drop has been demanded is a failure to obtain a
converged solution.
PRV/PSV static vs. stagnation pressure

The control pressure for a PRV or PSV can be either static or stagnation.
The default selection is static, as this is the most frequent application in
industry. See Chapter 8 for more information.
Action if setpoint not achievable

It is possible that the control valve cannot achieve the specified control
setpoint during steady-state and/or transient conditions. There are
several actions the valve can take in such cases. By default, the Always
Control During Steady-State is selected. For the steady-state portion of
the analysis this selection overrides all the others. This option is
physically unrealistic but useful for steady-state modeling purposes. The
reason this option is physically unrealistic is that the valve can add
8/10/2007 4:10:28 PM Page 233

pressure to maintain control, thus acting like a pump. If this occurs, the
user is clearly warned in the output. The Always Control option is useful
because it gives a clear indication to the user of how far away the valve
is from being able to control.
For transient analyses, if the Always Control is used and the valve fails,
the transient analysis will terminate.
For flow control valves, other options for valves which cannot achieve
control are to have the valve Open Fully or Close. By default the valve
fully opens. Once the control valve loses control during steady-state,
whether open or closed, Impulse will solve the model again.
Pressure control valves are more complicated. These valves have two
potential actions when control cannot be achieved. The first action is
when the control cannot be achieved due to insufficient upstream
pressure. The second is due to excessive downstream pressure. The
default behavior for pressure reducing valves is to Open Fully due to
insufficient upstream pressure, and close due to excessive downstream
pressure. The defaults for pressure sustaining valves are the opposite.
You can override the default behavior by clearing the Use Default
Actions selection and specifying the desired actions.
AFT Impulse will determine if the valve can maintain control based on
the Loss When Fully Open value. If this fully-open pressure drop is
greater than the pressure drop which balances the system, the control
valve fails. In other words, the control valve can’t open wide enough
(becoming less restrictive) to achieve control, so it fails. If you enter data
in the Open Percentage table on the Optional tab, you can select to have
the data taken directly from the table for the 100% open case. You can
also choose to not have any losses associated with the valve in the full-
open state by selecting None. Note that when a control valve fails open
this fully-open loss will be used for the valve.
It is possible for the valve to control properly during the steady-state,
and then lose control during the transient part of the run. In fact, it can
lose control and regain it repeatedly.
Special Conditions
Control valves have two special conditions. To open fully and not
control, or to close.

manual (si) g.doc
Open Percentage Table

The Optional tab allows data to be entered for special control valve
characteristics. Specifically, the valve Cv and Flow Area can be
Specified vs. the Open Percentage of the valve. This data does not affect
AFT Impulse’s flow solution. However, engineers frequently desire to
know the valve’s open percentage during operation to ensure it meets
design requirements. The Valve Summary (in the Output window)
always displays the valve Cv, and if data is specified for open percentage
and flow area it will also display open percentage and flow area at the
operating point.
Transient Data
The Control Valve junction is the only one with two transient tabs. The
first is the valve reaction to inability to achieve control, and the second
is to the control point. More information on transient data, including
event transients, is given earlier in this chapter and in Chapter 10.
Failure Transient
Control valves have what might be called inherent event logic. That is,
rather than respond to events at some other location, the failure
transients respond to events of the valve itself. Specifically, the events
are the valve's loss of control and possibly regaining of control.
The failure transient is the valve action when it loses control. If the valve
is set to close upon failure, the closure transient will be used. This is a K
or Cv table vs. time. If the valve were to reopen, the opening transient
would be used. The time scale for the closing or opening of a valve is
relative to when the failure occurs. That is called time zero in a relative
sense.
You can elect to not enter data for the closing or opening transient. If the
valve loses control, the valve action is assumed to be instantaneous. If
the failure action is to fully open, then the closing transient data will
have no meaning and not be used. In a sense the opening transient will
have meaning, but in reality when the valve loses control it is already
fully open by that point, so no transient will occur. The valve will just
stay fully open.
The closing and opening transients are thus intended for use when the
valve is set to close upon failure to control.
8/10/2007 4:10:28 PM Page 235

Control Transient
The other type of transient data supported by the Control Valve junction
is the Control Transient. This is an optional transient you can apply to
the control point. For instance, a change in control pressure for a
regulator valve can be modeled.
Control transients support events, although only a single event can be
modeled. Thus the regulator pressure could be modified when some
event occurred in the system.
Dead End Specifications window

The Dead End Specifications window is shown in Figure 6.14. The Dead
End junction must always have one connecting pipe. This junction type
is one way to define flow in a pipe as truly zero.
Figure 6.14 The Dead End Specifications window
In steady flow, dead ends have no flow and thus play no role in the
hydraulic solution to the system. However, in transient flow pipes that
terminate in dead ends can have flow. Thus dead ended lines can and do

manual (si) g.doc
influence the transient response of the system. Unless you know
otherwise for a specific system, dead ends should always be included in
the model.
The Dead End Specifications window follows the first of the two basic
Specifications window formats, displaying the connecting pipes in a
fixed format. The only required information is the connecting pipe and
the elevation.
Gas Accumulator Specifications window

The Gas Accumulator Specifications window is shown in Figure 6.15.
The Gas Accumulator junction allows up to twenty-five connecting
pipes. This junction type is useful for modeling a gas volume designed
as a surge protection device (Figure 6.16). The Gas Accumulator may be
integral with the pipe system, separated by an orifice, or by a short
connector pipe. The waterhammer theory for a gas accumulator is given
in Chapter 9. During steady flow the gas accumulator typically acts a
lossless branch junction, although it can be made to act like a pressure
junction. This is discussed later in the gas accumulator section.
The Gas Accumulator Specifications window follows the second of the
two basic Specifications window formats. A table displays the
connecting pipe information. The table grows in size to accommodate up
to twenty-five connecting pipes. After you add a fifth pipe a scroll bar
appears, allowing you to review all the pipes in the table.
Initial conditions and polytropic process

AFT Impulse needs to know the initial volume of gas. This can be
determined in one of two ways. First, the Actual Initial Volume can be
entered. If the accumulator is passive, the initial volume depends on the
initial pressure at the accumulator. Thus the initial pressure must be
known in order specify the Actual Initial Volume. One way to do this is
to first run the model in Steady Only mode to determine the pressure.
Some gas accumulators have an active design (with compressors) which
maintain a specified volume. In such a case, the initial volume is known
and independent of the gas pressure, and can thus be entered directly.
The second way is to enter the gas volume at some user specified
pressure. This is called “Volume at Specified Pressure”. When selected,
8/10/2007 4:10:28 PM Page 237

the user must specify a Reference Gas Volume and Reference Gas
Pressure. Frequently this will be the gas volume at atmospheric pressure.
When the steady-state is run, the gas volume will be determined from the
steady-state pressure, assuming the gas volume ratio is inversely
proportional to the gas pressure ratio.
The gas volume will change during the transient according to the
thermodynamic law. The law requires a polytropic constant to relate the
pressure to the gas volume. An isentropic process should use a
polytropic constant equal to the specific heat ratio of the gas (for air it is
1.4). An isothermal process should use a polytropic constant of 1.
Typically the process falls in between the two and an average polytropic
constant can be assumed (for air the average is 1.2).
The gas in the accumulator is assumed to behave as an ideal gas.
Figure 6.15 Gas Accumulator Specifications window

manual (si) g.doc
Maximum and minimum volumes

If the gas is constrained in some way to a maximum or minimum
volume, this can be entered in the optional fields. If a maximum or
minimum exists, AFT Impulse will not allow the gas volume to exceed
this value. If no data is entered, the volume can grow as large or small as
necessary.
Interface with pipe system

The Gas Accumulator can connect into the pipe system in three ways: it
can be integral with the pipe, it can be separated by an orifice, or it can
be connected by a short pipe.
These options are provided in the Flow Restrictor and Connector Pipe
areas. If neither option is used, the accumulator is assumed to be
attached directly to the pipe with no hydraulic loss.
If the Flow Restrictor is used, the accumulator is assumed to be
separated from the pipe system by an orifice, which causes a hydraulic
loss as the liquid flows in and out of the accumulator. The loss is
specified by entering a flow area and discharge coefficient. Further, this
loss can be different for inflow vs. outflow.
PA
Orifice
Connector
m& A Pipe
LC
Figure 6.16 Schematic of a gas accumulator
8/10/2007 4:10:28 PM Page 239

Lastly, the accumulator can be connected by a short pipe which is
considered to be short in relation to the other pipes in the system after
they are sectioned. If the pipe is short, it can be assumed to react
instantaneously with the accumulator. The short pipe accounts for liquid
inertia and friction. If a connector pipe exists which cannot be
considered short, it should be modeled as an actual AFT Impulse pipe.
The Flow Restrictor and Connector Pipe can both exist in the same
accumulator. That is, a flow restrictor can be modeled at the top of a
connector pipe, and the effect of both will be included.

As mentioned previously, the gas accumulator acts like a branch junction
during steady flow. However, the user can specify that it behave like a
pressure junction. Why would one want to do this? The most obvious
case is where the gas accumulator is modeling a pressurizer type of
component in a closed system. A pressurizer is sometimes called an
expansion tank or accumulator. The main purpose is to provide thermal
expansion volume when the temperature of the process fluid changes.
Sometimes these tanks are open to the atmosphere, in which case a
Surge Tank would be a better choice than a gas accumulator.
When a pressurizer is modeled as a gas accumulator, there will typically
be no other pressure type junctions in the system. In fact, the pressurizer
is the component which controls the static pressure on the system during
steady-state operations. However, during transients the pressure will
change. Thus there is a need to allow a gas accumulator to act as a
pressure junction during the steady-state, and then like a normal gas
accumulator during the transient. This is the function of the Initial
Pressure for Zero Steady-State Flow on the Optional tab (see Figure
6.17).
When the accumulator is a pressurizer during steady-state, it has zero net
flow. The pressure entered will be the steady-state static system
pressure. If the gas accumulator is not in fact arranged as a pressurizer
and data is entered in the field for Initial Pressure for Zero Steady-State
Flow, AFT Impulse will warn the user before running the transient
simulation.

manual (si) g.doc
Figure 6.17 Gas Accumulator Specifications window
Graphing Gas Accumulator data

You can track the gas volume, liquid flowrate, and pressure in the
accumulator by selecting these options in the Junction section of the
Transient Control window. These parameters can be graphed or
reviewed in the Output window.
General Component Specifications window

The General Component Specifications window is shown in Figure 6.18.
The General Component junction type is internal to the system and
requires two connecting pipes. This junction type allows you to model
the irrecoverable loss that occurs through equipment that may not be
8/10/2007 4:10:28 PM Page 241

explicitly represented on the Toolbox and to specify loss factors as a
function of a flow.
The General Component Specifications window follows the first of the
two basic Specifications window formats, displaying the connecting
pipes in a fixed format. The flow direction through the junction is
determined by the defined directions of the connecting pipes. Consistent
with AFT Impulse's convention, the loss factor base area is referenced to
the upstream flow area.
Figure 6.18 The General Component Specifications window
When you select Variable K Factor or Resistance Curve, the General

Component Specifications window makes additional features available.
Using these new features you can input loss factors or pressure drops
that vary with flow. To enter these factors, you can specify polynomial
constants, fit a curve to available data, or use interpolated x-y data.
When a variable loss is specified, AFT Impulse modifies the loss factor
in the Solver to agree with the solution. You can choose any of the
optional flow and pressure parameters provided, and you can specify the
most convenient units.

manual (si) g.doc
No transient data can be entered for a General Component. From a
transient point of view, the junction is a static element. For this reason
the user should use General Components with caution. Where possible
all static element hydraulic losses should be entered into pipes and not
junctions.
Liquid Accumulator Specifications window

The Liquid Accumulator Specifications window is shown in Figure 6.19.
The Liquid Accumulator junction allows up to twenty-five connecting
pipes. This junction type is useful for modeling a liquid chamber that has
more elasticity than the pipe itself. This may be a liquid-solid tank or a
condenser box, for example.
Figure 6.19 Liquid Accumulator Specifications window
The Liquid Accumulator Specifications window follows the second of

the two basic Specifications window formats. A table displays the
connecting pipe information. The table grows in size to accommodate up
to twenty-five connecting pipes. After you add a fifth pipe a scroll bar
appears, allowing you to review all the pipes in the table.
8/10/2007 4:10:28 PM Page 243

Initial conditions and elastic process

The relationship between liquid accumulator volume and pressure is
assumed to be linear. The amount of volume change for a given amount
of pressure change is given by the elasticity of the accumulator. This
input will need to be obtained from test data or approximated based on
mechanical correlations.
The Initial Volume of the tank is a required input. This volume will
change during the transient according to the linear law.
Pump Specifications window

The Pump Specifications window is shown in Figure 6.20. The Pump
junction type requires one or two connecting pipes. This junction type
allows you to model the pressure (head) that is added to the system by a
pump, and the pump response to a trip or startup. The pump can be of
either the centrifugal or positive displacement kind. It also provides
access to curve fitting tools. Data can also be interpolated between the
provided data.
The Pump Specifications window follows the first of the two basic
fixed format. A flow direction through the junction is adopted from the
defined directions of the connecting pipes.
Because all pumps are different, AFT Impulse does not provide any
pump types. You must obtain the proper pressure and flow data that
describes the pump from the manufacturer or from test data. You will
want to save frequently-used pumps into the Component database.
In performing curve fitting or specifying pump constants, you can work
with pressure or head and with volumetric or mass flow rate. A variety
of units are provided in dropdown lists. These features allow you to
enter your pump information in the most convenient and meaningful
format.
Pump Model
The basic definition of the pump is performed on the Pump Model tab.
You can model a pump with a pump curve or an assigned flow. When

manual (si) g.doc
using a fixed flow, the total pressure/head developed by the pump is
calculated. This is useful for modeling positive displacement pumps.
Figure 6.20 The Pump Specifications window. Top shows the

Pump Model tab, and bottom shows Variable Speed.
The pump operating speed can be entered as a percent of full speed. It is

assumed that the pump curve is entered at full speed. You can use a
8/10/2007 4:10:28 PM Page 245

speed that is lower or higher than full speed. Impulse will adjust the full
speed pump curve using standard affinity laws relating speed ratios to
pressure rise ratios and flow rate ratios.
Submerged pump
If a pump is submerged, there is no upstream pipe. Thus only one
connecting pipe is needed for the discharge. This is modeled by ticking
the check box and entering the pump suction conditions.
Do not allow backwards rotation

For pumps modeled using four quadrant methods, the model allows the
pump to spin backwards. If there is an anti-reverse spin device on the
pump, then one can model this by selecting the “Do Not Allow
Backwards Rotation” option.
Check valve at pump discharge

AFT Impulse offers a dedicated check valve junction, but frequently the
check valve is close to the pump discharge and the pipe length between
them is very small. This can result in models with unacceptable run
times if a check valve junction is used for transient models.
You can model a check valve at the pump discharge assuming negligible
pipe length between them. The valve model is relatively simple, with
instantaneous closure once backflow occurs. There is also an input area
to account for valve closure starting at some other velocity than zero.
You also can enter re-opening criteria based on pressure.
Pump configurations
Pump data can be entered for multiple configurations. This is available
only for pumps modeled as pump curves. The default is a single
configuration.
A pump configuration is a pump with a specific impeller trim and
operating speed. Multiple impeller trims and operating speeds can be
specified as part of the pump, and then a particular combination can be
chosen.
These parameters are set in the Auxiliary Input tab.

manual (si) g.doc
During transients, the pump speed will vary as the transient conditions
affect the pump. The resulting pump speed can be graphed in the Graph
Results. The response of the controller during the transient is assumed to
be instantaneous.
Figure 6.21 Pump Configuration window Pump Data tab above

where raw data is entered and curve fit or
interpolated, Composite Graph tab below.
8/10/2007 4:10:28 PM Page 247

Figure 6.21 shows the Pump Configuration window. Raw data can be
entered, imported from file, or pasted from the clipboard. Then a
polynomial curve can be fitted to the data or the data can be interpolated.
Data for head, NPSH, and efficiency or power can be entered. Also data
for the pump’s end of curve and NPSH constant for variable speed can
be entered.
If multiple configurations are entered, they are displayed on the Pump
Specifications window in dropdown lists for selection (see Figure 6.22).
Multiple configuration
selections displayed
Figure 6.22 Pump Specifications window with multiple

configurations available.
Controlled pressure or flow (variable speed)

You can optionally model a controlled pump. The pump will then vary
the speed to deliver the flow rate or pressure you set (either on the
suction or discharge side). The resulting speed will be reported in the
Pump Summary in the General Results section on the Output window.
Alternatively, when using a controlled pressure, it can be controlled only
if it exceeds some level high or low. Otherwise the pump will operate on
its regular pump curve.

manual (si) g.doc
NPSHR and NPSPR (optional in pump configuration)

The net positive suction head required (NPSHR) or net positive suction
pressure required (NPSPR) can be expressed as a function of mass or
volumetric flow rate. This data is optional and is not required to
completely define the pump. Both the required and available values are
given in the Pump Summary in the General Output section.
For variable speed pumps, you can supply an affinity exponent to adjust
the full speed NPSHR/NPSPR to that at the alternate speed. According
to the affinity laws, the head ratio and speed ratio are related by the
square law. However, the NPSHR ratio does not necessarily obey this
affinity law. This exponent allows you to account for this.
Note that in the Elevation area a data field is provided for the reference
elevation for NPSH calculations. This may be the pump suction
elevation or pump centerline.
Efficiency and power usage (optional in pump

configuration)
The pump efficiency or power usage can be specified. This does not
affect the pump performance during steady-flow, but allows for quick
assessment of power used at the operating point and the proximity to
BEP (Best Efficiency Point). If efficiency or power data is supplied, the
power, efficiency and percent of BEP at the operating point is given in
the Output Pump Summary.
During transient flow, the efficiency/power usage is not optional if four
quadrant pump modeling is used. In this case, the efficiency/power data
will affect the results.
Transient Data
As discussed previously, pumps can be modeled as a pump curve or as
an assigned flow. Using assigned flow pumps in steady flow modeling is
useful for sizing a centrifugal pump. In AFT Impulse, however, that is
not appropriate during transient modeling. When modeling a centrifugal
pump transient, a pump with a pump curve is essential. Positive
displacement pumps act as constant flow devices, and the assigned
flowrate feature allows one to adequately model such pumps.
8/10/2007 4:10:28 PM Page 249

Pumps with curves
Once a pump curve is entered, several options are available for transient
modeling. Modeling a pump with a curve is usually associated with a
centrifugal pump.
No transient from pump, no back flow

If a pump remains at 100% speed during a transient, and the flow is
always positive, then no transient model is required. The appropriate
Transient Model selection is therefore “None”. In such a case, the pump
performance will be based on the provided head curve which only
applies to positive flow. Of course, if there is a check valve after the
pump which prevents backwards flow, this is the best choice.
No transient from pump, but back flow possible

If a pump remains at 100% speed during a transient, but backwards flow
is possible, the head curve data is insufficient. Therefore four quadrant
data can be used. As in other cases, the appropriate four quadrant data
selection is the one with the closest specific speed. This Transient Model
option is called “No Transient – Four Quadrant”. This case is only
possible when there is no check valve after the pump.
Known speed transients, no back flow

An estimate of the pump speed change with time can be entered in the
transient data area (see Figure 6.23). The speed change can be a time-
based or event-based transient. The appropriate Transient Model is
“Without Inertia”.
This model uses the provided speed data and the pump homologous laws
to determine the pump performance. Since the only pump performance
data is for forward flow, only forward flow or zero flow can be modeled.
As before, if there is a check valve after the pump which prevents
backwards flow, this is the best choice when the pump speed is known.

manual (si) g.doc
Figure 6.23 A speed transient can be entered for a pump with a

curve
Pump startup transient with possible reverse flow

If the speed for startup is known or can be approximated, and reverse
flow is possible (no check valve exists), then a four quadrant model can
be used. With the speed known, we use the four quadrant data FH only.
Other data for moment of inertia and initial synchronous speed in rpm is
not required.
Chapter 9 offers more information on the parameters and equations, and
the user is encouraged to consult one of the excellent texts for more
background on this method (Chaudhry, 1987, Chapter 4, Swaffield, et
al., 1993, pp. 154-166, Wylie et al., 1993, Chapter 7).
Pump transients with inertia

There are four pump models in AFT Impulse that account for pump
inertial effects and predict the speed over time for either. Two allow
reverse flow and/or speed, while two only work with forward or zero
flow. These are:
8/10/2007 4:10:28 PM Page 251

• Trip With Inertia and No Back Flow or Reverse Speed
• Startup With Inertia and No Back Flow or Reverse Speed
• Trip With Inertia – Four Quadrant
• Startup With Inertia – Four Quadrant, Known Motor Torque/Speed
Predicting the response of a pump during a trip or startup requires
additional data for the pump. Use of the homologous pump laws is
essential. If backflow is possible, "four quadrant" pump curve data is
required. The first model mentioned above assumes the presence of a
check valve which prevents backflow. The second model also assumes
there is no backflow, although no check valve is assumed. In such cases,
the pump will operate in the first quadrant. Thus four quadrant data is
not required.
The third and fourth models allow backflow and possibly reverse speed,
and thus require four quadrant data. All four models will be discussed in
the following sections.
Trip with inertia and no back flow or reverse speed

When no backflow or reverse speed occurs through the pump, the
normal pump curve data for flow vs. head can be used to model the
pump response during a trip. Pump power data is also required as a
function of flow. Data for the pump rotating moment of inertia and the
initial synchronous speed in rpm is required.
This model will use the pump head curve and power curve to determine
the pump torque, and use the moment of inertia to predict the pump
speed decay over time.
See Chapter 9 for more information on the methodology.
Startup with inertia and no back flow or reverse speed

When no backflow occurs through the pump, and the pump motor torque
vs. speed is known, the normal pump curve data for head vs. flow and
power vs. flow can be used to model the pump response during a startup.
Data for the pump rotating moment of inertia and the 100% operation
synchronous speed in rpm is required.
This model will use the pump head curve and power curve to determine
the torque from the fluid, balance this with the torque from the motor,
and use the moment of inertia to predict the pump speed rise over time.

manual (si) g.doc
Trip with inertia – four quadrant

This model allows the pump to flow and/or spin backwards, and requires
pump data on the head and torque characteristics for reverse flow and
speed. Data for the pump rotating moment of inertia and the initial
synchronous speed in rpm is required.
In addition, dimensionless curves of FH and FB are required. The
parameters FH and FB are properties of the pump, and it is typically
assumed that for pumps of similar specific speed they are invariant.
Figure 6.24 shows data for a pump with a specific speed of 0.46 as given
by Wylie, et al., (1993, pp. 147-148).
Figure 6.24 Pump with four quadrant data for pump trip modeling
AFT Impulse provides data of FH and FB vs. θ for pumps of twenty-one

specific speeds. The references come from Donsky (1961) as given by
Wylie, et al. (1993), Brown & Rogers (1980), and Thorley (1996). See
Chapter 9 for a list of the available selections. The conventional
assumption is that you should use four quadrant data for a pump with the
8/10/2007 4:10:28 PM Page 253

closest specific speed. At lower specific speeds, Brown & Rogers (1980)
question this assumption.
If you should need to construct data for other sets, please consult the
three textbooks referenced below for additional information. Note that
the textbooks do not give a method for constructing these data sets, but
they contain references that you can pursue.
Chapter 9 offers more information on the parameters and equations, and
the user is encouraged to consult one of the excellent texts for more
background on this method (Chaudhry, 1987, Chapter 4, Swaffield, et
al., 1993, pp. 154-166, Wylie et al., 1993, Chapter 7).
Startup with inertia - four quadrant, known motor

torque/speed
When backflow can occur through the pump, and the pump motor torque
vs. speed is known, four quadrant modeling is required to model a
startup. The options for four quadrant data are the same as just discussed
for pump trip cases. Data for the pump rotating moment of inertia and
the 100% operation synchronous speed in rpm is required.
This model will use the four quadrant data to determine the torque from
the fluid, balance this with the torque from the motor, and use the
moment of inertia to predict the pump speed rise over time.
Estimating the pump inertia

Chapter 9 offers a method described by Wylie et al. (1993) to estimate
the moment of inertia, I, for a pump and the entrained liquid.
Event modeling – including trips followed by restarts or vice

versa
All pump inertial modeling is a result of events. The Time option is
disabled. By default, a single event is selected and the event parameter is
itself time. The default criteria is zero time, which means that the pump
trips or starts on the first step of the simulation. If you want the pump to
trip or start at some later time which is known, you can just change the
time criteria for the event. Pumps can also be tripped or started based on
any other of the available event criteria such as pressure or flow.

manual (si) g.doc
Pumps also support dual events, both cyclic and sequential. When the
first event is a pump trip, the second event will always be a pump start,
and vice versa. This allows a pump to be tripped and then restarted at
some later time, for example. The restart can occur while the pump is
still spinning down.
If the first event is a pump trip using "Trip With Inertia and No Back
Flow or Reverse Speed", then the restart model will be "Startup With
Inertia and No Back Flow or Reverse Speed", and vice versa. This means
that motor torque vs. speed data is required to model the restart.
On the other hand, if the first event is a pump trip using "Trip With
Inertia - Four Quadrant", then the restart model will be "Startup With
Inertia - Four Quadrant, Known Motor Torque/Speed", and vice versa.
As before, this means that motor torque vs. speed data is required to
model the restart.
Controller transient
If the pump has a variable speed controller (see Figure 6.25), a control
flow or pressure transient can be specified. Again, this transient can be
time-based or event-based (see Figure 6.26).
Figure 6.26 transient essentially increases the speed of the pump to
achieve 300 m3/hr of flow.
Note that the resulting pump speed transient with time can be saved to
the output file (see Transient Control window in Chapter 5). If saved, the
speed can be plotted vs. time.
Pumps as fixed flow

When a pump is modeled as a fixed flow, then the flowrate can be varied
over time. This flow transient can be time-based or event-based. The
transient data might look like that in Figure 6.26.
8/10/2007 4:10:28 PM Page 255

Figure 6.25 A pump with a variable speed controller can have a

transient on the control point.
Modeling pumps as an assigned flow will be most appropriate for

positive displacement type pumps. The transient can represent the
starting or stopping of a pump, or it can represent the pulsating transient
flow that can occur in reciprocating pumps. More information on
transient data, including event transients, is given earlier in this chapter
and in Chapter 10.

manual (si) g.doc
Figure 6.26 A control point transient can be entered for a pump.
Special Conditions
Pump junctions in AFT Impulse have two types of Special Conditions.
The first is to turn the pump off and have no flow through it. This is
called "Pump Off No Flow". The second type of Special Condition turns
the pump off but allows flow to go through the pump. This is called
"Pump Off With Flow Through".
Why would one want to use one Special Condition rather than the other?
The simple answer is that the first Special Condition is more appropriate
for positive displacement type pumps and the second for centrifugal
pumps.
Here's why. When a positive displacement pump is turned off and has a
pressure difference across it, it will usually act like a closed valve and
not allow flow to go through it. Thus the first Special Condition would
be most appropriate. For instance, assume one wants to model the
transient that occurs during the startup of a positive displacement pump.
The best way to do this would be to use the first Special Condition with
8/10/2007 4:10:28 PM Page 257

no flow, and then input a flowrate transient in which the first data point
is zero flow.
On the other hand, when a centrifugal pump is turned off and has a
pressure difference across it, in the absence of other valving which
prevents flow the pump will usually allow flow to go through it. Thus
the Special Condition that allows through flow would be most
appropriate. For example, assume one wants to model the transient
during the startup of a centrifugal pump. One would use the second
Special Condition, and, in conjunction with a pump curve entered at
100% speed, input a speed transient with the initial speed as zero.
Representing Multiple Pumps With One Junction

If multiple identical pumps exist in parallel or closely spaced in series,
rather than modeling each as a separate junction they can all be
collapsed into a single junction. When doing so, intermediate piping is
ignored. This is often advantageous in that short pipes around the pump
which can cause long run times can be ignored.
To use this feature, specify the number of pumps at the location and
whether they are in parallel or in series.
Viscosity corrections
Viscosity corrections to standard pump curves have been published in
the Hydraulic Institute Standards. The method, based on the 14th
edition, 1983, is employed. The method itself consists of a semi-
logarithmic diagram relating flow rate, head rise and viscosity. The
details of this diagram have been digitized into numerical relationships
inside AFT Impulse 4.0. The corrections to head rise, flow rate and
efficiency, referred to as CH, CQ and CE, respectively, are given in the
Pump Summary output.
The viscosity corrections are assumed to be valid during transients.
Graphing pump data

You can track the pump speed by selecting this option in the Junction
section of the Transient Control window. This parameter can be graphed
or reviewed in the Output window.

manual (si) g.doc
Relief Valve Specifications window

The Relief Valve junction type can have one or two connecting pipes.
This valve stays closed until it cracks open, at which time it relieves to
an external pressure or to another pipe system.
The Relief Valve Specifications window, shown in Figure 6.27, follows
the first of the two basic Specifications window formats, displaying the
connecting pipes in a fixed format.
Figure 6.27 Relief Valve Specifications window
Three relief valve configurations: internal, exit and inline

The Relief Valve junction can be an internal relief valve, an exit relief
valve, or an inline relief valve. If internal, the valve relieves into another
pipe which was previously isolated. In this case the valve can only have
two connecting pipes. Before cracking open the internal relief valve acts
like a regular valve which is closed.
A Relief Valve junction that is at an exit relieves to an external ambient
pressure such as the atmosphere. An exit relief valve only has one
8/10/2007 4:10:28 PM Page 259

connecting pipe. Relief valves that are located at an exit must also be
supplied with an associated back pressure.
An inline Relief Valve junction discharges to an external ambient
pressure, and is located between two pipes. When closed, the valve is a
lossless connection and allows flow between the pipes.
Specifying losses
You specify the losses for the Relief Valve on the Valve Model tab in
the Specifications window. For convenience, you can specify the
constant loss characteristics of a Relief Valve as a valve coefficient (CV),
as a loss factor (K), or as a variable Cv where Cv depends on pressure.
Chapter 8 details the relationship between K and CV.
Transient valve cracking

Transient data for a relief valve is only used for the constant Cv and
constant K models. The variable Cv model does not use pre-defined
transient data.
The transient data for a relief valve is somewhat different from other
junctions. The transient data does not refer to the absolute time in the
model. Time zero is the moment of cracking, not the start of the model.
You specify the transient relief valve data by clicking the Transient Data
button. A dialog window will open where you can enter the data. Once
entered, the data will display in the Transient Input Data table in the
window.
A zero value of Cv indicates a closed valve. For a lossless valve, enter a
very large Cv thus approximating infinity.
To simulate a closed valve with one of the K factor methods, use a K
factor of –1 (negative one).
Transients with variable Cv

For typical passive relief valves, it is not possible to pre-specify the
transient valve behavior. The valve Cv depends on the pressure
difference across the valve, as the pressure difference is what causes the
valve to open. When the pressure difference is below the cracking
pressure, the valve remains closed. When it rises above the cracking
pressure, it opens. As the pressure difference increases, it opens further

manual (si) g.doc
until it is fully open. Any further pressure increases do not increase the
Cv, and so the fully open Cv is used for all higher pressures.
In general the cracking pressure can be specified as either an absolute
pressure/head or a pressure/head difference. When working with the
variable Cv model, the cracking pressure must be entered in
pressure/head difference. In addition, the first Cv data must equal zero,
and the corresponding pressure/head to zero Cv must be the same as the
cracking pressure/head difference. See the Relief Valve Theory section
in Chapter 9 for more information.
Reservoir Specifications window

The Reservoir Specifications window is shown in Figure 6.28a-b. The
Reservoir junction type allows you to connect up to twenty-five pipes.
One connecting pipe is required. The Reservoir junction is convenient
for specifying a fixed pressure in your system. During a transient, you
can vary the liquid surface elevation in a prescribed way.
A Reservoir junction applies a defined pressure at the junction location
in the model. When solving a pipe flow system, a Reservoir causes the
rest of the system to distribute the flow in a manner consistent with the
defined pressure.
The Reservoir Specifications window follows the second of the two
basic Specifications window formats. A table on the Loss Coefficients
tab displays the connecting pipe information. This table grows to
accommodate up to twenty-five pipes. After you add a fifth connecting
pipe, a scroll bar appears, allowing you to review and enter loss factors
for all pipes in the table.
The pipe table displays the reference positive flow direction of each
connecting pipe. To enter loss factors, select the cell in the table and edit
the value. Each pipe can use a different loss model or custom value.
The depth or elevation for each connecting pipe can be entered in the
table. The depth is the distance below the liquid surface level.
8/10/2007 4:10:28 PM Page 261

Figure 6.28a The Reservoir Specifications window Reservoir

Model tab.
If a pipe returns to the reservoir above the liquid surface, the depth is
entered as a negative number, or the elevation is specified higher than
that of the liquid surface. Pipes that discharge above the liquid surface
are assumed to have liquid free fall to the liquid surface. AFT Impulse
applies the proper boundary condition for pipes with negative depths. If
the pipe is above the liquid surface, the only appropriate condition is for
the fluid to be flowing from the pipe into the reservoir. Fluid cannot flow
from the reservoir into the pipe as the pipe is above the liquid surface. If
this occurs, AFT Impulse will assume the fluid flowing into the pipe is
the same as the reservoir fluid, solve the system, and then give a warning
in the output.

manual (si) g.doc
Figure 6.28b Individual pipe depths (or elevations) and loss

coefficients may be specified in the Reservoir
Transient Data
On the transient data tab the surface elevation can be changed with time.
Spray Discharge Specifications window

The Spray Discharge junction must have at least one connecting pipe,
and it allows up to four. This junction offers a flexible way of modeling
a flow exiting the system through a nozzle or spray device. The flow
occurs across a user specified hydraulic loss such as a discharge
coefficient. This can be varied with time to open or close the Spray
Discharge junction.
8/10/2007 4:10:28 PM Page 263

The Spray Discharge Specifications window, shown in Figure 6.29,
follows the second of the two basic Specifications window formats. A
table displays the connecting pipe information. The pipe table grows in
size to accommodate up to four connecting pipes.
The exit pressure to which the Spray Discharge junction discharges must
be entered.
Two geometries can be modeled. The first is a normal, one hole spray,
where the discharge flow area of the hole is entered. The second is a
sparger which, in principle, works in the same way as a spray. For the
sparger, the flow area of a single hole and the number of holes are
entered. This assumes that the holes are hydraulically close in proximity
and are the same area.
Figure 6.29 The Spray Specifications window

manual (si) g.doc
Loss Model
The discharge coefficient for the junction is entered in the appropriate
box. The physical area of the exit flow can also be entered. These two
parameters are used to calculate the discharge flow rate, which depends
on the difference between the internal pressure in the system and the
specified exit pressure.
The K Fire Sprinkler model uses the K value as defined by the fire
sprinkler industry. Whereas the standard K loss value is dimensionless,
the K Fire Sprinkler has units associated with it. See Chapter 8 for more
details.
Transient Data
On the transient data tab the discharge flow area or K data can be varied
with time to simulate the opening or closing of the junction. More
information on transient data, including event transients, is given earlier
in this chapter and in Chapter 10.
Graphing Spray Discharge data

You can track the discharge flowrate and total mass or volume
discharged at the spray discharge junction by selecting this options in the
Junction section of the Transient Control window. This can be graphed
or reviewed in the Output window.
Surge Tank Specifications window

The Surge Tank Specifications window is shown in Figure 6.30. The
Surge Tank junction allows up to twenty-five connecting pipes. This
junction type is useful for modeling a small vessel designed as a surge
protection device (see Figure 6.31). The Surge Tank may be integral
with the pipe system, separated by an orifice, or by a short connector
pipe.
The Surge Tank Specifications window follows the second of the two
basic Specifications window formats. A table displays the connecting
pipe information. The table grows in size to accommodate up to twenty-
five connecting pipes. After you add a fifth pipe a scroll bar appears,
allowing you to review all the pipes in the table.
8/10/2007 4:10:28 PM Page 265

Figure 6.30 Surge Tank Specifications window, with a variable

height and flow restriction orifice
Tank geometry
The Tank Height is the physical top of the tank in relation to the
elevation. If there is a Connector Pipe, then the Tank Height is relative
to the top of the Connector Pipe. This height is used in the case of
overflow. If the surge tank overflows, spillage will occur and the liquid
will exit the pipe system not to be recovered. If you do not want to
model the overflow, specify a very tall tank or do not enter a Tank
Height.
If the Tank Cross-Sectional Area is specified as Constant, the surge tank
is assumed to be a vertical tank with constant cross-sectional area. If the
tank cross-sectional area changes with height, this can be entered by
selecting the variable option and entering the cross-sectional variation in
the table.
The Tank Cross-Sectional Area is used for the purpose of determining
the liquid volume. It will determine how high the liquid will rise or drop,
since a given volume of liquid will flow into the tank during a time step.

manual (si) g.doc
Interface with pipe system

The Surge Tank can connect into the pipe system in three ways: it can be
integral with the pipe, it can be separated by an orifice, or it can be
connected by a short pipe.
These options are provided in the Flow Restrictor and Connector Pipe
areas. If neither option is used, the surge tank is assumed to be attached
directly to the pipe with no hydraulic loss.
If the Flow Restrictor is used, the surge tank is assumed to be separated
from the pipe system by an orifice, which causes an hydraulic loss as the
liquid flows in and out of the surge tank. The loss is specified by
entering a flow area and discharge coefficient. Further, this loss can be
different for inflow vs. outflow.
LL Orifice
Connector
m& T Pipe
LC
Figure 6.31 Surge tank schematic
Lastly, the surge tank can be connected by a short pipe which is

considered to be short in relation to the other pipes in the system after
they are sectioned. If the pipe is short, it can be assumed to react
instantaneously with the surge tank. The short pipe accounts for liquid
inertia and friction. If a connector pipe exists which cannot be
considered short, it should be modeled as an actual AFT Impulse pipe.
The Flow Restrictor and Connector Pipe can both exist in the same surge
tank. That is, a flow restrictor can be modeled at the top of a connector
pipe, and the effect of both will be included.
8/10/2007 4:10:28 PM Page 267

One-Way Surge Tank

One drawback of surge tanks is that they must be located in low pressure
parts of the system because high pressure locations require a very tall
surge tank (a water surge tank located where the pipe system is at 690
kPa must have a liquid surface at least 70 meters higher than the pipe).
Unless there is a check valve connecting the surge tank to the pipe. Since
the tank can only flow out, it is only useful to limit low transient
pressures by flowing liquid into the system when the pressure drops.
However, when the pressure rises, the check valve closes and the tank
cannot accept any fluid and thus decrease the system pressure.
Since the surge tank liquid is isolated from the pipe system during the
initial steady-state calculation, it is required to specify the liquid surface
level in an otherwise optional input field. The Optional tab offers an area
to specify the initial liquid height. This is discussed below in “Special
features for steady flow”.
Such a design is called a One-Way Surge Tank, and it can be modeled by
defining the check valve characteristics.
Transient Data
On the transient data tab the surface pressure can be varied with time to
simulate a tank pressurization. More information on transient data,
including event transients, is given earlier in this chapter and in Chapter
10.

As mentioned previously, the surge tank acts like a branch junction
during steady flow. In such cases the steady-state liquid height in the
surge tank is calculated (it is the steady-state EGL at the junction).
However, the user can specify that it behave like a pressure/reservoir
junction. Why would one want to do this? One important case is where
the surge tank is modeling a pressurizer type of component in a closed
system. When open to the atmosphere, a pressurizer is sometimes called
an expansion tank or head tank. The main purpose is to provide thermal
expansion volume when the temperature of the process fluid changes.
Sometimes these tanks are closed, in which case a Gas Accumulator
would be a better choice than a surge tank.

manual (si) g.doc
On other occasions, one may want to model a fixed steady-state liquid
level at a location, but is interested in the liquid level changes in a finite
size tank.
These two possibilities have different implications which will be
explored here, along with the standard use where the surge tank acts like
a branch. The both cases the input is on the Optional tab (see Figure
6.32).
Figure 6.32 Surge Tank junctions allow the user to specify an

liquid level for special steady-state behavior.
Surge tank as a branch during steady-state

When a Surge Tank is used as a surge suppression device as in pumped
systems, it will act like a branch during steady flow. There will be no
flow in or out of the tank, and the surface level will equal the local
gradeline of the system. Thus the surface level is not an input, but is
calculated. This is the default behavior of the Surge tank junction.
During the transient, flow will go in and out of the tank, and the liquid
level will change.
8/10/2007 4:10:28 PM Page 269

Surge tank as a pressurizer during steady-state
When a pressurizer is modeled as a surge tank, there will typically be no
other pressure type junctions in the system. In fact, the pressurizer is the
component which controls the static pressure on the system during
steady-state operations by virtue of its liquid surface level. However,
during transients the liquid surface level will change. Thus there is a
need to allow a surge tank to act as a pressure junction during the steady-
state (in which the liquid level is fixed), and then like a normal surge
tank during the transient (in which the liquid level varies). This is the
function of the Initial Liquid Height for Steady-State Flow on the
Optional tab (see Figure 6.32).
When the surge tank is a pressurizer during steady-state, it has zero net
flow. The liquid level entered will be the steady-state static system
pressure. A flow balance exists at the junction because the net flow is
zero.
During the transient, flow will go in and out of the tank, and the liquid
level will change.
Surge Tank as a finite size reservoir during steady-state

When the surge tank is not a pressurizer, but acts like a reservoir, it will
have a net flow imbalance during steady-state just like Reservoir
junctions do. This is how pressure junctions work (see Chapter 12).
Once the transient begins, the liquid surface level will vary as for any
surge tank.
Difference between a Reservoir and Surge Tank

A Reservoir junction is assumed to be sufficiently massive in volume
that surface level changes do not occur during the simulation. A good
physical example of this is when the liquid source is a large lake.
However, sometimes the liquid source comes from a finite size tank, and
the liquid surface level will in fact change during the simulation. To
track the change, the geometry of the source tank must be known. This
can be entered for a Surge Tank, but not a Reservoir.
Model as dipping tube vessel

A dipping tube vessel is a combination surge tank and gas accumulator
(see Figure 6.33). As long as the liquid level remains below LA in Figure

manual (si) g.doc
6.33, it acts like a conventional surge tank. If the liquid level rises above
LA , gas gets trapped in the upper region and the device acts like a gas
accumulator.
The dipper tube capability is entered on the Optional tab (see Figure
6.32).
LA
LL Orifice
Connector
m& T Pipe
LC
Figure 6.33 Dipping tube vessel geometric representation. When

the liquid level rises above LA, it acts like a gas
accumulator.
Graphing Surge Tank data

You can track the surge tank flowrate, total mass or volume, and liquid
surface level pressure by selecting these options in the Junction section
of the Transient Control window. These parameters can be graphed or
reviewed in the Output window.
Tee/Wye Specifications window

The Tee/Wye Specifications window is shown in Figure 6.34. The
Tee/Wye junction type requires three connecting pipes. To increase
clarity, the window allows you to arrange the pipes with respect to an
image. Four different tee/wye models are available.
The Tee/Wye Specifications window does not strictly follow either of
the two basic Specifications window formats. Instead, the three
8/10/2007 4:10:28 PM Page 271

connecting pipes are arranged in the Arrangement area using provided
lists and images.
Loss factors
There are two Loss Models available for Tee/Wye junctions: Simple and
Detailed. If you choose the Simple type of Tee/Wye, no loss factors will
be calculated. This is equivalent to a Branch with no loss factors entered
for the connecting pipes. In some systems the losses associated with a
Tee/Wye are small compared to the overall loss in the system. To avoid
unnecessary calculations you should choose the Simple type of Tee/Wye
unless you need the actual losses.
Figure 6.34 The Tee/Wye Specifications window
The Detailed type of Tee/Wye will calculate the loss factors as they
depend on the flow split as well as the geometry. Because of the
dependence on the flow split, loss factors usually cannot be calculated
before performing an analysis. AFT Impulse incorporates the best
available models to calculate tee and wye loss factors during the actual
solution. These calculations account for the percentage of the flow split

manual (si) g.doc
and the angle of connection. This offers a significant advantage over
hand calculated losses.
If you have specific loss factors that you want to use for a tee/wye
junction, you should use a Branch junction instead.
The Tee/Wye junction functions the same during the steady-state and
transient.
Valve Specifications window

The Valve Specifications window is shown in Figure 6.35. The Valve
junction type requires two connecting pipes, unless you specify it as an
exit valve, in which case only one connecting pipe is allowed. This
junction type allows you to model the irrecoverable loss that occurs
through a flow control component. You also have the ability to change
the loss with time to open or close the valve in full or in part.
Figure 6.35 The Valve Specifications window
8/10/2007 4:10:28 PM Page 273

The Valve Specifications window follows the first of the two basic
fixed format. A flow direction through the junction is adopted from the
defined directions of the connecting pipes. Consistent with AFT
Impulse's convention, the loss factor base area is referenced to the
upstream flow area shown as the Base Area. Valve junctions are
typically internal to the system, with two connecting pipes.
Valves that are located as exit flow control mechanisms are specified as
exit valves. These valves require an associated back pressure definition.
Exit valves can only be connected to a single upstream pipe. To specify
an exit valve, tick the Exit Valve checkbox. Then specify the exit
pressure.
Numerous handbook loss factors for generic valve configurations are
given in the Valve Data Source area by selecting Handbook Data.
If you input Cv vs. open percentage on the Optional tab, you can enter
the actual open percentage and AFT Impulse will calculate the Cv.
For more information on the loss models for these valves, refer to
Chapter 8.
Transient Data
On the transient data tab a transient Cv or K table can be entered to
simulate the opening or closing of the valve. For a closed valve, the K
factor is infinite. To account for this, enter a K factor of –1 (negative
one) in the transient data table. More information on transient data,
including event transients, is given earlier in this chapter and in Chapter
10.
Special Conditions
The Special Condition for a valve always closes it.
Vacuum Breaker Valve Specifications window

The Vacuum Breaker Valve Specifications window is shown in Figure
6.36. The Vacuum Breaker Valve junction type (also called an air inlet
valve) requires two connecting pipes.

manual (si) g.doc
The Vacuum Breaker Valve Specifications window follows the first of
the two basic Specifications window formats, displaying the connecting
pipes in a fixed format. The flow direction through the junction is
determined by the defined directions of the connecting pipes.
Figure 6.36 Vacuum Breaker Valve Specifications window
The Vacuum Breaker Valve offers protection for low pressure

conditions. When the internal pressure drops below the cracking
pressure (usually atmospheric), it opens to admit air (or whatever gas
you specify). This keeps the pressure from dropping much below the
atmospheric pressure. When the internal pressure rises above the
atmospheric pressure, the air is discharged. To prevent large pressure
spikes, the discharge flowrate is frequently kept lower than the incoming
flowrate by using a smaller discharge flow area for outflow than inflow.
A vacuum breaker junction can connect with one pipe or two. When
connected to one pipe, the valve is located at the end of the pipe. When
8/10/2007 4:10:28 PM Page 275

connected to two pipes, the valve is “inline” and allows air to flow in
and out at that location.
During steady-state a vacuum breaker valve with one connecting pipe
acts as a dead end. If connected to two pipes, it acts as a lossless valve
allowing liquid to flow between the two pipes.
By default, the vacuum breaker junction allows air to flow in and out.
Alternatively, it also model only air inflow or outflow.
Three-Stage Valve Modeling

AFT Impulse supports modeling of a special type of vacuum breaker
valve called a three-stage valve. Three stage valves work the same as
regular valves when gas is flowing in, but have two orifice sizes when
gas flows out. Pressure difference or volumetric flowrate criteria specify
which orifice to use. When the actual value is less than the specified
value the normal Outflow CdA is used. When the actual value is higher,
the intermediate orifice CdA is used. This intermediate orifice CdA
values are typically much smaller than the normal outflow CdA.
Graphing Vacuum Breaker Valve data

You can track the amount of air in the system and the flowrate by
selecting these options in the Junction section of the Transient Control
window. These parameters can be graphed or reviewed in the Output
window.
Volume Balance Specifications window

The Volume Balance Specifications window is shown in Figure 6.37.
The Volume Balance junction type is internal to the system and requires
two connecting pipes.
The Volume Balance Specifications window follows the first of the two
a fixed format. The flow direction through the junction is determined by
the defined directions of the connecting pipes.

manual (si) g.doc
Figure 6.37 The Volume Balance Specifications window
The Volume Balance junction is useful for modeling pipelines with a

moving interface between fluids of different density. Such a scenario is
in reality a transient application, because the mass in the system is not
steady with time. AFT Impulse assumes the interface does not move
during the transient.
Volume Balance junctions should be used in conjunction with the
variable physical property model (see System Properties window).
Global Edit windows

The Global Pipe Edit and Global Junction Edit windows, accessible
through the Edit menu, assist you in making large scale changes to your
model.
8/10/2007 4:10:28 PM Page 277

Global Pipe editing

The Global Pipe Edit window, shown in Figures 6.38 and 6.39, allows
you to edit data for multiple pipes at the same time. Here is how you use
it:
1. Open the Global Pipe Edit window from the Edit menu.
2. Select the pipes you want to edit from the Pipe list on the left (see
Figure 6.38).
3. Click the Select Pipe Data button. This will open the Pipe
Specifications window (see Figure 6.1) in which you can enter the
data you want to change on each of the selected pipes.
4. Enter the pipe data you want to change and click the OK button in
the Pipe Specifications window.
5. The Global Pipe Edit window now lists the parameters to change in
a list on the right (see Figure 6.39). Select the parameters you want
to change.
6. Click the Apply Selections button.
Note that you cannot change some data, like the pipe ID number, using
this global method.

manual (si) g.doc
Figure 6.38 The Global Pipe Edit window allows you to change
data for many pipes at the same time. This is how the
window appears when first opened.
Global Junction editing

The Global Junction Edit window, shown in Figures 6.40-6.44, allows
you to edit data for multiple junctions at the same time.
Junction data falls into two broad categories. Common data such as
elevation and initial pressure which apply to all junctions, and junction
specific data, such as a pump curve, which are specific to certain types
of junctions. You can switch between the two categories by selecting the
option in the upper left (see Figure 6.40).
8/10/2007 4:10:28 PM Page 279

Figure 6.39 The Global Pipe Edit window after selecting the
“Select Pipe Data” button. Data can now be applied.
Editing common junction data

You can open the Global Junction Edit window for common data editing
in two ways. First, you can choose Common Data from the Global
Junction Edit menu selection on the Edit menu, or you can choose one of
the other specific junction types on the menu and then choose the
Common Data option in the upper left (see Figure 6.40).

manual (si) g.doc
Choice of editing
common or
specific data
Figure 6.40 Global Junction Edit window. This is how the window
appears when first opened to edit Common data.
Here is how you globally edit junction common data:

1. Open the Global Junction Edit window.
2. Select the Common Data option in the upper left (see Figure 6.40).
3. Select the junctions you want to edit from the Junction list on the
left (see Figure 6.40). Note that all junctions in the model are shown,
no matter what the junction type.
4. Click the Select Common Junction Data button. This will open the
Common Junction Data Edit window (see Figure 6.41) in which you
can enter the data you want to change on each of the selected
junctions.
5. Enter the junction data you want to change and click the OK button
in the Pipe Specifications window.
8/10/2007 4:10:28 PM Page 281

6. The Global Junction Edit window now lists the parameters to change
in a list on the right (see Figure 6.42). Select the parameters you
want to change.
Note that you cannot change some data, like the junction ID number,
using this global method.
Figure 6.41 Common Data Edit window for global junction editing.
Editing specific junction data

You can open the Global Junction Edit window for specific data editing
in two ways. First, you can choose the junction type from the Global
Junction Edit menu selection on the Edit menu, or you can choose
Common Data on the menu and then choose the Specific Data For option
in the upper left (see Figure 6.43), and then choose the junction type
from the dropdown list.

manual (si) g.doc
Figure 6.42 The Global Junction Edit window after selecting the
“Select Common Junction Data” button. Data can now
be applied.
Here is how you globally edit specific types of junctions:

1. Open the Global Junction Edit window.
2. Select the Specific Data For option in the upper left (see Figure
6.43).
3. Select the type of junction you want to edit in the dropdown list in
the upper left (see Figure 6.43). Here Spray Discharge is selected.
4. Select the junctions you want to edit from the Junction list on the
left (see Figure 6.43). Note that only junctions of type selected in the
upper left are shown Spray Discharge junctions in this case).
5. Click the Select Spray Discharge Data button (the button name will
change based on the type of junction selected). This will open the
appropriate junction Specifications window, in this case the Spray
Discharge Specifications window.
8/10/2007 4:10:28 PM Page 283

Option for
specific types of
junctions.
Junction type
selected in list.
Figure 6.43 The Global Junction Edit window allows you to

change data for specific types of junctions (Spray
Discharge shown here)
6. Enter the junction data you want to change and click the OK button
in the Spray Discharge Specifications window (or whatever window
type is displayed).
7. The Global Junction Edit window now lists the parameters to change
in a list on the right (see Figure 6.44). Select the parameters you
want to change.

manual (si) g.doc
Figure 6.44 The Global Junction Edit window allows you to

change data for specific junction types (Spray
Discharge junctions shown here)
8/10/2007 4:10:28 PM Page 285


manual (si) g.doc
CHAPTER 7
Customizing AFT Impulse
This chapter explains how to customize AFT Impulse in ways that

increase your productivity. To accelerate repetitive operations you can
customize the way the Workspace looks and behaves and the way output
is formatted. Changes to the program's appearance and functionality are
saved to the IMPULSE4.INI file in your application data folder m (e.g.,
C:\Documents and Settings\Your Account\Application Data\Applied
Flow Technology\AFT Impulse\4.0).
You can also build custom databases that are read in during startup and
used as if they were native to the program. These databases are saved to
the IMP_USER4.DAT file in your application data folder.
If you pursue extensive customizing of AFT Impulse, you are strongly
advised to make regular backup copies of the two files (IMPULSE4.INI
and IMP_USER4.DAT). If the files ever become corrupted you could
lose all the custom data.
Tools are also provided to allow creation of network databases to share
frequently used data and report formats across a local or wide area
network.
Parameter and Unit Preferences

The Parameter and Unit Preferences window, shown in Figure 7.1, is
accessed from the Options menu. The features in this window allow you
to specify the default parameters with which you would like to work. By
changing any of the parameters in this window you do not lock yourself
out from certain features. Rather, when you encounter lists or options

manual (si) g.doc
during assembly of your model, AFT Impulse offers your default
parameters as a first choice. This way you do not have to repeatedly
change the parameter in the Specifications windows.
Parameter Preferences
The first parameter you can set is the Use Most Recent Pipe Size Data
feature. When this is selected, AFT Impulse remembers your most recent
pipe material size setting and sets new or unspecified pipes to this pipe
material and size.
You can also set a default pipe material and default friction model. You
can change the material and loss model of each individual pipe at any
time, but this will be the first choice.
Figure 7.1 Parameter and Unit Preferences window sets default

values to use.
There are Default Design Factors. The function of design factors is

discussed in more depth in Chapter 8. If you want to set up defaults that

manual (si) g.doc
Chapter 7 Customizing AFT Impulse 289
are applied to your pipes and junctions, here is where you do it. If you
want to change the design factors for existing pipes or junctions, see the
Global Pipe Edit and Global Junction Edit topics in Chapter 6.
In the upper right you can specify default parameters for flow rate and
pressure head.
Another parameter on this tab is the Default Elevation. The Default
Elevation is especially useful when the majority of your junctions all
reside at the same elevation. When you specify a default elevation, all
the junctions you create will show the default elevation, saving you from
having to enter the same elevation repeatedly. AFT Impulse defaults to
working without a default elevation specified.
Note: The inputs here merely specify defaults for pipes and junctions
not yet placed in the model. If you want to change data for existing pipes
and junctions, use the Global Pipe Edit and Global Junction Edit
windows.
Unit Preferences
You can set the units you prefer to use in AFT Impulse on the Unit
Preferences tab (see Figure 7.2). These units are selected as a first choice
in the unit selection boxes. You can choose any of the units available. To
set a new default, choose the unit type from the list on the right and click
the Set As Preferred Unit button.
You also can customize the units used by AFT Impulse. For each
parameter, AFT Impulse can use many different units. However, if only
a few are ever typically used, then the others can be removed. This will
decrease the number of items in the unit selection boxes, making it easier
to choose the one desired. If a preferred unit is removed from those
available to Impulse, a new one will have to be chosen before leaving
this area. You also must have one set of units available for each
parameter.
To customize the units, first choose the parameter to modify on the list
on the left. Then, among the unit types on the right, check or uncheck the
boxes for the units you want or do not want to see.
You can select or unselect all displayed units by clicking the appropriate
button. Also, English and SI units can be selected as a group for
parameters by clicking the appropriate button.
8/10/2007 4:10:28 PM Page 289

Figure 7.2 Using the Unit Preference tab in the Parameter and
Unit Preferences window, you can set the default units
you want to work with and choose to show only the
units you want to work with.
Tip: To choose only English or SI units for all parameters, go to the Unit
Preferences tab, choose the Default Unit System then click the Apply
Default Units button. This will use AFT Impulse’s default units. This
will also change the preferred units.
Setting preferred units

To set a particular unit as the preferred unit, choose it in the list on the
right, and click the Set as Preferred Unit button.

manual (si) g.doc
Description of selected unit

A more complete description of the selected unit is shown at the lower
right of the window in the Unit Description field.
In the lower left of the Parameter and Unit Preferences window there are
two Database checkboxes. If these boxes are checked, your Parameter
and Unit Preferences parameters are set up as determined by the
database to which you are connected. This is referred to as an active
database. To make them inactive, uncheck the boxes or change one of
the Parameter and Unit Preferences settings controlled by the database.
You can specify database connections separately for parameters and
units.
If the checkboxes are unchecked, but enabled, you are connected to a
database but the settings are not being passed to the Parameter and Unit
Preferences window. The databases are thus inactive. To make them
active, check the boxes then click the OK button.
If the checkboxes are disabled, there are no connected databases.
Later in this chapter we’ll discuss how databases are configured and
administered through your local or wide area network.
Command buttons
There are eight buttons at the bottom of the Parameter and Unit
Preferences window. Impulse has built-in default parameters, units and
settings which you can choose by clicking the Impulse Default button.
You can also develop your own settings, tailored to your project or
industry, and have these used by default (instead of Impulse’s defaults).
To make your own default, first select the parameters, units and settings
you would like to use, then click the Set As Default button. Your settings
will be saved and will be used each time any new project is initiated. If
you make changes to the settings, and want to get back to your defaults,
click the User Default button. The default settings are updated only
when you click Set As Default.
You can save the parameter and unit settings to a file by pressing the
Save Preferences button and entering a file name. These setting are
loaded again by pressing the Load Preferences button and choosing the
file name. The format files you create can be placed on a network for
8/10/2007 4:10:28 PM Page 291

sharing among a group or company, or incorporated into a company-
wide database, allowing standardized reporting.
If you have made changes that you don’t want to keep, click the Cancel
button. Click OK to use the settings you have defined.
Workspace Preferences
The Workspace Preferences window is accessed from the Options menu.
The features offered in this window allow you to customize the
appearance and behavior of the Workspace.
Pipes and Junctions

You can customize the pipe thickness and style, and the junction size, on
the Pipes and Junctions tab, shown in Figure 7.3.
Pipe Line Options

Normally, all pipes on the Workspace are displayed using the same line
thickness. This thickness value is set using the dropdown list at the top
of the Pipe Line Options area. A sample of how the pipe will be
displayed is shown immediately below the dropdown list.
There are two additional options that will vary the line thickness
according to the pipe diameter. If you choose the Vary Thickness With
Pipe Diameter option, the pipe line thickness will be varied according to
the pipe physical diameter. The smallest diameter pipe in the model will
be displayed as a line 1 pixel thick and the largest diameter pipe will be
displayed as a line the thickness you specify in the Pipe Line Thickness
dropdown list. All other pipes will be shown using a thickness
somewhere in between, according to their diameter.
Similarly, the Vary Thickness Using option will vary the pipe line
thickness from the smallest diameter pipe in the model to the largest.
However, in this option you can specify the range of pipe diameters and
line thicknesses to use. If a pipe is smaller than the minimum diameter
specified it will be shown using the minimum thickness. Accordingly, if
a pipe is larger than the maximum diameter, the maximum thickness will
be used. This option is useful to keep a few very large or very small
pipes in the model from skewing the display.

manual (si) g.doc
Figure 7.3 The Pipes and Junctions tab in the Workspace

Preferences window.
By varying the pipe thicknesses you can easily see the feeding pipes, the
large flow area pipes, or the small supporting flow pipes.
Each individual pipe allows you to modify its thickness, which overrides
the default size set here.
Special Conditions options

You can change the pattern of a closed pipe by choosing the style (e.g.
dots or dashes) from the dropdown list in the Closed Pipe/Junction
Options area. All pipes, whether closed using the Special Conditions
option, or closed as a result of another object, will be displayed using
this style.
8/10/2007 4:10:28 PM Page 293

The junctions in a closed flow path are outlined using this style. An
example of the closed pipe style is shown below the list box. If you
choose Solid, no distinction will be no distinction made between a
closed and open pipe.
Action When Drawing Selection Right To Left

By default this option is turned ON, and when ON the Selection
Drawing tool on the Workspace will select all objects completely or
partially within the selection box when the selection box is drawn from
right to left.
Pipe endpoint adjustments

By default, pipes snap to the center of junction icons. You can disable
this here if you desire.
Modifying icon size

The default size of the junction icons on the Workspace can be changed
to 25%, 50%, 75%, or 100% of the full size. Click the option button next
to the desired size in the Junction Icon Size area; the sample junction
will change, reflecting your choice. Reducing the junction size may free
up more Workspace area and present a more useful or appealing
perspective of your models. Each individual junction allows you to
modify its size, which overrides the default size set here.
Action when dragging junctions

Normally, when a junction is dragged around the Workspace, the pipes
stay attached unless the CTRL key is pressed during the movement. You
can reverse this action (pipe will detach while dragging unless the CTRL
key is pressed) by checking the Leave Connected Pipes in Place check
box. This is useful when you are rearranging the model.
Auto increment labels

When new pipes or junctions are added to the Workspace, they are
assigned a new number that is greater than the highest pipe or junctions
number by one. The increment can be changed from this default to a
higher value.

manual (si) g.doc
Icon source
Limitations exist in Windows 95, 98 and Me that can affect AFT
Impulse’s ability to display large models on the Workspace. The
limitations relate to what are called GDI resources. The practical result
is that models that display more than 150-250 junctions may have
trouble. In such cases, it is recommended the Icon Source be changed to
the Large File Icon Method. This allows you to use all icons that are
displayed on the Toolbox, plus it offers the ability to rotate icons. The
Toolbox Icons Only option only allows display of icons that are shown
on the Toolbox. When either of these options are chosen, the Workspace
Icon area on each junction Specifications window is disabled. Using
different color junctions is therefore not available.
Windows NT, 2000 and XP do not experience these limitations and the
user should thus always use the Complete Icon Set.
Display Options
From the Display Options tab you can change the symbols used, text
content, and location of the text on the Workspace (see Figure 7.4).
8/10/2007 4:10:28 PM Page 295

Figure 7.4 The Display Options tab in the Workspace Preferences

window.
Workspace grid
A grid can be displayed on the background for modeling convenience.
The grid can be fine, medium or coarse. The grid lines can be solid or
dashed. And finally, the pipes and junctions can be snapped to the grid
you specify.
The grid can be included in printouts by specifying its inclusion in the
Print Preview/Special window.
Workspace symbols
Special symbols can be displayed on the Workspace adjacent to pipes
and junctions to denote special behavior. For example, an X (or symbol
of your choice) can represent a Special Condition. If you want to use

manual (si) g.doc
these symbols, click the “Use” checkbox and specify the desired symbol.
By default all symbols are selected for use.
There are special symbols displayed on the Workspace for transient
related data. These symbols can be modified here. By default, a "T"
symbol is displayed if a junction has transient data. A "#" symbol means
that a Transient Special Condition has been specified. And a "*" symbol
means that the junction is to have no reflections.
Pipe direction arrows

Each pipe has a reference direction, and this direction can be displayed
in one of two ways. First, an arrow can be displayed on the Workspace
pipe. Second, an arrow can be displayed next to the pipe number. If an
arrow is shown on the pipe itself, one of several styles can be selected.
Displaying name and/or ID numbers

You can set the default number and/or name display setting for junctions
and pipes in the Pipe/Junction Display area. This setting will be used for
all new pipes and junctions. You can change the setting on a pipe-by-
pipe or junction-by-junction basis in the Specification window or by
using Global Edit.
Setting the default junction label position

The default label position for all junctions on the Workspace can be
adjusted in the Junction Label Location area. Click the option button at
the desired location around the sample junction icon. Note that the
junction (and pipe) labels can be moved individually
Allowable Workspace label movements

By default, pipe and junction labels Workspace can be moved to new
locations. You can change this functionality to only allow movements
when holding the shift key.
Popup menu
The popup menu can be disabled (it is enabled by default) here. The
popup menu is a context sensitive menu available on the Workspace.
8/10/2007 4:10:28 PM Page 297

Special Conditions Graphics
By default, a junction with a Special Condition has a red square drawn
around it. That can be disabled here.
Background Picture Scaling

Background pictures loaded on the workspace can be scaled by
specifying a scale multiplier here.
Colors and Fonts

The color of pipes, junction labels, etc., can be set on the Colors and
Fonts tab (see Figure 7.5). To change the color of an item such as the
Workspace background, select the Workspace Background from the list
and click the Set Color button. You also can just double-click the
Workspace Color itself. The color of individual pipes can be assigned in
each pipe window. This will override the default pipe color specified
here.
In addition to the color area, the junction and pipe font is displayed. To
change fonts, click the Change Fonts button to open the Fonts window
then, choose a new font or font style. A sample is displayed above the
Change Font button.
Sample Workspace
The tab for Sample Workspace shows you a sample of how the
Workspace pipes and junctions will appear if you click the OK button.
In the lower left of the Workspace Preferences window there is a
Database checkbox. If this box is checked, your Workspace Preferences
parameters are set up as determined by the database to which you are
connected. This is referred to as an active database. To make it inactive,
uncheck the box or change one of the Workspace Preferences settings
controlled by the database.

manual (si) g.doc
Figure 7.5 The Colors and Fonts tab in the Workspace

Preferences window.

database but the settings are not being passed to the Workspace
Preferences window. The database is thus inactive. To make it active,
check the box then click the OK button.
Later in this chapter we’ll present a discussion of how databases are
Command buttons
There are eight buttons at the bottom of the Workspace Preferences
window. Impulse has built-in default parameters, units and settings
which you can choose by clicking the Impulse Default button. You also
can develop your own settings, tailored to your project or industry, and
8/10/2007 4:10:28 PM Page 299

have these used by default (instead of Impulse’s defaults). To make your
own default, first select the settings you would like to use, then click the
Set As Default button. Your settings will be saved and will be used each
time any new project is initiated. If you make changes to the settings,
and want to get back to your defaults, click the User Default button. The
settings are updated only when you click Set As Default.
You can save the parameter and unit settings to a file by pressing the
Save Preferences button and entering a file name. These setting are
loaded again by pressing the Load Preferences button and choosing the
file name. The format files you create can be placed on a network for
sharing among a group or company, or incorporated into a company-
wide database, allowing standardized reporting.
If you have made changes which you don’t want to keep, click the
Cancel button. Click OK to use the settings you have defined.
Toolbox Preferences
The Toolbox Preferences window is accessed through the Options menu.
Figure 7.6 shows the layout of the Toolbox Preferences window. The
features in this window allow you to customize the Toolbox in the
manner that is most productive for you.
Once you have changed the appearance and functionality of the Toolbox,
you can set your changes as the user default values by clicking the Set
As Default button. From then on, AFT Impulse will always load with
your default Toolbox settings.
Customizing Toolbox contents

When you open the Toolbox Preferences window you can see two large
areas labeled Current Toolbox and Unused Tools. By dragging the
junction type icons from the Current Toolbox to the Unused Tools area
you can remove these junction types from the Toolbox. This may be
useful if you use only a limited number of junction types and don't want
the unused types displayed on the Toolbox.

manual (si) g.doc
Figure 7.6 Toolbox Preferences window allows you to customize

the Workspace Toolbox.
You can also reorder the icons on the Toolbox by dragging and dropping
them to the desired locations on the Current Toolbox.
Toolbox shortcuts
You can show or hide Toolbox shortcut information.
8/10/2007 4:10:28 PM Page 301

In the lower left of the Toolbox Preferences window there is a Database
checkbox. If this box is checked, your Toolbox Preferences parameters
are set up as determined by the database to which you are connected.
This is referred to as an active database. To make it inactive, uncheck
the box or change one of the Toolbox Preferences settings controlled by
the database.
database but the settings are not being passed to the Toolbox Preferences
Later in Chapter 7 a lengthy discussion is given of how databases are
General Preferences
The General Preferences window (Figure 7.7) allows you to:
• Specify automatic model saving
• Options for opening models with scenarios
• What content to display on the Status Bar
• Require a check of the Waterhammer Assistant before running
models
• Enable Tip of the Day feature
• Number of data points to show in input tables
• Whether or not to confirm before clearing output results
• Setting parameters for “button flashing” for curve fit buttons on
junction Specifications windows
• Default locations for saving models
• Common database locations

manual (si) g.doc
• Backup folders where your IMP_USER4.DAT and IMPULSE4.INI
files are copied (this is useful for keeping backups should you
change computers)
• Specify how AFT Impulse handles transient junction data that does
not cover the entire time range of the simulation (either maintain the
data as constant at the last data point or extrapolate). Maintain is the
default option.
Figure 7.7 General Preference window
Customizing the Output

AFT Impulse allows you to save your output preferences as the user
default. To do this, change the Output Control window to your preferred
arrangement, then click the Set As Default button. From then on this
arrangement will be the default used when any new project is initiated.
8/10/2007 4:10:28 PM Page 303

If you make changes to the Output Control window and you want to
return to your user default setup, click User Default. Clicking Impulse
Default activates the original defaults used by AFT Impulse. No updates
to the defaults are made unless you click Set As Default.
You can save frequently-used output formats to file for later reuse. For
example, for checking modeling errors you may wish to create an Output
Control file named “ALL.DAT” that has all parameters, which you can
quickly access when needed. You may have another named
“FINAL.DAT” for final reports. The Output Control format files can be
located on a PC network for access by all engineers to allow for standard
reporting between different projects and groups.
More extensive discussion on using Output Control to customize your
output is given Chapter 4.
Specifying graph style preferences

When preparing graphs in the Graph Results window, you may prefer to
work with certain colors and styles. The Customize Graph window
allows you to save the style preferences using the features on the System
tab. You can later recall these preferences using the features in the same
area.
Auxiliary Graph Formatting window

The Auxiliary Graph Formatting window allows you to more easily
change graph colors and curve thickness in the Graph Results window.
These preferred settings can be saved as the default for future use. In
addition, these preferences are used in all graphs throughout AFT
Impulse. This window is opened from the Options menu or Toolbar.
Specifying Visual Report preferences

The Visual Report window content is specified in the Visual Report
Control window. To open the Visual Report Control window, choose it
from the View menu or Toolbar. Through the Visual Report Control
window you can choose the parameters you want to see in the Visual
Report window.

manual (si) g.doc
The parameters and text locations are automatically saved with the
model. However, you may want to save the settings for use with a
different model. To save your preferred layouts to an AFT Impulse
visual layout file, click the Save Options button and assign a file name.
The default file extension is .ivs. Later you can recall your customized
layouts for rapid preparation of the Visual Report window.
The many features in the Visual Report Control window are described
further in Chapter 4.
Customizing Pipe Fittings & Losses

The Pipe Fittings & Losses window presents organized lists of pipe loss
fittings. This list is customizable. You can add your own fittings to the
base list provided with AFT Impulse.
The standard losses provided with AFT Impulse are located in the
IMPULSE4.DAT file in the same directory as the AFT Impulse
executable file. You should not change anything in this file.
The Fittings & Losses Edit window (Figure 7.8) can be opened from the
Database menu.
Building custom databases

One of the more powerful features of AFT Impulse is that it provides
extensive customization capabilities. AFT Impulse provides the tools to
build custom databases for fluid properties, pipe materials, and all
junction types except Volume Balance. AFT Impulse incorporates these
databases internally and offers them to you through lists.
There are two types of databases that may be customized: local and
external (Note: the AFT Default Internal Database is not included here
since it may not be altered). The local database is located in the
IMP_USER4.DAT file in the application data directory.
8/10/2007 4:10:28 PM Page 305

Figure 7.8 Fittings & Losses Edit window allows you to and edit
fitting and loss data that is used in pipes.
In AFT Impulse you can create multiple external or local custom

databases. The Database Manager window on the Database menu
simplifies this process.
When the databases are placed on a network, the data management
features described below can be implemented.
Specific, quality-checked data can be entered into databases located on a
local or wide area network of PC’s. This data is not accessible to the end
users through AFT Impulse itself, so it is guarded data.
Once placed on a network, users can connect to any database found in
the Database Manager. Users can connect to as many databases as they
want, and get access to all data in databases using Database Manager.
Adding custom fluid properties to the Fluid Database

You can open the Fluid Database window from the Database menu or
from within the System Properties window. Figure 7.9 shows the layout
of the Fluid Database window. This window allows you to review

manual (si) g.doc
currently defined custom fluids, to change the parameters that define the
fluids, and to enter new fluids. All fluid properties are assumed to be a
function of temperature only.
Figure 7.9 Fluid Database window
Once you have entered a custom fluid, it will appear in the Fluid List
area of the System Properties window whenever you use AFT Impulse.
You can add new fluids to the database by clicking the Add New Fluid
button and opening a new window (see Figure 7.10). The constants
provided allow you to describe the variation of density, dynamic
viscosity, bulk modulus of elasticity, and vapor pressure as they depend
on temperature or solids concentration for newtonian fluids. In addition,
there is an option to describe the variation of density, viscosity, rigidity
coefficient and yield stress for a Bingham Plastic non-newtonian fluid as
it depends on solids concentration. This window provides the tools to
perform polynomial curve fits of fluid data; or, if you already have
appropriate constants, you can input them directly. In addition, the raw
data can be interpolated
8/10/2007 4:10:28 PM Page 307

Figure 7.10 Add New Fluid window
Density, viscosity and bulk modulus are required for all fluids. Vapor
pressure is required if cavitation modeling is desired.
The Fluid Database window also allows you to open the Change Fluid
Data window, which is essentially the same window as the Add New
Fluid window. For easy editing, data for the fluid you wish to change is
entered for you in the appropriate areas. The original curve fit data you
entered is also displayed.
The Add New Fluid/Change Fluid Data windows displays cross-plotting
of your custom curves against the curve fit data points. You can also
import data from text files into these windows.

manual (si) g.doc
Adding custom pipe materials to the Pipe Material

Database
You can open the Pipe Material Database window from the Database
menu. Figure 7.11 shows the Pipe Material Database window. Using this
window, you can easily add to the user database the geometric
information for any pipe material in the length units of your choice.
Figure 7.11 Pipe Material Database window
After you have entered a custom material, each time you open a Pipe
Specifications window the material will be available to you for quick
selection. This minimizes the need to work with handbooks.
Pipe materials typically have a range of nominal sizes and a specific
type, class, or schedule that describes the actual dimension. AFT
Impulse works with the inner diameter of the pipe, since that is the
parameter that directly affects the fluid mechanics.
You can also add multiple Friction Data Sets for each pipe material.
This allows you to keep, for example, different friction values that may
apply to the same pipe under different conditions. These conditions
could be related to the age of the pipe or the type of fluid being carried.
¾ To add custom pipe materials to the user database, choose the entry
location you desire in the list on the left (Pipe Materials, Sizes and
8/10/2007 4:10:28 PM Page 309

Types). Then click the New button below. For example, to add an
entirely new material, click the Pipe Materials selection in the list then
click the New Material button. To add a new size for an existing material
(say Steel), click Steel in the list then click the New Size button. To add
a new type to the Steel 1 inch size, click Steel 1 inch in the list then click
the New Type button.
When adding a new material, you also will be asked to specify a nominal
size, type of material, diameter, wall thickness, friction data set, and
friction value to define the first data. From then on you can enter
additional nominal sizes and types. The name descriptors you choose for
the material, size, and type are at your discretion. They can contain text
mixed with relevant numbers if you desire.
In addition to manual entry of the custom pipe materials, you also can
import pipe material data from a file. When you click the Import From
File button, a dialog box will appear that is self explanatory. Just follow
the directions in the dialog box. By choosing to import from a file, you
will be able to have a separate copy of the data for later access.
Friction Data Sets

A friction data set can be associated with the entire pipe material, or
with specific sizes or types. For instance, a friction data set associated
with Steel (i.e., the top level) might be called “25 year usage” and have a
roughness value of 0.127 mm. This data set is then available to all sizes
and types of steel. However, if “25 year usage” is associated only with
the 1 inch (2.66 cm ID) of Steel, it will apply only to the pipe schedules
that are 1 inch (2.66 cm ID).
You can also associate friction data sets with a specific type (i.e.,
schedule). For instance, you can make a “25 year usage” and associate it
only with 1 inch (2.66 cm ID) schedule 40 Steel pipe and assign its
roughness value as 0.152 mm. Then all Steel pipe would use 0.127 mm
for “25 year usage” friction while the different value for 1 inch (2.66 cm
ID) schedule 40 of 0.152 mm would supercede the value of 0.127 mm
associated with all Steel pipe.
Finally, among multiple friction data sets you can specify a default data
set.

manual (si) g.doc
Pipe Physical Properties
The Pipe Material Database window allows you to assign physical
properties of the pipe. The density and Poisson ratio of the pipe can be
entered as constants, and the modulus of elasticity as a function of
temperature.
Adding custom junction data to the Component

Database
To simplify modeling and eliminate repetitive data entry, AFT Impulse
allows you to save frequently-used junction data to a custom database.
Once saved, the custom equipment data can be immediately recalled
from the Database List in the Specifications window. Using this tool,
you can easily build your own database of pumps, valves, or other
components supplied by specific manufacturers. If transient data is
entered, it too is saved to the database.
¾ To save a Custom junction to the database, open the junction

Specifications window and enter all required data. Close the window and
choose Add Component to Database from the Database menu. You will
be prompted to enter a name for the custom component. Enter a
descriptive name and close the window. If you open the junction again,
the custom name will be selected in the Database List.
AFT Impulse accepts custom data for all junction types except Volume
Balance. Volume Balances are not included because there is no custom
information for Volume Balance and thus nothing to store.
Editing the Component Database

You can delete or rename any existing database component by opening
the Component Database window from the Database menu (see Figure
7.12). To edit an existing database entry, select the item in the list and
click the Edit Data button. This opens the data into a specifications
window where the data can be modified.
8/10/2007 4:10:28 PM Page 311

Figure 7.12 Component Database window shows junction items in

database and allows editing
Database Manager
Accessed from the Database menu, Database Manager allows you to –
• Connect to and disconnect from databases
• Create and delete databases
• Review the content of databases
• Move contents between databases
Setting available databases

To work with a database you must first make it available to AFT
Impulse. Referring to Figure 7.13, clicking on the Edit Available List
button and choosing Add Engineering Database from the dropdown
menu will open a standard file dialog box with which you may browse to
and select a database. A database may be removed from the list of
available databases by selecting it and clicking on the Remove Database

manual (si) g.doc
button. Contents of a database may be reviewed by selecting a database
from the list and clicking on the Review Content button.
Figure 7.13 Available Databases in the Database Manager
The AFT DEFAULT INTERNAL DATABASE and the AFT IMPULSE

LOCAL USER DATABASE may not be removed and will always
appear in the list of available databases. External shared databases will
also always appear in the available database list. These databases are
distinguished from other external databases by being listed in the
DATABASE.LIB file residing in the AFT Impulse directory.
Connecting and disconnecting to a database

To use a database within a model it must first be made available, as
described above, and connected using the features in the Connect To
Database tab. To connect a database, select it from the list of available
databases and click on Add to Connections (see Figure 7.14). Similarly,
to disconnect a database, select it from the list of Connected Databases
and click on Disconnect Database. The sections contained in the selected
database will appear in the list to the right, Database Selections, from
8/10/2007 4:10:28 PM Page 313

which the sections to be used in the model may be selected; i.e.
components, fluids or pipe materials.
Figure 7.14 Connect To Database tab of the Database Manager

allows you to connect to and disconnect from
databases, review their contents and select the
sections of a database to be used.
The contents of a connected database may be reviewed by selecting it

and clicking on Review Content.
Clicking on Set As Default will establish the list of connected databases
as the default list, while clicking on User Default will set the
connections to the default list.
Editing databases
The Edit Database tab of Database Manager allows you to create and
delete databases, and edit their contents by copying sections or items
from one database to another. The tab is divided into a Source Database
area on the left and a Changing Database area on the right (see Figure
7.15).

manual (si) g.doc
To edit a database, first choose a source database by selecting one of the
four options:
• Currently available external
• Data From Current Model – components, fluids and materials made
in the current model using the Component Database, Fluid Database
or Pipe Material Database
• Local User Database – IMP_USER4.DAT located in your Windows
directory, where component, fluid and materials data created by the
user is stored by default
• Other External – by selecting this option and clicking on the Select
Database, you are able to browse and select a database file from any
available drive and directory
Select the database to
be used as the source
Click here to select an

external database as
the source
File name, title and

available sections
within the source are
listed here
Selected items may be

moved or copied from
the source to the
database being
changed
Figure 7.15 Edit Database tab allows you to select sections from a
specified database and add them to an existing or new
database.
The file and descriptive name of the selected database are displayed as
well as the sections available in that database, i.e. fluids, components or
pipe materials.
8/10/2007 4:10:28 PM Page 315

The database to be changed and what changes are to be made are
specified in the Changing Database portion of the Edit Database tab.
• Select Database to Edit… – browse to and select an existing
database file
• Create New Database… - specify a location, file name and
descriptive name for a new database. The file and descriptive name
of the selected or created database appears below (Figure 7.16).
• Delete database – deletes the database displayed
• Review content – displays the content of the database listed
• Delete section – deletes from the listed database the section selected
from the Database Sections list
Figure 7.16 After specifying a database file name, you will be

prompted for a description which will then appear in
the various lists of the Database Manager when this
database is selected.
Benefits of shared databases

When the Impulse databases are placed on a network, powerful data
management procedures can be implemented. Specific, quality-checked
data can be entered into databases located on a local or wide area
network of PC's. This data is not accessible to the end users through
AFT Impulse itself, so it is guarded data.
Once the data is placed on a network, users can connect to any database
listed in DATABASE.LIB by selecting the database in the Connect
Database window on the Database menu. Users can connect to as many
databases as they want, and get access to all data in databases listed in
DATABASE.LIB.

manual (si) g.doc
In addition, special modeling control files can also be included in your
network databases. These include:
• Output Control format files
• Model Data Control format files
• Workspace Preference format files
• Parameter Preference format files
• Unit Preference format files
• Visual Report Control format files
• Toolbox Preference format files
If the user connects to a database with these files and selects to be
connected to any of these control files, the model will inherit the control
formatting specified in the control files. This means that special report
layouts, special graphical representations and colors, and other special
controls can be decided upon by the database constructor, and those
control features can be selected into local models by end users. This is
highly useful for creating customized reports for certain customers or
project requirements.
By way of example, if the Output Control of Figure 4.15 is specified as
coming from the “New Mile High Stadium Project” database, it would
appear as shown in Figure 7.17.
If you uncheck the Database checkbox or change certain data in the
Output Control window, the database will remain connected but change
to an inactive state. When the database is inactive, none of its data flows
through to the model. To bring the Output Control data back into sync
with the connected database, recheck the Database checkbox in Output
Control.
This entire custom database concept lends itself naturally to the idea of
specific project databases. Each project has differing requirements that
depend on the nature of the project and the end customer. Data for each
project can be assembled into separate databases, allowing faster and
more accurate transmission of the controlled data and desired reporting
formats to engineers.
Maintaining a network database allows commonly used data to be
controlled, which may be helpful in instances where quality control is
strict.
8/10/2007 4:10:28 PM Page 317

Database
connection
Figure 7.17 Output Control window with database connected and

active.
Creating an enterprise-wide, network database system
Overview
Users of AFT Impulse on a network can access common databases to
simplify model verification, improve analysis and reporting consistency,
and eliminate redundant work.
To accomplish this there is some initial setup that must be done. First,
the data for components, materials, formats, etc., needs to be created and
saved to various data files. Secondly, these files must be made available
to the users of AFT Impulse by creating a file named DATABASE.LIB
in the IMPULSE folder. Thirdly, the users need to connect to the new
databases listed and select the sections to use. Each of these broad steps
is discussed below. To maintain control and integrity of the data, it
would be advised that one person is given the responsibility to create and

manual (si) g.doc
maintain the network databases. The discussion below will be given
from the viewpoint of this person.
Figure 7.18 shows the relationship between Internal Databases and
External Databases as well as the database file listing that AFT Impulse
uses to find these databases. An External Database is any database file
that is created by the user and listed in the IMP_DBUSER.LIB residing
in the users local application data directory. The user can make these
available and unavailable, and connect and disconnect from these
databases using Database Manager. The database files themselves may
be located on any local or mapped drive. These database files are also
distinguished by being under the local user’s complete control. External
Shared Databases, on the other hand, are common to all AFT Impulse
users on the network, therefore the local user cannot remove them from
the list of available databases. However, the user can choose to connect
or disconnect from these databases using Database Manager. Finally, to
complete the discussion, there are two special databases, the Default
Internal Database and the Local User Database. These files are AFT
Impulse Internal Databases which cannot be disconnected or made
unavailable.
Creating database files

The first thing to do is create the files that contain the information to be
shared. These files will be referred to as database files and can contain
information about pipe materials, components (e.g. pumps, valves, etc.),
fluids, Output Control formats, Visual Report settings, etc. The
following are several ways to create database files:
• Use the Edit Database feature in the Database Manager, as described
previously. Data can be taken from the current model or other
databases.
• Save the control formats to separate files. This can be done on the
Output Control window, Visual Report control window, etc.
• Copy and paste sections from one file into another file using a text
editor.
8/10/2007 4:10:28 PM Page 319

Database Groups and Database

Impulse Internal Databases
AFT Default Internal Database
IMPULSE4.DAT
IMP_USER4.DAT
AFT Impulse Local User Database
External Shared Databases

(DATABASE.LIB) Vail Resort Materials
VailMaterials.dat
Development Output Control Settings
Development Output.dat
Final Report Output.dat

Final Report Output Format
Visual Report Format.dat
Visual Report Format

External Databases
(IMP_DBUSER.LIB)
My Custom Output Settings
My Report Output.dat
My VR Format.dat
My Custom Visual Report Format
Figure 7.18 An example of the relationship between External

Databases and External Shared Databases.
This discussion will focus on the first and second methods. Let’s assume
that the person responsible for the network database has already created
a new pipe material. Also, assume that the goal is to create a network-
wide database to contain this material, as well as two Output Control
settings and a Visual Report format.
The information for the material is copied from the Local User Database
(IMP_USER4.DAT) to a new file using the Database Editor in the
Database Manager (see Figure 7.15). For the source database choose
Local User Database and then select Pipe Materials. Then click Create
New Database and enter a file name, VailMaterials.dat, for this example.
Then enter a description of the database. This description is how AFT
Impulse users will refer to the database and should concisely describe
what is in the database (for this example we’ll use “Vail Resort
Materials”). Then click the Copy button to copy the pipe materials from

manual (si) g.doc
the Local User Database to this new database file. The pipe material
database file has now been created and is ready to be used.
Now we need to create two Output Control settings, one with a wide
variety of parameters for use during the project development phase and
another, more concise set, with the pertinent data for a final report.
Again, there are many ways of creating these database files. For the
development report format, let’s create a database file directly. Open the
Output Control window and select the pipe and junction output
parameters needed. Then click Save Control Format and enter the file
name (Development Output.dat in this case). You may want to open the
file in Notepad or other text editor and change the file heading (first line
of the file) from the default to something more descriptive,
“Development Output Control Settings”, for example.
Now, for the settings for a final report, let’s create the file using the
Database Editor in the Database Manager (see Figure 7.15). Open the
Output Control window again, select the parameters needed for the final
reports and click OK. Then open the Database Manager and go to the
Edit Database tab (see Figure 7.15). Click Data From Current Model and
then choose the Output Control section. Then click Create New
Database and enter a file name, Final Report Output.dat, for this
example. Then enter a description of the database, something descriptive
like “Final Report Output Format”. Then click the Copy button to copy
the Output Control format from the local model to the new database file.
To complete the database files, we need to create a Visual Report
format. To do this a model must first be run; any model can be used in
this case. After the model is run and a converged answer is obtained, go
to the Visual Report Control window and select the parameters to show
on the Visual Report. Then click Save Options, then enter the name of
the file, for example, Visual Report Format.ivs. Again, you may want to
open the file in Notepad or other text editor and change the file heading
(first line of the file) from the default to something more descriptive,
“Visual Report Format”, for example.
Sharing database files using DATABASE.LIB

Now that the database files for Pipe Materials, Output Controls Settings
and a Visual Report Format have be created, they need to be made
available to the users of AFT Impulse. Depending on the scope of use,
these database files can be placed into one of two groups. If the database
is to be used locally by one user, it is put in the External Database group.
8/10/2007 4:10:28 PM Page 321

On the other hand, if the database is to be common among many users
and shared over a network it is put into the External Shared Database
group.
Continuing the above example, the goal is to share these database files,
so ultimately they need to be External Shared Databases and listed in the
DATABASE.LIB file in the Impulse directory on the network. However,
since Impulse does not write directly to the network files for security
reasons, it is easiest to temporarily treat these files as local External
Database files which will be listed in the IMP_DBUSER.LIB file located
in the application data folder of the local machine.
It is recommended that External Shared Databases files be copied to the
network server and placed in a subfolder of the AFT Impulse folder (e.g.
AFT Products\AFT Impulse\Databases). The files may, however, be
located anywhere on the network that is accessible to the users. For our
example, we are going to put all the database files on the network in a
subfolder called Databases. It is easier to copy the files to the Databases
folder before they are made available in AFT Impulse. This will help
establish the correct paths to the files as the files are made available
using the Database Manager.
Once the database files are in their final locations, open the Database
Manager and click the Edit Available List button on the Connect to
Database tab (see Figure 7.13), and select Add Engineering Database.
Each of the new database files we have created needs to be added to the
Available Databases. Once added, the description of the database you
just added appears in the Available Databases list. When you have added
all the new database files, click Close and the External Database listing
file (IMP_DBUSER.LIB) in your application data folder is automatically
updated.
Now we need to discuss how the paths should be setup. When you add a
database, AFT Impulse will write the full path of the database to the
External Database listing (IMP_DBUSER.LIB). That means that if you
browsed to the database file using a mapped drive letter then the path
will include that drive letter. However, if other users try to access the
databases but have the AFT Impulse folder mapped to a different drive
letter they will not be able to locate the database files. Other options
include using either relative paths or the computer name in the path (e.g.
\\Network Server\Eng Apps\AFT Products\AFT
Impulse\Databases\filename.dat). You can automatically have AFT
Impulse use the computer name when you add the file by choosing

manual (si) g.doc
Network Neighborhood, then the computer and then browse to the
database file. If you would rather use relative paths, you can manually
edit them in the database listing file. All database files must be relative
to the AFT Impulse folder where the executable is located.
When the database files were added above, they became External
Databases and were listed in IMP_DBUSER.LIB. Now we need to move
this listing from the local machine to the AFT Impulse folder on the
network so that all users can access the database files. Once the listing
file is moved to the network, it must be renamed DATABASE.LIB.
Once this is done, all users can now access the database file we have
created.
Connecting to the external shared databases

Now that the databases have been placed on the network and made
available to all AFT Impulse users, each user must now connect to the
databases and make the sections active as needed. When a user opens the
Database Manager, the External Shared Databases are listed in the
Available Databases section on the Connect To Database tab (See Figure
7.14). Choose a database from this list and click Add to Connections and
the database will now appear in the Connected Databases. Then choose
the section(s) from the list on the right-hand side to become active. A
database can be connected without having any sections selected and
active. Finally, to have these connections in place every time you start
AFT Impulse, click Set As Default. To apply these settings to a model
made previously, open the model, go to Database Manager, click User
Default and, if you have set the default, the connections and setting will
now be applied to the model.
8/10/2007 4:10:28 PM Page 323


manual (si) g.doc
CHAPTER 8
Steady-State Hydraulic Theory and Loss

Models
AFT Impulse employs standard matrix solution techniques for solving

steady-state pipe network problems. The solution techniques use the
basic engineering concepts of mass and momentum balance to determine
the steady-state flow distribution and pressure losses in the network.
This chapter presents an overview of the steady-state solution technique
and the steady-state solution control parameters available to you. This
chapter also describes each of the junction types as it relates to the
solution scheme. Many junction types have standard, handbook loss
models. The nature of the loss models that AFT Impulse uses and
references for those loss models are also given.
The Steady-State Solver

AFT Impulse's Steady-State Solver is designed to achieve a balanced
flow and pressure solution in a pipe network. Before using the Steady-
State Solver in an application, you must have at least a basic
understanding of pipe fluid mechanics. A review of basic pipe fluid
mechanics can be obtained from standard undergraduate fluid mechanics
textbooks (Fox and McDonald 1985; White 1979).
AFT Impulse makes use of standard matrix solution techniques (Jeppson
1976). The method is known as the H-Equation method, where H, the
piezometric head, is solved for at each junction by forcing continuity of
flow through each connecting pipe. Simultaneously, the head loss across

manual (si) g.doc
each pipe is updated based on the flow balance information. The flow
rate and head are solved in an inner-outer loop algorithm, where the flow
is guessed, the head loss is calculated consistent with that guess, and the
flow is updated according to the new pressure drop information. The
Newton-Raphson method is employed to refine each successive solution,
resulting in a sparse square matrix that is solved during each solution
pass.
The concepts of pressure and hydraulic grade line (HGL, also called
piezometric head) are related but use different frameworks for
considering pipe system behavior. The HGL includes both the static and
elevational effects of pressure. The relationship between the two is given
by Equation 8.1:
P
HGL = +z (8.1)
ρg
where z is the elevation. Each concept offers certain advantages.
The solution technique makes use of the continuity and one-dimensional
momentum equations. AFT Impulse uses pressure and mass flow rate as
the unknown parameters because they are more fundamental in nature
than the head and volumetric flow rate used by Jeppson and thus extend
more naturally to mechanical and chemical systems.
In the following discussion, subscripts denote values at junctions. Thus,
Pi represents the pressure at junction i. Double subscripts denote values
along pipes connecting two junctions. Thus, m & ij represents the mass
flow rate in the pipe connecting junctions i and j.
Application of the law of mass conservation to each junction yields:
n
∑ m&
j =1
ij =0 (8.2)
where n is the number of pipes connected to junction i. Equation 8.2

states that the net mass flow rate into each junction must sum to zero.
The basic equation for pipe pressure drop due to friction can be
expressed with the Darcy-Weisbach equation:
L ⎛1 2 ⎞
ΔPf = f ⎜ ρV ⎟ (8.3)
D⎝2 ⎠

manual (si) g.doc
Chapter 8 Steady-State Hydraulic Theory and Loss Models 327
where ΔPf is the frictional pressure loss. The total pressure change
between junctions is given by the momentum equation in the form of the
Bernoulli equation:
1 1
P1 + ρV12 + ρgz1 = P2 + ρV2 2 + ρgz2 + ΔPf (8.4)
2 2
The static and stagnation pressure are related as follows:
1
Po = P + ρV 2 (8.5)
2
Substituting Equation 8.5 into 8.4
Po,1 + ρgz1 = Po,2 + ρgz2 + ΔPf
Solving for the frictional pressure drop for a constant area pipes yields:
Po,i − Po, j + ρg ( zi − z j ) = ΔPf (8.6)
where i and j denote upstream and downstream junction values.

The definition of mass flow rate is:
m& = ρAV (8.7)
Combining Equations 8.3 and 8.6 and substituting for velocity, V, using
Equation 8.7 gives the mass flow for each pipe:
1/ 2
⎛ Po,i − Po, j + ρg ( zi − z j ) ⎞
⎜ ⎟ = m& ij (8.8)
⎜ R′ij ⎟
⎝ ⎠
where Rij′ is the effective flow resistance in the pipe and the subscript ij
refers to the pipe connecting junctions i and j.
⎛ f ij Lij ⎞ 1
Rij′ = ⎜ + K ij ⎟ (8.9)
⎜ Dij ⎟ 2 ρA 2
⎝ ⎠ ij
Note that R ′ is not the same as R, which is the resistance used in the
Output Control window and is expressed by the following equation:
ΔH = RQ 2
8/10/2007 4:10:28 PM Page 327

Substituting Equation 8.8 into Equation 8.2 results in the equation to be
applied to each junction i:
n 1/ 2
⎛ Po,i − Po, j + ρg ( zi − z j ) ⎞
∑ ⎜
⎜ R′ij
⎟
⎟
=0 (8.10)
j =1 ⎝ ⎠
where n is the number of pipes connected to junction i.
To be completely general, Equation 8.10 should be written for junction i:
n 1/ 2
⎛ Po,i − Po, j + ρg ( zi − z j ) ⎞
∑ ⎜
⎜ R′ij
⎟
⎟
= m& i, Applied (8.11)
j =1 ⎝ ⎠
to allow for application of boundary condition flow rates to a boundary
junction node.
Equation 8.11 as applied to each junction in the network represents the
system of equations that need to be solved to determine the stagnation
pressure at each junction. To solve this system, the Newton-Raphson
method is employed. In the Newton-Raphson method, new values for
each unknown are calculated based on the previous value and a
correction that uses the first derivative of the function.
In this instance the function would be of the form:
n 1/ 2
⎛ Po,i − Po, j + ρg ( zi − z j ) ⎞
Fi = ∑ ⎜
⎜ R′ij
⎟
⎟
− m& i, Applied (8.12)
j =1 ⎝ ⎠
The method involves finding all the junction stagnation pressures, Po,i ,
that cause all of the Fi to go to zero, thus satisfying Equation 8.11 at all
junctions.
When applied to a system of equations, the Jacobian matrix contains all
the required derivative information to employ the Newton-Raphson
technique. The Jacobian, JF, is given by:

manual (si) g.doc
⎡ ∂F1 ∂F1 ∂F1 ⎤

⎢ ∂P ∂P L ∂P ⎥
⎢ o,1 o,2 o, n ⎥
⎢ ∂F2 ∂F2 ∂F ⎥
⎢ L 2 ⎥
J F = ⎢ ∂Po,1 ∂Po,2 ∂Po, n ⎥
⎢ ⎥
⎢ M ⎥
⎢ ∂Fn ∂Fn ∂F ⎥
⎢ L n ⎥
⎣⎢ ∂Po,1 ∂Po,2 ∂Po, n ⎦⎥
v
The column matrix Po contains the stagnation pressure at each junction,
v
and column matrix F contains the F values (Equation 8.12) at each
v
junction. The updated solutions for Po are obtained from the following
Newton-Raphson equation:
v v v
Po, new = Po,old − J F−1F (8.13)
Traditional implementations of the Newton-Raphson method require

initial guesses at the solution before the Steady-State Solver can begin.
This often results in the situation where you have to know the solution
ahead of time in order to solve for it. To avoid placing this burden on the
user, AFT Impulse uses a proprietary method for generating first
guesses. This method is usually sufficient to get the solution going in the
right direction, especially for systems that don't have highly non-linear
features.
However, there are times when AFT Impulse's first guess does not lead
to a converged solution. In these cases, AFT Impulse provides you with
the option of specifying your own initial guesses of flow and pressure.
Stagnation vs. static pressure boundaries

With two exceptions (to be discussed), all pressure-type boundary
conditions in AFT Impulse are stagnation. This works very well for
things such as storage tanks, cooling ponds, lakes, etc., where the
volume associated with a pressure is large and will not change
(significantly) with time. These boundaries have no velocity associated
with them, and using stagnation pressure is thus appropriate. These
boundary conditions are most clearly rendered in AFT Impulse by use of
a Reservoir junction.
8/10/2007 4:10:28 PM Page 329

2 3 4
Figure 8.1 Reservoir with one connecting pipe
Consider the reservoir in Figure 8.1. When the liquid in a reservoir flows
into connecting pipes, the static pressure immediately drops due to the
increase in velocity (see Figure 8.2).
Po,2
P2
ΔPloss
1
ρV2 2
Pressure
Po,4
2 1
ρV42 P4
2
Location 2 Location 3 Location 4

Distance (x)
Figure 8.2 Steady-state pressure profile along horizontal pipe

shown in Figure 8.1.

manual (si) g.doc
It is tempting to say that because there is no velocity in the reservoir,
then it does not matter whether the reservoir boundary pressure is
considered as a static or stagnation pressure because they are equal. This
is a misconception. The pressure boundary condition in an AFT Impulse
model is applied at location 2 in Figure 8.1, not location 1. At location 1
the static and stagnation pressures are the same, but they are not the
same at location 2. As shown in Figure 8.2, they differ by the amount of
the dynamic pressure (1/2)ρV22.
The appropriateness of either stagnation or static pressure as a boundary
condition depends on whether there is a velocity associated with the
pressure that is specified. Is the pressure in a vessel or large (lower
velocity) header, or is it in a pipe somewhere in the middle of a system?
If the first, stagnation is more appropriate; if the second, static.
Pressure boundary conditions with no velocity are more common across
all piping industries, and are usually appropriate. However, this is not
always the case.
When to use static pressure

Now let's consider the application of a true static pressure boundary
condition. A static boundary condition has a unique velocity associated
with it. From Equation 8.4 it is clear that application of the static
pressure is not sufficient to specify location 1. The velocity and
elevation is also required.
In AFT Impulse, elevation data is entered for junctions. The pipes adopt
the elevation of the junction to which they are connected.
But where does the velocity information for Equation 8.4 come from?
One might respond that it can be obtained from the flow rate. But this
raises another question: what if the flow rate isn’t known? Put another
way, what if the flow rate is what we are solving for?
When the user supplies a stagnation pressure boundary condition, AFT
Impulse can use Equation 8.6. Here no velocity information is required.
A stagnation pressure boundary can thus be supplied to multiple pipes
that might have different velocities. That is why the same reservoir
junction can connect with multiple pipes.
However, if the user specifies a static pressure as the boundary
condition, to apply Equation 8.4 a unique velocity must be supplied.
8/10/2007 4:10:28 PM Page 331

Thus it is only possible to connect a static pressure to a single pipe with
a unique velocity.
Important: A static pressure boundary has a unique velocity associated
with it, and can thus connect to only one pipe.
Where would one find a static pressure boundary? The best example is
when the boundary condition is inside a pipe. A pipe system model can
start and end anywhere it is convenient for the user. It may be convenient
to not start the model at the physical boundary (such as a tank) but at a
particular location in the pipe system. This could be, for example, at the
location of a pressure measurement. Or it could be at the boundary of the
pipe system for which your company is responsible, with another
company responsible for what is on the other side of that boundary.
If one needs to model such a situation, the Assigned Pressure junction
allows one to model either a static or stagnation pressure. The default
stagnation pressure allows connection of up to 25 pipes. If static
pressure is chosen, only one connecting pipe is allowed.
Static pressure at pressure control valve

Another example of static pressure is at pressure control valves. For
pressure control valves (i.e., PRV’s and PSV’s) the default control
pressure is static pressure. The reason is that the measured pressure that
provides feedback to the controller will typically be a static pressure
measurement. You have the option of modeling pressure control valves
as either static or stagnation pressure.
Open vs. closed systems

In order to model a closed system, only one pressure junction is used in
the model. Typically this would be either a Reservoir or Assigned
Pressure. As elaborated in Chapter 12, pressure type junctions are an
infinite source of fluid and do not balance flow. How then can one be
used to balance flow in a closed system?
To answer this question, it is worth considering how AFT Impulse views
a closed system model. AFT Impulse does not directly model closed
systems, and in fact does not even realize a closed system is being
modeled.

manual (si) g.doc
Consider the system shown in Figure 8.3. This is an open system. Fluid
is taken from J1 and delivered to J4 and J5. Because AFT Impulse’s
steady-state solution engine solves for a mass balance in the system, all
flow out of J1 must be delivered to J4 and J5. Because the flow is
steady-state, no fluid can be stored in the system; what goes in must go
out.
J4 J5
P3
J1 J2 J3
P1 P2
P4
Figure 8.3 Open system - Flow out of J1 equals the sum of J4

and J5
Now consider the systems in Figure 8.4. The first system appears to be
closed, while the second appears to be open. If the same boundary
condition (i.e., surface elevation and pressure) is used for J1, J11 and
J12 in the second system, to AFT Impulse it will appear as an identical
system to the first system. The reason is that AFT Impulse takes the first
system and applies the J1 reservoir pressure as a boundary condition to
pipes P4, P9 and P10. The second system uses three reservoirs to apply
boundary conditions to P4, P9 and P10. But if the reservoirs all have the
same elevation and pressure, the boundary conditions are the same as J1
in the first system. Thus the same boundary condition is used for P4, P9
and P10 in both models, and they appear identical to AFT Impulse.
8/10/2007 4:10:28 PM Page 333

J8 J9 J6
P8 P7
P9 P6
J1 J5 J3
P4 P3
P10 P11
J4 J2 J10
P1 P2
J11 J8 J9 J6
P9 P8 P7
J12
P6
P4
J1 J5 J3
P3
P10 P11
J4 J2 J10
P1 P2
Figure 8.4 The first system is closed, and the second open. In
both systems the flow into P10 is the sum of P4 and
P9. If the second system has the same conditions at
J1, J11 and J12, the two system will appear identical
to AFT Impulse.
But how is the flow balanced at J1 in the first system? Looking at the
first system, one sees that to obtain a system balance, whatever flows
into P10 must come back through P4 and P9. Because there is overall
system balance by the Steady-State Solver, it will give the appearance of
a balanced flow at the pressure junction J1. If there is only one boundary
(i.e., junction) where flow can enter or leave the pipe system, then no

manual (si) g.doc
flow will enter or leave because there isn’t anywhere for it to go. Thus
the net flow rate will be zero at J1 (i.e., it will be balanced). But
recognize that AFT Impulse is not applying a mass balance to J1
directly. It is merely the result of an overall system balance.
Verifying network solutions

As discussed previously, AFT Impulse uses a Newton-Raphson matrix
method to obtain a system level mass balance. When the Steady-State
Solver is finished, all branching sections are balanced for flow. Once
solved, it is a simple matter to go back and sum up all mass flows at each
branch to verify that these equations are satisfied. In fact, AFT Impulse
does just that.
After the Steady-State Solver converges, AFT Impulse loops over each
junction and adds up the inflow and outflow of mass and compares the
sum to the solution tolerances. If it appears that any junction is out of
tolerance, a warning is given in the output.
In the Solution Balance Summary table AFT Impulse provides additional
information on the balance at each junction. This table can be displayed
by selecting the option in the Output Control window General Output
tab. Figure 8.5 shows an example of the Solution Balance Summary
table.
Pressure drop calculation methods

AFT Impulse can model both Newtonian and non-Newtonian fluid
behavior. A non-Newtonian fluid is one where the viscosity depends on
the fluid dynamics. For example, as the velocity changes a non-
Newtonian fluid’s viscosity can change.
On the other hand, a Newtonian fluid’s viscosity does not depend on the
fluid dynamics, and is a function only of the thermodynamic state (i.e.,
temperature and pressure). Most industrial fluids follow the Newtonian
fluid behavior model, although there are important industrial
applications where the fluids are non-Newtonian.
8/10/2007 4:10:28 PM Page 335

Figure 8.5 Solution Balance Summary table shown in the Output

window offers a balance report of all junctions in the
model.
Pressure drop calculation methods for Newtonian fluids

AFT Impulse offers the Darcy-Weisbach loss model approach as the
default method for describing pipe frictional losses. Standard textbooks
give additional discussion on this method. AFT Impulse uses the Moody
approach to calculating friction factors; note that the Moody approach
differs from the Fanning friction factor by a factor of 4.

manual (si) g.doc
Except for laminar flow, there are no precise explicit friction factor
correlations available. Like all the more precise methods, AFT Impulse
uses an iterative formulation of friction factor. All methods of friction
factor calculation must address the rather poorly understood area of
laminar to turbulent transition.
Roughness-based methods
Roughness-based methods use a pipe roughness value to calculate the
pressure drop. This can be in the form of an absolute roughness (which
has units of length) or a relative roughness that ratios the pipe roughness
to its diameter.
Laminar flow
For laminar flow AFT Impulse uses the standard laminar equation:
64
f = Re < 2300 (default) (8.14)
Re
Note that the laminar friction factor does not depend on the user
roughness.
The System Properties window allows you to change the default laminar
transition Reynolds Number (see Chapter 5).
Turbulent flow
The Colebrook-White iterative friction factor equation is used to obtain
friction factors in the turbulent flow regime.
−2
⎡ ⎛ε 9.35 ⎞⎤
f = ⎢1.14 − 2 log⎜ + ⎟⎥ Re > 4000 (default)
⎜ D Re f ⎟
⎣⎢ ⎝ ⎠⎦⎥
The System Properties window allows you to change the default
turbulent transition Reynolds Number (see Chapter 5).
Transition flow
For Reynolds Numbers in the transition flow regime, a linear
interpolation is used between the laminar and turbulent values.
8/10/2007 4:10:28 PM Page 337

Hydraulically smooth
The friction factor in the turbulent range is calculated using Colebrook-
White with roughness = 0 while in the laminar range it continues to be
calculated using Equation 8.14 above.
Hazen-Williams method
AFT Impulse also offers the Hazen-Williams method of specifying
irrecoverable loss information. The Steady-State Solver converts the
Hazen-Williams factor to a Darcy-Weisbach friction factor (Walski
1984, 37). This allows a consistent solution approach to be used for all
pipe system models, while retaining the flexibility of the two approaches
to account for losses.
17.25
CHW =
f (VD) 0.081
0.54
where V is in ft/s and D is in ft.
Resistance
The pipe resistance relates head loss to volumetric flow rate. In equation
form, the head loss is:
ΔH = RQ 2
Using standard relationships, the friction pressure drop is related to
volumetric and mass flow rate as follows:
Rgm& 2
ΔPf = RρgQ 2 =
ρ
MIT Equation for crude oil

The MIT Equation is appropriate for crude oil and is given by the
following equation (Pipe Line Rules of Thumb Handbook):
dP 139.212 fQ 2 ( ρ ρ water )
=
L D5

manual (si) g.doc
1
f = 2
⎡ ⎛ Re ⎞⎤
4 ⎢0.86859 ln⎜ ⎟⎥
⎣ ⎝ 1964
. ln( Re) − 38215
. ⎠⎦
10,059Qν
Re =
(
d ν 2 − 818.56 )
where:
dP = pressure drop (psid)
L = length (miles)
f = friction factor, MIT
Q = volumetric flow rate (barrels/hour)
ρ = density (lbm/ft3)
ρwater = density of water (62.3 lbm/ft3)
D = diameter (inches)
Re = Reynolds number, MIT
ν = kinematic viscosity (Seconds Saybolt Universal)
Miller Turbulent method

The Miller Turbulent method is appropriate for light hydrocarbons and
is given by the following equation (Pipe Line Rules of Thumb
Handbook):
d 2.5 ( dP L) ⎡ ⎛ d 3(ρ ρ ⎤
water )( dP L) ⎟
⎞
Q= ⎢log⎜ + 4.35⎥
⎜ ⎟
5.9113( ρ ρ water ) ⎢⎣ ⎝
0.5
μ2 ⎠ ⎥⎦
where:
dP = pressure drop (psid)
L = length (miles)
Q = volumetric flow rate (barrels/hour)
ρ = density (lbm/ft3)
ρwater = density of water (62.3 lbm/ft3)
8/10/2007 4:10:28 PM Page 339

d = diameter (inches)
μ = dynamic viscosity (centipoise)
Frictionless pipes
Frictionless pipes are convenient for connecting junctions that may not
have a physical pipe between them. There are, however, some
limitations where a frictionless pipe may be located in a model. An
example is a frictionless pipe that connects two Assigned Pressure
junctions. Such a pipe would have an infinite flow rate. In such cases
AFT Impulse identifies inappropriate placement of frictionless pipes and
informs the user when a model is run.
Pressure drop calculation methods for non-Newtonian fluids
Duffy method for pulp and paper stock

The Duffy method includes 20 different stock selections to obtain the
constants necessary to apply the method (TAPPI 1988):
Vmax = K ′C σ
Vwater = 122
. C 1.4
If V < Vmax, then
dH
= FKV α C β D γ
L
where:
dH /L = head loss per length (m water / 100 m pipe)
C = pulp consistency (% dryness)
Vmax = max velocity (for Duffy Equation) in above equation
(m/s)
Vwater = water velocity in above equation (limit at which pulp
effects are negligible), (m/s)
D = inner pipe diameter (mm)

manual (si) g.doc
F = correction factor for temperature or pipe roughness
(dimensionless)
K’, K, σ, α, β , γ = constants depending on pulp type
(dimensionless)
If V > Vmax but less than Vwater, then use above head loss equation with V
= Vmax. If V > Vwater then just use normal friction calculation for water.
The constants used for the specific type of paper stock are displayed in
the General Results section of the output. In addition, you can optionally
display in the output the Vmax and Vwater terms, the dryness coefficient
entered, and the Safety Factor either entered or calculated.
Brecht & Heller method for pulp and paper stock

The Brecht & Heller method includes eight different stock selections for
the constants (Ingersoll-Dresser Pumps, 1995). The Brecht & Heller
method is recognized to be more conservative than Duffy. That is, the
pressure drop predictions are higher.
D 0.205Vρ
Re′ =
C1.157
3.97
f =
Re′1.636
fV 2 LK
dH =
D
where:
Re’ = pseudo Reynolds Number (dimensionless)
D = inner diameter (feet)
V = average velocity (ft/s)
ρ = stock density (lbm/ft3)
C = pulp consistency (% dryness)
f = pseudo friction factor (dimensionless)
K = constant depending on pulp type (dimensionless)
L = length (feet)
8/10/2007 4:10:28 PM Page 341

The constants used for the specific type of paper stock are displayed in
the General Results section of the output. In addition, you can optionally
display in the output the special Reynolds Number and Friction Factor
terms.
Power Law non-Newtonian

This correlation is taken from Darby (2001), pp. 166-167. Note that
Darby uses the Fanning friction factor, which is different from the
Moody friction factor used in AFT Impulse by a factor of 4. The
equations below have been modified from Darby accordingly.
Defining:
K = constant with units of dynamic viscosity
n = dimensionless constant
Then:
α
f = (1 − α ) f L +
( fT −8 + fTr −8 )1 8
where:
64
fL =
Re pl
0.2728n −1 2
fT =
Re pl1 (1.87 + 2.39n )
f Tr = 7.16 × 10 −4 exp(− 5.24n ) Re pl

(0.414 + 0.757 n )
1
α=
1 + 4−Δ
Δ = Re pl − Re plc
8 D nV (2 − n )ρ
Re pl =
K (2(3n + 1) n )n

manual (si) g.doc
Re plc = 2100 + 875(1 − n )
This friction factor can then be used in Equation 8.3.
Bingham Plastic non-Newtonian

This correlation is taken from Darby (2001), pp. 168-169. Note that
Darby uses the Fanning friction factor, which is different from the
Moody friction factor used in AFT Impulse by a factor of 4. The
equations below have been modified from Darby accordingly.
Defining:
Rc = coefficient of rigidity
Sy = yield stress
Then:
(
f = f L m + fT m )1 m
40000
m = 1 .7 +
Re
where:
64 ⎛ He He 4 ⎞
fL = ⎜ 1+ − ⎟⎟
Re ⎜⎝ 6 Re 3 f 3 Re 7 ⎠
DVρ
Re =
Rc
He is the Hedstrom number:
D 2 ρS y
He =
Rc 2
10 a
fT =
Re 0.193
( (
a = −5.88 1 + 0.146 exp − 2.9 × 10−5 He ))
8/10/2007 4:10:28 PM Page 343

Non-Newtonian flow through non-pipes
At your option in the System Properties window, you may apply a
correction to hydraulic losses at elements other than pipes. This includes
junctions and pipe fittings & losses.
The correction is as follows:
f
K non − newtonian = K newtonian non − newtonian
f newtonian
where the friction factor is for the upstream pipe.
Design factors
In each pipe you can specify a Design Factor for the pipe friction. This is
a multiplier that is applied to the friction factor calculated with the
preceding methods.
Steady Solution Control parameters

AFT Impulse provides you control over a number of parameters that
influence numerical convergence:
• Pressure solution tolerance
• Mass flow rate solution tolerance
• Flow relaxation
• Pressure relaxation
• Maximum iterations
• Matrix method
The default values for the solution control parameters perform well in
the majority of cases. Modify the default parameters only when
necessary or when you are comfortable with numerical convergence
issues.

manual (si) g.doc
Solution tolerance specification

The nature of tolerance specification for pressure and mass flow rate is
the same. A discussion of tolerance specification will therefore cover
both areas.
AFT Impulse uses an iterative approach to obtain pipe flow solutions; it
can obtain a converged solution to as many significant digits as you want
(within the computer's numerical capability). However, for practical
engineering problems, each digit after about the fourth significant digit
generally has low confidence associated with it because of the
uncertainties in the input data. The more digits, the lower the
confidence. Further, obtaining those additional digits in your solution
requires additional computer resources, which typically means longer
run times. At a certain point, obtaining additional significant digits
becomes counterproductive.
Therefore, the Steady-State Solver needs to know when to stop iterating
and assume that a sufficient solution has been obtained. Because you
know the nature of your pipe system much better than the Steady-State
Solver, you are in a better position to specify the appropriate level of
convergence. You do this by setting the tolerances in the Steady Solution
Control window.
AFT Impulse bases the solution tolerance specifications on absolute
change or relative change of the unknown parameter. As the Steady-
State Solver iterates, AFT Impulse continuously updates the solution.
Each time a solution update is obtained, AFT Impulse compares the new
solution to the old solution for each individual pipe and junction.
Presumably, when the correct solution is being approached, the
corrections made during each iteration become smaller and smaller.
When all corrections for an iteration are smaller than the tolerance you
specify, AFT Impulse considers the solution converged.
The absolute change method is generally more reliable for iterative
solvers like AFT Impulse because it is less sensitive to the magnitude of
the solution, whether close to zero or very large. However, in principle,
specifying absolute tolerance requires some knowledge of the final
solution. If you specify mass flow convergence as sufficient when the
flow no longer changes by 1 lbm/sec each iteration, then your solution
will be compromised if the magnitude is also near 1 lbm/sec. In this case
it would be better to set the tolerance 3 or 4 orders of magnitude below
the solution.
8/10/2007 4:10:28 PM Page 345

In programming terms, the logic for absolute tolerance looks as follows:
If (For All Junctions) Pj , new − Pj ,old < TOLabs Then

Convergence = True
Else
Convergence = False
End If
Relative tolerances, on the other hand, check only the relative change of
the number. That is, each successive change is divided by the number
itself. This removes the requirement to be knowledgeable about the final
solution, but can give trouble for problems where one of the solutions is
too close to zero (because the change is divided by a number close to
zero, making a very large number).
In programming terms, the logic for relative tolerance looks as follows:
Pj , new − Pj , old
If (For All Junctions) < TOLrel Then
Pj , new
Convergence = True
Else
Convergence = False
End If
Table 8.1 shows a tolerance example for pressures. Notice how relative
change does not have units, but absolute change has units (of psia).
The default in AFT Impulse is relative tolerance because experience
indicates it is the most robust.
AFT Impulse 4.0 offers options for combining absolute and relative
tolerance. You can tell AFT Impulse to assess convergence based on
whether either an absolute or relative criteria is satisfied, or whether
both absolute and relative criteria are satisfied.

manual (si) g.doc
Table 8.1 Example of relative and absolute change for pressure at
a junction. AFT Impulse looks at the maximum change
for all junctions and checks that it is less than the
tolerance.
Pressure Relative Abs olute
Iteration # (psia) Change Change (psia)
1 100.0 — —
2 60.0 0.6667 40.00
3 72.0 0.1667 12.00
4 66.0 0.0909 6.00
5 67.0 0.0149 1.00
6 66.9 0.0015 0.10
Figure 8.6 Tolerance input on Steady Solution Control window.
Figure 8.6 shows the tolerance input area in the Steady Solution Control
window. Table 8.2 shows an example of how the four different criteria
would be applied to flow rate iterations.
8/10/2007 4:10:28 PM Page 347

Table 8.2 Example of four tolerance methods for a mass flow rate
solution. Absolute tolerance = 0.0005 lbm/s, Relative
tolerance = 0.0001.
Iteration Mass Flow Relative Absolute Meets Abs. Meets Rel. Meets Abs. Meets Abs.
# Rate (lbm/s) Change Change Tolerance ? Tolerance ? OR Rel. AND Rel.
(lbm/sec) Tolerance ? Tolerance ?
1 20.0000 — — — — — —
2 12.0000 0.666667 8.000000 NO NO NO NO
3 14.0000 0.142857 2.000000 NO NO NO NO
4 13.5000 0.037037 0.500000 NO NO NO NO
5 13.6000 0.007353 0.100000 NO NO NO NO
6 13.5900 0.000736 0.010000 NO NO NO NO
7 13.5913 0.000096 0.001300 NO YES YES NO
8 13.5907 0.000044 0.000600 NO YES YES NO
9 13.5909 0.000015 0.000200 YES YES YES YES
How tolerances relate to solution accuracy

The tolerances you set do not have very much to do with how close you
are to the true solution. Do not make the mistake of concluding that the
tolerance you specify means AFT Impulse’s solution is that close to the
true solution. As in Table 8.2, the tolerance is how close the current
iteration values are to the previous iteration values. That is all it means.
If there is concern that the AFT Impulse results may not be close enough
to the true answer, you have the option of lowering the tolerances,
rerunning the models, then comparing the results to the previous run.
After performing such a comparison, if you are satisfied that the results
are not changing in a significant way then you can feel confident in the
model predictions.
Relaxation
Another feature of iterative solvers is called relaxation. In short,
relaxation slows the Steady-State Solver's progress toward a solution and
in the process reduces the ability of poorly behaved (non-linear)
components to destabilize the solution. AFT Impulse implements a
proprietary adaptive relaxation scheme that reduces the relaxation
dynamically in response to highly non-linear features of the problem.
However, you are free to override AFT Impulse's adaptive approach.
When applied to the flow solution, the relaxation is used as follows:

manual (si) g.doc
m& new = r (m& new − m& old ) + m& old

where r is the relaxation. The solution after each iteration is modified by
weighting it with the previous solution and applying the relaxation
factor. The relaxation must always be greater than zero and less than or
equal to 1. The march toward a solution is thus relaxed, because the
impact of the new solution is minimized by the magnitude of the
relaxation parameter. In essence, by specifying small relaxation values,
you reduce the speed of the march while causing the march to go more
smoothly. This is sometimes the only practical approach for highly
non-linear problems.
A relaxation value of 1 is the same as specifying no relaxation at all.
Each new solution is used directly for the next iteration step. A
relaxation of zero is not accepted because no updates would occur to the
unknown parameters and no solution would be obtained. See Table 8.3
for an example.
Table 8.3 Example of different relaxation values. Old Flow Rate is

that at the previous iteration. Ideal New Flow Rate is that
determined by the matrix solution. Actual New Flow Rate
is what is actually accepted for the new iteration.
Relaxation Old Flow Rate Ideal New Flow Actual New Flow
Value (lbm/s) Rate (lbm/s) Rate (lbm/s)
1 10 20 20
0.5 10 20 15
0.2 10 20 12
0.1 10 20 11
0.05 10 20 10.5
0 10 20 10
There are two relaxation parameters in AFT Impulse 4.0 (See Figure
8.7): Flow rate relaxation and pressure relaxation.
Flow Rate Relaxation

For poorly-behaved systems, a flow rate relaxation of 0.1 is advised.
Experience shows that using flow rate relaxation factors between 0.5 and
1 has limited usefulness. If a flow rate relaxation of 1 does not work, you
should generally try flow rate relaxation parameters between 0.1 and 0.5.
If these still do not work, there are probably other areas where your
8/10/2007 4:10:28 PM Page 349

model requires modification. In extreme cases, a flow rate relaxation
factor between zero and 0.1 may lead to convergence.
Pressure Relaxation
Because of the nature of the matrix solver, experience indicates that if
pressure relaxation is used, it should always be set to 0.5 or 1.0.
Anything else will usually destabilize the solution. A pressure relaxation
value of 0.5 usually works best.
Avoiding false convergence

To avoid a false convergence it is important to keep the tolerance at least
one order of magnitude smaller (two orders or more is recommended)
than the relaxation. A false convergence can occur when the change
from the old value to the new value is small enough to be within the
tolerance because it is restricted by the relaxation.
Figure 8.7 Other tab on Steady Solution Control window.

manual (si) g.doc
Maximum Iterations
The Maximum Iterations parameter in the Steady Solution Control
window determines when the Steady-State Solver considers the model
nonconvergent; it does not directly affect the solution. There may be
many reasons why a model does not converge. You want to be sure that
the Steady-State Solver does not continue indefinitely searching for a
solution that cannot be obtained because of an input error or ill-behaved
system.
Most properly defined models will converge in 50,000 iterations. For
larger models, more than 50,000 iterations may occasionally be required,
especially when relaxation is being used. AFT Impulse's default of
50,000 iterations is sufficient for most models.
Matrix Method
AFT Impulse’s network solution method requires solution of matrices
thousands of times. The default method is Gaussian Elimination, which
is very robust and usually provides the fastest convergence.
For larger systems, Gaussian Elimination with Pivoting and the LU
Decomposition methods can be better and are therefore provided.
Solution Progress window and iteration history

The Solution Progress window (Figure 8.8) shows you how the current
state of the solution compares to the tolerances you specify. It also gives
information on your Steady Solution Control settings and gives you
helpful feedback on what stage the Steady-State Solver is in.
When the Solution Progress window is displayed during execution, you
have the option of pausing or canceling the Steady-State Solver in the
middle of a run. If you pause, you can change solution control
parameters during a run and resume execution with the new parameters.
You also can check the iteration history.
8/10/2007 4:10:28 PM Page 351

Figure 8.8 The Solution Progress window shows you the status
of the Steady-State and Transient Solvers. This
includes showing how far out of tolerance the Steady-
State Solver is, and when the steady solution has
converged.
The Iteration History window shows you a complete history of all

iterations, including the pipe or junction that was most out of tolerance.
If a model is not converging, knowing the pipe or junction most out of
tolerance can be helpful in troubleshooting the model.
Warnings in the solution

A number of warnings may be given in the output when you run AFT
Impulse's Steady-State Solver. These warnings are intended to alert you
about potential problems with the model or solution. Some warnings
may prove to be harmless upon closer inspection, but you should never
ignore them.

manual (si) g.doc
There are different warning levels depending on the importance of the
warning.
1. WARNING – A warning that actually says WARNING offers
important information about the converged solution or the model
itself. The information should always be reviewed.
2. CAUTION – A cautionary warning is of less importance than a full
warning. The included information may or may not be of interest to
the user.
3. Informational – The informational warning is for information only
and is not serious.
A complete list of the warnings you may receive, with a brief
explanation given of each, is given in the Help System. Whenever any
warnings occur that are WARNING level, the General Results section of
the Output Window will be displayed with the text in red.
Modeling irrecoverable losses

Chapter 6 discusses in detail the features offered by AFT Impulse to
model irrecoverable losses. Chapter 3 discusses the basic conventions
used by AFT Impulse for defining losses.
As discussed in Chapter 3, AFT Impulse allows you to specify the base
area for modeling losses at junctions. By default the upstream pipe area
is used.
Junction pressure losses are modeled as point losses (and point additions
for pumps). They are not smoothed out over a length. Pipe losses, both
frictional losses and additional minor losses, are smoothed out over the
pipe length and are more accurately referred to as distributed losses.
It is important to note that AFT Impulse does not include the frictional
part of the loss when assigning a loss factor. Frictional lengths in fittings
are commonly lumped into the adjacent pipes. An example is a large
radius elbow. The loss due to the bend is the loss factor K; the friction in
the elbow is an independent effect. The magnitude of this distinction is
frequently negligible.
8/10/2007 4:10:28 PM Page 353

Loss models and reference material

Table 8.4 lists the sources for the loss models used in AFT Impulse. The
losses implemented directly in the code were chosen on the basis of ease
of use. Many loss factor types are functions of the flow, and thus too
general to be easily incorporated.
Table 8.4 Loss model references

Junction Type References
Bend Crane 1988, A-29
Crane 1988, A-26 and
Area Change
Idelchik 1994, 208, 216
Orifice Idelchik 1994, 218, 220
Idelchik 1994, chapter 7
Tee/Wye
and Miller
Ingersoll-Rand 1970 Crane
Valve
1988, A-29
Crane (1988) offers good general purpose correlations for modeling

irrecoverable losses in pipe systems. A well prepared pipe flow analyst
would be advised to have copies of all of these references.
Another lesser known source of loss factor information is Idelchik's
Handbook of Hydraulic Resistance (1994). This reference is
indispensable for the engineer who must make detailed hydraulic
assessments of pipe systems in which so-called minor losses play a
significant roll. The reference is voluminous in scope and full of tables,
charts and equations for calculating loss factors for almost any pipe
arrangement.
Miller's Internal Flow Systems (1990) is another reference offering good
general purpose hydraulic data.
AFT Impulse models component losses according to the following
equation:
1
ΔPloss = K ρV 2 (8.18)
2
where K is commonly referred to as the loss factor. Other loss models
and their relationship to K are described at the end of this chapter.

manual (si) g.doc
Design factors
Each junction that allows modeling of pressure losses allows input of a
Design Factor. The design factor is multiplied by the K factor
determined by methods discussed in this chapter.
Many junction types allow modeling of K factors as well as other
pressure loss methods. The Design Factor is also applied to these other
methods as a multiplier on the pressure loss.
Area change
Two standard area change loss geometries are available: the conical
transition and the abrupt transition.
The conical expansion correlation (Crane 1988, A-26) is:
2
⎛ μ ⎞⎛ Aup ⎞
K up = 2.6⎜ sin ⎟⎜⎜1 − ⎟
⎟ (μ < 45 degrees)
⎝ 2 ⎠⎝ Adown ⎠
2
⎛ Aup ⎞
K up = ⎜⎜1 − ⎟
⎟ (μ > 45 degrees)
⎝ Adown ⎠
The conical contraction correlation (Crane 1988, A-26) is:
⎛ μ ⎞⎛ A ⎞
0.8⎜ sin ⎟⎜1 − down ⎟
⎝ 2 ⎠⎜⎝ Aup ⎟
⎠
K up = (μ < 45 degrees)
2
⎛ Adown ⎞
⎜ ⎟
⎜ Aup ⎟
⎝ ⎠
⎛ A ⎞ μ
0.5⎜1 − down ⎟ sin
⎜ Aup ⎟⎠ 2
⎝
K up = (μ > 45 degrees)
2
⎛ Adown ⎞
⎜ ⎟
⎜ Aup ⎟
⎝ ⎠
The abrupt expansion correlation (Idelchik 1994, 208) is:
8/10/2007 4:10:28 PM Page 355

2
⎛ Aup ⎞
K up = ⎜⎜1 − ⎟
⎟
⎝ Adown ⎠
The abrupt contraction correlation (Idelchik 1994, 216) is:
0.75
⎛ A ⎞
0.5⎜1 − down ⎟
⎜ Aup ⎟
⎝ ⎠
K up =
2
⎛ Adown ⎞
⎜ ⎟
⎜ Aup ⎟
⎝ ⎠
Bend
90 degree bends
The three Bend loss correlations are all for turbulent Reynolds Numbers
(Crane 1988, A-29).
The K values for a smooth, flanged bend are provided in Table 8.5.
Table 8.5 K values for 90 degree smooth, flanged bends
r/d K r/d K
1 20 f T 8 24 f T
1.5 14 f T 10 30 f T
2 12 f T 12 34 f T
3 12 f T 14 38 f T
4 14 f T 16 42 f T
6 17 f T 20 50 f T

manual (si) g.doc
Table 8.6 Pipe friction factors used for Crane formulas
Nominal Friction Nominal Friction Nominal Friction

Size Factor f T Size Factor fT Size Factor fT
½” 0.027 2” 0.019 8”-10” 0.014
¾” 0.025 2 ½”, 3” 0.018 12”-16” 0.013
1” 0.023 4” 0.017 18”-24” 0.012
1 ¼” 0.022 5” 0.016
1 ½” 0.021 6” 0.015
The standard threaded elbow is given by:

K = 30 f T
where fT is the turbulent friction factor given in Table 8.6.
The Mitre bend is given by:
K = 60 f T
Non-90 degree bends

The two non-90 degree bend loss correlations are for turbulent Reynolds
Numbers (Crane 1988, A-29).
A smooth, flanged bend is given by:
⎛ r ⎞
K = (n − 1)⎜ 0.25πf T + 0 .5 K ⎟ + K
⎝ D ⎠
where n is the number of 90 degree bends, K is the loss factor for one 90
degree bend (Table 8.5), and fT is given by Table 8.6.
The Mitre bend is given by Table 8.7.
8/10/2007 4:10:28 PM Page 357

Table 8.7 K values for Mitre bends
μ (deg.) K
0 2 fT
15 4 fT
30 8 fT
μ
45 15 f T
60 25 f T
75 40 f T
90 60 f T
Valve
Standard valve loss models are used from Idelchik, Miller and Crane.
Orifice
The orifice loss factors are all for turbulent Reynolds Numbers. The
sharp-edged orifice shown in Figure 8.9, is given by the following
(Idelchik 1994, 218):
2⎡ 2
⎛ Aup ⎞ ⎛ ⎞
0.375 ⎤
⎢0.707⎜1 − Aorifice ⎟ ⎛ Aorifice ⎞⎥
K up = ⎜ ⎟ + ⎜⎜1 − ⎟
⎜ Aorifice ⎟
⎝ ⎠
⎢ ⎜
⎝ Aup ⎟⎠ ⎝ Adown ⎟⎠⎥
⎢⎣ ⎥⎦
Figure 8.9 Sharp-edged orifice
The round-edged orifice, shown in Figure 8.10, is given by the

following (Idelchik 1994, 220):

manual (si) g.doc
2⎡ 0.75 2
⎛ Aup ⎞ ⎛ ⎞ ⎛ Aorifice ⎞
K up = ⎜ ⎟ ⎢ K ′⎜1 − Aorifice ⎟ + ⎜⎜1 − ⎟
⎜ Aorifice ⎟
⎝ ⎠
⎢ ⎜ Aup ⎟⎠ ⎝ Adown ⎟⎠
⎣⎢ ⎝
0.375 ⎤
⎛ Aorifice ⎞ ⎛ A ⎞ ⎥
+2⎜⎜1 − ⎟ K ′⎜1 − orifice ⎟
Adown ⎟⎠ ⎜ Aup ⎟⎠ ⎥
⎝ ⎝ ⎥⎦
where:
K ′ = 0.03 + 0.47 • 10
(− 7.7r / Dorifice )
For other orifice configurations, see Chapter 4 of Idelchik's Handbook of

Hydraulic Resistance (1994).
Figure 8.10 Round-edged orifice
Tee/Wye
The loss factors calculated for tee and wye junctions involve
complicated correlations that depend on the flow split, the ratio of flow
areas, and the angle of the connecting pipes. The models used by AFT
Impulse are taken mostly from chapter 7 of Idelchik's Handbook of
Hydraulic Resistance (1994). A few of the more important equations
that AFT Impulse uses will be presented here, and references will be
given for the others.
8/10/2007 4:10:28 PM Page 359

All tees and wyes have either diverging or converging flow streams.
Two sets of equations are required for each configuration.
In the following discussion, the parameter Kc,s represents the loss factor
based on the combined flow stream flow area, and velocity for the side
branch. Similarly, Kc,st is for the straight-through pipe (the run).
Figure 8.11 depicts the nomenclature. K factor subscripts refer to base
area reference and location in the Tee. For example, Kc,s represents the K
factor for the side branch (s) referenced to the combined (c) pipe flow
area.
c = combined flow st = straight-through flow s = side branch
Ac Ast Ast Ac
As As
Figure 8.11 Idelchik’s nomenclature for diverging and converging

flow through Tee/Wye junction
In the AFT Impulse Tee/Wye window, a simpler and more general

nomenclature is used. AFT Impulse refers to the a, b and c pipes. This
allows AFT Impulse to cover all the possibilities in Table 8.10 without
specifying exactly which pipe is the combined flow pipe. In some cases,
you may not know all the flow directions in a system, and might not be
able to specify which is the combined flow pipe.
AFT Impulse’s input method allows you to specify which pipe is the
branch pipe. Idelchik calls this the “s” pipe for “side branch”. AFT
Impulse calls it the c pipe. If the tee angle is 90 degrees, it is not
important which pipe you call the a pipe and which the b pipe. The
distinction between a and b takes on more importance for non-90 degree
pipes. In this case, AFT Impulse allows you to specify the angle of the
branch pipe.
Figure 8.12 shows the Tee/Wye Specification window. The α angle
itself is specified on the Tee/Wye Model tab, where there are other
options for how to model the tee.

manual (si) g.doc
Figure 8.12 Tee/Wye Specifications window allows you to specify

the a, b and c pipes. The c pipe is always the branch
pipe. This general method means you do not need to
know the flow direction ahead of time.
Diverging Case
Side branch (Idelchik 1994, 418):
ΔPs ⎡ ⎛ V ⎞2 ⎛ Vs ⎞ ⎤
K c, s = ⎢ s
= λ 1 + ⎜⎜ ⎟⎟ − 2⎜⎜ ⎟⎟ cosα ⎥
2 ⎢ ⎝ Vc ⎠ ⎥
ρVc / 2 ⎣ ⎝ Vc ⎠ ⎦
where λ is found from Table 8.8.
8/10/2007 4:10:28 PM Page 361

Table 8.8 Diverging case λ values
As / Ac Qs / Qc λ
≤ 0.35 ≤ 0.4 1.1 - 0.7 Qs / Qc
≤ 0.35 > 0.4 0.85
> 0.35 ≤ 0.6 1.0 - 0.6 Qs / Qc
> 0.35 > 0.6 0.6
Straight run, constant area (Idelchik 1994, 419):

2
ΔPst ⎛Q ⎞
K c, st = = τ st ⎜⎜ s ⎟⎟
ρVc2 / 2 ⎝ Qc ⎠
where τst is given in Table 8.9:
Table 8.9 Diverging case τst values
As / Ac Qs / Qc τ st
≤ 0.4 0 - 1.0 0.4
> 0.4 ≤ 0.5 2 (2 Qs / Qc - 1)
> 0.4 > 0.5 0.3 (2 Qs / Qc - 1)
Converging Case
Side branch (Idelchik 1994, 417):
ΔPs
K c, s =
ρVc2 / 2
⎛ ⎛ Q A ⎞2 A ⎛ Qs ⎞
2
Ac ⎛ Qs ⎞
2 ⎞
= A′⎜⎜1 + ⎜⎜ s c ⎟⎟ − 2 c ⎜⎜1 − ⎟⎟ − 2 ⎜⎜ ⎟⎟ cosα ⎟
⎟⎟
⎜ ⎝ Qc As ⎠ Ast ⎝ Qc ⎠ As ⎝ Qc ⎠
⎝ ⎠
Straight run, constant area (Idelchik 1994, 417):

2 2
ΔPst ⎛ Q ⎞ A ⎛Q ⎞
K c, st = = 1 − ⎜⎜1 − s ⎟⎟ − 2 c ⎜⎜ s ⎟⎟ cos α
ρVc2 / 2 ⎝ Qc ⎠ As ⎝ Qc ⎠

manual (si) g.doc
References for all tee and wye junction correlations are given in Table
8.10.
Table 8.10 References for Tee/Wye junction correlations
Geometry Type Direction Reference

Constant area, sharp Diverging tee Branch Idelchik 1994, 451
Diverging tee Run Idelchik 1994, 453
Converging tee Branch Idelchik 1994, 429-432
Converging tee Run Idelchik 1994, 429-432
Splitting tee From branch Miller 1990, 323*
Combining tee To branch Miller 1990, 316*
Constant area, r/D = 0.1 Diverging tee Branch Idelchik 1994, 459-461
Diverging tee Run Idelchik 1994, 453
Area sum, sharp Diverging tee Branch Idelchik 1994, 452
Diverging tee Run Idelchik 1994, 454-455
Wye Diverging wye Branch Idelchik 1994, 451
Converging wye Run Idelchik 1994, 473
*Estimates have been made on available data
K and CV (valve coefficient)

The definition for CV is as follows (Lyons 1982, 133):
Q
CV =
ΔPvalve ( 62.4 / ρ )
where Q is in gpm, ΔP is in psid, and ρ is in lbm/ft3.
8/10/2007 4:10:28 PM Page 363

It can be shown that K and Cv are related by the following approximate
expression (Lyons 1982):
A2
Kvalve = 1460 2
Cv
where A is in in2 and CV2 is in its normal units (always gpm2/psi).
K for fire sprinklers

The definition of K value for fire sprinklers is:
Q
K sprinkler =
ΔP
where Q and ΔP can be in the units of your choice. The preference in the
USA is gpm for flowrate and psid for pressure. In Europe it is frequently
liter/min for flow rate and bars for pressure. Note that Ksprinkler is not
dimensionless, but has units associated with it. Also, note that sprinkler
vendor values for Ksprinkler implicitly assume that water is the fluid, so
there is a built-in density of 62.3 lbm/ft3 or 1000 kg/m3. If this sprinkler
is used for something other than water, the Ksprinkler value will need to be
adjusted.
Also note that if the spray is into the atmosphere, as is typical, the ΔP is
the same thing as the gage pressure. That is why the previous equation is
stated by some vendors as just P in the denominator, rather than ΔP.
When stated as just P, the pressure is gage.
Note: The fire sprinkler K is not the same as the dimensionless K value
as given by Equation 8.18 and used extensively throughout AFT
Impulse.
The Ksprinkler is related to the discharge coefficient. It can be shown that

the relationship is given by
ρ water
CD A = K sprinkler
2

manual (si) g.doc
Pumps
Pumps can operate at variable speeds. Affinity laws (also called
homologous laws) allow an estimation of the pump performance at
speeds other than the design speed.
Before discussing the affinity laws, it should be noted that the affinity
laws are an approximation. In many cases pump manufacturers will have
performance data for a number of speeds. These data should be used if
available. When not available, the affinity laws can be used with an
understanding that an approximation is being made.
The affinity laws for pumps are as follows:
2
ΔH1 ⎛ s1 ⎞
=⎜ ⎟
ΔH 2 ⎜⎝ s2 ⎟⎠
(8.19)
Q1 s1
=
Q2 s2
where s is the speed. If the pump data is input as a polynomial, the speed
will affect the curve as follows:
2 3 4
ΔH1 = a + bQ1 + cQ1 + dQ1 + eQ1
2 3 4
ΔH 2 = s 2 ΔH1 = s 2 a + s 2bQ1 + s 2 cQ1 + s 2 dQ1 + s 2eQ1
2 3 4
Q2 ⎛Q ⎞ ⎛Q ⎞ ⎛Q ⎞
ΔH 2 = s 2 a + s 2b + s 2 c⎜ 2 ⎟ + s 2 d ⎜ 2 ⎟ + s 2 e⎜ 3 ⎟
s ⎝ s ⎠ ⎝ s ⎠ ⎝ s ⎠
3 4
Q 2 Q
ΔH 2 = s a + sbQ2 + cQ2 + d 2 + e 23
2
s s
(8.20)
where ΔH1 is the input curve that typically represents 100% speed and
ΔH2 is the head rise at speed s. In this form, the speed s will be a decimal
and not a percent. That is, 50% speed will result in the appropriate s of
0.5. AFT Impulse determines this automatically in the Steady-State and
Transient Solvers.
If you model the pump as a controlled flow or discharge pressure, AFT
Impulse will backsolve Equation 8.20 for the speed s and include that in
the output Pump Summary.
8/10/2007 4:10:28 PM Page 365

Chempak thermophysical property database

The optional Chempak Database offers approximately 700 fluids to the
user. In addition, it offers non-reacting mixture calculations. Information
on the theory and calculation methods used in Chempak is given in
Appendix D.
Accuracy options
When performing interpolations for properties, a 2 point or 4 point
interpolation scheme can be used. The 2 point scheme implemented in
AFT Impulse is the Standard Accuracy option. The 4 point scheme is the
high accuracy option.
ASME steam tables database

ASME Press in References). These properties are both pressure and
temperature dependent.
AFT Standard fluids

The AFT Standard Database offers about 10 common fluids to the user.
Information on how to use these fluids within AFT Impulse are given in
Chapter 5.
Viscosity
The dynamic viscosity in AFT Impulse is assumed to be a function of
temperature only. In general this is a good assumption, although systems
at very high pressure usually start to show some pressure dependence.
These properties are modeled as polynomial curve fits of temperature.

manual (si) g.doc
CHAPTER 9
Theory of Waterhammer and Solution

Methodology
This chapter discusses the mathematical models used to predict

waterhammer events.
Description of waterhammer
Please see Chapter 3 for a summary description of waterhammer in a
simple one-pipe system.
Instantaneous waterhammer
With few exceptions (discussed at the end of the chapter), it is safe to
calculate the maximum possible waterhammer pressure surge by using
the instantaneous waterhammer equation. The instantaneous
waterhammer equation assumes that the transient event occurs either
instantaneously or rapidly enough such that it is in effect instantaneous.
In such a case, it can be shown (Wylie, et al., 1993, pp. 4) by use of the
mass and momentum equation that the pressure transient is given by the
following equation:
ΔP = − ρ aΔV (9.1)

manual (si) g.doc
where:
ρ = density
a = wavespeed
By adding the pressure surge to the existing static pressure, one can
obtain the maximum theoretical pressure in the pipe. However, in some
cases, such as transient cavitation, the pressure can exceed the
instantaneous prediction. The engineer should therefore be cautious in
its application.
Here’s an example. Assume flow in a pipe filled with water is
instantaneously stopped by a valve closure. Assume the density is 62
lbm/ft3, the wavespeed is 4000 feet per second, and the initial velocity is
10 feet per second. Also assume that before the transient the static
pressure is 50 psig.
lbm ft ft 1 ft 2 lbf − s 2
Pmax = 62 * 4000 *10 * * + 50 psig
ft 3 s s 144 in 2 32.2 lbm - ft
Pmax = 535 + 50
Pmax = 585 psig
Wavespeed
When a transient event is initiated in a pipe system, the remainder of the
system must adjust to the new conditions. In order to adjust, the
existence of the event must be communicated to the rest of the system.
This communication takes place at the wavespeed of the fluid. The
wavespeed is somewhat analogous to the sonic speed of the liquid.
However, the wavespeed is affected by the pipe structure.
The relationship between these parameters is expressed in the following
equation (Wylie, et al., 1993, pp. 27):

manual (si) g.doc
Chapter 9 Theory of Waterhammer and Solution Methodology 369
K ρ
a2 =
1 + c1[(K E )(D e )]
where:
a = wavespeed
K = fluid bulk modulus of elasticity
ρ = fluid density
E = pipe modulus of elasticity
D = pipe inner diameter
e = pipe wall thickness
c1 = correction factor
The sonic speed of the liquid is given by the (square root of the)
numerator. The denominator contains the reduction from sonic speed
caused by the pipe properties.
The correction factor, c1, is a function of how the pipe is restrained.
Some examples from Wylie, et al., 1993, page 27-28 are:
1. Thin-walled pipe anchored at upstream end only
c1 = 1 − μ 2
2. Thin-walled pipe anchored throughout
c1 = 1 − μ 2
3. Thin-walled pipe free to expand throughout
c1 = 1
4. Thick-walled pipe anchored at upstream end only
2e
c1 = (1 + μ ) + D (1 − μ 2)
D D+e
5. Thick-walled pipe anchored throughout
c1 =
2e
D
(
(1 + μ ) + D 1 − μ 2
D+e
)
8/10/2007 4:10:28 PM Page 369

6. Thick-walled pipe free to expand throughout
2e
c1 = (1 + μ ) + D
D D+e
7. Circular tunnel
2e
c1 = (1 + μ )
D
Given the previous, it should be apparent that in any given pipe the
wavespeed can only be approximated.
Communication time in pipes

Transient events communicate their effect through the pipe at the
wavespeed of the pipe and fluid. When transient events occur, they are
communicated upstream (or downstream) so conditions can adjust to the
new conditions. No response to the changing conditions can be obtained
until the communication wave travels to the other end of the pipe and
back. This communication time is thus given by the following:
L
Δt = 2 (9.2)
a
Any event that occurs with a time frame shorter than this is equivalent to
an instantaneous event.
Method of Characteristics
The analysis of waterhammer in liquid pipe systems is based on the
transient mass and momentum equations. For a complete development of
the equations please see one of the many excellent references (Wylie, et
al., 1993, Chaudhry, 1987, and Swaffield et al., 1993). Solving the
equations based on a prescribed relationship between time step and
distance step is the essence of the Method of Characteristics which will
be discussed in depth in the following sections.

manual (si) g.doc
Momentum equation
Following from Wylie, et al., 1993, page 22, the momentum equation
can be expressed as
1∂ P ∂V fV V
+ + g sin(α ) + =0 (9.3)
ρ∂ x ∂t 2D
where:
P = pressure
V = velocity
ρ = density
x = distance along pipe
t = time
g = acceleration due to gravity
D = diameter
f = friction factor
α = angle of pipe slope
Mass continuity equation

Following from Wylie, et al., 1993, page 27, the continuity equation can
be expressed as
a2 ∂ V ∂ P
+ =0 (9.4)
g ∂ x ∂t
where:
a = wavespeed
Equations 9.3 and 9.4 are two partial differential equations with the two
unknowns P and V and the two independent variables x and t.
Application of the Method of Characteristics will convert these two
partial differential equations to four ordinary differential equations as
follows. To maintain flexibility with g-level, the development will leave
the parameter P in the equations rather than convert to head, H, as in
Wylie, et al., 1993.
8/10/2007 4:10:28 PM Page 371

The equations are identified as L1 and L2 .
1∂ P ∂V fV V
+ + g sin(α ) + = L1 (9.5)
ρ∂ x ∂t 2D
∂V ∂ P
ρa 2 + = L2 (9.6)
∂ x ∂t
Combine these two equation linearly using a parameter, λ, (Wylie, et al.,
1993, pp. 27)
L1 + λ L2 = 0
or,
⎛ 1 ∂ P ∂ P⎞ ⎛ ∂ V ∂V⎞
λ⎜ + ⎟+⎜ρ + λρ a 2 ⎟
⎝λ ∂ x ∂ t ⎠ ⎝ ∂ t ∂ x⎠
ρ fV V
+ ρ g sin(α ) + =0 (9.7)
2D
If
dx 1
= = λ a2 (9.8)
dt λ
then
1
λ=±
a
and
dx
= ±a
dt
From calculus
dP ∂ P dx ∂ P
= + (9.9)
dt ∂ x dt ∂ t
dV ∂ V dx ∂ V
= + (9.10)
dt ∂ x dt ∂ t
Substituting Equation 9.8 into 9.7

manual (si) g.doc
⎛ ∂ P dx ∂ P ⎞ ⎛ ∂ V ∂ V dx ⎞
λ⎜ + ⎟+⎜ρ +ρ ⎟
⎝ ∂ x dt ∂ t ⎠ ⎝ ∂ t ∂ x dt ⎠
ρ fV V
+ ρ g sin(α ) + =0 (9.11)
2D
and substituting Equations 9.9 and 9.10 into 9.11
⎛ dP ⎞ ⎛ dV ⎞ ρ fV V
λ⎜ ⎟+⎜ρ ⎟ + ρ g sin(α ) + =0
⎝ dt ⎠ ⎝ dt ⎠ 2D
or
1 dP dV ρ fV V
± +ρ + ρ g sin(α ) + =0 (9.12)
a dt dt 2D
Now multiply Equation 9.12 by (a dt), which also equals dx
ρ fV V
± dP + ρ adV + ρ g sin(α )dx + dx = 0 (9.13)
2D
The gravity term contains the term α, and the term can be converted as
follows
⎛ dz ⎞
ρ g sin(α )dx = ρ g ⎜ ⎟dx = ρ gdz (9.14)
⎝ dx ⎠
where dz is the vertical change in pipe elevation. Therefore, substituting
9.14 into 9.13
ρ fV V
± dP + ρ adV + ρ gdz + dx = 0
2D
Further substitution of the mass flow rate for velocity yields
a f
± dP + dm& + ρ gdz + dx m& m& = 0 (9.15)
A 2 ρDA2
Integrating Equation 9.15 with the positive sign along the C+
characteristic from point A to P (Figure 9.1) yields
8/10/2007 4:10:28 PM Page 373

PP m& P zP xP
a f
∫ dP + A ∫ dm& + ρ g ∫ dz + 2 ρDA2 x
∫ m& m& dx = 0 (9.16)
PA m& A zA A
where the subscript P denotes the current solution point at x = i at time t

= Δt, and subscript A denote the values at point x = i -1 at the previous
time step, t = 0. A similar equation can be written for the C-
characteristic from point B to P.
The term on the right before the equal sign in Equation 9.16 can be
integrated assuming a second order accurate representation of m& to
obtain
(PP − PA ) + a (m& P − m& A ) + ρ g (z P − z A ) +

A
(9.17)
fΔx
m& P m& A = 0
2 ρDA2
t = 5Δt
t = 4Δt
t = 3Δt
t = 2Δt
P
t = Δt
C+ C-
A B
t=0
x=0 x = i-1 x=i x = i+1 x=L
Figure 9.1 Method of Characteristics grid

manual (si) g.doc
Similarly
a
− (PP − PA ) + (m& P − m& A ) + ρ g (z P − z A )
A
(9.18)
fΔx
+ m& P m& A = 0
2 ρDA2
Introducing two convenient parameters:
Impedance:
a
B=
A
Resistance:
f Δx
R=
2 ρDA2
Equations 9.17 and 9.18 then become
(PP − PA ) + B(m& P − m& A ) + ρ g (z P − z A ) + Rm& P m& A =0
− (PP − PA ) + B(m& P − m& A ) + ρ g (z P − z A ) + Rm& P m& A = 0

Substituting more general variables for A, B and P, and simplifying the
equations yields
Pi, new = C P − BP m& i, new (9.19)
Pi, new = CM + BM m& i, new (9.20)
where:
8/10/2007 4:10:28 PM Page 375

C P = Pi −1,old + Bm& i −1, old − ρ g ( zi − zi −1 )
C M = Pi +1,old − Bm& i +1, old + ρ g ( zi − zi +1 )
BP = B + R m& i −1, old
BM = B + R m& i +1,old
Equations 9.19 and 9.20 are referred to as the compatibility equations

(Wylie, et al., 1993, pp. 42). Since all the parameters in CP, CM, BP, and
BM are known, the only unknowns in 9.19 and 9.20 are Pi,new and m & i, new .
Thus there are two equations and two unknowns.
For example, to solve for the pressure at an interior pipe point, the flow
rate can be eliminated and the pressure solved for directly:
C B + C M BP
Pi, new = P M (9.21)
BP + BM
This value for pressure can be substituted back into equation 9.19 and
9.20 to solve for flow rate, or the flow rate can be solved for directly:
C − CM
m& i, new = P (9.22)
BP + BM
The compatibility equations allow a solution of all interior points in a
pipe. However, the pipe endpoints are solved by applying specific
boundary conditions. The boundary condition relationships build on the
previous methods, and offer the final needed information to generate a
complete solution.
Transient vapor cavitation

When the rarefaction pressure waves drop as low as the vapor pressure
of the fluid, the fluid experiences local vaporization and vapor cavities
form. There are several approaches to model this. AFT Impulse
incorporates one of the more simple models called the Discrete Vapor
Cavity Model, or DVC Model (Wylie, et al., 1993, pp. 66-69).
To apply the method, consider what happens when the fluid pressure
drops to the vapor pressure. When a solution for pressure at a node is

manual (si) g.doc
below the vapor pressure, the pressure is then set to the vapor pressure.
When this happens there is not enough flow to maintain continuity, so a
cavity forms. This results in an imbalance of flow into and out of the
node.
Assuming the cavity occurs discretely at the node location, the volume is
given as follows:
Pnew = Psat
Vvapor , new = Vvapor ,old
(
+ − m& up, new − m& up, old + m& down, new + m& down, old ) 2Δρt
(9.23)
If no vapor existed previously, the old volume is zero. When the vapor
volume becomes negative, the cavity collapses and the pressure again
rises above vapor pressure. In this case the conventional methods
(Equations 9.19, 9.20) again are applicable.
The DVC method is well understood and documented for interior pipe
points. However, application of the method to general boundary
conditions in multi-pipe systems is limited. The theory for many of the
junctions used in AFT Impulse is developed in Walters, 1991.
Pipe sectioning
The relationship between pipe sectioning and time step is given by
L
a= (9.24)
n Δt
where n is the number of sections and L/n is the length of each pipe
section. If the transient event occurs within some particular time step,
the number of sections, n, must be chosen such that it is consistent with
the time step in Equation 9.24.
Specifying the number of sections for multi-pipe systems is more
complicated. Each pipe has its own wavespeed and length. For each pipe
the following must be satisfied:
8/10/2007 4:10:28 PM Page 377

Li
ni = (9.25)
ai Δt
In Equation 9.25 as applied to each pipe in the system, the time step Δt
must be the same for all pipes. The length, L, is a given for each pipe and
thus cannot be changed. The wavespeed, a, is also a given for each pipe,
although it is known with less certainty than is the length. The free
parameter is then the number of sections, n.
In any pipe system there will be one pipe that is the controlling pipe.
The controlling pipe is that pipe which has the least number of sections,
sometimes only one. Once the controlling pipe is selected, the time step
is determined by solving Equation 9.24 for Δt. Then the number of
sections in the remaining pipes is obtained from Equation 9.25.
However, this presents a dilemma. Because the number of sections in
remaining pipes is derived from Equation 9.25, typically the number of
sections, n, will not be a whole number. Since partial sections in a pipe
cannot be modeled, an alternative must be found.
Because the wavespeed for each pipe is the least certain parameter, an
uncertainty in wavespeed up to 15% is possible. Therefore, the
wavespeed in each pipe is allowed to depart slightly from its original
calculated value in order to cause n to be a whole number.
Therefore, Equation 9.25 becomes
Li
ni =
(1 ± ψ )ai Δt
where ψ is the accepted uncertainty in wavespeed up to ±0.15 (Wylie,
et al., 1993, pp. 54).
Two other things should be noted about pipe sectioning. Because each
pipe section is explicitly solved along the characteristic lines, breaking
pipes into continuously smaller sections offers no improvement in
accuracy. That is, using twice the number of sections in a pipe will not
yield a more accurate prediction (see Chapter 12). This is in contrast to
other kinds of finite difference methods where more increments offers
improved accuracy.
The other item of note is the effect of breaking the model into additional
increments. According to Equation 9.24, when the number of sections is

manual (si) g.doc
doubled, the time step must cut in half. This results in increasing runtime
by a factor of 4 for each doubling of the number of sections.
The Section Pipes window, discussed in Chapter 5, automates the
process of obtaining acceptable roundoff error when assigning sections.
Effects of flow resistance

The resistance to flow inside the piping system is a result of friction and
irrecoverable losses through valves and fittings. Without resistance the
pressure waves would never damp out.
It is beneficial to include all of the resistance in the system. However,
properly locating the resistance is not nearly as important in
waterhammer analysis as it is in conventional steady flow analyses.
What this means is that it is usually more efficient to “lump” the
resistances together into a single pipe, rather than use a separate junction
to model the resistance for each individual component.
For example, it is usually best to lump the valve resistance into the
adjacent pipes unless there are valve position changes during the
simulation. This is true of all types of local resistances (i.e., valves and
fittings).
The effect on the results of lumping a resistance is usually minor, while
the benefit of faster computation time is major. The faster computation
results from the fact that the user can use less pipe sections to model the
same system.
Assigned Pressure theory

With an assigned pressure, the pressure is known at the pipe endpoint.
The unknown is then the flow rate, which can be obtained by solving the
appropriate compatibility equation.
If the pressure is known at the upstream end of the pipe, the negative
compatibility equation is used (Equation 9.20).
P − CM
m& i, new = known
BM
8/10/2007 4:10:28 PM Page 379

Conversely, if the pressure is known at the downstream end, the positive
C P − Pknown
m& i, new =
BP
If the known pressure changes with time, then a current pressure is
obtained for the current time step and used in the equations.
Assigned Pressure vapor cavitation theory

Assigned Pressure junctions cannot reach vapor pressure because the
pressure is specified and is therefore always above vapor pressure.
Reservoir theory
A reservoir is similar to an assigned pressure. With the known reservoir
height, the pressure can be obtained at the junction and the flow can be
solved directly (Wylie, et al., 1993, pp. 43):
Pknown = Pi, new = H R ρg + Psurface
Then the flow rate is obtained by employing the compatibility equations.

If the pressure is known at the upstream end of the pipe, the negative
P − CM
m& i, new = known
BM
C P − Pknown
m& i, new =
BP
If the reservoir height changes with time, then a current height is

manual (si) g.doc
Reservoir vapor cavitation theory

Reservoir junctions cannot reach vapor pressure because the pressure is
specified and is therefore always above vapor pressure.
Assigned Flow theory

With an assigned flow, the flow rate is known at the pipe inlet or outlet
(Wylie, et al., 1993, pp. 43). The unknown is then the pressure, which
can be obtained by solving the appropriate compatibility equation.
If the flow rate is known at the upstream end of the pipe, the negative
Pi, new = CM + BM m& known
Pi, new = C P − BP m& known
If the known flow rate changes with time, then a current flow rate is
Assigned Flow vapor cavitation theory

When the calculated pressure at an Assigned Flow junction drops below
vapor pressure, a vapor cavity forms at the junction.
Assuming the specified flow is an inflow, the vapor volume calculation
is as follows:
Vvapor , new = Vvapor ,old
(
+ − m& up, new − m& up, old + m& down, new + m& down, old ) 2Δρt
The m
& up terms are known because they are specified terms for inflow
junctions. For outflow junctions the m & down terms are known. Similar to
a pipe interior node, when the vapor volume is negative, the cavity
collapses and the fluid pressure then rises above the vapor pressure.
8/10/2007 4:10:28 PM Page 381

Dead End theory

A dead end is similar to an assigned flow, except the flow rate is zero
(Wylie, et al., 1993, pp. 44). The unknown is then the pressure, which
can be solved by calling the appropriate compatibility equation.
If the dead end is at the upstream end of the pipe, the negative
compatibility equation is used (Equation 9.20) with zero flow.
Pi, new = C P
Conversely, if the dead end is at the downstream end, the positive
compatibility equation is used (Equation 9.19) with zero flow.
Pi, new = CM
Dead End vapor cavitation theory

When the calculated pressure at a Dead End junction drops below vapor
pressure, a vapor cavity forms at the junction.
Assuming the dead end is at the end of a pipe, the vapor volume
calculation is as follows:
(
Vvapor , new = Vvapor , old + − m& up, new − m& up, old ) 2Δρt
since the m & down terms are always zero. For dead ends at the beginning
of a pipe, the m& up terms are zero. Similar to a pipe interior node, when
the vapor volume is negative, the cavity collapses and the fluid pressure
then rises above the vapor pressure.
Branch theory
At a branch, there can be multiple pipes (Wylie, et al., 1993, pp. 51-53).
An additional relationship is needed, and that relationship is the
conservation of mass. The total mass flow in and out of the junction
must sum to zero.
In addition, the branch has a single pressure solution, Pj, and this
solution is common to all inflowing pipes.

manual (si) g.doc
The compatibility equation is written for each junction. For the pipes
flowing into the branch, the positive equation is used (Equation 9.19)
C P − Pj
m& i, new =
BP
where Pj is the branch junction pressure, as yet unknown.
For the pipes flowing out of the branch, the negative equation is used
(Equation 9.20)
Pj − C M
m& i, new =
BM
Now, sum all of the pipe flow rates into the junction
∑ m& i,new = SC − S B Pj = 0 (9.26)
where
C C
SC = ∑ P + ∑ M (9.27)
BP BM
1 1
SB = ∑ +∑ (9.28)
BP BM
In general, there may be a known flow into the branch (a flow sink) or
into the branch (a flow source). Therefore rewrite Equation 9.26 as
∑ m& i , new = SC − S B Pj = − m& Applied

Now solve for the junction pressure
SC + m& Applied
Pj = (9.29)
SB
With the pressure known, substitute back into the compatibility
equations to obtain the flow rate into the branch coming from each
connecting pipe.
The applied flow rate may vary with time.
8/10/2007 4:10:28 PM Page 383

Branch vapor cavitation theory

When the calculated pressure at a Branch junction drops below vapor
pressure, a vapor cavity forms at the junction.
Assuming a flow source/sink is specified, the vapor volume calculation
is as follows:
⎜
( )
⎛ ∑ − m& in, new − m& in, old ⎞
⎟
(
Vvapor , new = Vvapor , old + ⎜ + ∑ m& out, new + m& out, old ⎟ Δt
)
⎜ ⎟ 2ρ
⎝ ( )
⎜ − m& B, new + m& B, old ⎟
⎠
where m
& B is the branch source/sink term defined as positive flow into
the branch. The m
& in terms are obtained from the positive compatibility
equation, while the m
& out terms are obtained from the negative equation.
Similar to a pipe interior node, when the vapor volume is negative, the
cavity collapses and the fluid pressure then rises above the vapor
pressure.
Tee/Wye theory
The Tee/Wye junction is a static element, and its solution method is
essentially the same as the Branch junction. The only difference is that
the loss factors are updated dynamically similar to steady flow.
Valve theory
A valve junction has a pressure loss across the valve. When the valve
closes this loss becomes infinite (Wylie, et al., 1993, pp. 57). With two
connecting pipes, the upstream and downstream pipe pressures are
obtained from the compatibility equations (Equations 9.19, 9.20):
Pi, new,up = C P,up − BP,up m& i, new,up
Pi, new, down = C M , down + BM , down m& i, new, down

From a mass balance:

manual (si) g.doc
m& i, new,up = m& i, new, down = m& valve
The pressure drop is related to the flow rate according to the following:
Pi, new,up − Pi, new, down = Rvalvem& valve 2
where
Rvalve = Valve Resistance
Substituting the compatibility equations (Equations 9.19, 9.20) in and
eliminating the pressures
(CP,up − BP,up m& valve )
( )
− C M , down + BM , down m& valve = Rvalvem& valve2
Rearranging terms,
m& valve 2 −
1
Rvalve
(CP,up − BP,up m& valve )
+
1
Rvalve
(CM ,down + BM ,down m& valve ) = 0
Using the quadratic equation, the solution for positive flow through the
valve is
(
m& valve = − 1 Rvalve BP,up + BM , down )
( ) (
+ 1 Rvalve 2 BP,up + BM , down 2 + 4 Rvalve C P,up − C M , down )
For negative flow, the solution is
(
m& valve = 1 Rvalve BP ,up + BM , down )
( ) (
− 1 Rvalve 2 BP ,up + BM , down 2 − 4 Rvalve C P,up − CM , down )
After solving for flow rate, m
& valve , the upstream and downstream
pressures can be obtained from the compatibility equations.
If the valve closes then 1/Rvalve = 0 and
8/10/2007 4:10:28 PM Page 385

m& valve = 0
Exit valves
If the valve is an exit valve, then the downstream pressure is known and
m& valve = − 1 Rvalve BP ,up
(
+ 1 Rvalve 2 BP,up 2 + 4 Rvalve C P,up − Pexit )
Valve vapor cavitation theory
When the calculated pressure at the upstream or downstream side of a
valve junction drops below vapor pressure, a vapor cavity forms on that
side of the junction.
Vapor cavities can form on either side of the valve or on both sides.
Vvapor ,up, new = Vvapor ,up,old
(
+ − m& up, new − m& up, old + m& valve, new + m& valve, old ) 2Δρt
Vvapor , down, new = Vvapor , down, old
(
+ − m& valve, new − m& valve, old + m& down, new + m& down, old ) 2Δρt
As an example, when a vapor cavity forms downstream of a valve, the
downstream pressure become fixed at the vapor pressure. The
calculation then is similar to an exit valve, which also flows to a fixed
pressure.
m& valve = − 1 Rvalve BP ,up
(
+ 1 Rvalve 2 BP ,up 2 + 4 Rvalve C P,up − Psat )
When a vapor cavity occurs on both sides of the valve, the pressure is at
vapor pressure on both sides and thus there is no pressure drop across
the valve. Accordingly, the flow rate goes to zero.

manual (si) g.doc
pressure.
Infinite Pipe theory

An infinite pipe junction is a special junction that accepts pressure
waves but never reflects any waves (Wylie, et al., 1993, pp. 121-122).
This junction is useful in modeling the part of a pipe system which is so
long that its communication time results in no interaction with the rest of
the system. A discussion of when this junction should be used is given in
Chapter 12.
Note: In steady flow, the infinite pipe junction is usually a known flow,
but can also be a known pressure. Thus the Infinite Pipe capability is
offered through the Assigned Flow or Assigned Pressure junction. It is
specified by setting the transient special condition to No Reflections
(Infinite Pipe).
The pressure and flow at an infinite pipe is calculated by extrapolating

the known pressures and flows at nearby pipe stations at previous time
steps. Calculations are therefore performed at an “infinity point”, one
artificial station beyond the physical pipe end. The values at the infinity
point are then used for each successive calculation.
Pinfinity, new = 3Pi, − Δt − 3Pi +1,−2Δt + Pi + 2,−3Δt
m& infinity,new = 3m& i, − Δt − 3m& i +1,−2Δt + m& i + 2, −3Δt
Assuming that the infinite pipe is at the downstream end, the current
pressure can be solved using the positive characteristic compatibility
equation (Equation 9.19),
Pi, new = C P − BP m& i, new
and the negative characteristic compatibility equation (Equation 9.20)
Pi, new = CM + BM m& i, new (9.30)
8/10/2007 4:10:28 PM Page 387

For Equation 9.30, the normal CM or BM cannot be calculated because the
values past the end of the pipe are not known. This is where the infinity
values are used. Instead, the following are used:
C M = Pinfinity,old − Bm& infinity,old − ρg ( zi − zi +1 ) (9.31)
BM = B + R m& infinity, old (9.32)
In Equation 9.31, use the following approximation which is perfectly

accurate for straight pipes
zi − zi +1 = zi −1 − zi
Also, in equations 9.31 and 9.32 it is assumed that the pipe property
values B and R remain valid for the infinity point, which is a good
assumption unless the pipe properties change.
Infinite Pipe vapor cavitation theory

The Infinite Pipe junction is designed to only pass pressure waves out
through the junction, but never to allow any reflections. If the junction
drops to vapor pressure, a cavity would occur. But it would cause
reflections. Therefore, vapor cavities at these junctions are ignored and
the pressure is allowed to fall below the vapor pressure. This maintains
the function of the junction.
In such cases, the user may want to change the model.
Spray Discharge theory

Spray discharge junctions (shown in Figure 9.2) model spray locations in
a system. The spray junction discharges to a known pressure through an
orifice or nozzle of some kind.

manual (si) g.doc
m& spray
Pexit
Figure 9.2 Spray discharge schematic
The relationship for pressure drop across the nozzle is as follows

1
ΔP = ρV 2
2
1
ΔP = Pj , new − Pexit = sgn( m& spray ) m& spray 2
(
2 ρ C D Aspray )
2
where, for clarity, m

& spray is defined positive flowing out of the system.
In addition, the spray junction is similar to a branch junction, so the

following equation applies with the flowrate defined positive out.
SC − m& spray
Pj , new =
SB
where SC and SB are given by Equations 9.27 and 9.28.
Combining the preceding equations and eliminating flow rate,
1 1 S
sgn (m& ) m& 2 + m& + Pexit − C = 0
C D 2 Aspray 2 2 ρ SB SB
or
2 2 ρC D 2 Aspray 2
sgn( m& )m& + m&
SB
2 ρC D 2 Aspray 2 (S B Pexit − SC )
+ =0
SB
8/10/2007 4:10:28 PM Page 389

This is a quadratic equation, and the mass flowrate can be obtained by
use of the quadratic formula.
Note: If the spray junction pressure falls below the exit pressure, reverse
flow will occur. If the exit pressure is the atmosphere, the reverse flow
will be air. However, Impulse assumes the inflow is at the pipe system
density. This can lead to erroneous results if ignored. Reverse flow can
be determined by graphing the flow at the spray junction or looking at
the summary of maximum and minimum flows in the Output window.
Spray Discharge vapor cavitation theory

When the calculated pressure at a Spray Discharge junction drops below
vapor pressure, a vapor cavity forms at the junction. This is similar to a
Branch junction.
The vapor volume calculation is as follows
⎜
(
⎛ ∑ − m& in, new − m& in, old ) ⎞
⎟
(
Vvapor , new = Vvapor , old + ⎜ + ∑ m& out, new + m& out, old ) ⎟ Δt
⎜ ⎟ 2ρ
⎝ (
⎜ + m& spray , new + m& spray , old )
⎟
⎠
where m
& spray is the spray outflow term. The m& in terms are obtained
from the positive compatibility equation, while the m
& out terms are
obtained from the negative equation.
pressure.
Pump theory
The pump junction can model the effects of a pump with a known
change in pump speed (Wylie, et al., 1993, pp. 59-60), or with inertial
data where the pump speed is calculated.

manual (si) g.doc
Modeling pump as a known speed

Assuming that the pump curve can be modeled by a quadratic, then
ΔH = a + bQ pump + cQ pump 2
where a, b and c are obtained from curve fits of vendor data. This curve
assumes the pump operates at 100% speed. During the transient, the
speed will change.
The affinity law for pumps given by the following equations
ΔH
= constant1
α2
Q
= constant2
α
where α is the pump speed. Applying these relationships the pump curve
is transformed
ΔH Q pump Q pump 2
= a′ + b′ + c′
α2 α α2
ΔH = α 2 a′ + αb′Q pump + c′Q pump 2
Converting to pressure rise
⎛ αb′ c′ ⎞
ΔP = ρg ⎜α 2 a′ + m& pump + m& pump 2 ⎟
⎜ ρ ρ2 ⎟
⎝ ⎠
ΔP = α 2 a + αbQ pump + cQ pump 2 (9.33)
where the primes on the constants were dropped after they were divided
through by ρg.
By use of the two compatibility equations (Equation 9.19, 9.20),
Pup, new = C P − BP m& pump
Pdown, new = CM + BM m& pump
8/10/2007 4:10:28 PM Page 391

and substituting these equations into Equation 9.33
(C M + BM m& pump ) − (C P − BP m& pump )
= α 2 a + αbm& pump + cm& pump 2
This is a quadratic equation for the pump flow rate which can be solved
by the quadratic formula
⎛ BP,up + BM , down − bα ⎞
Q pump = ⎜⎜ ⎟
⎟
⎝ 2c ⎠
⎡ ⎛ 2
(
⎢ ⎜ 4b α a + C P ,up − C M , down
× ⎢1 − 1 −
)⎞⎟1 2 ⎤⎥
⎢⎣ ⎝
⎜ (
BP,up + BM , down − bα 2 ) ⎟
⎠
⎥
⎥⎦
With the solution for the pump flow rate, the pressure upstream and
downstream of the pump can be obtained from Equations 9.19 and 9.20.
Pumps with higher order curve fits

If the pump is modeled with a third or fourth order curve, the full speed
pump curve is:
ΔH = a + bQ pump + cQ pump 2 + dQ pump 3 + eQ pump 4
and with the speed included,
ΔH Q pump Q pump 2 Q pump 3 Q pump 4

= a′ + b′ + c′ + d′ + e′
α2 α α2 α3 α4
Combine this with the compatibility equation as in the previous section
to obtain:
(CM + BM m& pump )− (CP − BP m& pump ) = α 2a + αbm& pump

d e
+ cm& pump 2 + m& pump 3 + m& pump 4
α α 2
This equation is then solved by Newton-Raphson iteration to obtain the

flow rate, m
& pump .

manual (si) g.doc
The above models are accessed using the Without Inertia option.
Backflow through the pump – one quadrant

If the pump cannot supply enough head to overcome the system
requirements at some time during the transient, backflow will occur
through the pump. When this occurs, the pump curve loses meaning.
To address this situation, AFT Impulse makes an approximation when
using the Without Inertia model. In the case of backflow it is assumed
that the head rise for all negative flow rates is equivalent to the head rise
at zero flow for the given curve at the given speed. With the head rise
known, the negative flow rate through the pump is obtained by using
both compatibility equations with the known head rise. After the flow
rate is known, the pump suction and discharge pressures are calculated.
To avoid this approximation, a four quadrant model can be used.
Backflow through the pump – four quadrants

The model “Startup – Four Quadrant Known Speed” uses the four
quadrant pump model, which allows backflow through the pump for
pump startup modeling with known speed. If the pump operates at full
speed and backflow is possible, the model “No Transient – Four
Quadrant” can be used.
Flow rates greater the pump maximum – one quadrant

It is possible for the system transients to cause flow rates through the
pump greater than the maximum flow from the curve. In such cases, the
Without Inertia model assumes the head rise is zero and calculates the
flow rate from the compatibility equations.
To avoid this approximation, a four quadrant model can be used.
Flow rates greater the pump maximum – four quadrants

The model “Startup – Four Quadrant Known Speed” uses the four
quadrant pump model, which allows flows greater the pump maximum
to be modeled for pump startups with known speed. If the pump operates
at full speed and flows greater the pump maximum are possible, the
model “No Transient – Four Quadrant” can be used.
8/10/2007 4:10:28 PM Page 393

Submerged pumps
Pumps can also be submerged in a reservoir. In such cases the upstream
pressure is known, and the flow rate is obtained from the following:
⎛ BM , down − bα ⎞
m& pump = ⎜⎜ ⎟⎟
⎝ 2c ⎠
⎡ ⎛
⎢ ⎜
× ⎢1 − 1 −
(
4b α 2 a + PR − CM , down )⎞⎟1 2 ⎤⎥
⎢⎣ ⎝
⎜ (
BM , down − bα 2 ) ⎟
⎠
⎥
⎥⎦
Where PR is the pressure at the pump suction.
Pump with known speed vapor cavitation theory

When the calculated pressure at the upstream side of a Pump junction
drops below vapor pressure, a vapor cavity forms on that side of the
junction.
Usually vapor cavities form on the upstream side of the pump only, but
in cases where there is backflow through the pump cavities can also
form on the downstream side.
Vvapor ,up, new = Vvapor ,up,old
(
+ − m& up, new − m& up, old + m& pump, new + m& pump, old ) 2Δρt
Vvapor , down, new = Vvapor , down, old
(
+ − m& pump, new − m& pump, old + m& down, new + m& down, old ) 2Δρt
As an example, when a vapor cavity forms upstream of a pump, the
upstream pressure becomes fixed at the vapor pressure. The calculation
then is similar to a submerged pump, which also flows from a fixed
pressure.

manual (si) g.doc
⎛ BM , down − bα υ ⎞
m& pump = ⎜ ⎟
⎜ 2c ρ 2 ⎟
⎝ ⎠
⎡ ⎛
× ⎢⎢1 − ⎜1 −
(
4b α 2 a + Psat − CM , down )⎞⎟1 2 ⎤⎥
⎢⎣ ⎝
⎜ ( BM , down − bα ρ 2) ⎟
⎠
⎥
⎥⎦
Similar to a pipe interior node, when the vapor volume is negative the
pressure.
Modeling pump with unknown speed

There are four pump models in AFT Impulse 4.0 that account for pump
inertial effects and predict the speed over time. These are:
• Trip With Inertia and No Back Flow or Reverse Speed
• Startup With Inertia and No Back Flow or Reverse Speed
• Trip With Inertia – Four Quadrant
• Startup With Inertia – Four Quadrant, Known Motor Torque/Speed
Predicting the response of a pump during a trip or startup requires
additional data for the pump. Use of the affinity laws for pumps is
essential. If backflow is possible, "four quadrant" pump curve data is
required. The first two models mentioned above assume the presence of
a check valve which prevents backflow, or just that backflow never
occurs. In such cases, the pump will usually operate in the first quadrant.
Thus four quadrant data is not required.
The third and fourth models allow backflow, and thus require four
quadrant data. All four models will be discussed in the following
sections.
Trip With Inertia and No Back Flow or Reverse Speed

When no backflow or reverse speed occurs through the pump, the
normal pump curve data for flow vs. head can be used to model the
pump response. Pump power data is also required as a function of flow.
8/10/2007 4:10:28 PM Page 395

The method is presented by Watters (1979, p. 169). The unbalanced
torque is given by:
dω
T = −I (9.34)
dt
where T is the torque on the pump, I is the moment of inertia of the
pump and entrained liquid, and dω/dt is the change in angular speed with
time.
The pump power is related to the pump torque by:
P = Tω (9.35)
The pump power is related is a function of the flow as specified by the
user. By rearranging and integrating Equation 9.34 one obtains
1 dω
I ∫ dt = − ∫
T
Assuming an average value of torque over the time step one obtains:
Δt 1
= − Δω
I T
With the speed known at the previous time step, one can solve for the
new speed as
Δt
ω new = ω old − T
I
Adding in the definition of torque from Equation 9.35
Tnew + Told Δt
ω new = ω old − (9.36)
2 I
The old torque is known from the previous step, and thus the new speed
can be solved for by iteration. The mass flow rate and head are obtained
from Equations 9.19 and 9.20 (with the pressure converted to head) and
the pump curve supplied by the user.

manual (si) g.doc
Startup With Inertia and No Back Flow or Reverse Speed

This model uses a similar approach to that just discussed, but adds motor
torque into the calculation. This is described, for example, in Brown &
Rogers (1980).
dω
Tmotor − T fluid = I
dt
Integrating over a time step:

ω new = ω old +
⎛ Tmotor , new + Tmotor , old T fluid , new + T fluid , old ⎞ Δt
⎜ − ⎟
⎜ 2 2 ⎟ I
⎝ ⎠
(9.37)
The “old” motor torque and fluid torque at the previous time step are
known, and the values at the new time step are both functions of the new
speed. The fluid torque is related to the power through Equation 9.35,
and the power can be obtained from the new flow.
The motor torque is entered by the user as a function of speed (see
Figures 9.3 and 9.4), and is therefore known at the new speed. This
process thus requires iteration.
8/10/2007 4:10:28 PM Page 397

Figure 9.3 Motor torque vs. speed data is used for startup
modeling
This calculation uses the user provided pump data for head vs. flow and
power vs. flow, and thus only works for positive or zero flows. The mass
flow rate and head are obtained from Equations 9.19 and 9.20 (with the
pressure converted to head) and the pump curve supplied by the user.
The typical shape of a motor torque vs. speed curve is shown in Karassik
(2001, p. 8.25, Figure 26), which is reflected in the Figure 9.4 data.

manual (si) g.doc
Figure 9.4 Example graph of motor torque vs. speed
Trip With Inertia – Four Quadrant

The method used for modeling a pump trip is described in most
waterhammer textbooks (Chaudhry, 1987, Chapter 4, Swaffield, et al.,
1993, pp. 154-166, Wylie et al., 1993, Chapter 7). From the conventional
method, four dimensionless parameters are defined:
ΔH ΔP
h= =
ΔH R ΔPR
T
β=
TR
(9.38)
Q m&
v= =
QR m& R
N
α=
NR
8/10/2007 4:10:28 PM Page 399

where the R subscript refers to rated conditions, usually those at the best
efficiency point. In summary, h is the ratio of the pump head to the rated
head, β is the ratio of the pump torque to the rated torque, v is the ratio
of the flowrate to the rated flowrate, and α is the ratio of the
synchronous speed to the rated synchronous speed.
Figure 9.5 Pump window with pump data entered for specific
speed of 0.46.
Two further dimensionless parameters are created to relate these

parameters:
h
FH (θ ) =
α + v2
2
(9.39a)
β
FB (θ ) =
α 2 + v2
where

manual (si) g.doc
v
θ = π + tan −1 (9.39b)
α
Some references choose to define θ slightly differently:
α
θ = tan −1
v
Figure 9.6 Variation of FH and FB for a pump with specific speed

of 0.46 (see Wylie, et al., 1993).
AFT Impulse allows you to select either definition as the basis for the FH
and FB data in Equation 9.39a.
The parameters FH and FB are properties of the pump, and it is typically
assumed that for pumps of similar specific speed they are invariant (see
Brown & Rogers for a discussion questioning this assumption). Figures
9.5 and 9.6 show data for a pump with a specific speed of 0.46 as given
by Wylie, et al., (1993, pp. 147-148).
AFT Impulse provides data of FH and FB vs. θ for pumps of twenty-one
specific speeds. See Table 9.1 for a list of all specific speeds available.
8/10/2007 4:10:28 PM Page 401

If you should need to construct data for other sets, please consult the
three textbooks referenced above for additional information. Note that
the textbooks do not give a method for constructing these data sets, but
they contain references that you can pursue.
Table 9.1 Four quadrant data available in AFT Impulse

-- Specific Speed -- Reference
(-) (gpm units) (metric units)
0.3 810 15.7 Brown & Rogers, Partial, From Nevada Test, 1980
0.39 1060 20.5 Kittredge, 1956
0.42 1140 22.1 Brown & Rogers, from Bureau of Reclamation, 1980
0.46 1270 24 Donsky, 1961
0.71 1935 37.4 Kittredge, 1956
0.79 2160 41.8 Thorley, 1996
1.21 3300 64 Thorley, 1996
1.36 3725 71.9 Thorley, 1996
1.44 3940 76.1 Thorley, 1996
1.61 4400 85.1 Brown & Rogers, from Bureau of Reclamation, 1980
1.83 5000 97 Thorley, 1996
1.9 5200 101 Thorley, 1996
2.32 6340 123 Thorley, 1996
2.48 6790 131 Thorley, 1996
2.54 6940 134 Thorley, 1996
2.78 7600 147 Donsky, 1961
3.2 8760 170 Thorley, 1996
4.94 13500 261 Donsky, 1961
With data for FH and FB, the compatibility equations (9.19, 9.20) are
used as follows:
ΔPpump = C M − C P + (BM + BP )m& pump (9.40)
But from Equations 9.38 and 9.39,
(
ΔPpump = PR α 2 + v 2 FH ) (9.41)
m& pump = vm& R
Substituting Equations 9.41 into 9.40 obtains:

manual (si) g.doc
( )
C P − CM − vm& R (BP + BM ) + PR α 2 + v 2 FH = 0 (9.42)
Since the value of FH depends on θ, which depends on α and v (see

Equation 9.39b), iteration is required to find the appropriate value from
the tabulated data supplied (e.g., Figure 9.5).
We need one more equation to solve for all parameters in Equation 9.38.
We can get this from the pump torque. The change in rotational speed
with torque was given in Equation 9.34. Defining several more terms,
2π
ω = NR α
60
T0
β0 =
TR
T
β=
TR
where NR is the rated speed in rpm, and the "0" subscript refers to
conditions at the beginning of the time step.
We can then write,
2π 2π
dω = N R dα = N R (α − α 0 )
60 60
Equation 9.34 can be written
T + T0 dω
T= = −I (9.43)
2 dt
Substituting these terms into Equation 9.43 obtains
NR π
β + βo − I (α − α 0 ) = 0 (9.44)
TR 15Δt
Using Equations 9.38 and 9.39 in Equation 9.44 obtains
(α 2 + v 2 )FB + β o − I NT R 15πΔt (α − α 0 ) = 0 (9.45)

R
8/10/2007 4:10:28 PM Page 403

Equations 9.42 and 9.45 are solved simultaneously using a Newton-
Raphson iterative method. Details are given in Wylie, et al., 1993, pp.
149-152.
Startup With Inertia – Four Quadrant, Known Motor

Torque/Speed
For pump startup where backwards flow occurs, four quadrant data can
be used along with the known motor torque vs. speed to calculate the
time to reach full speed. This model is especially useful for cases where
a pump is being started with backflow already established.
The methodology is similar to that just discussed Trip With Inertia –
Four Quadrant section. However, Equations 9.43, 9.44 and 9.45 must be
modified to include the motor torque. Including the motor torque
changes these equations as follows:
Tm + Tm,0 T f + T f ,0 dω
Tm − T f = − =I
2 2 dt
where the m subscript refers to the motor torque and the f subscript
refers to the fluid torque. Equation 9.44 becomes:
NR π
β f + β f , o − β m − β m, o − I (α − α 0 ) = 0
TR 15Δt
and Equation 9.45 becomes:
(α 2 + v 2 )FB + β f ,o − β m − β m,o − I NT R 15πΔt (α − α 0 ) = 0

R
Estimating the pump inertia

Obtaining data for rotational moment of inertia is difficult and
frequently requires estimation. Wylie et al., 1993, reports a method for
estimating the moment of inertia, I, for a pump and the entrained liquid.
The moment of inertia for the pump impeller including entrained water
is estimated as:
( )⎝ N
0.9556
⎛ P ⎞
I p = 1.5 107 ⎜⎜ ⎟ (9.46)
3⎟
⎠

manual (si) g.doc
where P is the power in kW, N is the rotational speed in rpm and I is in
kg-m2.
An estimate of the motor rotational inertia is
1.48
⎛P⎞
I m = 118⎜ ⎟ (9.47)
⎝N⎠
The total moment of inertia is the sum of the two.
Applied Flow Technology provides an Excel spreadsheet which uses
these correlations to predict the inertia.
Liquid Accumulator theory

A liquid accumulator, also called a lumped capacitance, models a liquid-
filled container in a system (Wylie, et al., 1993, pp. 124-125). The
interaction with the pipe system occurs as the container stretches and
contracts in response to transient pressures.
The liquid accumulator is similar to a branch, and thus the same
equations are involved.
SC + m& A, new
Pj , new =
SB
where SC and SB are given by Equations 9.27 and 9.28. The relationship
between accumulator pressure and volume is assumed to be linear,
ΔP
K′ =
ΔV V
where K’ is the effective bulk modulus of the container, and has units of
pressure. The change in volume can be related to flow rate as follows:
(
ΔV = − m& A, new + m& A, old ) 2Δρt
Therefore,
Pj , new = Pj , old −
(
K ′Δt m& A, new + m& A,old )
2Vρ
8/10/2007 4:10:28 PM Page 405

Substituting and eliminating flow rate
2Vρ
SC + Pj , old − m& A, old
Pj , new = K ′Δt
2Vρ
SB +
K ′Δt
Liquid Accumulator vapor cavitation theory

When the calculated pressure at a Liquid Accumulator junction drops
below vapor pressure, a vapor cavity forms at the junction.
The vapor volume calculation is as follows:
⎜
( )
⎛ ∑ − m& in, new − m& in, old ⎞
⎟
(
Vvapor , new = Vvapor , old + ⎜ + ∑ m& out, new + m& out, old ⎟ Δt
)
⎜ ⎟ 2ρ
⎝ ( )
⎜ − m& A, new + m& A, old ⎟
⎠
where m & A is the accumulator flow out of the junction. The m& in terms
are obtained from the positive compatibility equation, while the
m& out terms are obtained from the negative equation.
pressure.
Relief Valve theory

Relief valve junctions stay closed until the cracking pressure is reached,
at which point it will open and relieve the pressure. The valve can
relieve either to an outside ambient pressure or into another pipe in the
system. There are therefore two models required to address each case.
Internal relief valve

An internal relief valve is similar to a regular valve except that it
remains closed until the cracking pressure is exceeded. One significant
difference is if the Variable Cv model is used. This is distinct to the
relief valve junction.

manual (si) g.doc
Exit relief valve

An exit relief valve is like a regular exit valve. When the cracking
pressure is exceeded it relieves to the exit ambient.
SC − m& exit , new
Pj , new =
SB
The pressure drop across the exit is
1
ΔP = ρV 2
2
or,
K
ΔP = Pj , new − Pexit = m& exit , new2
2
2 ρAexit
Substituting
SC − m& exit , new K
− Pexit = m& exit , new2
SB 2
2 ρAexit
This is a quadratic equation in which m
& exit , new can be obtained from the
quadratic formula.
1 2K 2 Kρ ⎛ S ⎞
m& exit , new2 + m& exit , new + ⎜⎜ Pexit − C ⎟⎟ = 0
S B Aexit 2 Aexit 2 ⎝ SB ⎠
If using the variable Cv model, the K factor above itself becomes a
function of ΔP, which requires iteration to solve for the mass flow rate.
Inline Relief Valve

When this valve opens, it is similar to a Spray Discharge junction with
two pipes connected. When it is closed it acts like a branch junction with
two pipes. If the Variable Cv model is used, the solution requires
iteration.
8/10/2007 4:10:28 PM Page 407

Variable Cv model
For passive relief valves the valve Cv typically depends on the pressure
difference across the valve, as the pressure difference is what causes the
valve to open. When the pressure difference is below the cracking
pressure, the valve remains closed. When it rises above the cracking
pressure, it opens. As the pressure difference increases, it opens further
until it is fully open. Any further pressure increases do not increase the
Cv, and so the fully open Cv is used for all higher pressures.
When working with the variable Cv model, the cracking pressure is
entered in pressure/head difference (see Figure 9.7). In addition, the first
Cv data must equal zero (see Figure 9.8), and the corresponding
pressure/head to zero Cv must be the same as the cracking pressure/head
difference (shown as 700 kPa in both Figures 9.7 and 9.8).
In Figure 9.8, the first data point represents the valve closed at the
cracking pressure, with any greater pressure difference resulting in a
positive Cv, which means that flow will occur. The maximum pressure
difference in the table (770 kPa in Figure 9.8) represents the pressure
difference to fully open the valve. The Cv of 150 thus represents the Cv
when the valve is fully open. Any pressure differences greater than 770
kPa will result in a Cv of 150, the maximum value.
Figure 9.7 Variable Cv model must use pressure/head difference

manual (si) g.doc
Relief Valve vapor cavitation theory
Internal relief valves

Internal Relief Valve junction vapor cavitation is identical to that of a
regular valve. Initially the valve is closed and acts like a regular valve. If
it cracks open, then it acts like an open valve.
Exit relief valves

Exit Relief Valve junction vapor cavitation acts like a dead end junction.
Inline relief valves

Inline Relief Valve junction vapor cavitation acts like a branch junction
with two pipes.
8/10/2007 4:10:28 PM Page 409

Figure 9.8 First data point must have Cv of zero with the
pressure/head loss equal to the cracking
pressure/head difference (as shown in Figure 9.7)
Check Valve theory

A check valve is identical to a regular valve, except that it closes when
reverse flow occurs.
Check Valve vapor cavitation theory

The check valve vapor cavitation model is identical to a regular valve,
except that it closes when reverse flow occurs.

manual (si) g.doc
Control Valve theory

We can use the methods for known flow or pressure to calculate the
conditions at a control valve. AFT Impulse flow and pressure control
valves allow the control setting to be changed with time. Whether
constant or changing, the flow or pressure is known and can be used to
solve the equations.
Flow control valves

With the known flowrate, one can substitute into Equations 9.19 and
9.20 and solve for the upstream and downstream pressures, respectively,
at the valve. These can then be compared to the failure settings to see if
loss of control occurs. For instance, if the upstream pressure was less
than the downstream, that would indicate loss of control.
Once the valve loses control, it acts like a regular valve as discussed
earlier in the chapter.
Pressure control valves

Pressure control can be specified on the upstream or downstream side of
the valve. A discussion of one these control locations will apply to the
other. Let's consider the downstream pressure control case – a Pressure
Reducing Valve (PRV).
With the known pressure downstream, one can substitute the known
pressure into Equation 9.20 and solve for the flowrate through the valve.
This flowrate can then be substituted into Equation 9.19 to solve for the
upstream pressure. The upstream pressure can then be compared to the
control pressure and failure settings to see if loss of control occurs. For
instance, if the upstream pressure was less than the control pressure, that
would indicate loss of control.
Once the valve loses control, it acts like a regular valve as discussed
earlier in the chapter.
Pressure drop control valves

The flow through a constant pressure drop is obtained by subtracting
Equation 9.20 from 9.19 and solving for flowrate.
8/10/2007 4:10:28 PM Page 411

C − CM − ΔP
m& i, new = P
BP + BM
Surge Tank theory

A surge tank is an open tank to the atmosphere (Figure 9.9) and is used
as a surge suppression device (Wylie, et al., 1993, pp. 128-129).
The flow balance on the surge tank junction is similar to a branch
junction, with the deficit equaling the flow out of the surge tank, m
&T .
Similar to a branch, it can be shown that for all connecting pipes to the
junction
SC + m& T , new
Pj , new =
SB
LL Orifice
Connector
m& T Pipe
LC
Figure 9.9 Surge tank schematic
The optional connector pipe and orifice effect is incorporated by the

lumped inertia relationship as follows:
Pj , new = Psurface + ρgLL, new − C1 − C2 m& T , new

manual (si) g.doc
where C1 and C2 are defined in conjunction with lumped inertia. Refer
to Equations 9-50 and 9-51 later in this chapter.
The new tank liquid height can be related to the old height by the
following:
m& T , new + m& T , old
LL, new = LL, old − Δt
2 ρAT
Substitution yields
SC + m& T , new ⎛ m& T ,old Δt ⎞

− Psurface + ρg ⎜⎜ − LL, old + ⎟
SB ⎝ 2 ρAT ⎟⎠
m& T , newΔt
+g − C1 − C2 m& T , new = 0
2 AT
Solving for m
& T , new
SC ⎛ m& Δt ⎞
− Psurface − ρg ⎜⎜ LL,old − T , old ⎟⎟ + C1
SB ⎝ 2 ρAT ⎠
m& T , new =
⎛ gΔt 1 ⎞
− ⎜⎜ C2 + + ⎟⎟
⎝ 2 AT S B⎠
This flow rate can be back substituted to obtain the junction pressure and
surge tank liquid height.
Surge Tank vapor cavitation theory

Any short-lived vapor cavities that develop at a Surge Tank junction are
assumed to have a negligible effect compared to the surge tank itself.
Conceptually, vapor bubbles would rise by buoyancy to the surge tank
surface and evaporate at the surface.
Therefore, vapor cavitation is ignored at surge tanks.
8/10/2007 4:10:28 PM Page 413

Gas Accumulator theory

A gas accumulator (shown in Figure 9.10) is an enclosed tank filled with
gas and has a free surface. It is used as a surge suppressor (Wylie, et al.,
1993, pp. 125-128).
The flow balance on the accumulator junction is similar to a branch
junction, with the deficit equaling the flow out of the accumulator, m
& A.
Similar to a branch, it can be shown that for all connecting pipes to the
junction,
SC + m& A, new
Pj , new =
SB
The potential connector pipe and orifice effect is incorporated by the
lumped inertia relationship as follows
Pj , new = PA, new − C1 − C2 m& A, new
where C1 and C2 are defined in conjunction with lumped inertia. Refer

to Equations 9-50 and 9-51 later in this chapter.
PA
Orifice
Connector
m& A Pipe
LC
Figure 9.10 Gas accumulator schematic

manual (si) g.doc
The gas volume and pressure in the accumulator can be related by the
thermodynamic law
PAV An = C A
where n is the polytropic constant (1.4 for isentropic air, 1.0 for
isothermal) and CA is a constant.
Therefore,
PA, newV A, newn = C A

and
(
V A, new = V A, old − m& A, new + m& A,old ) 2Δρt
By substitution
SC + m& A, new
= PA, new − C1 − C2 m& A, new
SB
or
SC + m& A, new
PA, new = + C1 + C2 m& A, new
SB
Combining equations and simplifying,
(C3 + C4m& A,new )(C5 + C6m& A,new )n = C A

where
S
C3 = C + C1
SB
1
C4 = + C2
SB
Δt
C5 = V A, old + m& A, old
2ρ
Δt
C6 =
2ρ
8/10/2007 4:10:28 PM Page 415

This is a non-linear equation that can be solved using Newton-Raphson
iteration. Newton-Raphson has the following form:
F ( xi )
xi +1 = xi −
F ′( xi )
where xi is the current value of x, xi+1 is the new value, F is the function
of x to drive to zero, and F´ is the derivative.
In this case
( )( )
F = C3 + C4 m& A, new C5 + C6 m& A, new n − C A
When the value of m & A, new has been found that causes F to equal zero,
the equation will be satisfied. To solve by Newton-Raphson the
derivative of F is given by,
dF
dm& A, new
= F′ =
C7
C5 + C6 m& A, new
(
+ C4 C5 + C6 m& A, new n )
where
C7 = − nC AC6
By iteration, m& A, new is calculated. This value can then be back-
substituted to obtain the new junction pressure and new gas volume.
Gas Accumulator vapor cavitation theory

Any short-lived vapor cavities that develop at a Gas Accumulator
junction are assumed to have a negligible effect compared to the
accumulator itself. Conceptually, vapor bubbles would rise by buoyancy
to the accumulator surface and evaporate into the accumulator gas. Their
effect on the system would thus be negligible compared to that of the
accumulator gas. Therefore, vapor cavitation is ignored at gas
accumulators.
Vacuum Breaker Valve theory

A vacuum breaker valve (sometimes called an air inlet valve) allows air
to flow into the pipe during low pressure events, and expels the air when

manual (si) g.doc
the pressure rises again. Typically the cracking pressure is atmospheric.
The theory is summarized in Wylie, et al., 1993, pp. 130-131, and
detailed here.
There are four cases that can occur – two for inflow and two for outflow.
The inflow can be subsonic or sonic, as can the outflow. These make up
the four cases. The flowrate equations are given as follows:
Sonic inflow
(γ +1) (2(1− γ ))
C d Ain Po ⎛ γ − 1⎞
m& = γ ⎜1 + ⎟ (9.48a)
RTo ⎝ 2 ⎠
Subsonic inflow
2γ (γ +1) γ ⎤
C A P 2γ ⎡⎢⎛ Pi ⎞ ⎛ Pi ⎞
⎥
m& = d in o ⎜ ⎟ − ⎜⎜ ⎟⎟ (9.48b)
RTo γ − 1 ⎢⎜⎝ Po ⎟⎠ P
⎝ o⎠ ⎥
⎣ ⎦
Sonic outflow
(γ +1) (2(1−γ ))
C A P ⎛ γ −1⎞
m& = d out i γ ⎜1 + ⎟ (9.48c)
RTL ⎝ 2 ⎠
Subsonic outflow
−2 γ − (γ +1) γ
C d Aout Pi 2γ ⎡⎢⎛ Pi ⎞ ⎛ Pi ⎞ ⎤
⎥ (9.48d)
m& = ⎜ ⎟ − ⎜⎜ ⎟⎟
RTL γ − 1 ⎢⎜⎝ Po ⎟⎠ ⎝ Po ⎠ ⎥
⎣ ⎦
where:
Pi = pressure in the pipe at the junction node
Po = atmospheric pressure
To = atmospheric temperature
TL = liquid temperature
8/10/2007 4:10:28 PM Page 417

CdAin = flow area for inflow
CdAout = flow area for outflow
R = gas constant
γ = specific heat ratio cp/cv
Equation of state
Besides the compatibility equations (9.19 and 9.20), we need another
equation to form a complete set. The ideal gas equation of state is used:
PV = mRT
Substituting terms one obtains
Pi ⎛
RTL ⎝
⎜Vold +
Δt
2
( ⎞
Qdown, new + Qdown,old − Qup, new − Qup, old ⎟ = )
⎠
⎛
⎜ mold +
Δt
2
( ⎞
m& gas, new + m& gas, old ⎟ )
⎝ ⎠
Eliminating the volumetric flowrate using the mass flowrate obtains
Pi ⎛ 2Vold
⎜⎜
RTL ⎝ Δt
+
1
ρL
(
m& down, new + m& down,old − m& up, new − m& up, old )⎞⎟⎟ =
⎠
⎛ 2mold ⎞
⎜ + m& gas, new + m& gas, old ⎟
⎝ Δt ⎠
Using Equations 9.19 and 9.20 to eliminate mass flowrate yields:
Pi ⎛ 2Vold 1 ⎛ Pi − C M C −P ⎞⎞
⎜ + ⎜⎜ + & down, old − P i − m& up,old ⎟⎟ ⎟ =
m
RTL ⎜⎝ Δt ρ L ⎝ BM BP ⎠⎠
⎟
⎛ 2mold ⎞
⎜ + m& gas, old + m& gas, new ⎟
⎝ Δt ⎠
(9.49)
Now we can use Equations 9.48 to eliminate m
& gas, new as follows:

manual (si) g.doc
Sonic Inflow
The pressure at the junction can be solved directly from Equation 9.49 as
follows:
⎛ 1 1 ⎞
Pi 2 ⎜⎜ + ⎟
⎝ ρ L BM ρ L BP ⎟⎠
⎛ C CP m& d m& u 2V ⎞
+ Pi ⎜⎜ − M − + old − old + old ⎟
⎟
⎝ ρ L BM ρ L BP ρL ρL Δt ⎠
⎛ 2m ⎞
− RTL ⎜ old + m& gasold + m& sonic ⎟ = 0
⎝ Δt ⎠
where m& sonic is the sonic inflow which can be calculated from Equation
9.49a based completely on input data.
Subsonic Inflow
Equation 9.49 is modified using Equation 9.48b and solved iteratively
using Newton-Raphson. Defining Pr = Pi/Po obtains:
⎛ 1 1 ⎞
Pr 2 Po 2 ⎜⎜ + ⎟⎟
⎝ ρ L BM ρ L BP ⎠
⎛ 2V C C m& d m& u ⎞
+ Pr Po ⎜⎜ old − M − P + old − old ⎟
⎟
⎝ Δt ρ L BM ρ L BP ρL ρL ⎠
⎛ 2m ⎞
− RTL ⎜ old + m& gasold ⎟
⎝ Δt ⎠
⎛ 2γR ⎞ 2 γ
+ Cd Ain PoTL ⎜⎜
(γ − 1)T
⎟⎟ Pr (
− Pr (γ +1) γ = 0 )
⎝ o⎠
Defining terms for Newton-Raphson solution yields
8/10/2007 4:10:28 PM Page 419

(
C1 Pr a − Pr b)0.5 + C2 Pr 2 + C3Pr + C4 = F
0.5C1 (Pr a − Pr b ) (aPr a −1 − bPr b −1 )+ 2C2 Pr + C3 =
−0.5 dF
dP r
2
a=
γ
γ +1
b=
γ
2γR
C1 = Cd Ain PoTL
(γ − 1)To
⎛ 1 1 ⎞
C2 = Po 2 ⎜⎜ + ⎟⎟
ρ
⎝ L MB ρ B
L P⎠
⎛ 2V C C m& d m& u ⎞
C3 = Po ⎜⎜ old − M − P + old − old ⎟
⎟
⎛ 2m ⎞
C4 = − RTL ⎜ old + m& gasold ⎟
⎝ Δt ⎠
where the function F is solved for Pr.
Subsonic Outflow
Equation 9.49 is modified using Equation 9.48d and solved iteratively
using Newton-Raphson. Defining Pr = Pi/Po obtains:

manual (si) g.doc
⎛ 1 1 ⎞
Pr 2 Po 2 ⎜⎜ + ⎟⎟
⎛ 2V C CP m& d m& u ⎞
+ Pr Po ⎜⎜ old − M − + old − old ⎟
⎟
⎛ 2m ⎞
− RTL ⎜ old + m& gasold ⎟
⎝ Δt ⎠
+ Pr (C D Aout Po )
γ −1
Pr [
2γRTL − 2 γ
− Pr − (γ +1) γ = 0 ]
Defining terms for Newton-Raphson solution yields
(
C1 Pr − a − Pr − b )0.5 + C2 Pr + C3 + C4 P1 = F
r
(
0.5C1Pr Pr − a − Pr −b )−0.5 (− aPr −a −1 + bPr −b−1 )
1 dF
+ C 2 − C4 =
2 dPr
Pr
2
a=
γ
γ +1
b=
γ
2γRTL
C1 = Cd Aout Po
γ −1
⎛ 1 1 ⎞
C2 = Po 2 ⎜⎜ + ⎟⎟
⎛ 2V C CP m& d m& u ⎞
C3 = Po ⎜⎜ old − M − + old − old ⎟
⎟
⎛ 2m ⎞
C4 = − RTL ⎜ old + m& gasold ⎟
⎝ Δt ⎠
where the function F is solved for Pr.
8/10/2007 4:10:28 PM Page 421

Sonic Outflow
Using Equation 9.48c, the pressure at the junction can be solved directly
from Equation 9.49 as follows:
⎛ 1 1 ⎞
Pi 2 ⎜⎜ + ⎟⎟
⎛ 2V C CP m& d m& u ⎞
+ Pi ⎜⎜ old − M − + old − old ⎟
⎟
⎛ (γ +1) (2(1−γ )) ⎞
⎛ γ −1⎞
+ Pi ⎜ C D Aout γRTL ⎜1 + ⎟ ⎟
⎜ ⎝ 2 ⎠ ⎟
⎝ ⎠
⎛ 2m ⎞
− RTL ⎜ old + m& gasold ⎟ = 0
⎝ Δt ⎠
Lumped inertia pipe with orifice theory

AFT Impulse uses a pipe with lumped inertia (Wylie, et al., 1993, pp.
122-124) as a connector for some elements, although it is not a complete
element by itself. Such an element may also include an orifice (or a
check valve, which will be treated as an orifice) as shown in Figure 9.11.
1
Δz
2
orifice 3
Figure 9.11 Lumped inertia with friction schematic

manual (si) g.doc
Summing forces acting on the lumped section, 2,
dV2
F1 − F3 − F friction − Forifice + Fg = ρA2 L2
dt
or,
dm& 2
F1 − F3 − F friction − Forifice + Fg = L2
dt
Now,
F1 = P1 A2
F3 = P3 A2
fL2 m& 2 2
F friction = A2
D2 2 ρA2 2
m& 2 2
Forifice = A2
2 ρC D 2 Ao 2
Fg = ρA2 gΔz
Substituting terms and dividing by area A2
fL2 1 1 L dm& 2
P1 − P3 − m& 2 2 − & 2 2 − ρgz = 2
m
D2 2 ρA2 2 2 ρC D 2 Ao 2 A2 dt
Taking average values over time, and denoting current time by a prime,
P1′ + P1 P3′ + P3 fL2 1

− − m& 2′ m& 2
2 2 D2 2 ρA2 2
1 L (m& ′ − m& 2 )
− & 2′ m& 2 − ρgz = 2 2
m
2 ρC D 2 Ao 2 A2 Δt
Collecting terms and simplifying
P1′ − P3′ = C1 + C2 m& 2′

where
8/10/2007 4:10:28 PM Page 423

2 L2
C1 = P3 − P1 + 2 ρgz − m& 2 (9.50)
A2 Δt
fL2 1 1 2 L2
C2 = m& 2 + m& 2 + (9.51)
D2 ρA2 2 2 2 A2 Δt
ρC D Ao
If a connector pipe exists but no orifice then
2 L2
C1 = P3 − P1 + 2 ρgz − m& 2
A2 Δt
fL2 1 2 L2
C2 = m& 2 +
D2 ρA2 2 A2 Δt
If an orifice exists but no connector pipe

C1 = P3 − P1
1
C2 = m& 2
ρC D 2 Ao 2
If neither exist then
C1 = 0
C2 = 0
Interpreting cavitation results

Transient cavitation is a highly complex phenomenon. AFT Impulse uses
the Discrete Vapor Cavity (DVC) Model, which is a fairly simple
representation of the phenomenon (Wylie, et al., 1993, pp. 67-69). When
vapor pockets form in a system, the fluid is no longer continuous and
thus the wavespeed decreases. Neglecting the wavespeed change
compromises the results to some degree.
In addition, the DVC Model assumes that the vapor cavities form at the
computing stations, rather than being distributed along a pipe. This also
is an approximation. The DVC Model implicitly assumes that the
cavities are small compared to the volume of the pipe section and that
they do not move during the simulation. If the cavities grow as much as
10% of the computing volume, the results are further compromised. If

manual (si) g.doc
the pipe is at an angle or vertical, the vapor bubbles can move due to
buoyancy. The importance of this can be checked by comparing the
predicted bubble lifetime to the time it would take bubbles to move
between computing stations (assuming a typical bubble velocity).
A result of the DVC Model is that when cavities form and collapse,
numerous non-physical pressure spikes occur which complicates data
interpretation. The interpretation thus requires considerable engineering
judgment.
Here are some general guidelines analysts can follow:
1. The first one or two pressure spikes after cavity collapse tend to be
the most reliable. Fortunately, these spikes are usually the worst and
therefore sufficient engineering data is available to draw a
conclusion.
2. After the first one or two spikes, the non-physical pressure spikes
frequently begin. Sharp spikes that appear to occur for only a single
time step can usually be ignored. Rather, attention should be focused
on the major trends of the pressure results. These tend to be where
more pressure solutions are grouped together, rather than isolated
spikes.
3. Pay attention to the maximum vapor volumes in relation to the
computing volume. If the vapor volume percentage grows to more
than 10% then be more cautious in the interpretation.
4. The Method of Characteristics method without cavitation is a
reliable and accurate solution method. However, after cavitation
occurs the reliability of the predictions decreases significantly. The
reason for offering the DVC Model is that, even with its
deficiencies, it offers significantly better information than would be
obtained if cavitation were to be ignored. Used with the above
cautions, the DVC Model in AFT Impulse can be used successfully
to model many systems that experience cavitation.
Instantaneous waterhammer equation exceptions

The instantaneous waterhammer equation predicts the theoretical
maximum pressure rise in a pipe. It was discussed in Chapter 3, and here
we will discuss cases where pressure surge can exceed that predicted by
the equation.
8/10/2007 4:10:28 PM Page 425

Wylie, et al., (1993, pp. 4) gives the instantaneous waterhammer
equation as:
ΔP = − ρ aΔV (9.52)
where:
ρ = density
a = wavespeed
To apply this equation, one would typically determine the steady-state
velocity and assume the flow stops suddenly. The -ΔV is thus equal to
the steady flow velocity.
Following is a description of cases where the pressure surge can exceed
that predicted by the instantaneous waterhammer equation.
Cavitation
When a positive pressure surge occurs, it is always followed by negative
pressure surge that drops the pressure. If the drop in pressure causes
cavitation, then the pressure spike that occurs when the cavity collapses
can cause the local pressure to exceed that predicted by Equation 9.52.
The potential to cavitate can be assessed by comparing the pressure rise
in Equation 9.52 to the steady-state pressure. If the pressure surge is
greater than the steady-state pressure, then cavitation is possible. If it is
less than the steady-state pressure and the fluid has a low vapor pressure
(like water), then it will probably not cavitate. If the fluid has high vapor
pressure, this will need to be subtracted from the steady-state pressure
and then compared to the pressure surge.
Pipe size changes

When using Equation 9.52, the case considered is usually a valve
closure. Equation 9.52 is frequently applied with the ΔV obtained based
on the pipe diameter at the valve. However, if smaller diameter pipes
exist upstream that carry the same flowrate, their ΔV will be larger
because they have less area. This will result in a larger pressure surge in
the smaller pipe.

manual (si) g.doc
To be safe, Equation 9.52 needs to be applied to the pipe with the largest
ΔV potential in the entire system.
Multiple pressure surge sources

If there is potential for multiple elements to generate pressure surge at
the same time, the surge pressures can interact and add together, thus
exceeding the individual surge pressures.
Line pack due to friction

In systems with high friction, pressure can increase above the
instantaneous values because of something called line pack. Even
systems with modest friction can experience this. In brief, the friction in
the system prevents the flow from coming to a complete stop in cases
such as a closing valve. With the closed valve, the fluid continues to
move towards the closed valve, albeit at a lower velocity, increasing
pressures above the instantaneous prediction. See Wylie (1993) for more
information.
Elevation changes
If the pressure surge propagates to an area of the system at a lower
elevation, the peak pressure can be higher because of the higher static
pressure.
8/10/2007 4:10:28 PM Page 427


manual (si) g.doc
CHAPTER 10
Modeling Time and Event Based

Transients
In general, transients can be initiated in one of two ways. They can be

initiated according to a prescribed time schedule or according to certain
events that happen in the system. These two categories are referred to as
time-based and event-based transients.
Event-based transients break down further into three categories: Single
event, cyclic dual events and sequential dual events. The behavior of
time-based and the three event-based transients are described in this
chapter.
Time-based transients
Time-based transients occur in accordance with the time of the transient
simulation. The start and stop times of the simulation are specified in the
Transient Control window. The initiation of time-based transients are
pre-specified before a model is run.
For example, consider a valve closure transient. Assume the valve starts
to close at two seconds into the simulation, and the time it takes for the
valve to close is one second. After that, the valve stays closed. If the
initial valve Cv is 250, the transient would appear as in Figure 10.1.
In the "Initiation of Transient" area there are four options. For time-
based transients, the option is specified as Time. In the Transient Data
area the data is entered. Here the Cv profile of the valve will follow the
profile entered and start to close at 2 seconds.

manual (si) g.doc
Figure 10.1 Time-based transient specifies the transient behavior

before simulation is run.
Event-based transients
Event transients are initiated when some user specified criteria is met.
For example, a valve can start to close when a certain pressure is reached
at a point in the system. If the pressure is never reached, the transient is
never initiated.
Event transients are one of three types. The three types will be discussed
in the following sections.
Single event transients

To specify a Single Event transient, select Single Event in the Initiation
of Transient area of the junction window (see Figure 10.2). An "Event"
tab will appear where the event criteria is entered. In Figure 10.2, the

manual (si) g.doc
Chapter 10 Modeling Time and Event Based Transients 431
criteria is such that the transient will be initiated when the outlet static
pressure in Pipe 2 exceeds 1000 kPa. If the pressure at this location
never reaches 1000 kPa, the transient specified in the Transient Data
area will not occur.
Figure 10.2 Single event transient specifies how a junction will

respond if and when some condition is satisfied.
The data in the Transient Data table has a slightly different meaning than
does a time-based transient. Here time zero is relative to the time at
which the event criteria is first met. For example, if Pipe 2 reaches 1000
kPa at 3.65 seconds, the valve will start its transient at 3.65 seconds. The
valve will close (i.e., Cv = 0) at 1 second after the event initiation, or
4.65 seconds.
Dual event transients: cyclic and sequential

Dual event transients contain two different transients for the junction.
8/10/2007 4:10:28 PM Page 431

Cyclic events
Cyclic dual events can repeatedly switch from one to the other based on
the behavior of the model. For example, these could be high and low
pressures at some location. When the low pressure setting is reached a
valve transient is initiated which closes the valve. The pressure may then
build up to a point where the high pressure setting is reached. In this
case a second valve transient is initiated which opens the valve. Should
the pressure drop to the low pressure setting again, the valve will again
close. The two settings will continue to cycle the valve on and off until
the end of the simulation.
Cyclic dual events are specified by selecting Dual Event Cyclic in the
Initiation of Transient area of the junction window (Figure 10.3). When
selected, two event tabs appear. Here the first and second event can be
specified. In this case the first event is a pressure at Pipe 2 inlet less than
500 kPa. The transient that is initiated when this event occurs is in the
table of the First Transient tab in Figure 10.3. Here the valve closes one
second after the event.
A second event, not shown, is specified on the Second Event tab, with
the accompanying transient specified on the Second Transient tab, also
not shown. Typically, the first data of the second transient should match
the final data of the first transient. In other words, if the valve is closed
as in Figure 10.3 by setting the Cv value to 0, the first data point of the
second transient should be zero and then proceed to some higher value
which represents the opening of the valve.
More often than not the two events will specify the same parameter at
the same location. For example, if the first event specified a pressure at a
certain pipe, the second event would also specify pressure at the same
pipe location. However, this is not required. You can specify the first
event as pressure and the second as flowrate, for example, or you could
specify the two events to be at different locations in the pipe system.

manual (si) g.doc
Figure 10.3 Dual Event Cyclic selection for transient initiation
Sequential events
Sequential dual events progress from the first event to the second and no
further. For example, these could be two high pressure settings. When
the first setting is reached the valve opens partially, and when the second
setting is reached, which would likely be at a higher pressure than the
first event, the valve opens the rest of the way. Similar to a single event
transient, if the pressure drops and then rises again, the events would not
be initiated again.
Sequential dual events are specified by selecting Dual Event Sequential
in the Initiation of Transient area of the junction window (Figure 10.3).
Similar to Cyclic Dual Events, two event tabs appear where the first and
second event can be specified. The transients that are initiated when the
events occur are specified in the table of the First Transient and Second
Transient tab.
8/10/2007 4:10:28 PM Page 433

Similar to Cyclic Dual Events, the two events do not have to be for the
same parameter (e.g., pressure) or at the same location.
Junctions with inherent event logic

There are three junction types in AFT Impulse which have built-in, or
inherent, event logic. These are the check valve, relief valve and vacuum
breaker valve. The user does not need to specify the nature of the events,
and in fact is not allowed to.
The inherent event logic is very similar to the Dual Event Cyclic logic
described previously.
Check valve
The check valve has two built-in events. The first is that it closes when
backflow starts to occur. The second is that it reopens when sufficient
pressure differential occurs. If the user does not enter a closing or
opening transient in the check valve window, these transients are
assumed to be instantaneous.
Relief valve
The relief valve has two built-in events. The first is that it opens when
the cracking pressure is reached. The second is that it closes again when
the pressure falls back below the cracking pressure. If the user does not
enter an opening or closing transient in the relief valve window, these
transients are assumed to be instantaneous.
Vacuum breaker valve

The vacuum breaker valve is very similar to the relief valve, and has two
built-in events. The first is that it opens when the set pressure is reached.
The second is that it closes again when the pressure rises above the set
pressure and all gas is expelled.

manual (si) g.doc
Thought experiment to further clarify event transients

To further clarify the capabilities of event transient modeling, here we
will discuss how one could use a regular AFT Impulse Valve junction to
simulate a check valve.
As discussed previously, check valves have inherent event logic. That
logic assumes that the valve closing transient is initiated when backflow
begins. Once a certain pressure differential exists across the closed
check valve, the valve reopens.
One could use a regular valve junction to simulate a check valve as
follows. First, the valve would need to be specified as Dual Event Cyclic
in the Initiation of Transient area. The first event would be a flowrate
that is less than or equal to zero at the inlet of the valve's discharge pipe.
A Cv transient could then be entered to simulate how the check valve
closes.
The second event would be a differential pressure across the check valve
that is greater than or equal to the check valve's reopening pressure
differential. A reopening Cv profile could be entered in the second
transient data table.
With the above considerations, the valve junction and check valve
junction would respond identically.
Other transient features
Absolute vs. relative transient data

In the Transient Data area (see Figure 10.1, 10.2 or 10.3) it can be seen
that there are two options: "Absolute Values" and "Relative To Steady-
State Values". In the case of a valve junction, the first option allows you
to put in actual Cv values (for example) vs. time. The second option
allows you to put in a percentage of the steady-state Cv value. While not
shown here, the steady-state value is entered on the Loss Model tab.
For junctions other than valves, different parameters are specified as the
transient parameter. For example, the transient could be the steady-state
pump speed, pump flowrate, spray discharge area, or boundary pressure.
8/10/2007 4:10:28 PM Page 435

Repeat transient
For transients that are periodic, you can specify that the transient repeats
itself. This is specified by selecting the "Repeat Transient" checkbox
(see Figure 10.1, 10.2 or 10.3). When Repeat Transient is selected, the
first and last data points must match. After the transient reaches the final
data point, it returns to the first data point and starts over.
Add time offset

There may be times you want to shift the time data points in the transient
data table. For instance, assume you have a valve specified as a time-
based transient where the valve starts to close after five seconds. Further
assume that you wish to change the transient so that it starts to close
after seven seconds. In such a case, you want to shift all time values by
two seconds. If you select the Edit Table button (see Figure 10.1, 10.2 or
10.3), there is menu choice for Add Time Offset. If you select this, you
are prompted to enter an offset time, and all times in the table will shift
by the specified time.
Graphing the transient data

It is helpful to see the transient data in graph form, and this is easily
done by selecting the Show Graph button (see Figure 10.1, 10.2 or 10.3).
Event messages
Events that are initiated during the simulation are displayed in the two
Event Messages area of the Output window (see Figure 10.4). The first
is a list of event messages sorted by junction. The second is a list of
event messages sorted by time. If no events occur, the Event Messages
tab will be hidden.

manual (si) g.doc
Figure 10.4 The Event Messages area shows a listing of all events
that occur during the simulation sorted by junction
and time.
Transient indicators on the Workspace

To help the user quickly see which junctions are specified with transient
data, a "T" is prepended to the junction number on the Workspace.
Figure 10.5 shows an example where it is seen the junction J7 has
transient data.
8/10/2007 4:10:28 PM Page 437

Figure 10.5 Junctions that have transient are shown with an

adjacent "T" on the Workspace as J7 is shown here.
Transient data in Model Data

All transient junction data can be viewed in the Model Data window
junction area on the Transient Data tab (see Figure 10.6). In addition, if
events are specified the event data will also be displayed.

manual (si) g.doc
Figure 10.6 All junction transient data is shown in the Model Data
window on the Transient Data tab.
8/10/2007 4:10:28 PM Page 439


manual (si) g.doc
CHAPTER 11
Reducing Surge Pressures
This chapter provides information on different options for reducing

surge pressures in piping systems.
General considerations
Surge pressures that are excessively high can burst pipes and damage
equipment. Surge pressures that reduce the local pressure can result in
pipes being crushed (due to atmospheric pressure exceeding the internal
liquid pressure), cavitation and liquid column separation (which can then
result in large pressure spikes when the cavity collapses), and
subatmospheric pressures that are unacceptable for drinking water
pipelines.
The surge pressures result from changing conditions in the system,
which change the fluid velocity. Options include slowing down the
system changes and thus reducing velocity or providing locations which
can deliver extra fluid when the pressure decreases or accept extra fluid
when the pressure rises.
If using a surge suppression device, it is generally desirable to locate the
device as close as possible to the cause of the transient. The practicality
of the different options depend highly on the particular case of interest.
Methods for surge reduction are discussed in detail in Swaffield et al.
(1993, Chapter 6), Wylie et al. (1993, Chapter 10), and Chaudhry (1987,
Chapter 10).

manual (si) g.doc
Larger pipe sizes

If at the design stage, one can consider using larger pipes in the area of
concern. This will reduce fluid velocity and hence pressure surge.
Flexible hose
Flexible hose can sometimes be used in small systems, which reduces
wavespeed and hence pressure surge.
Slower or modified valve closure profiles

If the pressure surge is due to valve closure, the magnitude of the surge
can frequently be reduced by closing the valve more slowly. Another
option is to close the valve in a different manner. Swaffield et al. (1993,
pp. 204-206) discuss closing the valve more quickly at first and then
more slowly at the end. Examples are given where the first 80% of
closure occurs over the first 20% of closure time, with the final 20%
closure occurring over the remaining 80% of the time.
Parallel valves
Rather than use one large valve to stop the flow, perhaps two or more
valves can be used in parallel. The timing of the valve closures is
staggered, thus allowing a portion of the flow to be stopped with each
valve.
Relief valves
Relief valves can be used to avoid high pressure situations. In addition,
inward relief valves can be used for low pressure situations.

manual (si) g.doc
Chapter 11 Reducing Surge Pressures 443
Vacuum breaker valves for low pressures

Vacuum breaker valves, also called air inlet valves, open to permit air
into the system during low pressure conditions. These are usually
located at high elevation points along the pipeline.
Surge tank
Surge tanks can be an effective way to reduce surge pressures if the
steady-state system pressures are relatively low. High steady-state
pressures result in high hydraulic gradelines. Since the liquid level in a
surge tank rises to the local hydraulic gradeline, they would have to be
very tall if located in a high pressure region.
Gas accumulator
Gas accumulators can be effective in spreading out the wave front of a
transient and thus reducing peak pressures. Gas accumulators change the
frequency response of the system and can in fact amplify pressure surge
if not sized and located properly.
8/10/2007 4:10:28 PM Page 443


manual (si) g.doc
CHAPTER 12
Special Topics: Including

Troubleshooting Models
This chapter discusses techniques that can be implemented to generate

models using AFT Impulse, and provides direction when you are having
problems obtaining convergence in your model.
The philosophy of computer modeling

Some might say that a “good” computer model is one that accurately
predicts the response of a system; but this is only partly true. The
practicing engineer generally must weigh the accuracy of the model
against the risk, schedule, and/or budget involved. Through experience,
an engineer learns where to focus resources and when an answer is good
enough to base important decisions on.
Although AFT Impulse is designed to speed up the modeling process, it
should never be used as a “black box.” Your model input and output
should always be carefully reviewed.
To make the most effective use of AFT Impulse, you should follow these
guidelines:
1. Be aware of what AFT Impulse can and cannot do and the
assumptions it makes.
2. Learn and use the tools in AFT Impulse that are appropriate for the
specific analysis you want to perform. Detailed and accurate answers

manual (si) g.doc
demand detailed and accurate input. Quick and dirty answers require
less detailed input.
3. Learn and use the tools in AFT Impulse that allow you to process
information in the formats that mean the most to you. Because pipe
flow applications vary so much within different industries, AFT
Impulse is customizable for the way you want to do your pipe flow
work. Understanding the tools will help you gain greater conceptual
control over your pipe flow analyses and reduce modeling errors.
4. Check your input data. Use the features available in AFT Impulse to
review your results for input errors. Pay attention to the information
in the output printout. Don't neglect warnings in the output. Perform
hand analysis checks of results.
5. Bound your problem. Instead of trying to calculate the exact
resistance in a network, or the right inertia for a pump, use your
judgment, guided by data, on how much variation might occur. By
running your model at the upper and lower limits, you can obtain
answers, for example, on the maximum and minimum possible
pressure surge in the system. If the results at the maximum and
minimum are acceptable, your analysis is probably finished. If they
are not acceptable, you can further refine the input model.
Troubleshooting steady-state models

What should you do if your model will not run for steady-state, or does
run and gives strange results? There are numerous possible causes of
such problems. This section identifies common problems and offers
strategies you can use to troubleshoot models yourself.
Use the Model Data window

Besides its usefulness for documentation purposes, the Model Data
window is a powerful tool for finding input errors in your model.
Look at the model in Figure 12.1. If you suspected an input error in one
of the pipes, the task of looking for the error in each individual pipe
input window would be tedious.
Now look at the Model Data window for the pipes in Figure 12.2. Here it
is much easier to scan down columns of data looking for input errors. In

manual (si) g.doc
Chapter 12 Special Topics: Including Troubleshooting Models 447
this case, one of the pipe lengths was accidentally entered in miles rather
than feet.
Figure 12.1 Workspace layout of a pipe system with an input error

in one of the pipes (error not shown).
Poor pump curve fits

The Pump Specification window allows curve fits of pump data up to
fourth order. Depending on the number of data points entered, pump
curve fits that look like the one in Figure 12.3a are possible.
AFT Impulse interprets flow rates past the inversion as acceptable
operating points, which can lead to strange results or non-convergence.
You should always take care that your pump curves cross the flow rate
axis as shown in Figure 12.3b. If you attempt to generate a curve fit like
that shown in Figure 12.3b, AFT Impulse will warn you. In addition, it
will warn you when you try to run the model.
8/10/2007 4:10:28 PM Page 447

Figure 12.2 Model Data window display for the model in Figure
12.1. From this perspective it is much easier to see
that one of the pipes (#4) has length data entered in
miles rather than feet. This is not automatically an
input error, as AFT Impulse accepts pipe input with
mixed engineering units. However, more frequently
than not the user intends to use consistent units, so
this probably is an error.
Use the Output window Sort feature

The Sort Output feature in the Output window is discussed in Chapter 5.
If a model runs but the results look incorrect, use the Sort feature in the
Output Window to look for extremes in velocity, pressure drop or other
parameters. An input error (like incorrect diameter) may be easier to see
by looking at its effect on the results.

manual (si) g.doc
12
10
Pressure (psid) 6
a
4
0
0 50 100 150 200
Flow (ft3/s ec)
12
10
8
Pressure (psid)
b 6
0
0 50 100 150 200
Flow (ft3/s ec)
Figure 12.3 Top graph shows a third or fourth order pump curve
fit with an inversion at 150 ft3/sec. AFT Impulse
interprets flow rates to the right of this inversion as
acceptable which can cause convergence problems or
erroneous results. All pump curve fits should include
sufficient data such that the curve crosses the zero
pressure rise axis as shown in the bottom curve.
8/10/2007 4:10:28 PM Page 449

Tee/Wye junction complexity

As discussed in Chapter 8, tee/wye junctions contain powerful and
sophisticated models for determining the pressure losses at flow splits
and converging flow streams. This power comes at the cost of additional
model complexity. The complexity can lead to convergence problems in
models with numerous tees. Here are a couple of recommendations.
Tee/Wye junctions have two loss options: Simple and Detailed. The
Simple loss model ignores losses at the tee. The Detailed model uses the
correlation summarized in Chapter 8. The potential convergence
problems occur when using the Detailed loss model.
The potential problem is not with the Detailed tee model itself. Rather, it
is due to multiple tees coupling together in such a way that AFT Impulse
cannot converge on all of the flow splits.
You can quickly and easily check if a convergence problem is being
caused by tees. Using the Global Junction Edit window, you can change
all tees to the Simple loss model and rerun the model. If it now
converges, then the problem was due to the use of Detailed tee modeling.
Experience has shown that this problem can frequently be resolved by
using absolute tolerance criteria in the Steady Solution Control window
rather than relative tolerance. Tolerance criteria is discussed at length in
Chapter 8.
You also have the option of just using the Simple (lossless) tee model
for all tees. This is how tees have been traditionally handled, so lossless
tees do have some utility in engineering calculations.
Lower the flow rate relaxation

The Steady Solution Control window allows you to change the
numerical relaxation parameter for both flow rate and pressure. A full
explanation of relaxation is given in Chapter 8. It is recommended that
you use only 0.5 or 1.0 for pressure relaxation. But lower values of flow
rate relaxation can be very helpful.
A good first step is to lower the flow rate relaxation to 0.05 or 0.1. By
lowering the flow rate relaxation you improve AFT Impulse's ability to
obtain convergence. The cost is slower run times. In some cases,
accepting the slower run times that come with flow rate relaxation is the
only means of obtaining convergence. You can lower the relaxation

manual (si) g.doc
values as low as you want, but it is recommended it not be lowered
below 0.01.
Try absolute tolerance

Most models converge more reliably with relative tolerances, which is
the default in AFT Impulse’s Steady Solution Control window. But some
models converge much better with absolute tolerance. What this usually
means is that there is at least one element of the model that is close to
the real answer in absolute terms but does not lock in on a relative
(percentage) basis.
If a model will not converge with relative tolerance, you can try
changing to absolute tolerance. Even better, try the option to converge
on either absolute or relative, which will offer the most flexibility.
An explanation of absolute and relative tolerance is given in Chapter 8.
Make initial flow rate guesses for pipes

If you do not provide any initial guesses, AFT Impulse will make its own
guesses. On occasion, the AFT Impulse guesses are far away from
reality. For some models, providing a crude initial guess for all the pipe
flow rates in the system can help the model converge.
The crude guesses can be as simple as selecting all the pipes and
initializing all of them to 1 kg/sec, even though some operate at 0.1
kg/sec and others at 10 kg/sec.
Review the Iteration History

The Iteration History window is discussed in Chapter 8. You can use this
to see which pipes or junctions are farthest out of tolerance, which may
direct your troubleshooting efforts toward these elements.
Change boundary conditions

Changing the boundary pressures or flow rates can sometimes give you
insight into what is giving the model trouble.
8/10/2007 4:10:28 PM Page 451

Turn off parts of the model

Using the Special Conditions feature, you can turn off elements in your
model and, in effect, shut down entire sections. Then you can focus on
getting selected sections to run to convergence.
Break a large model into submodels

AFT Impulse provides convenient features for deleting parts of your
model and merging parts back in. A section of your model can be
quickly deleted by selecting it on the Workspace and choosing Delete
from the Edit menu. If the section is instead Cut, it can be Pasted into a
new model of its own for testing or holding. At a later time, the removed
sections can be added back into the current model by using the Merge
feature on the File menu.
The advantage of breaking your model into sub-models is that you can
investigate the convergence characteristics of each part of your model
independently. If you can get each sub-model to converge, you should be
able to get the original model to converge also. If one of your sub-
models will not converge, it is doubtful your original model will
converge either.
Control valve failure issues in steady-state

AFT Impulse offers modeling of PRV’s, PSV’s and FCV’s. Based on the
interaction with the pipe system, control valves can end up in a situation
where they cannot control to the desired control points. When this
occurs, it is an indication that the desired control point cannot be
obtained unless the control valve acts like a pump and adds pressure to
the system. To diagnose which control valves (if any) need to add
pressure to reach their control point, set their failure state to Always
Control (Never Fail).
It is very important to grasp that when a valve fails to control and the
failure action is set to Always Control (Never Fail), the resulting
solution in the Output is only useful for diagnostics and does not
represent a physically meaningful answer. Why is that? Because in order
to maintain the control pressure or flow the valve had to add pressure. If
a control valve is operating in this mode, you will be clearly notified in
the Output window, both in the Warnings and in the Valve Summary.

manual (si) g.doc
The following failure options are available for control valves:
1. Maintain control by adding pressure (useful for diagnostics only)
2. Fail open or fail closed if upstream pressure is insufficient
3. Fail open or fail closed if downstream pressure is excessive
(applicable to pressure control valves only)
In such cases where failure to control exists, one or more valves may fail
simultaneously. Once failed, AFT Impulse will not return a valve to its
control capability.
Infinite pipe junctions

An infinite pipe junction is a special junction that accepts pressure
waves but never reflects any waves (Wylie, et al., 1993, pp. 121-122).
This junction is useful in modeling the part of a pipe system which is so
long that its communication time results in no interaction with the rest of
the system.
This feature is offered as part of the Assigned Flow and Assigned
Pressure junctions. During steady-state a fixed flow or pressure exists at
the junction, but during transient flow both the pressure and flow change
in a manner consistent with the fact that no waves are reflected. The
theory for an infinite pipe is given in Chapter 9.
Judicious use of infinite pipe junctions can significantly reduce model
run time. The reason is that if there is a pipe which is so long it will not
reflect any waves during the simulation, the extra computation for that
pipe can take a lot of time. The infinite pipe junction is a way of
avoiding any computations for the pipe but still maintaining its effect on
the system.
Application example
A real life example is shown in Figure 12.4. The system represents part
of a tank farm for refined petroleum products. Gasoline flows from a
refinery some 100 km away into the tank farm. The supply pipeline is
represented by pipe P1. Pipes P2-P9 are relatively short local pipes at the
tank farm with changing diameters. This model is for gasoline, and the
gasoline comes into junction J10 and flows initially to the tank at J106.
The transient of interest is when the J201 valve is opened and the J104
8/10/2007 4:10:28 PM Page 453

valve is closed. This will divert the incoming gasoline from J106 to
J212. The total scale of the tank farm piping is less than a mile.
Figure 12.4 Example of where an infinite pipe junction can be

used. Here J1 is specified as an infinite pipe. Use of an
infinite pipe reduces the model run time by a factor of
100.
The transient lasts for about 30 seconds. During the opening and closing
of the valves, the transient waves propagate out to J106 and J212, as
well as to the supply junction J1. Using the wavespeed of gasoline it can
be determined that any waves that propagate to pipe P1 will not reflect
back to the tank farm within 30 seconds, and thus will not impact the
transient. However, during the transient procedure, gasoline still flows
from junction J1 into the system.

manual (si) g.doc
One approach is to model pipe P1 as the full 100 km length. Thus all the
pipe stations along 100 km must be calculated for the entire 30 second
simulation. A second approach is to specify pipe P1 as very short (13
feet long in this case, chosen for modeling convenience based on other
considerations) and to specify J1 as an infinite pipe junction.
Both approaches will give the same transient results. However, the first
approach would take over 100 times as long to run (a 1.9 hour run vs. a 2
minute run when using an infinite pipe). Again, the difference in runtime
is that with an infinite pipe, no computation is required along the 100 km
pipe except for two stations at the entrance.
How pressure junctions work in steady-state

Pressure junctions (e.g., Reservoirs or Assigned Pressures) in AFT
Impulse are an infinite source or sink of fluid. This means that they can
draw or discharge as much fluid as is necessary to maintain the specified
pressure. Some engineers have difficulty grasping this concept, and
misuse pressure junctions in AFT Impulse. This section offers a physical
example of how pressure junctions work to clarify for those having
difficulty with the concept.
Consider the physical pumping system shown in Figure 12.5. This
system pumps water from a tank near the shore of Lake Michigan to a
tank on a nearby hill.
An engineer is tasked with sizing the pump for this system. The engineer
has an idea of the discharge head needed for the pump, and so builds an
AFT Impulse model as shown in Figure 12.6.
8/10/2007 4:10:28 PM Page 455

Discharge
Tank (J2)
Transfer
Pipeline
Transfer
Supply Pump (J3)
Tank (J1)
Shoreline
Elevation
Lake Michigan
Figure 12.5 Physical pumping system to transfer water from

shoreline supply tank (J1) to discharge tank (J2) at top
of hill
J2
P3
J1 J3 J4
P1 P2
Figure 12.6 Model to size pump attempting to represent system in

Figure 12.5, but actually represents the system in
Figure 12.7.
The Figure 12.6 model is attempting to represent the physical system in

Figure 12.5, but it in fact represents the physical system in Figure 12.7.

manual (si) g.doc
Transfer Discharge
Pipeline Tank (J2)
Transfer
Supply Pump (J3)
Tank (J1)
Shoreline
Elevation
Lake Michigan (J4)
Figure 12.7 Physical system represented by model in Figure 12.6.
Why? Because the reservoir junction (J4), which is an infinite source of

fluid, has been located between the pump (J3) and the discharge tank
(J2). The J4 reservoir isolates the supply tank and pump (J1 and J3) from
the discharge tank (J2). The J4 reservoir is very much like placing Lake
Michigan between the supply and discharge piping.
If the engineer changes the surface elevation in J4, it is like changing the
surface elevation of Lake Michigan. It will change the flow rate in the
pipes, but no matter how high the water level of Lake Michigan, it does
not change the fact that the pumping system is isolated from the
discharge tank. The flow rate in the piping from J4 to J2 is entirely
dependent on the difference in surface elevations between J2 and J4, and
is not influenced by the J3 pump in any way.
How then should the physical system in Figure 12.5 be modeled? It
should be modeled as shown in Figure 12.8.
8/10/2007 4:10:28 PM Page 457

J2
P2
J1 J3
P1
Figure 12.8 Proper model of physical system shown in Figure

12.5.
If the goal is to size the pump at J3, the model is Figure 12.8 can be used
with the pump modeled as an assigned flow pump.
The role of pressure junctions in steady-state

In modeling a generalized pipe network, it is possible to construct
models that do not have a unique solution. A common occurrence of this
is when a model contains one or more sections completely bounded by
known flow rates. A series of examples with accompanying explanation
will clarify situations where this occurs. The length of this topic is
proportional to the frequency with which engineers get confused over
the issue. It became apparent that an exhaustive discussion would be
very helpful for some engineers. Do not dismiss this topic too quickly
because of its length.
Before beginning, it will be helpful to consider an aspect of the
philosophy of computer modeling. At the risk of stating the obvious, it
must be recognized by the user that a computer model cannot calculate
anything that cannot, in principle, be calculated by hand. All the
computer does is accelerate the calculation process. Some users expect
the model to generate independent information and become frustrated
when the software requests additional information from them. But if the
user was doing the calculation by hand that same information would still
be required. The difference is that in the case of the hand calculation, the
engineer would be forced to think through why the information is
needed.

manual (si) g.doc
However, when the engineer interacts with AFT Impulse, he/she does
not need to think at the same depth as in the hand calculation. So when
AFT Impulse asks for the additional information, it is not as apparent
why it is required. This disconnect in thinking can cause frustration for
the user. Fully understanding the concepts in this topic will greatly
reduce the frustration for some users, and will improve the quality of
models for all users.
Examples
The simplest example is the system shown in Figure 12.9. In AFT
Impulse terms, the system has two assigned flow junctions.
L = 30 m
D = 5 cm
Inlet f = 0.02 Outlet
Q = 0.05 m3/s Q = 0.05 m3/s
El = 0 m El = 0 m
Figure 12.9 Model with two assigned flows. This model does not
have a unique solution.
Obviously, the flow in the pipe is known, but, what is the pressure at the
inlet? At the outlet? It cannot be determined because there is no
reference pressure. The reference pressure is that pressure from which
other pressures in the system are derived. There can be one or more
reference pressures, but there always has to be at least one.
The model in Figure 12.9 can be built with AFT Impulse. If you try to
run this model, AFT Impulse will inform you that it cannot run it
because of the lack of a reference pressure. In AFT Impulse there are
numerous junctions that can act as a reference pressure: the reservoir,
assigned pressure, spray discharge, pump with submerged option, and
valve and relief valve after cracking (with the exit valve option) behave
as reference pressures because that is their nature. Three other junction
types, surge tank, gas accumulator, and assigned flow (when the steady-
state flow is zero) have a special feature that allows them to act as
reference pressures. This feature is found on each junction's Optional
tab, and is discussed at length in Chapter 6.
There are a couple other things worth noting about Figure 12.9. First is
that the model in Figure 12.9 has redundant boundary conditions. When
8/10/2007 4:10:28 PM Page 459

one flow is specified in a single pipe, the other end must have the same
flow. Thus the second known flow does not offer any new information.
Because the conditions are redundant, there is no unique solution to the
model in Figure 12.9. In any AFT Impulse model, there is always at least
one unknown for each junction in the model – either pressure or flow
rate. Sometimes there are two unknowns such as with a pump. It is
possible the user may not know either the flow or the pressure at the
pump. In such cases, the user is required to specify the relationship
between flow and pressure. This relationship is called a pump curve, and
by specifying the relationship between the two, effectively one of the
two unknowns can be eliminated. Thus we always end up with one
unknown for each junction.
L = 30 m
Inlet D = 5 cm Outlet
a Q = 0.05 m3/s f = 0.02 P = 0 MPa(g)
El = 0 m El = 1.5 m
L = 30 m
D = 5 cm
Inlet Outlet
f = 0.02
b P = 0.1 MPa Q = 0.05 m3/s
El = 1.5 m El = 0 m
L = 30 m
Inlet D = 5 cm Outlet
c P = 0.1 MPa f = 0.02 P = 0 MPa(g)
El = 1.5 m El = 1.5 m
Figure 12.10a-c Top two models with one pressure and one
assigned flow, bottom model with two pressures.
All of these models have a unique solution.
Second, if the user was allowed to specify two flows as in Figure 12.9,
he/she could specify them with different flow rates. Clearly, they must
have the same flow or an inconsistency occurs. The basic reason the
inconsistency would be possible is because Figure 12.9 does not have a
unique solution.
Figure 12.10a-c shows three other model configuration possibilities. It is
not possible to specify inconsistent conditions for any of the Figure

manual (si) g.doc
12.10 models, and they always have a unique solution no matter what
input is specified by the user.
The four models in Figures 12.9 and 12.10a-c contain all logical
possibilities. In all four cases there are four things we want to know: the
pressure and flow at the inlet, and the pressure and flow in the outlet. In
each case we know two of these four. The problem is that in the Figure
12.9 model the lack of a reference pressure makes it impossible to
determine the inlet and outlet pressures, even though we know the flow.
The three models in Figure 12.10a-c all have at least one pressure, thus
all four of the desired parameters can be determined. Table 12.1
summarizes this. Also note in Table 12.1 that in the Figure 12.10c case
the flow can be determined, but it is by iteration.
Table 12.1 Four logical possibilities of boundary conditions for

single pipe systems
CASE Inlet Inlet Outlet Outlet Flow OK ? Comments
Pressure Flow Pressure
Fig. 12.9 X X No No reference pressure, cannot
calculate pressures
Fig. 12.10a X X Yes 1 known flow, can calculate
everything directly
Fig. 12.10b X X Yes 1 known flow, can calculate
everything directly
Fig. 12.10c X X Yes 2 pressures, can iterate for flow
As we consider multi-pipe systems, there are a host of other possibilities

that present themselves. All other configuration possibilities which lack
a reference pressure ultimately boil down to the same problem that
exists in Figure 12.9.
The pipe lengths, diameters and friction factors are not included in the
remaining models because they do not influence the main point of this
topic. It is assumed that these parameters can be obtained for each pipe
and the resulting pressure drop calculated with standard relationships.
Consider the model in Figure 12.11. With three boundaries having a
known flow, this clearly has the same problem as the model in Figure
12.9. No unique pressures at any location can be calculated.
8/10/2007 4:10:28 PM Page 461

Outlet
Q = 0.025 m3/s
El = 0 m
Inlet
Q = 0.05 m3/s Outlet
El = 0 m Q = 0.025 m3/s
El = 0 m
Figure 12.11 Three known flows at boundaries that lack a reference

pressure, similar to the model in Figure 12.9. This
model does not have a unique solution.
A pressure at any one of the three boundaries in Figure 12.11 would be

sufficient to allow a unique solution of the system. It is also possible that
there could be two pressures and one flow, or three pressures. As long as
there is at least one pressure, a unique solution exists.
Now consider the models in Figure 12.12. The Figure 12.12a model does
have one reference pressure, but in this case there is still a problem. The
Flow Control Valve (FCV) is controlling the flow to 0.05 m3/s, and the
downstream flow is demanding 0.05 m3/s. The section of the system
preceding the FCV can be solved (because there is a pressure upstream),
but the section after the FCV does not have a unique solution. Why? In
order to obtain a unique solution, the pressure drop across the FCV must
be known. But any pressure drop could exist and satisfy the conditions
of this model. It is thus not possible to determine the pressure at the
outlet flow demand because it depends on the FCV pressure drop which
is not known. The model in 12.12b does have a pressure downstream of
the FCV, and thus there is a unique pressure drop across the FCV and a
unique solution exists for the model in Figure 12.12b.
If you input the remaining data for the Figure 12.12a model and run it in
AFT Impulse, you will get the message shown in Figure 12.13. AFT
Impulse identifies the part(s) of the model where a known pressure is
needed.

manual (si) g.doc
J1 J2 J3
Inlet Outlet
a P = 0.1 MPa(g) Q = 0.05 m3/s
El = 3 m El = 0 m
FCV
Q = 0.05 m3/s
J1 J2 J3
Inlet Outlet
b P = 0.1 MPa(g) P = 0 MPa(g)
El = 1.5 m El = 1.5 m
FCV
Q = 0.05 m3/s
Figure 12.12a-b The model at the top does not have a reference
pressure after the Flow Control Valve so the
pressure drop across the FCV cannot be
determined. The top model does not have a unique
solution. The bottom model has a pressure
upstream and downstream, and a unique solution
exists.
8/10/2007 4:10:28 PM Page 463

Figure 12.13 AFT Impulse message when you try to run the model
shown in Figure 12.12a.
Sizing pumps with flow control valves

An analogous situation to Figure 12.12a is when the user is trying to size
a pump using the Pump with an assigned flow. The Pump with an
assigned flow is really just a reverse FCV. It controls the flow by adding
pressure, rather than by reducing it. Figure 12.14 shows this case and is
another example of a model without a unique solution, requiring more
than one pressure junction. If the user instead entered a pump curve for
this model, it could be solved.

manual (si) g.doc
Inlet Outlet
P = 0.1 MPa(g) Q = 0.05 m3/s
El = 1.5 m El = 0 m
Pump with Known (fixed) Flow
Q = 0.05 m3/s
Figure 12.14 A pump modeled as an assigned flow for sizing

behaves identically to the FCV model in Figure 12.12a,
for which no unique solution exists.
Increasing in model complexity, Figure 12.15 shows some additional

examples of models without a unique solution. In particular, Figure
12.15b has a pump with an assigned flow that can add as much pressure
as it wants, and the downstream FCV can then take out as much as it
wants. There are an infinite number of possible solutions.
While it would be highly unusual for a system to have two FCV’s in
series, it is worth considering the model in Figure 12.15b further. It is
common to have an FCV in series with a pump. How does one size the
pump in such a case? There are several ways to do this, one of which is
to change the pump from an assigned flow to an actual pump curve.
Trying multiple pump curves will guide you to the best pump.
But there is a better way. Consider for a moment the pressure drop
across the FCV in Figure 12.15b. In the installed system, is any pressure
drop acceptable? Usually not. It is typical to have pressure drop limits
based on the system design and the FCV itself. A common requirement
is a minimum pressure drop across the FCV. If no minimum pressure
drop exists, let’s say we choose a pump that results in the pressure drop
across the FCV being 70 Pa. If the system is built and the pump slightly
underperforms its published curve, the FCV will not be able to control to
its flow control point. Even if the pump does operate exactly on the
published curve, eventually fouling in the pipes will result in increased
pipe resistance (and pressure drop) and again the FCV will not be able to
control to its flow control point.
8/10/2007 4:10:28 PM Page 465

Inlet Outlet
P = 0.1 MPa(g) P = 0 MPa(g)
a El = 1.5 m
El = 1.5 m
FCV FCV
Q = 0.05 m3/s Q = 0.05 m3/s
Inlet Outlet
b P = 0.1 MPa(g) P = 0.15 MPa(g)
El = 1.5 m El = 1.5 m
Pump with Fixed FCV
Flow Q = 0.05 m3/s
Q = 0.05 m3/s
Figure 12.15 a-b Neither model above has a unique solution. The
top model has two flow control valves in series.
The bottom model has a pump modeled as an
assigned flow in series with an FCV. In both cases
either a third pressure junction is needed between
the two middle junctions, or one of the junctions
must be changed from a flow controlling device.
To avoid these kinds of problems in installed systems, a minimum

pressure drop across the FCV is typically required. This minimum
pressure drop provides the key to sizing the pump in Figure 12.15b.
Rather than model the FCV as a flow control valve, instead model it as a
pressure drop control valve (PDCV) set to the minimum pressure drop
requirement. Then a unique solution exists, the model will run and the
pump can be sized. Once the pump is sized and an actual pump curve
exists, the pump curve can be entered for the pump and the valve can be
switched back to an FCV. Remember that when we use the PDCV, we
are still getting the flow we want for the true FCV, because the pump is
now providing the control.
Think for a moment about what we just did. We went through a thought
process that many engineers have gone through during their hand
calculations. Specifically, we size the pump such that the FCV pressure
drop is minimized. If the FCV has a larger pressure drop than the
minimum, the pump must pump harder to overcome this excess pressure
drop. In short, it must use more energy and is thus less efficient. Figure
12.16 depicts this process.

manual (si) g.doc
Inlet Outlet
a P = 0.1 MPa(g) P = 0.15 MPa(g)
El = 1.5 m El = 1.5 m
Pump with FCV
Fixed Flow Q = 0.05 m3/s
Q = 0.05 m3/s
Inlet Outlet
b P = 0.1 MPa(g) P = 0.15 MPa(g)
El = 1.5 m El = 1.5 m
Pump with PDCV
Fixed Flow dP = 0.05 Mpa
Q = 0.05 m3/s
Inlet Outlet
c P = 0.1 MPa(g) P = 0.15 MPa(g)
El = 1.5 m El = 1.5 m
Pump with Curve FCV
dP = f(Q) Q = 0.05 m3/s
Figure 12.16a-c. The model at the top is the same as Figure 12.15b
and does not have a unique solution. To size the
pump, change the model to the second one shown
above, which uses a pressure drop control valve
(PDCV) rather than an FCV. Once the second
model is run, the pump is sized, a pump curve
exists, and the model at the bottom can be run
using an FCV.
Let’s make a slight addition in complexity to the model in Figure 12.16.

In this case we have two flow control valves in parallel (see Figure
12.17). To size this pump, we apply the process described in Figure
12.16. We choose one of the FCV’s, make it a PDCV, size the pump,
choose a pump with an actual pump curve and enter it into the model,
then return the PDCV to an FCV.
8/10/2007 4:10:28 PM Page 467

J6
P5^ ^P8
J5 J8
FCV FCV
Q = 100 gpm Q = 100 gpm
P4^ P7^
J1 J2 J3 J4 J7
P1> P2> P3> P6>
Pump – known flow

Q = 200 gpm
Figure 12.17 A pump with an assigned flow in series with two

parallel FCV’s. The model in its current form does not
have a unique solution. To size the pump, first
change the FCV at J8 to a PDCV.
This brings up the question: which FCV should you turn into a PDCV?
When FCV’s are put in parallel, frequently the pipe design has one of
the FCV’s further away from the pump than the others. Because of the
additional piping leading to this FCV, it will be the weakest link in the
chain of parallel FCV’s, by virtue of having the lowest pressure drop
across it. The most remote FCV should be chosen as the PDCV. If the
most remote FCV is chosen, when the minimum pressure drop required
is applied to the PDCV all other FCV’s in the parallel system will have
greater than the minimum pressure drop — thus satisfying the pressure
drop requirement for those FCV’s as well.

manual (si) g.doc
What if you do not know which FCV is the most remote? In this case
make your best guess, change it to a PDCV at the minimum required
pressure drop, run the model, then verify whether all other FCV’s have a
pressure drop that meets or exceeds the requirement. If not, then the
FCV with the smallest pressure drop as determined by the first run is in
reality the weakest FCV. Choose this FCV, change it to a PDCV, then
change the original PDCV back to an FCV since it is not the weak link.
If the pipe system leading to each FCV is truly identical, and all FCV’s
have identical pressure drops, any of the FCV’s will serve as the PDCV.
The preceding process can be extended to cases where there are three or
more FCV’s in parallel, and also cases with more than one pump in
parallel supplying the FCV’s.
In Figure 12.17 the most remote valve is J8, the one on the right. If the
minimum pressure drop is 0.05 MPa, change J8 to a PDCV with a 0.05
MPa drop. This results in the output shown in Figure 12.18a. Notice how
the pressure drop across the other FCV at J5 (6.837 psid) is greater than
0.05 MPa because it is closer to the pump. Also notice how the flow
through J8 is still 25 m3/hr even though it is not controlling flow. Once
again, the reason for this is that the pump is controlling the flow to 100
m3/hr, and with the J5 FCV controlling to 50 m3/hr, the excess must go
through J8. The pump is thus sized and its pump head rise is shown in
Figure 12.18b (39.06 meters).
Figure 12.18a-b. AFT Impulse results for the model in Figure 12.17
after J8 has been changed to a PDCV. Top results
show valve summary and bottom results show
pump size.
8/10/2007 4:10:28 PM Page 469

Pressure control valves

The problem of non-unique solutions can also occur when a Pressure
Reducing Valve (PRV) or Pressure Sustaining Valve (PSV) is used. A
PRV is used to control pressure downstream of the valve, while a PSV
controls pressure at the valve inlet. One thing that PRV’s and PSV’s
have in common with FCV’s is that the pressure drop cannot be known
ahead of time. The pressure drop depends on the balance of the pipe
system.
Consider the simple system shown in Figure 12.19. With the pressure
controlled downstream of the PRV, the pressure upstream is not known.
Without knowledge of the PRV’s pressure drop, the pressure upstream
of the PRV cannot be determined and thus no unique solution exists.
Inlet Outlet
Q = 0.05 m3/s P = 0 MPa(g)
El = 3 m El = 3 m
PRV
Pdown = 0.1 MPa(g)
Figure 12.19 If a PRV is downstream of a known flow the PRV

pressure drop cannot be calculated. This model does
not have a unique solution.
Figure 12.20 shows the possible cases with PRV’s and PSV’s, while
Table 12.2 comments on the six cases. In summary, at least one known
pressure is always needed on the side of a pressure control valve
opposite of the controlled side.

manual (si) g.doc
Inlet Outlet
a Q = 0.05 m3/s P = 0 MPa(g)
El = 3 m PRV El = 3 m
Pdown = 0.1 MPa(g)
Inlet Outlet
b P = 0.2 MPa(g) Q = 0.05 m3/s
El = 3 m El = 3 m
PRV
Pdown = 0.1 MPa(g)
Inlet Outlet
c P = 0.2 MPa(g) P = 0 MPa(g)
El = 3 m El = 3 m
PRV
Pdown = 0.1 MPa(g)
Inlet Outlet
d Q = 0.05 m3/s P = 0 MPa(g)
El = 3 m El = 3 m
PSV
Pup = 0.15 MPa(g)
Inlet Outlet
e P = 0.2 MPa(g) Q = 0.05 m3/s
El = 3 m El = 3 m
PSV
Pup = 0.15 MPa(g)
Inlet Outlet
f P = 0.2 MPa(g) P = 0 MPa(g)
El = 3 m El = 3 m
PSV
Pup = 0.15 MPa(g)
Figure 12.20a-f. Six possibilities for PRV and PSV configurations.

Cases a and e do not have unique solutions. See
Table 12.2 for comments.
8/10/2007 4:10:28 PM Page 471

Table 12.2 Summary of six possibilities for PRV and PSV
configurations.
CASE Inlet Inlet Outlet Outlet Flow OK ? Comments
Pressure Flow Pressure
Fig. 12.20a - PRV X X No No reference pressure upstream of
PRV, cannot calculate
Fig. 12.20b - PRV X X Yes 1 known flow, can calculate

everything directly
Fig. 12.20c - PRV X X Yes 2 pressures, can iterate for flow
Fig. 12.20d - PSV X X Yes 1 known flow, can calculate
everything directly
Fig. 12.20e - PSV X X No No reference pressure downstream
of PSV, cannot calculate
Fig. 12.20f - PSV X X Yes 2 pressures, can iterate for flow
If a pump is being sized in series with a PRV or PSV, modeling the

pump as an assigned flow will not permit a unique solution if the pump
is upstream of a PRV or downstream of a PSV. The reasons for this are
the same as those in the previous section on FCV’s.
In combination with the previous examples, it should be apparent how
the non-unique solution problem can occur in more complicated systems.
In all such cases, the reasoning will boil down to the same problems
already discussed.
Closing parts of a system

A model may be built which has sufficient information to obtain a
unique solution. Then a user closes a junction and suddenly the problem
has no unique solution. This is shown in Figure 12.21a-b. The top model
has a unique solution and will run fine. In the bottom model the valve is
turned off and the section of the system downstream of the valve has
only known flows. No reference pressure exists in this section because it
is isolated from the known pressures in the section on the left.

manual (si) g.doc
J1 J5
a
P1v P4^
J2 J3 J4 J7 J6
P2> P3> P6> P5>
Open
J1 J5
b
P1v P4^
J2 J3 XJ4 J7 J6
P2> P3> P6> P5>
Closed
Figure 12.21 a-b Top model has reference pressures at J1 and J2.
Bottom model has closed valve at J4 which
isolates the J5 and J6 assigned flows. There is no
reference pressure for J5 and J6 and no unique
solution exists for the bottom model.
Effect of pipe sectioning on transient run times

The transient run time for a simulation is related directly to the total
number of computations required. The number of computations depends
on the total number of time steps (shown on the Transient Control
window) and the total number of sections (which can be determined
from the display table in the Section Pipes window).
Run times also depend to a lesser degree on how frequently data is
written to disk, because of the time required for disk access. This is
determined by the Save Intermediate Output to File… in the Transient
Control window.
8/10/2007 4:10:28 PM Page 473

To satisfy the Method of Characteristics (MOC) requirement, the time
step and number of pipe sections are related by Equation 5.1. If more
sections are desired, then shorter time steps are required. For example, to
use ten sections in a pipe rather than five requires twice as many
computations because there are twice as many sections. However, the
time step must be reduced by half (Equation 5.3); thus to model the pipe
for the same time duration (say 1 second) requires twice as many time
steps. Therefore, using double the number of sections results in four
times the number of computations.
A related issue is that of the controlling pipe (see Chapter 5 for more
information on controlling pipes). If the controlling pipe is very short,
this requires more sections in all other pipes in the model. Hence it may
be prudent to either neglect the shortest pipe or identify a way to
combine it into a longer pipe, which will lead to a different controlling
pipe. The Section Pipes window will automatically combine pipes of
common diameter for you (see Figure 12.22).
For example, consider a case where the controlling pipe is 1 foot long
and the next shortest pipe is 10 feet long. If the controlling pipe can be
neglected or combined, then the 10 foot pipe becomes the controlling
pipe. This will result in a factor of 10 fewer pipe sections and factor of
10 fewer time steps, or a run time reduction by a factor of 100!
Artificial transients
An artificial transient occurs when the steady-state results do not match
the transient calculation initial step. During the first time step AFT
Impulse tries to bring the mismatched conditions to what it considers
steady conditions which generates a transient pressure. This transient is
not a true physical transient, but a false “artificial” one based on
mismatched data. If undetected, the artificial transient will interact with
the physical transient being modeled and corrupt the results.

manual (si) g.doc
Figure 12.22 The Section Pipes window alerts you to pipes that
could be combined and possibly reduce model
runtime.
Before AFT Impulse initiates the transient simulation, it performs a
“zero time step computation” and compares the predicted results with
the results from the steady-state solution. If the difference exceeds the
artificial transient criteria (0.5% by default), then AFT Impulse will
display a warning in the output. If Stop Run if Artificial Transient
Detected is selected, the simulation will be immediately terminated. This
feature is especially helpful for simulations that take a long time to run.
For example, it is wasteful to run a large model for an hour only to find
at the end that an artificial transient input error was present at the very
beginning, thus nullifying the results.
The artificial transient detection feature is an automatic way for the user
to identify problems of this nature. A manual process can also be used,
and it is instructive to consider the manual process to better understand
artificial transients.
In some cases where pipes are initially not flowing, the first transient
calculation is some very small but non-zero flowrate. This can appear to
8/10/2007 4:10:28 PM Page 475

be an artificial transient as compared to a flow of zero. To avoid these
types of spurious warnings, one can specify a flow rate threshold below
which artificial transients will be ignored.
Manually detecting artificial transients

For those interested in better understanding artificial transients, it is
instructive to consider how one can manually detect an artificial
transient.
Once the AFT Impulse model is created, run the model but do not use
any of the transient input (e.g., by setting the Transient Special
Conditions to None). This will, for example, make all valves stay at their
original positions, all pumps stay at their original speed, and all pressure
and flow boundary conditions steady. Then run the model for 10 to 100
time steps by specifying the appropriate Stop Time in the Transient
Control window. If no artificial transients exist, all pressure and flow
solutions will remain steady with time because nothing transient is
occurring. (It is easiest to see if the pressures or flows change by
graphing some of the results.) If pressures and/or flows change and
transients are evident, then an artificial transient exists. It is artificial
since no physical transients are being modeled, thus all transients that
occur are non-physical artificial ones.

manual (si) g.doc
APPENDIX A
Keyboard Modifiers and Shortcuts
The keyboard modifiers and shortcuts for the Workspace are

summarized here for convenient reference. These shortcuts are displayed
when you click the Shortcuts button on the Toolbox, and also display at
the bottom of the Toolbox in abbreviated form.
Selection Drawing Tool
Double-clicking tool
If you double-click the Selection Drawing Tool, the Selection Drawing
Tool remains active after you have completed the selection. It remains
active until you click it again a single time
Control key
If you hold down the CTRL key when completing the selection drawing
(just before releasing the mouse button), the Selection Drawing Tool
remains active. The CTRL key thus performs a similar function to
double-clicking the tool.
Shift key
CTRL key, you can also hold down the SHIFT key while drawing the
selection box and the enclosed pipes and junctions will be deselected.

manual (si) g.doc
Alt key
CTRL key, you can also hold down the ALT key while drawing the
selection box and the enclosed pipes and junctions will toggle. That is,
any pipes or junctions that were selected will become deselected, and
vice versa.
Drag left to right

If you select the objects by dragging left to right, you will select all
objects completely inside the box.
Drag right to left

If you select the objects by dragging right to left, you will select all
objects completely or partially inside the box (if this feature is enabled
in the Workspace Preferences window).
Pipe Drawing Tool
Double-clicking tool
If you Double-click the Pipe Drawing Tool it remains active until you
click it again a single time. This allows you to draw a series of pipes
without returning to the Toolbox each time.
Control key
If you hold down the CTRL key when completing the pipe drawing, the
Pipe Drawing Tool remains active, and you can draw a series of pipes
without returning to the Toolbox each time. The CTRL key thus
performs a similar function to double-clicking the tool.
Shift key
If you hold down the SHIFT key down while drawing pipes on the
Workspace, you can draw pipes that are perfectly horizontal or vertical.
This may improve the aesthetics of your model.

manual (si) g.doc
Appendix A Keyboard Modifiers and Shortcuts 479
Zoom Select Tool
Shift key
If you hold down the SHIFT key down while selecting the zoom area,
the zoom state will always be 100%.
Panning the Workspace

If the Workspace has been enlarged beyond the default one screen size
(Workspace scrollbars will be visible), then the Workspace can be
panned by holding down the CTRL key, pressing the mouse down on the
Workspace, and then dragging in the direction you want to move the
Workspace.
Dragging junctions from Toolbox
Control key
A junction can be “morphed” from one type to another. To morph a
junction, hold down the CTRL key while selecting a junction from the
Toolbox. Drop the junction onto an existing Workspace junction. The
junction type will change to the new junction type, and data that can be
kept will be copied into the new junction.
Shift key
If you hold down the SHIFT key while selecting a junction from the
Toolbox and then drop the junction onto an existing pipe, the pipe will
split into two pipes. The physical length of the original pipe will be
automatically halved, and the new pipe will be assigned the balance.
Thus the sum of the two pipe lengths will equal the length of the original
pipe. Any fittings/losses in the original pipe will be left in the original
pipe, and the new pipe will have no fittings/losses.
8/10/2007 4:10:28 PM Page 479

Inspecting pipes and junctions
Shift key
If you hold down the SHIFT key when you perform an inspection, only
the items not yet defined are displayed.
Control key
If you hold down the CTRL key when you perform an inspection, the
inspection window will show the output data for the pipe or junction.
Moving junctions with connected pipes
Control key
By default, when you move an existing junction icon, any connected
pipes will move along with the junction and retain the connection. If you
hold down the CTRL key as you release the junction, only the junction
will move. All pipe connections will be broken. This behavior can be
reversed such that connections will not be retained unless you hold down
the CTRL key. This reversal can be specified on the Workspace
Preferences window.
Pipe and junction specifications window
F2 function key
When in a pipe or junction specifications window you can toggle the
highlight feature by pressing the F2 function key. You can also toggle
this from the Options menu.
F5 function key
When in a pipe specifications window, if you press the F5 function key,
you will jump to the next highest numbered pipe. If you press the CTRL
key while pressing the F5 function key, you will jump to the next lowest
numbered pipe.

manual (si) g.doc
Appendix A Keyboard Modifiers and Shortcuts 481
The F5 function key also behaves this way for junction specifications
windows.
8/10/2007 4:10:28 PM Page 481


manual (si) g.doc
APPENDIX B
Limitations
The size of model you can create with AFT Impulse is limited by the
amount of RAM on your PC. When solving a system, AFT Impulse
creates a square matrix that is double-precision. The size of the matrix is
determined by the total number of tee/wye and branch junctions in your
model.
For example, if you have 100 tee/wye and branch junctions, a 100x100
array is created. With double precision, the amount of RAM needed is
16x100x100, or 160KB. A 1000x1000 array would require 16MB.
In addition to this memory requirements, a relatively small amount of
memory is required for other Steady-State Solver and modeling
activities. AFT Impulse itself requires approximately 15MB of RAM to
run.
Window 95, 98 and ME are limited to displaying a maximum of
approximately 350-400 junction icons as a result of the way in which
these operating systems handle graphics resources. Windows NT, 2000
and XP, on the other hand, has no such limit. One can determine the
amount of free graphics resources by selecting Help from within AFT
Impulse, then clicking on About AFT Impulse and then System Info.
Among the information displayed will be GDI Memory Available
(Graphical Device Interface). As this value approaches zero, Windows
95, 98 and Me will become unstable and, ultimately, crash.
It is important to note that this Windows 95, 98 and ME limitation does
not limit the number of junctions or, as a result, the model size, since
junctions may also be displayed as a solid box, an outline box or not at
all (the junctions are still there even if not displayed, and may be edited

manual (si) g.doc
by double clicking on them within the Model Data window). These
options are selected from the Display Pipes and Junctions item under the
View menu, which allows you to selectively specify how you want
junctions displayed. For example, by displaying all branches in a large
model as an outline, a large number of icons can be made available for
pump, valve, accumulator and other junction types where display of the
icon is of greater advantage.

manual (si) g.doc
APPENDIX C
Installation Issues
Customization files
As you customize AFT Impulse, two new files are created in your
application data folder (e.g., C:\Documents and Settings\Your
Account\Application Data\Applied Flow Technology\AFT Impulse\4.0):
IMPULSE4.INI
IMP_USER4.DAT
The IMPULSE4.INI file contains the information to customize the
features in AFT Impulse. The IMP_USER4.DAT contains all custom
database information. If you invest significant time in customizing AFT
Impulse, especially in the area of custom databases, you are strongly
advised to keep backup copies of these files.
Special file control features
What is the Control.aft File?

Control.aft is read as AFT Impulse begins to load and instructs AFT
Impulse where certain key files are located. Typically there is little need
to change the settings, however this file can be used to override the
internal defaults and customize your working environment.

manual (si) g.doc
Control.aft is typically located in the application folder. However, on
networks, it may be necessary to locate it in the Windows\System folder
of each client. If AFT Impulse finds the file in the Windows\System
folder it will use it and ignore any file in the application folder.
Control.aft can be edited using any text-based editor. Notepad, supplied
as part of the Windows installation, is a good choice. Be careful to avoid
changing any of the headings or key words. If these are changed AFT
Impulse will not find the important data.
Specifying where Chempak is located

Before AFT Impulse can use the Chempak Fluid Database it needs to
know where the data is located. This paths is specified in the control.aft
file.
When Chempak is installed the setup procedure automatically updates
any control.aft file specified with the path to the data.
You can manually edit the file by opening it in Notepad (or any text
editor) and adding the following lines to the [SEARCH PATHS] section
(replace C:\AFT Products\Chempak Data with the path information
based on your system):
ChempakDataPath = C:\AFT Products\Chempak Data
This will instruct AFT Impulse to look in the C:\AFT Products\Chempak
Data directory for the data.
Network installations
AFT Impulse 4.0 requires a number of auxiliary files to run properly.
Because of how 32-bit Windows operating systems work (Windows 9x,
NT/2000/XP), certain files must be registered in the Windows registry.
For software installed on a client PC this is accomplished directly during
setup. For software installed on a local or wide area network, however,
the installation becomes two-step process.
The first step is to install AFT Impulse on the network using the
“Network (auxiliary files on client)” in the setup program. This performs
a typical network installation except the auxiliary files are not installed
or registered.

manual (si) g.doc
Appendix C Installation Issues 487
The second step is for each client to run the Client Setup program
installed into a subfolder below AFT Impulse named “Network Client
Setup”. This process uses standard installation protocol whereby the
auxiliary files are copied to the client's SYSTEM folder only if they do
not exist, or are newer versions than those that already exist. The Client
Setup will also install the sample and verification files on the client and
will create shortcut links on the Start Menu referencing the applications
and files on the network server. Finally, it will update the
IMPULSE4.INI file with details of the setup. Although this option is not
a perfect solution, we believe it is the best because the client PC registry
points only at local files (as far as AFT Impulse is concerned).
Problems loading the Graphics Server

AFT Impulse uses a third party tool for graphing called the Graphics
Server. There are four files that are used to load and run the Graphics
Server:
• GRAPHS32.OCX
• GSW32.EXE
• GSWDLL32.DLL
• GSWAG32.DLL
Because of how the Graphics Server is written, multiple copies of
GSW32.EXE or GSWDLL32.DLL in your Path will cause problems
when your try to load AFT Impulse 4.0.
In practice, the most common problem is a network installation where all
AFT Impulse files have been installed into the network folder along with
AFT Impulse. Thus, GSW32.EXE and GSWDLL32.DLL will exist in
the AFT Impulse network folder. A client PC which happens to have
copies of either or both of these files on the local hard drive (usually in
the WINDOWS\SYSTEM folder on Windows 95/98 and
WINNT\SYSTEM32 on Windows NT) will fail to load properly. The
most likely reason copies exist on the local hard drive is another
application installed locally uses them.
A complete list of your options is given below, along with an
explanation of the impact of each option. In the following explanations,
the term conflicting files means the GSW32.EXE and GSWDLL32.DLL
8/10/2007 4:10:28 PM Page 487

files. Your WINDOWS\SYSTEM folder is just called your SYSTEM
folder.
1. Delete the conflicting files on your hard drive – This will solve
the problem for AFT Impulse 4.0, but may cause other software you
use to stop working. This is not a permanent solution in that if you
install other software in the future that installs copies of the
conflicting files into your SYSTEM folder, the current problem with
AFT Impulse will occur again.
2. Change the names of the conflicting files on your hard drive –
This will solve the problem for AFT Impulse 4.0, but may cause
other software you use to stop working. However, when you try to
run the other software and it fails to load because of these missing
files, you can change the names back to the original names. This will
then cause AFT Impulse 4.0 to stop working until you rename them
once again, but will allow you to run your other software that needs
these files. This will also require you to remember which files you
renamed. This is not a permanent solution in that if you install other
software in the future that installs copies of the conflicting files into
your SYSTEM folder, the current problem with AFT Impulse will
occur again.
3. Move the conflicting files on your hard drive into the application
folder that uses them – This will solve the problem for AFT
Impulse 4.0 and still allow the other software you use to run
properly. However, you will need to know which software uses the
conflicting files so you can copy the files into their folder. If you
remember what you do, you can move the files out of the SYSTEM
folder (or wherever conflicting files exist) and into a temporary
folder. When a software application that used to work fails because
these files are missing, you can copy the files into its folder at that
time. This is not a permanent solution in that if you install other
software in the future that installs copies of the conflicting files into
your SYSTEM folder, the current problem with AFT Impulse will
occur again.
4. Allow AFT Impulse 4.0 to automatically rename the files for you
– This is not a perfect solution, but it does have advantages. This
option enables a feature whereby AFT Impulse searches for the
conflicting files in both your SYSTEM folder and WINDOWS
folder before the full AFT Impulse application is launched. If it finds
them, it automatically renames them and thus only the copies in the

manual (si) g.doc
Appendix C Installation Issues 489
AFT Impulse network folder are found. When you exit AFT
Impulse, the files will be renamed back to their original names. The
advantages are, 1) you can continue to run your other software that
uses these files and AFT Impulse with no conflicts (as long as they
are not run at the same time) and, 2) even if you install other
software in the future, this option will continue to work for you. The
disadvantages are, 1) you are permitting AFT Impulse to change
files on your local hard drive every time it is run, which
understandably makes some users uncomfortable and, 2) you cannot
run AFT Impulse 4.0 at the same time as the other software which
needs these files.
5. Reinstall AFT Impulse with the Network option to run auxiliary
files on the client (recommended) – This option is the best long-
term option. To do this, first uninstall AFT Impulse from the
network and reinstall it with the Network option. See Information on
Software Installation for more help.
8/10/2007 4:10:28 PM Page 489


manual (si) g.doc
APPENDIX D
Chempak Technical Information
© Copyright 1998 Madison Technical Software Inc. Used with

permission.
The theoretical basis for Chempak is a lengthy topic and has been
included in the AFT Impulse 4.0 Help system. To learn more about
Chempak, open the Help system and search for Chempak.

manual (si) g.doc

manual (si) g.doc
APPENDIX E
Obtaining Technical Support
In order to receive technical support, the Licensing Agreement return

card must be signed by an authorized representative of your company
and returned to Applied Flow Technology within 30 days of purchase.
AFT Impulse is a sophisticated software application intended for trained
engineers, and in order to provide effective support, Applied Flow
Technology needs to know who is using the software.
AFT's three levels of support
1. Free Software Support for One Year

AFT provides free support for the first year to those who sign and return
the Licensing Agreement. This support is for resolving installation
problems and clarifying usage issues.
2. Annual Support, Upgrade and Maintenance (SUM)

After one year, AFT provides support to those sites that purchase an
annual SUM. This support will also be for resolving questions on how to
use specific features in the program.
3. Waterhammer Modeling Consulting Support

AFT offers specialized consulting support to users who have general
waterhammer analysis questions. For questions that do not pertain
strictly to AFT Impulse, but instead pertain to the general nature of how

manual (si) g.doc
to do waterhammer analysis, AFT offers high quality consulting services
by experienced engineers at an hourly rate. Call for an hourly rate quote.
Contacting AFT
Telephone support
Applied Flow Technology can be reached at:
(719) 686-1000 (Voice)
(719) 686-1001 (FAX)
Web site
You can download the latest maintenance releases of Impulse, find out
what new things are happening at AFT, and get the latest information by
visiting our website at:
http://www.aft.com
E-Mail support
AFT can also be reached by E-mail at:
support@aft.com or www.aft.com
Mail support
You can send mail to AFT at:
P.O. Box 6358
Woodland Park, CO 80866-6358 USA
You can send courier mail to AFT at:
400 W. Hwy. 24, Suite 201
Woodland Park, CO 80863 USA

manual (si) g.doc
References
ASME Press, ASME International Steam Tables for Industrial Use

(CRTD-Vol. 58), published by the American Society of Mechanical
Engineers, Three Park Avenue, New York, NY 10016
Brown R.J. , and D.C. Rogers, “Development of Pump Characteristics
from Field Tests”, Journal of Mechanical Design, ASME, vol. 102, pp.
807-817, Oct., 1980.
Carthew, G.A., Goehring, C.A., van Teylingen, J.E., "Development of
dynamic head loss criteria for raw sludge pumping", Journal Water
Pollution Control Federation (WPCF), Volume 55, Number 5, (May
1983).
Chaudhry, M. H., Applied Hydraulic Transients, 2nd edition, Van
Nostrand Reinhold Company, New York, NY, 1987.
Crane Co., Flow of Fluids Through Valves, Fittings, and Pipe, Technical
Paper No. 410, Crane Co., Joliet, IL, 1988.
Donsky, B., "Complete Pump Characteristics and the Effects of Specific
Speeds on Hydraulic Transients", ASME Journal of Basic Engineering,
pp. 685-699, Dec. 1961.
Fox, R. W., and A.T. McDonald, Introduction to Fluid Mechanics, 3rd
edition, Wiley & Sons, New York, NY, 1985.
Idelchik, I. E., Handbook of Hydraulic Resistance, 3rd edition, CRC
Press, Boca Raton, FL, 1994.
Ingersoll-Dresser Pumps, Cameron Hydraulic Data, Liberty Corner, NJ,
1995.
Jeppson, R. W., Analysis of Flow in Pipe Networks, Ann Arbor Science
Publishers, Ann Arbor, MI, 1976.

manual (si) g.doc
Karassik, I.J., J.P. Messina, P. Cooper and C.C. Heald, Pump Handbook,
3rd Edition, McGraw Hill, New York, NY, 2001.
Karney, B.W., and D. McInnis, "Efficient Calculation of Transient Flow
in Simple Pipe Networks", ASCE Journal of Hydraulic Engineering,
Vol. 118, No. 7, July, 1992.
Kittredge, C.P., “Hydraulic Transients in Centrifugal Pump Systems”,
ASME Journal of Basic Engineering, Vol. 78, No. 6, pp. 1307-1322,
Aug. 1956.
Lyons, J. L., Lyons' Valve Designers' Handbook, Van Nostrand
Reinhold, New York, NY, 1982.
Miller, D. S., Internal Flow Systems, 2nd edition, Gulf Publishing
Company, Houston, TX, 1990.
Pipe Line Rules of Thumb Handbook, Gulf Publishing, Houston, TX.
Swaffield, J. A., and Boldy, A. P., Pressure Surge in Pipe and Duct
Systems, Avebury Technical, Hampshire, England, 1993.
TAPPI, Generalized method for determining the pipe friction loss of
flowing pulp suspensions, TAPPI, TIS 0410-14, Atlanta, GA, 1988.
Tilton, J. N., Perry's Chemical Engineer's Handbook, Seventh Edition,
Chapter 6, McGraw-Hill, New York, NY, 1997.
Thorley, A.R.D., and A. Chaudry, “Pump Characteristics for Transient
Flow”, Pressure Surges and Fluid Transients, BHR Group, 1996.
Walski, T. M., Analysis of Water Distribution Systems, Van Nostrand
Reinhold Company, New York, NY, 1984.
Walters, T.W., "Analytical Approaches to Modeling Transient Vaporous
Cavitation in Multi-Pipe Fluid Systems", ASME FED-Vol. 116, New
York, NY, 1991.
Watters, G.Z, Modern Analysis and Control of Unsteady Flow in
Pipelines, Ann Arbor Science, Ann Arbor, MI, 1979.
White, F. M., Fluid Mechanics, McGraw Hill, New York, NY, 1979.
Wylie, E.B., V.L. Streeter & L. Suo, Fluid Transients in Systems,
Prentice Hall, Englewood Hills, New Jersey, 1993.

manual (si) g.doc
Glossary
accumulator A chamber in a pipe system that offers elastic relief to

pressure waves. Can be either liquid filled or gas filled. Gas
accumulators are used as surge suppression devices.
artificial transient A false transient pressure that is generated because
the initial conditions do not represent a steady-state condition.
BATCH.LOG File created to log events when a batch run is
performed.
Base scenario The top level scenario in a multiple scenario model.
Single scenario models always work in the base scenario.
Best Efficiency Point See BEP.
BEP Best efficiency point. Flow rate at which pump operates at
maximum efficiency.
Bingham plastic See non-Newtonian fluid.
Brecht and Heller method A model for pipe pressure drop in pulp
and paper stock flow. See non-Newtonian fluid.
cavitation A phenomenon that occurs when flowing liquids pass
through restricted areas resulting in reduced static pressures and
localized boiling. Can be very destructive to pipe systems. Transient
cavitation differs from the traditional cavitation that occurs in steady
flow systems. The concern in transient flow is that cavities form which,
when they collapse, produce a large pressure spike. This is sometimes
referred to as liquid column separation.
checklist The window that tracks the overall completeness of the
model.

manual (si) g.doc
communication time The time it takes a disturbance to propagate to
the nearest boundary and return.
compatibility equations The basic finite difference equations which
solve the mass and momentum balance.
controlling pipe The pipe with the fastest end-to-end communication
time in a pipe system. Its properties govern the choice of time step.
convergence The state of the solution when the correct answer has
been obtained.
custom database A customizable database that allows pipe system data
to be saved. Custom databases are read in during startup and used as if
they were part of AFT Impulse.
Database Manager The window that allows creation and connection
to network databases.
defined objects Objects whose properties and connectivity
requirements have been completed.
design alert A value assigned to a pipe parameter for which AFT
Impulse will warn the user when exceeded.
distributed loss A hydraulic loss that is modeled as occurring over a
finite distance in the pipe system (typical of pipe losses).
Duffy method A model for pipe pressure drop in pulp and paper stock
flow. See non-Newtonian fluid.
EGL Abbreviation for Energy Grade Line.
elevation The vertical coordinate system (in parallel with body force)
that is assigned to junctions. Elevation differences between junctions
make possible calculations of hydrostatic or gravity head pressure
changes.
Energy Grade Line The sum of the pressure head, elevation head and
velocity head. Quantity is relative to the elevational coordinate system
specified by the engineer.
exit valve In AFT Impulse, a valve that is located at the exit from a
pipe system. An exit valve passes fluid to an outside fixed pressure.
flow into system The flow rate that enters the pipe system through a
junction.
fluidhammer Same as waterhammer.

manual (si) g.doc
Glossary 499
fluid transient Same as waterhammer.
Graph Results window The graphical window that allows preparation
of plots.
head In AFT Impulse, synonymous with piezometric head and
Hydraulic Grade Line.
head loss The irrecoverable loss in head that occurs in a pipe system.
Head loss is equal to the frictional pressure loss converted into units of
length.
HGL Abbreviation for Hydraulic Grade Line.
history The information on how the solution scheme progresses.
Hydraulic Grade Line The sum of the pressure head and gravity head.
This quantity is relative to the elevational coordinate system specified by
the engineer.
hydraulic transient Same as waterhammer.
ID number The unique number associated with a pipe or a junction.
The ID number is always greater than zero and up to 30,000.
incompressible A fluid whose density changes are small enough that a
constant density can be safely assumed.
infinite pipe A junction which does not reflect and pressure or flow
waves. Available in Assigned Flow and Assigned Pressure junctions.
inspection A feature in AFT Impulse that allows you to obtain
read-only information about objects.
instantaneous waterhammer A waterhammer event that occurs so
rapidly that it is effectively instantaneous, thus yielding the highest
possible pressure surge.
iteration A single step of the Steady-State Solver, during which a new
solution is obtained based on the previous solution state.
Jump A feature that allows lateral changes to be made between
Specifications windows.
junction In AFT Impulse, a pressure/head solution point that connects
and balances flow from pipes.
lock To fix an object's position on the Workspace in order to prevent
accidental movement.
8/10/2007 4:10:28 PM Page 499

loss factor A parameter that relates the amount of irrecoverable
pressure loss through a component. Synonymous with K factor.
Model Data window The text-based window where the input for the
model is given in text form and input data can be entered.
Net Positive Suction Head (or Pressure) See NPSH.
Newtonian Fluid A fluid whose viscosity is independent of the
dynamics of the flow. Temperature or pressure dependence may exist.
Non-Newtonian Fluid A fluid whose viscosity is dependent on the
dynamics of the flow such as velocity.
NPSH Net Positive Suction Head. A property of pumps that relates the
supply head needed for proper performance.
NPSHA Net Positive Suction Head Available. See NPSH.
NPSHR Net Positive Suction Head Required. See NPSH
NPSP Net Positive Suction Pressure. A property of pumps that relates
the supply pressure needed for proper performance.
object A pipe system component (either a pipe or a junction) that is
created, exists, and can be viewed on the Workspace.
object status The state of the object, whether defined or undefined.
Output window The text-based window that shows the results of the
analysis.
piezometric head Synonymous with Hydraulic Grade Line.
pipe In AFT Impulse, a conduit for incompressible, steady-state fluid
flow. All pipes have constant diameter and are completely filled with
fluid.
point loss A hydraulic loss that is modeled as occurring at a single
point in the pipe system. Typical of junction losses.
positive flow direction See reference positive flow direction.
Power law See non-Newtonian fluid.
Pressure Reducing Valve A type of valve that maintains a constant
pressure/head at a location, independent of the flow rate.
primary window One of AFT Impulse's two input or three output
windows.

manual (si) g.doc
Glossary 501
PRV See Pressure Reducing Valve.
Pump configuration A particular pump which contains data for
rotational speed and impeller size.
reference positive flow direction The flow direction you assign to a
pipe, which by convention is positive. Flow in the opposite direction is
considered negative.
relaxation A parameter that affects how the iterative solution scheme
approaches a pipe flow solution.
relief valve A type of valve that is typically closed but opens to release
fluid when a set pressure occurs. Used as a protective device.
resistance A convenient term relating the head loss to flow rate (R =
H/Q2).
Scenario Manager Window that allows one to create and manage
multiple pipe flow model scenarios. These include different equipment,
sizes, and operating conditions.
Solver Either Steady-State or Transient. The Steady-State Solver is the
part of AFT Impulse that contains the pipe steady flow network solution
method. The Transient Solver contains the transient waterhammer
solution algorithms.
Specifications window The window in which an object's properties
are entered.
stagnation pressure Sum of static and dynamic pressures.
station A location in a pipe where a transient computation occurs.
Form computational reasons pipes are broken into sections. Because
computation is required at the pipe inlet and outlet, there will always be
one more station than there are sections.
steady-state The condition of a system in which the rate of change of
all parameters is negligibly small.
surge flow Same as waterhammer.
tolerance The value at which the Steady-State Solver should consider
that convergence has been obtained.
Toolbox The area on the Workspace window that offers tools for
building pipe flow models.
8/10/2007 4:10:28 PM Page 501

valve state The position of the valve control device which regulates the
flow through the valve.
Visual Report window The graphical window that allows integration
of the analysis results with the pipe system schematic.
waterhammer A transient phenomenon in pipe systems where a
rapidly changing operating condition causes large pressure surges.
wavespeed The speed at which a transient wave propagates in a pipe.
Workspace The area of the Workspace window where models are
visually assembled.
Workspace window The graphical window where model assembly
occurs and input data can be entered.

manual (si) g.doc
Change Fluid Data window 308
Index Database 4, 145, 366
Database editing 150
A
Aligning objects 81
Absolute roughness 62, 200, 201, 310,
337 Analysis type
Absolute tolerance See Steady steady-state vs. transient 189

Solution Control window Animate Using Output File See
Affinity laws 246 Animation
for pumps 365, 391, 395 Animate Using Solver See Animation
AFT DEFAULT INTERNAL Animation 53, 109, 160

DATABASE 313 Annotation Manager 79, 129
AFT Fathom Annotation tool 15, 78
Converting models from 10 Annotations 129
AFT Impulse Area Change
engineering assumptions 4 as Fittings & Loss 204
Example models 7 Area Change junction
Getting started 6 abrupt transition 221, 355
installation 5 conical transition 221, 355
New features in version 3.0 11 loss factors 222, 355
Overview 7 Specifications window 221
qualifications for use 4 Artificial transients 44, 159, 474
summary of capabilities 2 manual detection 476
using online help 7 Stop run when detected 159
AFT IMPULSE LOCAL USER ASME ’97 Water See ASME Steam
DATABASE 313 Tables
AFT Impulse window 14, 71 ASME Steam Tables 28, 139, 145,
AFT Mercury 147, 366
Converting models from 10 Assigned Flow junction
AFT Standard fluid 27, 139
8/10/2007 4:10:28 PM Page 503 c:\aft manuals\i4\impulse4 manual (si) g.doc

contrast with pressure-type junction loss factors 356
59
Bernoulli equation 327
example 459
Bingham Plastic See non-Newtonian
loss factors 62
Black and white junction icons 186
Special features for steady flow 224
Branch junction 16, 273
Specifications window 222
flow source or sink 228
Transient Data 224
loss factors 62
waterhammer theory 381
Specifications window 32, 227
Assigned Pressure junction
Transient Data 229
How pressure junctions work 455
loss factors 62
Brecht & Heller method See non-
periodic data 227 Newtonian
the role of pressure junctions 458 C
Transient Data 227 CAESAR II Force Files 113, 165
use in open and closed systems 332 Calculator 189
using static and stagnation pressure Cavitation 144, 308

332 transient 159
waterhammer theory 379 Check Valve
Atmospheric pressure 149 theory 410
Autosave 302 Check Valve junction
Auxiliary Graph Formatting window Delta Pressure to Re-Open 231
304
Forward Velocity to Close Valve
231
B inherent event logic 434
Base area 61, 221, 230, 242, 274, 360
loss factors 230
Specifying 220
Special Condition check valve 230
Batch runs 187
Bends
Transient Data 231

Index 505
Checklist 24, 136 Continuity equation
Chempak 27, 139, 146, 366 steady-state 326
accuracy options 366 transient 371
mixtures 28 Control Valve junction
specifying Chempak file locations action if setpoint not achievable 233
486
Control Transient 236
chopped sine wave 224, 227
failure to control issues 452
Clipboard 115
Failure Transient 235
Colebrook-White 337, 338
Open Percentage Table 235
Color Map See Visual Report window
sizing pumps with flow control valves
column separation See vapor cavitation 464
communication time 42, 65, 66, 370, Special Conditions 234
387, 453
compatibility equations 376, 380, 381,
static and stagnation pressure 332
382, 383, 385, 387, 391
Transient Data 235
Component Database 311
types 232
computing stations 154, 424
graphing 52
Control.aft file 485
in output file 95
controlling pipe 42, 44, 153, 378, 474
max/min results in output 99
Conventions See also Base area
numbering convention 154
flow entering and exiting 60
results in output window 101
Convergence 351
Connectivity 130, 132, 134
Copy 80, 126, 128, 130
graphical connection 131
Create Mixture window 146
in Junction Specification window
214 Crude oil 338
in Pipe Specifications window 198 CTRL key See Keyboard modifiers,
See Keyboard modifiers
model connection 132
Custom databases See Databases
Constant Pressure Drop Valve See
Pressure Drop Control Valve Customize Graph window 112
8/10/2007 4:10:28 PM Page 505

Customizing AFT Impulse cyclic dual events 432, See Event
transients
databases 305
files created 485
D
fluids 306
Darcy-Weisbach 201, 326, 338
graph styles 304
Database Manager 171, 306, 312
junction (component) databases 311
Benefits of shared databases 316
Model Data window 90
Connecting and disconnecting
Output window 95, 303 databases 313
Parameter Preferences 287 creating network databases 319
pipe fittings & losses 305 DATABASE.LIB 313, 318, 321
pipe materials 309 Editing databases 314
Toolbox Preferences 300 Enterprise network databases 318
Unit Preferences 289 IMP_DBUSER.LIB 319
Visual Report window 304 Making databases available 312
Workspace window 83, 292 sharing databases using
Customizing AFT Impulse See Visual DATABASE.LIB 321
Report Control window DATABASE.LIB 313, 318, 321, 323
Customizing AFT Impulse See Output Databases 29, 305
Control window
Benefits of shared databases 316
Customizing AFT Impulse See
Workspace Preferences window Component database 214, 244, 311
Customizing AFT Impulse See Connecting to external shared

Toolbox Preferences window databases 323
Customizing AFT Impulse See Creating an enterprise-wide network

Parameters and Unit Preferences database system 318
window custom junctions 311
Customizing AFT Impulse See Model fluid See AFT Standard fluid or
Data Control window Chempak
Customizing AFT Impulse See fluid database 145, 150, 306
Databases
fluid properties 150, 305
Cut 80, 130

Index 507
Friction Data Set 309, 310 Dynamic Pressure 331
IMP_USER4.DAT 287, 305, 315, Dynamic viscosity 144
320, 485
pipe materials 199, 309 E
special control files 317 Editing
Specifying common locations 302 on Workspace 80
Dead End junction EGL 108
Specifications window 236 Elbow
waterhammer theory 382 See Bend 356
Defining objects Elevation 108, 212
Checklist - Define All Pipes and default 289
Junctions 28, 139
Event messages 104, 436
List Undefined Objects window 29,
133 Event transients 430
object status 80 Example models 7
Delete 80, 130 Exit valve 386, See Valve junction
Density 144 Explicit friction factor 62, 202
Design Alerts See Pipe Specifications Extended Model Check 189

window
cross-plotting against results 108 F
Design factors Fanning friction factor 336
defaults 288 FCV See Flow Control Valve
junctions 219, 355 Find utility 177
pipes 206, 344 Fittings & Losses 63, See Losses, See
Losses
Discrete Vapor Cavity Model 376, 424
window 203
Distributed loss 62, 353
Flip workspace See Workspace
Drag-and-drop 3 window
Duffy method See non-Newtonian Flow Control Valve 232, 452, 462,
Duplicate 80, 126, 128, 130 463, 466
8/10/2007 4:10:28 PM Page 507

Flow rate relaxation See Steady non-Newtonian 340
Solution Control window
Relative roughness 337
Fluid database See AFT Standard fluid
Resistance 338
or Chempak
Frictional Losses See Losses
Fluid Database window 306, 308
Frictionless pipe 62, 202, 340
fluid transients
See waterhammer
G
fluidhammer
Gas Accumulator junction
See waterhammer
connector pipe 240
Forces
flow restrictor 239
Defining force sets 161
Graphing transient data 241
displaying transient force data in
Graph Results 112 in output file 161
exporting transient force data from Initial Gas Volume 237

Graph Results 165 lumped inertia 422
including friction in force balance Maximum and minimum volumes
112 239
including momentum in force balance orifice 239
112
Polytropic Constant 237
Fouling 201
Special features for steady flow 240
Friction Data Set 200, 309, 310
Friction factor 150, 201
surge reduction 443
changes during transient flow 159
Friction models 62, 335
Gaussian Elimination 351
Absolute roughness 337
With pivoting 351
Frictionless pipe 340
General Component junction
Hazen-Williams 338
loss factors 242
Hydraulically smooth 338
Miller Turbulent method 339
General Preferences 302
MIT Equation 338
Global Pipe Edit window 278, 279

Index 509
Glossary 497 Hydraulically smooth 62, 202
Graph Results window 8, 52, 72, 106, Hydrocarbons - roughness 339
304
Animation 53, 109 I
Import Graph Data 106 Icon size 294
Save Graph Data 106 ID numbers 20, 28, 126, 128, 151, 197,
Select Data 52 297
Select Graph Data 106 ID Reduction - Scaling 200
Graphics Server IMP_DBUSER.LIB 319, 322, 323
problems loading 487 IMP_USER4.DAT 287, 303, 305, 315,

320, 485
Gravitational acceleration 5, 149
impedance 375
Grid
Impulse Examples.hlp 7
On Workspace 23, 118, 296
IMPULSE4.INI 287, 303, 485
Group Manager 82
Incrementing Numbers 86
Grouping objects 80, 129
Infinite Pipe
Groups 81, 82
example 453
used in graphing 108
Special Conditions 183
H theory 387
Initial guesses
Hazen-Williams 202, 338
flow guess 205
H-Equation method 325
generation 218
HGL 108, 326
pressure guess 217
Hide Empty Columns
using for troubleshooting models
451
Highlight feature 30, 133, 194
Inspection feature 31, 134, 210, 212
Homologous pumps See Affinity Laws
Showing only undefined items 480
Hydraulic Grade Line See HGL
Installation 485
hydraulic transients
Network installations 486
See waterhammer
8/10/2007 4:10:28 PM Page 509

Intermediate Pipe Elevations 207 Initial pressure guess 217
Irrecoverable losses See Losses Jumping to next higher or lower
number junction 133
Iteration history 351, 451
Liquid Accumulator 243
J Name 212
Jacobian matrix 328 Notes 220
Jump 132, 193, 214 Number 212
Junction icons Pump 244
printing black and white 186 Relief Valve 259
Junction label position 297 Reservoir 261
Junction Results table 93 Special Conditions 219
Junction Specifications window 208 Spray Discharge 263
Area Change 221 Status 220
Assigned Flow 222 Surge Tank 265
Assigned Pressure 225 Tee/Wye junction 271
Branch 227 Vacuum Breaker Valve 274
Check Valve 229 Valve 273
connected pipes 214 Volume Balance 276
Control Valve 231 Junctions 59
Copy Data From Junction 214 Base area 220
Database list 214 comparison of types 208
Dead End 236 connectivity 130
design factor 219 creating 128
displaying name 219 customizing icons 300
Elevation 212 defining 132
format 210 displaying data in Output window
99
Gas Accumulator 237
displaying in Model Data 90
General Component 241
displaying in Visual Report 118

Index 511
displaying on Workspace 83 Loss factor 60, 62, 203, 210, 211, 221,
227, 230, 234, 242, 261, 272, 354
displaying transient data in Graph
Results 108 loss models 34, 35, 354
editing 80, 130 Losses 60, 62, 327, 336, 353, See
Fittings & Losses
finding 177
LU Decomposition 351
icons on Toolbox 79
Lumped inertia pipe with orifice
ID numbers 128
theory 422
keyboard shortcuts 479
loss factor 62
M
negative flow rates 210
Matrix Method 351
placing on Workspace 128
Merging models 184
pressure losses 353
with multiple scenarios 184
specifying data 132
Method of Characteristics 42, 370,
371, 474
K
reliability with cavitation 425
Keyboard modifiers 477
Miller Turbulent Method 202, 339
CTRL 73, 75, 125, 130
minor losses See Losses
SHIFT 73, 75, 78, 80, 125, 129
MIT Equation 202, 338
Mixtures 4, 28, 146, 366
L
Model Data Control window 90
Laminar flow 150, 337
database connections 92
Legend 117, 121
Scenario display 91
Liquid Accumulator junction
showing junction data 90
Elasticity 244
showing pipe data 90
Initial Volume 244
showing selected pipes and junctions
Specifications window 243 91
waterhammer theory 405 using saved format files in Database
Lock 80, 129 Manager 317
Model Data window 8, 40, 71, 87, 88
8/10/2007 4:10:28 PM Page 511

customizing 90 Brecht & Heller method 341
opening pipe and junction Duffy method 340
specifications windows from 132,
junction losses 344
193
Pipe Fittings & Losses 344
Pipe Detail Summary 89
Power Law 148, 342
Pipe Fittings & Losses 88
pulp and paper 340, 341
Pipe table 88
NPSH 144, 248, 249
print content 188
constant for variable speed 248
scenario display 91
NPSHR 249
transient junction data 438
NPSP 144
using for troubleshooting models
446 NPSPR 249
Viewing scenario differences 169 NPSP See NPSH
Model Status Light 136
modulus of elasticity 201 O
Momentum equation Object status 80, 134, 151, 208, 220
steady-state 326 One-dimensional flow 59
transient 371 Open Pipe/Jct Window 87, 132, 193
Moody friction factor 336 Orifice
morph 128, 479 loss factors 358

round edged 358
N sharp edged 358
Network databases 214, See Database Output Control window 25, 51, 95,
Manager 138, 140, 303, 335
Newton-Raphson method 3, 326, 328, database connections 100
335 Format and Action 98
transient solution for gas General Output Control 97
accumulators 416
Pipe and Junction parameters 95
non-Newtonian 148, 335, 340
Pipe transient parameters 97
Bingham Plastic 148, 343
scenarios 171

Index 513
Show Selected Pipes/Junctions 99 using saved format files in Database
Manager 317
using saved format files in Database
Manager 317 Paste 80, 126, 128, 130
Output file 47, 160 PDCV See Pressure Drop Control
Valve
transient data 94
Piezometric head 325
Output window 8, 48, 71, 93
Pipe Detail Summary
customizing 303
Event messages 436
Pipe Drawing tool 14, 20, 74, 125, 478
number formatting 98
Pipe Fittings & Losses
print content 188
customizing 305
setting defaults 100
design factors 206
updates 104
in Solver 353
Viewing scenario differences 169
Overview of AFT Impulse 7
Non-Newtonian flow 344
P specifying 62
Parallel Pipes Pipe flow direction 222
in Pipe Specifications window 206 Pipe Loss Data table 88
Parameter and Unit Preferences window Pipe Material Database window 199,
287 309
database connections 291 Pipe Results table 93
default roughness model 202 pipe sectioning 41, 377, 378, See
Section Pipes
Design Factors 288
effect on run time 473
elevation 289
Pipe Specifications window 37, 39,
setting defaults 291 194
setting preferred units 290 connected junctions 198
unit preferences 289 Copy Data From PIpe 198
Use Most Recent Pipe Size Data 288 Copy Previous 198
Design Alerts 205
8/10/2007 4:10:28 PM Page 513

displaying name 205 Pipes 20, 59
Fittings & Losses 203 connectivity 130
Fluid Properties 208 creating 125
friction model 201 defining 132
ID Reduction - Scaling 200 Design factors 206
initial flow 205 displaying data in Output window
99
jumping to next higher or lower
number pipe 133 displaying in Model Data 90
material 199 displaying in Visual Report 118
Notes 208 displaying on Workspace 83
optional input 205 displaying transient data in Output
window 101
Parallel Pipes 206
editing 129
pipe diameter 200
finding 177
Pipe Length 199
flow direction 60, 126, 128, 178,
Pipe Line Size and Color 206
198, 210
pipe modulus of elasticity 201
Global Editing 278
Pipe Name 197
handles 76, 129
Pipe Number 197
ID numbers 126
pipe Poisson ratio 201
length on Workspace 126, 199
pipe support 203
loss factor 62
pipe wall thickness 200
losses 60, 61, 62, 336, 338, 353
roughness model 201
number of connections 126
Status 208
Reverse direction 178
wavespeed 203
segmenting 76
pipe stations See computing stations
specifying data 132
Pipe Support 203
Splitting 128, 479
Pipe table
stretching 129
Point loss 62

Index 515
Poisson Ratio 201 controlled discharge pressure 248,
365
Popup menu 81
controlled flow 248, 365
positive flow direction 21
Controller transient 255
Power Law See non-Newtonian
efficiency and power usage 249
Pre-mixtures 146
Estimating the pump inertia 254
Pressure
Graphing pump data 258
defining system pressure 166, 458
in output file 161
Pressure Drop Control Valve 233, 466
Known speed transients 250
Pressure Reducing Valve 232, 452,
470 NPSHR 249
static and stagnation pressure problems from poor curve fits 447
discussion 332
pump configurations 246
Pressure relaxation See Solution
Pump Startup With Inertia 254
Control window
Pump trip with inertia – four quadrant
Pressure Sustaining Valve 233, 452,
253
470, 471
Pump trip with inertia and no
static and stagnation pressure
backflow 252
discussion 332
sizing with flow control valves 464
Primary windows 7, 71
Special Conditions 257, 258
Print Content 90, 93, 188
Print Preview 185
Submerged pump 246
Print Special 185
Transient Data 249
PRV See Pressure Reducing Valve
variable speed 245, 248, 365
PSV See Pressure Sustaining Valve
viscosity corrections 258
pulp and paper stock See non-
Newtonian waterhammer theory 390
Pump junction 464 pump speed 391
as assigned flow 255 Pumps 144, 365
check valve 246 efficiency 248, 249
four quadrant 253, 254, 399, 404
8/10/2007 4:10:28 PM Page 515

power usage 248, 249 Specifications window 259
Pump configurations 246 surge reduction 442
sizing method 455 Transient valve cracking 260
sizing with flow control valves 464 waterhammer theory 406
submerged 394 Renumber Increment 86
viscosity corrections 258 Renumber Wizard 86
Renumbering See Workspace window,
R See Workspace window
README.TXT 5 Repeat transient 436
Reference Information 97, 188 Reservoir junction 15, 60
Reference positive flow direction Difference between a Reservoir and
Surge Tank 270
See positive flow direction
How pressure junctions work 455
Reference Pressure 459
loss factors 62
References 495
Specifications window 29, 34, 261
Relative location 16
the role of pressure junctions 458
Relative roughness 62, 200, 202, 337
Transient Data 263
Relative tolerance See Steady Solution
Control window use in open and closed systems 332
Relaxation See Steady Solution waterhammer theory 380
Control window resistance 375, 379
using for troubleshooting models variable during transient 63
450
Resistance
Relief Valve junction
Hydraulic resistance friction model in
entering transient data 260 pipe 62, 202, 338
Exit 407 Resistance Curve 242
inherent event logic 434 Reverse Direction 178
Internal 406 Reynolds number
Internal vs. exit relief valves 259 effects during transient flow 159
loss factors 62, 230 transition 150, 337

Index 517
S in Output window 93
Scale/flip workspace See Workspace Show Object Status 151

window single event transients 430
Scaling See ID Reduction - Scaling Solution Balance Summary table 335
Scenario Manager 166 Solution Progress window 46, 93, 351
re-establishing broken links 176 iteration history 351
Scenario logic 172 View Output button 93
Viewing scenario differences 169 Solver
Scenarios discussion of Steady-State and
merging models with multiple Transient 46
scenarios 184 output file for transient data 47, 94
Model Data window ancestor display Recommendations on Steady Solver
91 190
same as parent junction 214 Steady-State 138, 189, 325, 344
same as parent pipe 198 Transient 190, 370
Section Pipes window 41, 139, 152, Transient Solver control 157
157, 190, 379, 473, 474
verification of steady-state 335
Select Flow Path 80, 81, 129
working with the steady and transient
Select Graph Data window 107 solvers 189
Select Special 80, 178 Sort output feature 93
Selection Drawing tool 14, 73, 80, 129, using for troubleshooting models
477 448
selecting objects completely or Sparger 264
partially in box 294
Special Conditions 180, 452
sequential dual events 433, See Event
Transients Ignore 184
Setup program 6 junctions 219
SHIFT key See Keyboard modifiers, No reflections - infinite pipe 183

See Keyboard modifiers, See Transient 183
Keyboard modifiers
Workspace appearance 293
Show list box
8/10/2007 4:10:28 PM Page 517

Specifications windows 87, 193 relative tolerance 346
accessing 193 relaxation 348
inspecting in 135 relaxation example 349
Jump feature 194 tolerance 345
Splitting pipes 128, 479 tolerance example 347
Spray Discharge junction Steam Tables See ASME Steam Tables
Graphing Spray Discharge data 265 Submerged pumps 394
in output file 161 Support 493
Specifications window 263 Annual Support 493
Transient Data 265 e-mail 494
waterhammer theory 388 surge flow
Stagnation pressure See waterhammer
at pressure control valves 332 surge suppression 441
explanation 329 Surge Tank junction 17
Static pressure as a branch during steady-state 269
at pressure control valves 332 as a finite size reservoir during
steady-state 270
explanation 329
as a pressurizer during steady-state
Status Bar 136
270
Steady Solution Control window 138,
connector pipe 267
139
Difference between a Reservoir and
absolute tolerance 346
Surge Tank 270
convergence 345
flow restrictor 267
false convergence 350
Graphing Surge Tank data 271
flow rate relaxation 349
in output file 161
Matrix Method 351
lumped inertia 422
maximum iterations 351
Model as dipping tube vessel 270
parameters 344
Only Allow Outflow 268
pressure relaxation 350
orifice 267

Index 519
Special features for steady flow 268 Simple, no loss 272
Specifications window 33, 265 Specifications window 271
surge reduction 443 waterhammer theory 384
Tank area 266 Terminology 1
Tank geometry 266 time step 378, 474, 475
Tank height 266 Time-based transients 429
Transient Data 268 Tip of the Day 302
waterhammer theory 412 Title 97, 117
surge transients Tolerance
See waterhammer balance verification 335
System pressure 166, 458 using for troubleshooting models
451
System Properties window 26, 138,
142 Toolbars 122
AFT Standard fluid database 145 Toolbox 72, 79, 125, 128
ASME Steam Tables 145 customizing See Toolbox
Preferences window
atmospheric pressure 149
shortcuts See Workspace
Chempak database 146
Toolbox Preferences window 300
creating mixtures 146
Current Toolbox/Unused Tools 300
fluid list 307
gravitational acceleration 149
scenarios 171
Transition Reynolds number 150
Manager 317
ToolTips 79
T
Transfer Results to Initial Guess 93,
Technical Support See Support 186
Tee/Wye junction 211 Transient Control window 44, 139,
convergence problems and 157, 190, 241, 255, 258, 265, 271,
recommendations 450 276, 429, 473
loss factors 62, 272, 359 artificial transients 474
8/10/2007 4:10:28 PM Page 519

Defining force sets 161 U
Estimated file size and run time 162 Undefined Objects window 29, 133,
Include in transient output file 160 151
Model transient cavitation 159 Undo 80, 130
Save Output to File Every Time Step Unit Preferences See Parameter and
158 Unit Preference window
Saving junction output 161 Unspecified fluid 27
Saving pipe output 160
scenarios 171
V
Vacuum Breaker Valve junction
specifying content for output file 95
Graphing Vacuum Breaker Valve
Start and stop time 157
data 276
Stop run if artificial transient detected
in output file 161
159
inherent event logic 434
variable resistance 64, 158
Transient Data
surge reduction 443
Absolute vs. relative 435
Valve coefficient 363
Repeat transient 436
Valve junction 19, 60
Transient output file See Output file
Exit valve 273
Transient special conditions See
Special Conditions loss factors 62, 356, 358
Transient vapor cavitation See vapor Special Conditions 274
cavitation
Specifications window 34, 35, 273
Transition flow 150, 337
Transient Data 274
Transition Reynolds number 150
Transport properties 366
vapor cavitation 44, 156
Troubleshooting steady-state models
transient 376, 424
446
vapor cavitation waterhammer theory
Turbulent flow 150, 337
assigned flow 381
assigned pressure 380

Index 521
branch 384 General Display 117
check valve 410 input-only mode 119
dead end 382 legend 117, 121
gas accumulator 416 Save Options 121
infinite pipe 388 Show Selected Pipes and Junctions
118
liquid accumulator 406
Visual Report Control 116
pump 394
Volume Balance junction
spray discharge 390
surge tank 413
valve 386
W
Vapor pressure 144
Walk Through Example 13
Variable fluid properties 143, 208
wall thickness 200
variable pipe resistance 63, 158
Water Data from ASME Steam Tables
Verification 28
Verification against published results Water properties from the ASME Steam
7 Tables See ASME Steam Tables
Verifying steady-state network waterhammer 1, 64
solutions 335
description 64
View Output button See Solution
Progress window instantaneous 65, 367
Viscosity corrections 258 Waterhammer Assistant 176
Visual Report Control window 116, wavespeed 38, 42, 64, 66, 152, 153,
304 203, 368, 370, 377, 378, 424
scenarios 171 assumption that it remains constant

5
Manager 317 Calculated 199
Visual Report window 8, 55, 71, 114, User Specified 200

304 Web site 494
Color Map 119 Workspace See Workspace window
Display Parameters 117 grid 296
8/10/2007 4:10:28 PM Page 521

Keyboard modifiers 477 Background Picture Scaling 298
Reverse Direction 21 changing junction icons 220
shortcuts 301, 477 colors and fonts 298
symbols 296 customizing 83, 292
Workspace Preferences window 129, Displaying name and/or ID numbers
151, 180, 183, 219, 292 297
Action When Dragging Junctions editing 80
294
finding pipes and junctions 177
Allowable Workspace Label
flipping 84
Movements 297
pipe lengths on 199
Background Picture Scaling 298
Popup menu 297
colors and fonts 298
relative locations 16
Renumber Wizard 86
display options 295
renumbering 85
displaying name and/or ID numbers
297 scaling 84
icon size 294 setting the junction label position
See Workspace Preferences
Pipe endpoint adjustments 294
window
pipe line options 292
size 83
Popup menu 297
Special Conditions Graphics 298
Sample Workspace 298
zoom feature 83
special conditions 293 Z
Special Conditions Graphics 298 Zoom feature 83, 115
symbols 296 Zoom Select tool 15, 77, 479
Transient indicators on the Zoom to Fit 115
Workspace 437
Manager 317
Workspace window 8, 14, 71, 72

Impulse 4.0 Users Guide SI PDF

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Impulse 4.0 Users Guide SI PDF

Hochgeladen von

Copyright:

Verfügbare Formate

TM

AFT Impulse version 4.0

Applied Flow Technology

8/10/2007 4:10:28 PM Page 1 c:\aft manuals\i4\impulse4 manual

© 2007 Applied Flow Technology Corporation. All rights reserved.

8/10/2007 4:10:28 PM Page 2 c:\aft manuals\i4\impulse4 manual

2. A Walk Through AFT Impulse...................................... 13

4. The Five Primary Windows.......................................... 71

5. Building and Running Models................................... 125

6. Pipe and Junction Specifications Windows ............ 193

7. Customizing AFT Impulse ......................................... 287

8. Steady-State Hydraulic Theory and Loss Models.... 325

9. Theory of Waterhammer & Solution Methodology .. 367

10. Modeling Time and Event Based Transients ......... 429

8/10/2007 4:10:28 PM Page iii c:\aft manuals\i4\impulse4 manual

11. Reducing Surge Pressures ..................................... 441

12. Special Topics: Troubleshooting Models .............. 445

Appendix A. Keyboard Modifiers and Shortcuts ......... 477

Appendix B. Limitations ................................................ 483

Appendix C. Installation Issues .................................... 485

Appendix D. Chempak Technical Information ............. 491

Appendix E. Obtaining Technical Support................... 493

References ...................................................................... 495

Glossary .......................................................................... 497

8/10/2007 4:10:28 PM Page iv c:\aft manuals\i4\impulse4 manual

8/10/2007 4:10:28 PM Page v

Converting models and databases from AFT Fathom or Mercury ......... 10

2. A Walk Through AFT Impulse...................................... 13

8/10/2007 4:10:28 PM Page vi c:\aft manuals\i4\impulse4 manual

B. Specify output control ................................................................. 25

8/10/2007 4:10:28 PM Page vii

Incorporating equivalent length data ......................................... 61

4. The Five Primary Windows.......................................... 71

8/10/2007 4:10:28 PM Page viii c:\aft manuals\i4\impulse4 manual

8/10/2007 4:10:28 PM Page ix

Show selected pipes/junctions................................................... 99

8/10/2007 4:10:28 PM Page x c:\aft manuals\i4\impulse4 manual

5. Building and Running Models................................... 125

8/10/2007 4:10:28 PM Page xi

The fluid databases ........................................................................ 145

8/10/2007 4:10:28 PM Page xii c:\aft manuals\i4\impulse4 manual

Exporting transient force results .................................................... 165

8/10/2007 4:10:28 PM Page xiii

Steady-state solver ......................................................................... 189

6. Pipe and Junction Specifications Windows ............ 193

8/10/2007 4:10:28 PM Page xiv c:\aft manuals\i4\impulse4 manual

Support for Partially Full Pipe................................................. 206

8/10/2007 4:10:28 PM Page xv

Special Conditions ......................................................................... 224

8/10/2007 4:10:28 PM Page xvi c:\aft manuals\i4\impulse4 manual

Do not allow backwards rotation............................................. 246

8/10/2007 4:10:28 PM Page xvii

Spray Discharge Specifications window .............................................. 263

7. Customizing AFT Impulse ......................................... 287

8/10/2007 4:10:28 PM Page xviii c:\aft manuals\i4\impulse4

Parameter Preferences.................................................................... 288

8/10/2007 4:10:28 PM Page xix

Database connections..................................................................... 302

8. Steady-State Hydraulic Theory and Loss Models.... 325

8/10/2007 4:10:28 PM Page xx c:\aft manuals\i4\impulse4 manual

Pressure drop calculation methods for Newtonian fluids .............. 336

8/10/2007 4:10:28 PM Page xxi

Bend ............................................................................................... 356