Impulse 3.0 Users Guide

AFT Impulse
TM
Users Guide
AFT Impulse version 3.0

Waterhammer Modeling in Piping Systems
Applied Flow Technology
2/24/2003
1b.doc
1:35 PM
Page 1
c:\aft manuals\i3\impulse3 manual
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 Users
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.
2003 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.
2/24/2003
1b.doc
1:35 PM
Page 2
Contents
Summary
Summary..................................................................................................iii
Detailed Contents ..................................................................................... v
Nomenclature....................................................................................... xxix
1. Introduction .................................................................... 1
2. A Walk Through AFT Impulse...................................... 15
3. Fundamental Concepts................................................ 59
4. The Five Primary Windows.......................................... 71
5. Building and Running Models................................... 121
6. Pipe and Junction Specifications Windows ............ 179
7. Customizing AFT Impulse ......................................... 271
8. Steady-State Hydraulic Theory and Loss Models.... 311
9. Theory of Waterhammer & Solution Methodology .. 355
10. Modeling Time and Event Based Transients ......... 413
2/24/2003
1b.doc
1:35 PM
Page iii
iv AFT Impulse 3.0 Users Guide
11. Reducing Surge Pressures ..................................... 425

12. Special Topics & Troubleshooting Models ............ 429
Appendix A. Keyboard Modifiers and Shortcuts ......... 461
Appendix B. Limitations ................................................ 467
Appendix C. Installation Issues .................................... 469
Appendix D. Chempak Technical Information ............. 477
Appendix E. Obtaining Technical Support................... 479
References ...................................................................... 481
Glossary .......................................................................... 483
2/24/2003
1b.doc
1:35 PM
Page iv
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 AFT Impulse 1.0 and 2.0........... 9
Pipe friction........................................................................................ 9
Infinite pipe junctions ........................................................................ 9
Transient junction data....................................................................... 9
Junction loss models .......................................................................... 9
Reservoir transients............................................................................ 9
Databases ......................................................................................... 10
2/24/2003
1:35 PM
Page v
c:\aft manuals\i3\impulse3 manual 1b.doc
vi AFT Impulse 3.0 Users 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
Whats new in version 3.0 ...................................................................... 11
General ............................................................................................. 11
Pipes and Junctions .......................................................................... 11
Workspace........................................................................................ 12
Output............................................................................................... 12
Graph Results................................................................................... 12
Visual Report ................................................................................... 13
Miscellaneous................................................................................... 13
2. A Walk Through AFT Impulse...................................... 15

Step 1. Start AFT Impulse ...................................................................... 15
The Workspace window ............................................................ 16
Step 2. Lay out the model ....................................................................... 17
A. Place the first reservoir ............................................................... 17
Objects and ID numbers ............................................................ 17
Editing on the Workspace.......................................................... 18
B. Place the first branch ................................................................... 18
C. Place surge tank........................................................................... 20
D. Place the second reservoir........................................................... 20
E. Add the remaining junctions........................................................ 21
F. Draw a pipe between J1 and J2.................................................... 22
Reference positive flow direction.............................................. 23
G. Add the remaining pipes ............................................................. 24
Step 3. Set the unit preferences .............................................................. 25
Step 4. Complete the first three checklist requirements ......................... 26
A. Specify Steady Solution Control ................................................. 27
B. Specify output control ................................................................. 27
C. Specify system properties............................................................ 28
Step 5. Define the model components (checklist item #4) ..................... 30
2/24/2003
1b.doc
1:35 PM
Page vi
Table of Contents vii

Object status .............................................................................. 31
A. Define Reservoir J1..................................................................... 31
The Inspection feature ............................................................... 33
B. Define Branch J2 ......................................................................... 34
C. Define Surge Tank J3 .................................................................. 35
D. Define Reservoir J4 and Branches J5 and J6 .............................. 36
E. Define Valve J7 ........................................................................... 36
F. Define Relief Valve J8................................................................. 37
G. Define Pipe P1............................................................................. 38
The Pipe Specifications window ............................................... 39
H. Define Pipes P2 through P7 ........................................................ 40
Reviewing input in Model Data window................................... 41
Step 6. Complete the last two checklist requirements............................ 43
A. Specify pipe sectioning ............................................................... 43
B. Specify transient control.............................................................. 45
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 ............................................................ 50
B. Graph the results.......................................................................... 51
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
Incorporating equivalent length data ......................................... 61
Convention for specifying junction base area.................................. 61
Specifying losses.............................................................................. 62
Modeling valve closures with K factors .......................................... 62
2/24/2003
1:35 PM
Page vii
viii AFT Impulse 3.0 Users Guide

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.............................................................. 69
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 Selection Drawing Tool ............................................................ 73
Selecting objects completely or partially inside the box........... 73
The Pipe Drawing Tool.................................................................... 74
Pipe handles and segmenting a pipe .......................................... 75
The Zoom Select Tool...................................................................... 77
The Annotation Tool........................................................................ 78
Junction icons................................................................................... 79
Editing features ................................................................................ 79
Selecting groups of objects........................................................ 80
Selecting objects in a flow path................................................. 80
Aligning objects......................................................................... 81
Creating and Managing Saved Groups ............................................ 81
Bookmarks on the Workspace.......................................................... 82
Last View ......................................................................................... 82
Panning the Workspace.................................................................... 82
2/24/2003
1b.doc
1:35 PM
Page viii
Table of Contents ix
Customizing features........................................................................ 82
Workspace Preferences ............................................................. 82
Adjusting the Workspace size ................................................... 83
Selective display of pipes and junctions.................................... 83
Scale/flip workspace.................................................................. 84
Renumbering the Workspace objects ........................................ 84
Using the Renumber Wizard ..................................................... 85
Background Graphic ........................................................................ 85
Specifications windows.................................................................... 86
The Model Data window ........................................................................ 86
Display and printing features ........................................................... 88
The Model Data Control window .................................................... 88
Pipe data display........................................................................ 88
Junction data display ................................................................. 88
Show selected pipes and junctions ............................................ 89
Database connections ................................................................ 89
The Output window ................................................................................ 90
Transient output files ....................................................................... 91
The Output Control window ............................................................ 92
Pipe and Junction parameters .................................................... 92
Pipe and Junction Transient parameters.................................... 94
General output ........................................................................... 94
Summaries ................................................................................. 94
Format and action ...................................................................... 95
Show selected pipes/junctions................................................... 96
Database connections ................................................................ 97
Command buttons...................................................................... 97
Output parameter descriptions................................................... 98
Steady flow results ........................................................................... 98
Transient flow results for pipes ....................................................... 99
Transient Max/Min results for pipes ....................................... 100
Summary Max/Min results ...................................................... 101
Transient flow results for junctions ............................................... 103
2/24/2003
1:35 PM
Page ix
x AFT Impulse 3.0 Users Guide

Transient event messages ........................................................ 103
Output window updates ................................................................. 104
The Graph Results window .................................................................. 104
The Select Graph Data window ..................................................... 105
Transient pipe results............................................................... 106
Transient junction results ........................................................ 106
Profile Along a Flow Path and EGL, HGL and Elevation Profile106
Animating profile results......................................................... 107
Other graph controls ................................................................ 110
The Customize Graph window....................................................... 110
The Auxiliary Graph Formatting window...................................... 111
The Visual Report window................................................................... 111
The Visual Report Control window............................................... 112
Display parameters .................................................................. 113
General display........................................................................ 113
Show selected pipes and junctions .......................................... 115
Color map ................................................................................ 115
Command buttons.................................................................... 118
Visual Report annotations.............................................................. 118
Annotations for the Workspace ............................................... 118
Creating Visual Report annotations ........................................ 119
The Toolbars......................................................................................... 119
5. Building and Running Models................................... 121

Creating objects .................................................................................... 121
Pipes ............................................................................................... 121
Junctions......................................................................................... 124
Morphing junctions ................................................................. 124
Splitting pipes.......................................................................... 124
Annotations .................................................................................... 125
Moving objects ..................................................................................... 125
Keyboard modifiers........................................................................ 126
Editing objects ...................................................................................... 126
2/24/2003
1b.doc
1:35 PM
Page x
Table of Contents xi
Connecting objects ............................................................................... 126
Defining objects.................................................................................... 128
Specifying required property data.................................................. 128
Highlighting required information ................................................. 129
Use status feature ........................................................................... 129
Satisfying connectivity requirements............................................. 129
Inspecting objects ................................................................................. 130
Using the Checklist............................................................................... 132
Analysis Type................................................................................. 133
Specify Steady Solution Control.................................................... 134
Specify Output Control .................................................................. 134
Specify System Properties.............................................................. 134
Define All Pipes and Junctions ...................................................... 134
Section Pipes .................................................................................. 135
Transient Control ........................................................................... 135
Steady Solution Control........................................................................ 135
Output Control...................................................................................... 136
System Properties ................................................................................. 138
System fluid property variation...................................................... 139
Density and dynamic viscosity....................................................... 140
Bulk modulus of elasticity ............................................................. 140
Vapor pressure and cavitation........................................................ 140
The fluid databases ........................................................................ 141
The AFT Standard Database ................................................... 141
The Chempak database............................................................ 141
Accuracy option....................................................................... 142
Viscosity models ............................................................................ 143
Non-Newtonian flow in non-pipe elements............................. 144
Atmospheric pressure..................................................................... 144
Gravitational acceleration .............................................................. 145
Transition Reynolds Numbers ....................................................... 145
Editing the AFT Standard fluid database....................................... 146
Finding object definition status ............................................................ 146
2/24/2003
1:35 PM
Page xi
xii AFT Impulse 3.0 Users Guide

Section Pipes......................................................................................... 147
Numbering convention................................................................... 149
Transient Control.................................................................................. 151
Start and stop time.......................................................................... 151
Save output to file ...................................................................... 152
Use variable pipe resistance........................................................... 152
Model transient cavitation.............................................................. 153
Stop run if artificial transient detected........................................... 153
Include in transient output file ....................................................... 154
Saving pipe output ................................................................... 154
Disadvantages to saving all stations ........................................ 154
Advantages to saving all stations............................................. 155
Saving junction output ................................................................... 155
Estimated file size and run time..................................................... 155
Defining pressures for the system ........................................................ 155
Using scenario manager........................................................................ 156
The Scenario Manager window ..................................................... 156
Creating, organizing and editing scenarios .................................... 157
Modifying individual scenarios ..................................................... 159
Passing changes to child scenarios................................................. 160
Re-establishing broken links ................................................... 161
Fast Scenario Changes ................................................................... 161
Special modeling features..................................................................... 162
Waterhammer Assistant ................................................................. 162
Workspace Find ............................................................................. 163
Reverse Direction........................................................................... 164
Select Special ................................................................................. 164
Special Conditions ......................................................................... 165
Special conditions with no transient data ................................ 167
Special conditions with transient data ..................................... 167
Pump special conditions .......................................................... 167
Special conditions only applied to junctions........................... 168
Transient special conditions........................................................... 169
2/24/2003
1b.doc
1:35 PM
Page xii
Table of Contents xiii

No reflections infinite pipe................................................... 169
Special Condition Ignore................................................................ 169
Merging models ............................................................................. 170
Merging models with multiple scenarios................................. 170
Print Preview/Special..................................................................... 170
Transfer Results to Initial Guesses ................................................ 172
Batch mode..................................................................................... 172
Running Scenarios in Batch Mode .......................................... 173
Print Format ................................................................................... 174
Extended Model Check.................................................................. 175
Math calculator .............................................................................. 175
Working with the steady and transient solvers..................................... 175
Steady-state solver ......................................................................... 175
Recommendations on Steady Solver ....................................... 176
Transient solver.............................................................................. 176
Artificial transients .................................................................. 176
6. Pipe and Junction Specifications Windows ............ 179

Highlight feature ............................................................................ 180
Jump feature................................................................................... 180
Pipe Specifications window ................................................................. 180
Common input parameters ............................................................. 183
Pipe number ............................................................................. 183
Pipe name ................................................................................ 183
Same As Pipe........................................................................... 184
Scenario same as parent........................................................... 184
Connected junctions ................................................................ 184
Pipe Model ..................................................................................... 184
Pipe length ............................................................................... 185
Pipe material ............................................................................ 185
Pipe diameter ........................................................................... 186
Pipe wall thickness .................................................................. 186
Pipe modulus of elasticity ....................................................... 187
2/24/2003
1:35 PM
Page xiii
xiv AFT Impulse 3.0 Users Guide

Pipe Poisson ratio .................................................................... 187
Friction Model ......................................................................... 187
Pipe Support ............................................................................ 189
Pipe Wavespeed....................................................................... 189
Additional Losses........................................................................... 189
Design Alerts.................................................................................. 191
Optional input ................................................................................ 191
Initial Flow Rate Guess ........................................................... 191
Workspace display................................................................... 192
Design Factor........................................................................... 192
Pipe Line Size and Color ......................................................... 192
Parallel Pipes ........................................................................... 192
Support for Partially Full Pipe................................................. 192
Intermediate Pipe Elevations ................................................... 193
Fluid Properties .............................................................................. 194
Notes .............................................................................................. 194
Status.............................................................................................. 194
Junction Specifications windows ......................................................... 194
Format #1: Junctions with one or two connecting pipes............... 197
Format #2: Junctions with more than two connecting pipes.......... 198
Parameters common to all junction Specifications windows ............... 199
Junction Number ............................................................................ 199
Junction Name................................................................................ 199
Junction Elevation.......................................................................... 200
Database List.................................................................................. 201
Same As Junction list..................................................................... 201
Scenario same as parent........................................................... 201
Pipe connectivity............................................................................ 201
Jumping to another junction........................................................... 202
Transient Data ................................................................................ 202
Repeat Transient...................................................................... 203
Transient Special Conditions................................................... 204
Time based and event based transients.................................... 204
2/24/2003
1b.doc
1:35 PM
Page xiv
Table of Contents xv
Initial Pressure/Head Guess ........................................................... 204
Displaying junction names and numbers ....................................... 206
Design Factor ................................................................................. 206
Changing the icon graphic ............................................................. 207
Changing the icon size ................................................................... 207
Specifying Base Area..................................................................... 207
Notes .............................................................................................. 207
Status.............................................................................................. 208
Area Change Specifications window.................................................... 208
Assigned Flow Specifications window................................................. 209
Transient Data ................................................................................ 211
Special features for steady flow..................................................... 211
Assigned Pressure Specifications window ........................................... 212
Transient Data ................................................................................ 214
Branch Specifications window ............................................................. 214
Transient Data ................................................................................ 216
Check Valve Specifications window.................................................... 217
Specifying losses............................................................................ 218
Forward Velocity to Close Valve................................................... 218
Delta Pressure to Re-Open ............................................................. 218
Transient Data ................................................................................ 218
Control Valve Specifications window.................................................. 219
Control Valve types ....................................................................... 220
PRV/PSV static vs. stagnation pressure .................................. 220
Action if control is lost .................................................................. 221
Open Percentage Table .................................................................. 222
Transient Data ................................................................................ 222
Failure Transient...................................................................... 222
Control Transient..................................................................... 223
Dead End Specifications window......................................................... 223
2/24/2003
1:35 PM
Page xv
xvi AFT Impulse 3.0 Users Guide

Gas Accumulator Specifications window............................................. 224
Initial conditions and polytropic process ....................................... 225
Maximum and minimum volumes.................................................. 226
Interface with pipe system ............................................................. 226
Graphing Gas Accumulator data.................................................... 229
General Component Specifications window ........................................ 229
Liquid Accumulator Specifications window ........................................ 231
Initial conditions and elastic process ............................................. 231
Pump Specifications window ............................................................... 232
Pump Model ................................................................................... 232
Submerged pump ..................................................................... 233
Do not allow backwards rotation............................................. 234
Check valve at pump discharge ............................................... 234
Pump configurations ...................................................................... 234
Controlled discharge pressure or flow (variable speed) ................ 234
NPSHR and NPSPR (optional in pump configuration) ................. 237
Efficiency and power usage (optional in pump configuration)...... 238
Transient Data ................................................................................ 238
Pumps with curves................................................................... 238
No transient from pump, no back flow.................................... 238
No transient from pump, but back flow possible..................... 238
Known speed transients, no back flow .................................... 239
Pump startup transient with possible reverse flow.................. 240
Pump transients with inertia .................................................... 240
Trip with inertia and no back flow or reverse speed ............... 240
Startup with inertia and no back flow or reverse speed .......... 241
Trip with inertia four quadrant ............................................. 241
Startup with inertia - four quadrant, known motor torque/speed243
Estimating the pump inertia..................................................... 243
Event modeling including trips followed by restarts or vice
versa......................................................................................... 243
Controller transient.................................................................. 244
2/24/2003
1b.doc
1:35 PM
Page xvi
Table of Contents xvii

Pumps as assigned flow........................................................... 246
Representing Multiple Pumps With One Junction......................... 247
Viscosity corrections...................................................................... 247
Graphing pump data....................................................................... 248
Relief Valve Specifications window .................................................... 248
Three relief valve configurations: internal, exit and inline............ 249
Specifying losses............................................................................ 249
Transient valve cracking ................................................................ 249
Transient Data ................................................................................ 250
Transients with variable Cv ........................................................... 250
Reservoir Specifications window ......................................................... 250
Transient Data ................................................................................ 253
Spray Discharge Specifications window .............................................. 253
Loss Model..................................................................................... 254
Transient Data ................................................................................ 255
Graphing Spray Discharge data ..................................................... 255
Surge Tank Specifications window ...................................................... 255
Tank geometry................................................................................ 256
Interface with pipe system ............................................................. 257
One-Way Surge Tank..................................................................... 258
Transient Data ................................................................................ 258
Surge tank as a branch during steady-state.............................. 259
Surge tank as a pressurizer during steady-state ....................... 259
Surge Tank as a finite size reservoir during steady-state ........ 260
Difference between a Reservoir and Surge Tank .................... 260
Model as dipping tube vessel ......................................................... 260
Graphing Surge Tank data ............................................................. 261
Tee/Wye Specifications window.......................................................... 261
Loss factors .................................................................................... 261
Valve Specifications window ............................................................... 263
Transient Data ................................................................................ 264
2/24/2003
1:35 PM
Page xvii
xviii AFT Impulse 3.0 Users Guide

Vacuum Breaker Valve Specifications window................................... 264
Graphing Vacuum Breaker Valve data .......................................... 266
Volume Balance Specifications window.............................................. 266
Global Edit windows ............................................................................ 267
Global Pipe editing......................................................................... 267
Global Junction editing .................................................................. 268
Editing specific junction data .................................................. 269
7. Customizing AFT Impulse ......................................... 271

Parameter and Unit Preferences ........................................................... 271
Parameter Preferences.................................................................... 272
Unit Preferences............................................................................. 273
Setting preferred units ............................................................. 275
Description of selected unit ........................................................... 275
Database connections..................................................................... 275
Command buttons .......................................................................... 276
Workspace Preferences......................................................................... 276
Pipes and Junctions ........................................................................ 276
Pipe Line Options .................................................................... 277
Special Conditions Options ..................................................... 278
Action When Drawing Selection Right To Left...................... 278
Modifying icon size ................................................................. 278
Action when dragging junctions.............................................. 279
Auto increment labels.............................................................. 279
Icon source............................................................................... 279
Display Options.............................................................................. 279
Workspace grid........................................................................ 279
Displaying source/sink indicator ............................................. 280
Special condition junction symbol .......................................... 281
Pipe direction arrows............................................................... 281
Displaying name and/or ID numbers ....................................... 281
Setting the default junction label position............................... 281
2/24/2003
1b.doc
1:35 PM
Page xviii
Table of Contents xix

Allowable Workspace label movements ................................. 281
Junction transient symbols ...................................................... 282
Colors and Fonts ............................................................................ 282
Sample Workspace......................................................................... 283
Command buttons .......................................................................... 284
Toolbox Preferences............................................................................. 284
Customizing Toolbox contents ...................................................... 285
General Preferences.............................................................................. 286
Customizing the Output........................................................................ 287
Specifying graph style preferences....................................................... 288
Specifying Visual Report preferences .................................................. 288
Customizing Additional Pipe Losses.................................................... 288
Building custom databases ................................................................... 290
Adding custom fluid properties to the Fluid Database .................. 292
Adding custom pipe materials to the Pipe Material Database ....... 294
Friction Data Sets .................................................................... 295
Pipe Physical Properties .......................................................... 296
Adding custom junction data to the Component Database ............ 296
Editing the Component Database................................................... 297
Database Manager ................................................................................ 297
Setting available databases............................................................. 298
Connecting and disconnecting to a database.................................. 298
Editing databases............................................................................ 300
Benefits of shared databases .......................................................... 302
Creating an enterprise-wide, network database system........................ 304
Overview ........................................................................................ 304
Creating database files ................................................................... 305
Sharing database files using DATABASE.LIB ............................. 307
Connecting to the external shared databases ................................. 309
8. Steady-State Hydraulic Theory and Loss Models.... 311
2/24/2003
1:35 PM
Page xix
xx AFT Impulse 3.0 Users Guide

The Steady-State Solver ....................................................................... 311
Stagnation vs. static pressure boundaries....................................... 315
When to use static pressure ..................................................... 317
Static pressure at pressure control valve ................................. 318
Open vs. closed systems ....................................................................... 318
Verifying network solutions ................................................................. 321
Pressure drop calculation methods ....................................................... 322
Pressure drop calculation methods for Newtonian fluids .............. 323
Roughness-based methods ............................................................. 323
Laminar flow ........................................................................... 323
Turbulent flow ......................................................................... 324
Transition flow ........................................................................ 324
Hydraulically smooth ..................................................................... 324
Hazen-Williams method................................................................. 324
MIT Equation for crude oil ............................................................ 324
Miller Turbulent method................................................................ 325
Frictionless pipes ........................................................................... 326
Pressure drop calculation methods for non-Newtonian fluids ............. 326
Duffy method for pulp and paper stock ......................................... 326
Brecht & Heller method for pulp and paper stock ......................... 327
Power Law non-Newtonian............................................................ 328
Bingham Plastic non-Newtonian.................................................... 328
Non-Newtonian flow through non-pipes ................................. 329
Design factors ....................................................................................... 329
Steady Solution Control parameters ..................................................... 329
Solution tolerance specification..................................................... 330
How tolerances relate to solution accuracy............................. 333
Relaxation ...................................................................................... 333
Flow Rate Relaxation .............................................................. 334
Pressure Relaxation ................................................................. 335
Avoiding false convergence .................................................... 335
Maximum Iterations ....................................................................... 336
Matrix Method ............................................................................... 336
2/24/2003
1b.doc
1:35 PM
Page xx
Table of Contents xxi

Solution Progress window and iteration history................................... 336
Warnings in the solution....................................................................... 337
Junction types and the Steady-State Solver.......................................... 338
Pressure specification junctions..................................................... 338
Flow specification junctions .......................................................... 339
Pressure modifying junctions......................................................... 340
Pipe joining junctions .................................................................... 340
Special junction.............................................................................. 340
Modeling irrecoverable losses.............................................................. 340
Loss models and reference material ..................................................... 341
Design factors ................................................................................ 342
Area change.................................................................................... 342
Bend ............................................................................................... 343
90 degree bends ....................................................................... 343
Non-90 degree bends ............................................................... 344
Valve .............................................................................................. 345
Orifice ............................................................................................ 345
Tee/Wye ......................................................................................... 346
Diverging Case ........................................................................ 348
Converging Case...................................................................... 349
K and CV (valve coefficient) .......................................................... 350
K for fire sprinklers........................................................................ 351
Pumps ................................................................................................... 352
Chempak thermophysical property database ........................................ 353
Accuracy options ..................................................................... 353
AFT Standard fluids ............................................................................. 353
Viscosity......................................................................................... 353
9. Theory of Waterhammer & Solution Methodology .. 355

Description of waterhammer ................................................................ 355
Instantaneous waterhammer ................................................................. 355
Wavespeed............................................................................................ 356
Communication time in pipes............................................................... 358
2/24/2003
1:35 PM
Page xxi
xxii AFT Impulse 3.0 Users Guide

Method of Characteristics..................................................................... 358
Momentum equation ...................................................................... 359
Mass continuity equation ............................................................... 359
Transient vapor cavitation .................................................................... 364
Pipe sectioning...................................................................................... 365
Effects of flow resistance ..................................................................... 367
Assigned Pressure theory...................................................................... 367
Assigned Pressure vapor cavitation theory .................................... 368
Reservoir theory ................................................................................... 368
Reservoir vapor cavitation theory .................................................. 368
Assigned Flow theory........................................................................... 368
Assigned Flow vapor cavitation theory.......................................... 369
Dead End theory ................................................................................... 369
Dead End vapor cavitation theory.................................................. 370
Branch theory ....................................................................................... 370
Branch vapor cavitation theory ...................................................... 371
Tee/Wye theory .................................................................................... 372
Valve theory ......................................................................................... 372
Exit valves...................................................................................... 373
Valve vapor cavitation theory ........................................................ 374
Infinite Pipe theory ............................................................................... 375
Infinite Pipe vapor cavitation theory.............................................. 376
Spray Discharge theory......................................................................... 376
Spray Discharge vapor cavitation theory ....................................... 378
Pump theory.......................................................................................... 378
Modeling pump as a known speed ................................................. 378
Pumps with higher order curve fits.......................................... 380
Backflow through the pump .................................................... 380
Flow rates past the pump maximum........................................ 381
Submerged pumps ................................................................... 381
Pump with known speed vapor cavitation theory .......................... 381
Modeling pump as with unknown speed........................................ 382
Trip With Inertia and No Back Flow or Reverse Speed ................ 383
2/24/2003
1b.doc
1:35 PM
Page xxii
Table of Contents xxiii

Startup With Inertia and No Back Flow or Reverse Speed............ 384
Trip With Inertia Four Quadrant................................................. 386
Startup With Inertia Four Quadrant, Known Motor Torque/Speed391
Estimating the pump inertia ........................................................... 391
Liquid Accumulator theory .................................................................. 392
Liquid Accumulator vapor cavitation theory ................................. 393
Relief Valve theory............................................................................... 393
Internal relief valve ........................................................................ 394
Exit relief valve.............................................................................. 394
Inline Relief Valve ......................................................................... 395
Relief Valve vapor cavitation theory ............................................. 395
Internal relief valves ................................................................ 395
Exit relief valves...................................................................... 395
Check Valve theory .............................................................................. 395
Check Valve vapor cavitation theory............................................. 395
Control Valve theory ............................................................................ 395
Flow control valves ........................................................................ 396
Pressure control valves................................................................... 396
Pressure drop control valves .......................................................... 396
Surge Tank theory................................................................................. 396
Surge Tank vapor cavitation theory ............................................... 398
Gas Accumulator theory....................................................................... 398
Gas Accumulator vapor cavitation theory...................................... 401
Vacuum Breaker Valve theory ............................................................. 401
Sonic inflow ................................................................................... 401
Subsonic inflow.............................................................................. 402
Sonic outflow ................................................................................. 402
Subsonic outflow............................................................................ 402
Equation of state............................................................................. 402
Sonic Inflow ............................................................................ 403
Subsonic Inflow....................................................................... 404
Subsonic Outflow .................................................................... 405
Sonic Outflow.......................................................................... 407
2/24/2003
1:35 PM
Page xxiii
xxiv AFT Impulse 3.0 Users Guide

Lumped inertia pipe with orifice theory............................................... 407
Interpreting cavitation results ............................................................... 409
Instantaneous waterhammer equation exceptions ................................ 410
Cavitation ....................................................................................... 411
Pipe size changes............................................................................ 411
Multiple pressure surge sources..................................................... 412
Line pack due friction .................................................................... 412
10. Modeling Time and Event Based Transients ......... 413

Time-based transients ........................................................................... 413
Event-based transients .......................................................................... 414
Single event transients ................................................................... 414
Dual event transients: cyclic and sequential .................................. 415
Cyclic events............................................................................ 416
Sequential events ..................................................................... 417
Junctions with inherent event logic ...................................................... 418
Check valve .................................................................................... 418
Relief valve .................................................................................... 418
Vacuum breaker valve.................................................................... 418
Thought experiment to further clarify event transients ........................ 419
Other transient features......................................................................... 419
Absolute vs. relative transient data ................................................ 419
Repeat transient.............................................................................. 420
Add time offset............................................................................... 420
Graphing the transient data ............................................................ 420
Event messages..................................................................................... 420
Transient indicators on the Workspace ................................................ 421
Transient data in Model Data ............................................................... 422
11. Reducing Surge Pressures ..................................... 425

General considerations ......................................................................... 425
Larger pipe sizes................................................................................... 426
Flexible hose......................................................................................... 426
2/24/2003
1b.doc
1:35 PM
Page xxiv
Table of Contents xxv

Slower or modified valve closure profiles............................................ 426
Parallel valves....................................................................................... 426
Relief valves ......................................................................................... 426
Vacuum breaker valves for low pressures............................................ 427
Surge tank ............................................................................................. 427
Gas accumulator ................................................................................... 427
12. Special Topics & Troubleshooting Models ............ 429

The philosophy of computer modeling................................................. 429
Troubleshooting steady-state models ................................................... 430
Use the Model Data window.......................................................... 430
Poor pump curve fits ...................................................................... 431
Use the Output window Sort feature.............................................. 434
Tee/Wye junction complexity........................................................ 434
Lower the flow rate relaxation....................................................... 434
Try absolute tolerance.................................................................... 435
Make initial flow rate guesses for pipes ........................................ 435
Review the Iteration History .......................................................... 435
Change boundary conditions.......................................................... 436
Turn off parts of the model ............................................................ 436
Break a large model into submodels .............................................. 436
Control valve failure issues in steady-state .......................................... 436
Infinite pipe junctions........................................................................... 437
Application example ...................................................................... 438
How pressure junctions work in steady-state ....................................... 439
The role of pressure junctions in steady-state ...................................... 442
Examples ........................................................................................ 443
Sizing pumps with flow control valves.......................................... 448
Pressure control valves................................................................... 453
Closing parts of a system ............................................................... 456
Effect of pipe sectioning on transient run times................................... 457
Artificial transients ............................................................................... 459
Manually detecting artificial transients.......................................... 460
2/24/2003
1:35 PM
Page xxv
xxvi AFT Impulse 3.0 Users Guide
Appendix A. Keyboard Modifiers and Shortcuts ......... 461

Selection Drawing Tool........................................................................ 461
Double-clicking tool ................................................................ 461
Control key .............................................................................. 461
Shift key................................................................................... 461
Alt key ..................................................................................... 462
Drag left to right ...................................................................... 462
Drag right to left ...................................................................... 462
Pipe Drawing Tool................................................................................ 462
Double-clicking tool ................................................................ 462
Control key .............................................................................. 462
Shift key................................................................................... 462
Zoom Select Tool ................................................................................. 463
Shift key................................................................................... 463
Panning the Workspace ........................................................................ 463
Dragging junctions from Toolbox ........................................................ 463
Control key .............................................................................. 463
Shift key................................................................................... 463
Inspecting pipes and junctions.............................................................. 464
Shift key................................................................................... 464
Control key .............................................................................. 464
Moving junctions with connected pipes ............................................... 464
Control key .............................................................................. 464
Pipe and junction specifications window ............................................. 464
F2 function key........................................................................ 464
F5 function key........................................................................ 464
Appendix B. Limitations ................................................ 467

Appendix C. Installation Issues .................................... 469
Customization files ............................................................................... 469
Special file control features.................................................................. 469
2/24/2003
1b.doc
1:35 PM
Page xxvi
Table of Contents xxvii

What is the Control.aft File?.......................................................... 469
Specifying where Chempak is located ........................................... 470
Special network installations (remote tokens) ............................... 470
Network installations............................................................................ 471
Problems loading the Graphics Server ................................................. 473
Appendix D. Chempak Technical Information ............. 477

Appendix E. Obtaining Technical Support................... 479
AFT's three levels of support................................................................ 479
1. Free Software Support for 180 Days.......................................... 479
2. Annual Support, Upgrade and Maintenance (SUM).................. 479
3. Waterhammer Modeling Consulting Support ............................ 479
Contacting AFT .................................................................................... 480
Telephone support.......................................................................... 480
Web site.......................................................................................... 480
E-Mail support ............................................................................... 480
Mail support ................................................................................... 480
References ...................................................................... 481

Glossary .......................................................................... 483
2/24/2003
1:35 PM
Page xxvii
xxviii AFT Impulse 3.0 Users Guide
2/24/2003
1b.doc
1:35 PM
Page xxviii
Nomenclature
a
wavespeed
cross-sectional flow area of a pipe
BM
parameter as defined after Equation 9.20
BP
CD
discharge coefficient
CM
CP
CV
valve coefficient
CHW
Hazen-Williams factor
diameter of a pipe
modulus of elasticity
pipe wall thickness
friction factor
fT
turbulent friction factor (used by Crane, 1988)
error value in mass balance
FB
Dimensionless pump torque parameter (see Chapter 9)
FW
Dimensionless pump head parameter (see Chapter 9)
gravitational constant
head ratio for pump
head
HGL
hydraulic gradeline (piezometric head)
moment of inertia of pump and entrained liquid
Jacobian matrix
loss factor
constant for Power Law fluid
KB
liquid bulk modulus of elasticity
Ksprinkler
fire sprinkler loss factor
length of a pipe
2/24/2003
1b.doc
1:35 PM
Page xxix
xxx AFT Impulse 3.0 Users Guide
m&
mass flow rate
general constant
constant for Power Law fluid
pump synchronous speed (rpm)
pressure, static
Po
pressure, stagnation
Pw
wetted perimeter
power
volumetric flow rate
radius
relaxation
R, R
resistance
Rc
coefficient of rigidity
Re
Reynolds Number
pump speed
SB
parameter as defined in Equation 9.27
SC
parameter as defined in Equation 9.28
Sy
yield stress
time
torque on pump
pump flow ratio to reference
velocity
distance along pipe centerline
elevation
, ,
angle
pump speed ratio
diameter ratio
torque ratio
method of characteristics multiplier
kinematic viscosity
roughness
2/24/2003
1b.doc
1:35 PM
Page xxx
Nomenclature xxxi
pump speed
density
dynamic viscosity
Subscripts
f
fluid
junction or pipe station at which solution is sought
junctions with pipes connecting to junction I
motor
rated conditions for a pump, usually BEP
stagnation
infinity, far away, ambient
new
current time for computation
old
most recent time for computation
2/24/2003
1:35 PM
Page xxxi
xxxii AFT Impulse 3.0 Users Guide
2/24/2003
1b.doc
1:35 PM
Page xxxii
CHAPTER 1
Introduction
Welcome to AFT Impulse 3.0 for Windows. AFT Impulse is a

graphical platform for modeling waterhammer and surge transients in
pipe networks.
AFT Impulses 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
2/24/2003
1b.doc
1:35 PM
Page 1
2 AFT Impulse 3.0 Users 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
2/24/2003
1b.doc
1:35 PM
Page 2
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.
2/24/2003
1:35 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 Unspecified (by Impulse) Fluids, for which data is
provided by the user.
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.
In the remainder of this Users 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
No chemical reactions
2/24/2003
1b.doc
1:35 PM
Page 4
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 Users Guide.
2/24/2003
1:35 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
30 MB
Floppy drive
Yes
CD-ROM drive
Yes
Monitor resolution
SVGA (800x600)
RAM
64 MB
MS-Windows
Windows 95, 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 3.0 directory.
3. Select Impulse 3.0 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 Users Guide. For an overview of the
model building process and AFT Impulses capabilities, read Chapter 2
and follow the instructions for building a simple model.
2/24/2003
1b.doc
1:35 PM
Page 6
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 Users 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.
2/24/2003
1:35 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.
2/24/2003
1b.doc
1:35 PM
Page 8
Converting models and databases from AFT Impulse 1.0 and 2.0
AFT Impulse 1.0 and 2.0 models can be opened directly by choosing
Open from the File menu. Once saved, they are saved in the version 3.0
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 3.0, 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 3.0 but in a different way. When
importing into version 3.0, 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 3.0. 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 3.0 convention of specifying the actual liquid surface level.
2/24/2003
1:35 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 and 5.0 models, and AFT Mercury 5.0 models can
be read directly by AFT Impulse 3.0 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 3.0 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 3.0 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.
2/24/2003
1b.doc
1:35 PM
Page 10
Whats new in version 3.0

General
New animation feature for graphing improves interpretation of

results
Pipes and Junctions
Pumps can be tripped or started based on events
Pumps can trip and then restart and vice versa
Pump four quadrant data sets increased from three to twenty-one
Anti-reverse rotation option for pumps
New pump startup method using motor torque and speed data
Pump pressure control on suction side implemented
New Special Condition Ignore feature for vacuum breaker valves,

surge tanks, gas accumulators, relief valves and liquid accumulators
simplifies locating surge suppression equipment
Reference gas volume for gas accumulators simplifies gas volume

initialization
Pump internal check valves can reopen
Pump can control pressure only when set pressure is exceeded high
or low
One pump can represent multiple pumps in parallel or in series
Vacuum Breaker Valves with one connecting pipe now allowed
Vacuum Breaker Valve can model dual vacuum break/air release, or

only vacuum break or air release
New feature allows modeling of three-stage air valves
One-Way Surge Tank modeling implemented
New relief valve pressure loss model with Cv vs. pressure difference
allows improved modeling of passive relief valves
New feature for inline relief valves
2/24/2003
1:35 PM
Page 11
Pipes can be modeled with diameter reduction because of scaling
Pipe size and schedule can be shown on the Workspace
Improved selection tool for junction icons
Pipes connected to vacuum breaker valves, exit valves, spray

discharge or assigned pressure junctions can drain or start the
transient partially full
Workspace
Annotation Manager simplifies working with annotations across

multiple scenarios
Renumber Wizard allows user to renumber pipes and junctions just

by clicking on them
Improved pipe segmenting features
Workspace pan feature
Toolbox shortcuts shown as text on Toolbox
Selecting objects from right to left selects all objects completely or

partially in selection box
Output
Event messages now separated into two lists one sorted by

junction and the other sorted by time
Graph Results
New animation feature allows user to replay transient with dynamic

graphics
Multiple profile plots can be cross-plotted using Workspace Groups
Cross-plotting for multiple junctions implemented
Pipe Outlet station can be saved into graph sets, and when model is
resectioned the outlet station is automatically updated
User has improved control over graph colors and curve thicknesses
and can now specify defaults
2/24/2003
1b.doc
1:35 PM
Page 12
Graph x-y data can be copied directly to the clipboard
Visual Report
User can now go directly to Visual Report from Solution Progress

window once run is complete
Visual Report window is automatically updated
More font flexibility
Miscellaneous
Support for output in languages other than English
New minimum absolute flowrate threshold for artificial transients

avoids artificial transients on high flowrate systems with some
sections initially at zero flowrate
Improved Global Junction Editing Tool allows pre-selection of

specific types of junctions
Brief explanation of Global Pipe Edit and Global Junction Edit

windows shown directly on window
More thorough use of junction and pipe names in all output
Archival of transient data for large models changed to improve

model loading time
2/24/2003
1:35 PM
Page 13
2/24/2003
1b.doc
1:35 PM
Page 14
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

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.
2/24/2003
1b.doc
1:35 PM
Page 15

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 waterhammer model on the Workspace using the
Toolbox tools. At the top of the Toolbox are four drawing tools. The
Selection Drawing tool, on the upper left, is useful for selecting groups
of objects on the Workspace for editing or moving. The Pipe Drawing
2/24/2003
1b.doc
1:35 PM
Page 16
Chapter 2 A Walk Through AFT Impulse 17

tool, on the upper right, 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

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.
2/24/2003
1:35 PM
Page 17

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 but less than 10,000.
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

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).
2/24/2003
1b.doc
1:35 PM
Page 18
Figure 2.2b Walk Through Model with one reservoir and one branch
placed
Figure 2.2c Walk Through Model after placing the Surge Tank
2/24/2003
1:35 PM
Page 19
C. Place surge tank

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).
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.
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).
Figure 2.2d Walk Through Model after placing the second Reservoir
2/24/2003
1b.doc
1:35 PM
Page 20
E. Add the remaining junctions

Next add the four remaining junctions - two branches, a regular valve
and a relief valve (Figure 2.2e).
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.
Add a valve junction and a relief valve junction 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.
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.
Figure 2.2e Walk Through Model after placing all junctions
2/24/2003
1:35 PM
Page 21
F. Draw a pipe between J1 and J2

Now that you have the eight junctions, you need to connect them with
pipes.
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.
Figure 2.2f Walk Through Model after drawing first pipe
2/24/2003
1b.doc
1:35 PM
Page 22

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).
Figure 2.2g Walk Through Model after placing first pipe
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.
The reference positive flow direction indicates which direction is
considered positive. 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.
2/24/2003
1:35 PM
Page 23
G. Add the remaining pipes

A quicker way to add a pipe is to draw it directly between the desired
junctions.
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 Relief Valve. Your model should now look similar to Figure
2.2h.
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.
2/24/2003
1b.doc
1:35 PM
Page 24
Figure 2.2h Walk Through Model with all pipes and junctions
placed
Step 3. Set the unit preferences

Open the Parameter and Units Preferences window from the Options
menu. This Walk Through model is in metric units. The default in AFT
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.
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.
2/24/2003
1:35 PM
Page 25
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

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.
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.
2/24/2003
1b.doc
1:35 PM
Page 26
Figure 2.4
The Checklist tracks the models 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.
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
2/24/2003
1:35 PM
Page 27

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.
You can model the fluid properties in one of four ways.
1. Unspecified fluid This fluid model allows you to directly type in
the density, viscosity, bulk modulus and vapor pressure.
2/24/2003
1b.doc
1:35 PM
Page 28

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.
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 is an optional add-on to AFT Impulse.
5. Water Data from ASME Steam Tables As its name implies, this
fluid model obtains water from ASME steam tables. This model is
pressure and temperature dependent.
2/24/2003
1:35 PM
Page 29
Figure 2.6
The System Properties window lets you enter physical

properties of the fluid
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 30

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
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.
A. Define Reservoir J1
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.
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 junctions name is the
junction type. The name can be displayed on the Workspace, Visual
Report or in the Output.
2/24/2003
1:35 PM
Page 31

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.
The Same As 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.
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.
2/24/2003
1b.doc
1:35 PM
Page 32
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.
2/24/2003
1:35 PM
Page 33
Inspection
window
Figure 2.8
Inspecting from the Workspace with right mouse button
B. Define Branch J2
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.
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.
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).
Click OK to accept the input and exit the window.
2/24/2003
1b.doc
1:35 PM
Page 34
C. Define Surge Tank J3

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.
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
2/24/2003
1:35 PM
Page 35
D. Define Reservoir J4 and Branches J5 and J6

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.
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.
Open the J6 Branch Specifications window. Enter an elevation of 50
meters. There is no imposed flow at this junction.
E. Define Valve J7
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).
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.
2/24/2003
1b.doc
1:35 PM
Page 36
Figure 2.10 Valve Specifications for J7 initiates the transient for the
system
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):
Time
0
Cv
2392.6
10
797.4
100
797.4
F. Define Relief Valve J8

Open the J8 Relief Valve Specifications window. Relief Valve
junctions can connect with one pipe if they relieve to an external source.
If they are internal relief valves they can connect to two pipes. This
2/24/2003
1:35 PM
Page 37

valve is an exit valve, so choose the Valve Type option as Exit. Enter an
elevation of 50 meters, a cracking pressure of 1.668 MPa, and an exit
pressure of 1 atm.
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 relief valve remains
closed until the cracking pressure is exceeded, at which point the valve
opens to relieve pressure.
Click the Transient Data tab and enter the data below for Cv.
Time
Cv
0
651.2
63
100
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.
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.
First open the Pipe Specifications window for Pipe P1 (Figure 2.11).
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.
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
2/24/2003
1b.doc
1:35 PM
Page 38

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
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.
2/24/2003
1:35 PM
Page 39
Figure 2.11 Pipe Specifications window for P1
H. Define Pipes P2 through P7

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
1
2
Length
(meters)
Diameter
(meters)
Friction
Factor
Wavespeed
(meters/sec)
1001.2
2000
1.5
1
0.012
0.013
2000
0.75
0.014
995
502.5
0.5
0.015
1000
502.5
0.5
0.015
1000
1001.2
0.014
996.3
2000.2
0.75
0.013
995.1
100.12
0.014
996.3
2/24/2003
1b.doc
1:35 PM
Page 40
996.3
995.3

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.
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 2.12. 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, or printed out
for review. Figure 2.13 shows an expanded view of the Transient Data
tab from Figure 2.12. Here all transient input data for the model is
shown.
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.
2/24/2003
1:35 PM
Page 41

Customize view
General data
Pipe data
Junction data
Figure 2.12 The Model Data window shows the input data in text
form
Figure 2.13 The Transient Data tab in the Model Data Junction data
area shows all transient data entered
2/24/2003
1b.doc
1:35 PM
Page 42
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
window helps divide the pipes into computation sections in a manner
which is consistent with the Method of Characteristics (MOC).
Select Section Pipes on the Analysis menu to display the Section Pipes
Window (Figure 2.14). 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:
ni =
Li
ai t
where n is the number of sections in pipe i, L is the length, and a is the

wavespeed. The t 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.
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.
2/24/2003
1:35 PM
Page 43

This model has been specified by Karney in such a way that the pipe
sectioning falls closely along whole numbers.
Figure 2.14 The Section Pipes window automates the sectioning

process and calculates the time step.
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
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.
2/24/2003
1b.doc
1:35 PM
Page 44

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.
Figure 2.15 The Transient Control window offers features to specify

the time span for the transient and what output data is
written
Select Transient Control on the Analysis menu to display the Transient

Control window (Figure 2.15). Enter zero for Start Time and 65 for Stop
Time.
2/24/2003
1:35 PM
Page 45

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 in the lower half of the window on the Pipe
Stations tab (see Figure 2.15). 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.
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.
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

Choose Run from the Analysis menu or click the arrow icon on the
toolbar. During execution, the Solution Progress window displays
(Figure 2.16). 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
2/24/2003
1b.doc
1:35 PM
Page 46

Solver, AFT Impulse uses the results to automatically initialize the
Transient Solver and then run it.
When the solution is complete, click the View Results 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, and printed out
on the printer.
Figure 2.16 The Solution Progress window displays the state of the
simulation
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
2/24/2003
1:35 PM
Page 47

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.
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 2.17) 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 2.18). 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).
2/24/2003
1b.doc
1:35 PM
Page 48

Customize view
General results
Pipe results
Junction results
Figure 2.17 The Output window displays output results in text form
2/24/2003
1:35 PM
Page 49
Figure 2.18 The Transient Output and Transient Max/Min results

tabs are in the Output window Pipe section area. These
show transient results for pipes.
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).
2/24/2003
1b.doc
1:35 PM
Page 50

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.
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.
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.
Select the Velocity parameter and use the Reorder scroll bar to move it
up to the top of the parameter list.
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.
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.
B. Graph the results

Change to the Graph Results window by choosing it from the
Windows menu, Toolbar, pressing Ctrl-G, clicking anywhere on the
window if its 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.
2/24/2003
1:35 PM
Page 51

Choose Select Graph Data from the View menu or the Toolbar to
open the Select Graph Data window (see Figure 2.19). 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 2.19 The Select Graph Data window controls the Graph
Results content
Click the Add >> button to add this station to the list on the right.
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 2.20).
This pressure/head history can be compared to that published in Karney
(1992). This is done in the Verifications documentation provided with
AFT Impulse.
2/24/2003
1b.doc
1:35 PM
Page 52
Figure 2.20 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 2.21). Select the X-Axis Units and HGL Units as meters.
Click the Show button.
2/24/2003
1:35 PM
Page 53
Figure 2.21 The Select Graph Data window allows selection of

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

Results window (see Figure 2.22). Push the Play button and AFT
Impulse will run through all of the results in pipes 1, 6 and 7 for 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 to fast, slow it down using the animation speed
control at the right.
2/24/2003
1b.doc
1:35 PM
Page 54
Figure 2.22 The Graph Results window allows animation of output

to be displayed. This is the animation shown at 8.1
seconds.
D. View the Visual Report

Change to the Visual Report window by choosing it from the Window
menu, Toolbar, pressing Ctrl-I, clicking anywhere on the window if its
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.
2/24/2003
1:35 PM
Page 55
Figure 2.23 The Visual Report Control window selects content for
the Visual Report window
Click the Visual Report Control button on the Toolbar (or View
menu) and open the Visual Report Control window, shown in Figure
2.23. 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 2.24).
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 2.24, 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, or saved to file.
2/24/2003
1b.doc
1:35 PM
Page 56
Figure 2.24 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 Users Guide for more detailed
information on each of the windows and functions.
2/24/2003
1:35 PM
Page 57
2/24/2003
1b.doc
1:35 PM
Page 58
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.
2/24/2003
1b.doc
1:35 PM
Page 59

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
2/24/2003
1b.doc
1:35 PM
Page 60
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 result in 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 100 feet long, and
has fittings with 20 feet of equivalent length, a design factor of 1.2 can
be used to account for this. This maintains the true physical length and
thus 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
2/24/2003
1:35 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
Additional 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 Additional Losses. This reduces the
number of pipes and model run time.
Modeling valve closures with K factors

In AFT Impulse 1.0, if you wanted to close a valve when using K factor
data, you had to specify extremely large values (10^10 or something like
that). In AFT Impulse 3.0, you can close a valve by specifying the 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 HazenWilliams factor, a hydraulically smooth roughness, an explicit friction
factor, 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.
2/24/2003
1b.doc
1:35 PM
Page 62
Comment on importance of hydraulic losses during

transients
Generally speaking, hydraulic losses (also called system resistance) from
pipes, fittings or equipment, has 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 Additional 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
2/24/2003
1:35 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 doesnt exist. What is meant by such statements is that
large, potentially destructive waterhammer doesnt 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 4000
foot (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 4000 feet/sec
(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).
2/24/2003
1b.doc
1:35 PM
Page 64

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 4000 feet/sec and the pipe is 4000 feet 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 supplys 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:
P = a V
(3.1)
where:
P =
pressure surge
density
wavespeed
V =
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.
2/24/2003
1:35 PM
Page 65

Heres 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.
ft
ft 1 ft 2
lbf s 2
Pmax = 62
* 4000 *10 *
*
+ 50 psig
s
s 144 in 2 32.2 lbm - ft
ft 3
Pmax = 535 + 50
lbm
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.
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:
t = 2
L
a
(3.2)
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
2/24/2003
1b.doc
1:35 PM
Page 66

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.
2/24/2003
1:35 PM
Page 67
V = Vsteady
V = Vsteady
V=0
a
V=0
a
Pinstantaneous
Psteady
Pinstantaneous
Psteady
Vsteady
x
Phase A: 0 < t < L / a
Phase B: L / a < t < 2L / a
V = Vsteady
V = Vsteady
V=0
V=0
P
Psteady
-Vsteady
P
Psteady
Pinstantaneous
x
Pinstantaneous
x
Vsteady
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
2/24/2003
1b.doc
1:35 PM
Page 68

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 steadystate 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.
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.
2/24/2003
1:35 PM
Page 69

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.
2/24/2003
1b.doc
1:35 PM
Page 70
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
2/24/2003
1b.doc
1:35 PM
Page 71

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), 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.
2/24/2003
1b.doc
1:35 PM
Page 72
Chapter 4 The Five Primary Windows 73
The Selection Drawing Tool

The drawing tool in the upper left corner 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.
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).
2/24/2003
1:35 PM
Page 73

Toolbars
Toolbox
Workspace
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 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 pipes starting point, hold down the left mouse button
while dragging the cursor to the pipe endpoint, then release the mouse
button.
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
2/24/2003
1b.doc
1:35 PM
Page 74

active, and you can draw a series of pipes without returning to the
Toolbox each time.
Selection Drawing Tool
Zoom Select Tool
Branch
Assigned Flow
Area Change
Tee/Wye
Control Valve
Dead End
Surge Tank
Pipe Drawing Tool

Annotation Tool
Reservoir
Assigned Pressure
Relief Valve
Valve
Check Valve
Gas Accumulator
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.
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
2/24/2003
1:35 PM
Page 75

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
Additional 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.
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. 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
2/24/2003
1b.doc
1:35 PM
Page 76

pipe to the desired location. By holding the SHIFT key the pipe segment
will move only at right angles.
Figure 4.5 The Pipe Segments window allows you to create, merge
and remove pipe segments.
The Zoom Select Tool

The second tool down from the upper left corner of the Toolbox 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%.
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%.
2/24/2003
1:35 PM
Page 77
The Annotation Tool

The second tool down from the upper right corner of the Toolbox 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 4.6a, 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 4.6a 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.
Annotations print out with the model. Figure 4.6b 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.
2/24/2003
1b.doc
1:35 PM
Page 78
Figure 4.6b 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.
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
2/24/2003
1:35 PM
Page 79

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
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 80

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. Also on this menu is the Adjust Pipes Automatically option,
which causes all pipe endpoints to adjust to the center of connected
junction icons.
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.
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.
2/24/2003
1:35 PM
Page 81

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.
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 82

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.
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
4.7).
Figure 4.7
Pipes and junctions can be selectively displayed on the

Workspace using the Display Pipes and Junctions
window on the View menu.
2/24/2003
1:35 PM
Page 83

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 4.8). 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
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 4.8
The Workspace can be scaled or flipped
Renumbering the Workspace objects

Objects on the Workspace can be renumbered by choosing Renumber
from the Edit menu (see Figure 4.9). 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
2/24/2003
1b.doc
1:35 PM
Page 84

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 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.
Figure 4.9
Pipes and junctions can be renumbered using the

Renumber window opened from the Edit menu
Using the Renumber Wizard

The Renumber Wizard 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.
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
2/24/2003
1:35 PM
Page 85

File menu by choosing Load Background Picture. At your option the
background graphic can be included in Workspace printouts.
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 objects 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.
The Model Data window

The Model Data window offers a text-based perspective on all
engineering information in the pipe flow model (see Figure 4.10). This
window is immensely useful for obtaining a global view of the model
and rapidly checking the input.
2/24/2003
1b.doc
1:35 PM
Page 86

Open Model
Data Control
General data
section
Customize view
Pipe data
section
Junction data
section
Figure 4.10 The Model Data window provides a convenient

summary of model element specifications
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 additional losses are displayed in the Pipe Loss Data table,
accessed with an adjacent tab. This table displays all fittings with K
factors grouped into 27 columns.
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.
2/24/2003
1:35 PM
Page 87

The transient data for each junction is shown on the Transient Data tab
(see Figure 4.10).
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.
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 Format 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 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 4.11). 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
junction type. Each type also includes the general data which may be
2/24/2003
1b.doc
1:35 PM
Page 88

selected/deselected. You can select/deselect the general data for all
junction types simultaneously by clicking Set Common.
Figure 4.11 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.
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.
2/24/2003
1:35 PM
Page 89

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.
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 4.12). 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 Results 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 4.13) lets
you specify the parameters, their units, and the order of their display in
the Output window (Figure 4.12).
Data in the Output window can be copied to the clipboard, saved to a
text file, exported to a file in a format for direct importing into a
Windows spreadsheet package, or sent to the printer. The printed 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 Format 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
flows and pressures equal to the steady-state results, which causes future
steady-state runs to converge much faster.
2/24/2003
1b.doc
1:35 PM
Page 90

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 doubleclick 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 4.12 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
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.
2/24/2003
1:35 PM
Page 91

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 4.13) allows you to customize
the parameters shown on the Output window. The control features are
grouped on seven 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 lefthand 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.
2/24/2003
1b.doc
1:35 PM
Page 92

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 4.13 The Output Control window selects parameters for the
Output window. Here the Pipe parameters (for steadyflow results) are shown.
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
up or down. The list will be reordered accordingly. The output is
presented in the order defined in this list.
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.
2/24/2003
1:35 PM
Page 93

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).
Pipe and Junction 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.
2/24/2003
1b.doc
1:35 PM
Page 94

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 userset 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
2/24/2003
1:35 PM
Page 95

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
2/24/2003
1b.doc
1:35 PM
Page 96

3. Click the None button for both pipes and junctions
4. Click the Select Special button for both pipes and choose the
Selected on Workspace option
5. Click the Select Pipes button then the OK button
6. Do the same thing for the junctions
7. Click the Output Control window Show 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 Impulses 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
2/24/2003
1:35 PM
Page 97

Recall Control Format button and choosing the file name. For example,
you may have a final report format that is always desired. You can recall
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.
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 dont 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 4.14 shows the steady flow results in tables for pipes and
junctions as specified in the Output Control window.
2/24/2003
1b.doc
1:35 PM
Page 98
Figure 4.14 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 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 4.15)
Figure 4.15 shows the transient flow results at the initial time step
(which is equal to zero here). The columns of data shown in Figure 4.15
are specified in the Output Control window.
Note that the pipe stations shown are those saved to the output file.
2/24/2003
1:35 PM
Page 99

Buttons to
increment or
decrement
currently
displayed time
Time step for
currently viewed
results
Time step
selection tool
slides up and
down
Figure 4.15 The Transient Output table in the Output window

shows output results for pipe stations and at specified
times.
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 4.16 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.
2/24/2003
1b.doc
1:35 PM
Page 100
Figure 4.16 The Transient Max/Min table details the maximum and
minimum values for the entire simulation.
Summary Max/Min results

There is a summary version of the Transient Max/Min table. The version
shown in Figure 4.16 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 4.17).
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
columns in the summary version of the Transient Max/Min table. See
Figure 4.18 for an example.
2/24/2003
1:35 PM
Page 101
Figure 4.17 Detailed and Summary versions are available for the
Transient Max/Min table
Figure 4.18 Summary version of the Transient Max/Min table
2/24/2003
1b.doc
1:35 PM
Page 102
Transient flow results for junctions

The Transient Data table in the Output window Junction area shows
selected parameter results for the entire simulation (Figure 4.19). The
parameters shown are selected in the Output Control window on the
Junction Transient tab.
Figure 4.19 The Transient Data table shows junction transient

results.
Transient event messages

When events occur at junctions during the transient simulation, a log of
all events is displayed in the Event Messages list. Events are first
displayed grouped by junction, and then displayed again in temporal
order. The log is shown in the General area of the Output window on the
Event Messages tab. A list of 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).
2/24/2003
1:35 PM
Page 103
Output window updates

Whenever you change the input model, all output windows and are
erased, as is the output file. This prevents you from changing your input
model, forgetting to rerun it, and mistaking the previous Output window
for the current model's results.
The Graph Results window

The Graph Results window, shown in Figure 4.20, 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.
Figure 4.20 The Graph Results window offers full-featured plot

generation
2/24/2003
1b.doc
1:35 PM
Page 104

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.
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 4.21). You can
create four types of graphs. Two types use time as the independent
variable, and the other two use pipe length.
Figure 4.21 The Select Graph Data window controls the Graph
Results window
2/24/2003
1:35 PM
Page 105

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 4.21, each pipe is shown in an expandable list. When a
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 4.22). Only groups which consist of a pipe sequence will be
displayed
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.
2/24/2003
1b.doc
1:35 PM
Page 106
Figure 4.22 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.
2/24/2003
1:35 PM
Page 107

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 4.23 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).
2/24/2003
1b.doc
1:35 PM
Page 108
Figure 4.23 The animation options display additional controls
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.
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.
2/24/2003
1:35 PM
Page 109

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
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.
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:
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.
2/24/2003
1b.doc
1:35 PM
Page 110
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.
The Visual Report window

The Visual Report will to help you visualize the model data and steadystate 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 4.24).
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
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 copied to the
clipboard for transfer into another Windows graphics program or sent to
the printer.
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.
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.
2/24/2003
1:35 PM
Page 111
Figure 4.24 Visual Report window integrates input data and output
results with model layout
The Visual Report Control window

The Visual Report Control lets you control the content and appearance
of the Visual Report window (see Figure 4.25). 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.
2/24/2003
1b.doc
1:35 PM
Page 112
Figure 4.25 The Visual Report Control window allows you to

configure the Visual Report display
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.
2/24/2003
1:35 PM
Page 113

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 4.26 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 4.24 or 4.26). 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
option is chosen while in input-only mode, all values will be converted
to a consistent set of units.
You can change the font 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, Xs, etc.). The number of a junction which has
special conditions set, by default, has an X preceding the J (e.g. XJ56).
2/24/2003
1b.doc
1:35 PM
Page 114

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
4.27). 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 xs to the left of the name. The xs
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 xs.
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.
Color map
The Color Map feature allows you to visualize magnitudes of various
parameters by coloring the pipes (see Figures 4.26 and 4.28). 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.
2/24/2003
1:35 PM
Page 115
Figure 4.27 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
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,
2/24/2003
1b.doc
1:35 PM
Page 116

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 4.28 The Visual Report Control Color Map allows you to
specify the display colors for pipes according to their
output values.
To avoid overlapping colors, only use less than (<) as the last entry. A
better map would be to use
1) >=60
2) >=20
2/24/2003
1:35 PM
Page 117

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 dont 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 4.28). 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 Recall 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 dont want to keep, click the
Cancel button. Click Show to apply your settings.
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
2/24/2003
1b.doc
1:35 PM
Page 118

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 annotation 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 4.29 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.
2/24/2003
1:35 PM
Page 119
Figure 4.29 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.
2/24/2003
1b.doc
1:35 PM
Page 120
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. Users who have not been exposed to pipe flow
analysis software may have difficulty grasping the network modeling
concepts initially.
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.
2/24/2003
1b.doc
1:35 PM
Page 121

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 but less than 10,000. 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.
2/24/2003
1b.doc
1:35 PM
Page 122
Chapter 5 Building and Running Models 123
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
Additional 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.
2/24/2003
1:35 PM
Page 123
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
junctions Specifications window you can change the ID number to any
desired value greater than zero but less than 10,000. 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 additional losses in the original pipe will be left in the original
pipe, and the new pipe will have no additional losses.
2/24/2003
1b.doc
1:35 PM
Page 124
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
2/24/2003
1:35 PM
Page 125
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.
2/24/2003
1b.doc
1:35 PM
Page 126

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, Doubleclick 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
2/24/2003
1:35 PM
Page 127

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
2/24/2003
1b.doc
1:35 PM
Page 128

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.
Satisfying connectivity requirements

In addition to entering the required information for each object in the
model, you must satisfy these connectivity requirements:
2/24/2003
1:35 PM
Page 129
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, a venturi 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).
2/24/2003
1b.doc
1:35 PM
Page 130
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.
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.
2/24/2003
1:35 PM
Page 131
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 the 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.
Figure 5.4
The six Checklist items help you track the status of

your model
2/24/2003
1b.doc
1:35 PM
Page 132
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.
Tip: If you double click the checkbox on the Checklist window or Status
Bar, the appropriate window will be opened. For example, doubleclicking Specify Output Control will open the Output Control window.
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.
2/24/2003
1:35 PM
Page 133
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
atmospheric pressure are entered. Defaults are assigned for gravity and
atmospheric pressure. You can choose to use an Unspecified Fluid, an
AFT Standard fluid 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
2/24/2003
1b.doc
1:35 PM
Page 134

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).
2/24/2003
1:35 PM
Page 135
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
2/24/2003
1b.doc
1:35 PM
Page 136

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
2/24/2003
1:35 PM
Page 137

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 steadystate 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,
2/24/2003
1b.doc
1:35 PM
Page 138

viscosity and bulk modulus cannot be edited as they depend on the
specified temperature (and pressure if Chempak).
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.
2/24/2003
1:35 PM
Page 139

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:
KB =
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
2/24/2003
1b.doc
1:35 PM
Page 140

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.
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 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.
2/24/2003
1:35 PM
Page 141
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 142
Figure 5.10 The Create Mixture window (opened from the System
Properties window) allows you to create predefined
mixtures for the model.
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. More extensive discussion of these viscosity models is
given in Chapter 8.
2/24/2003
1:35 PM
Page 143

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 nonNewtonian 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.
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 14.696 psia. 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 8500 feet you should
2/24/2003
1b.doc
1:35 PM
Page 144

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 32.174 ft/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.
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.
2/24/2003
1:35 PM
Page 145
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.
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
2/24/2003
1b.doc
1:35 PM
Page 146

Status off, perform your graphical editing, then turn it back on when you
need it.
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:
a=
x
t
(5.1)
where a is the wavespeed in the pipe, x is the length of the section and
t 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:
x =
L
n
(5.2)
The time step can be obtained by combining Equations 5.1 and 5.2:
t =
L
na
(5.3)
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:
t max =
L
a
(5.4)
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 t. Then the number of
sections in the remaining pipes is obtained from:
ni =
2/24/2003
Li
ai t
1:35 PM
(5.5)
Page 147

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:
ni =
Li
(1 )ai t
where is the accepted uncertainty in wavespeed up to 15% (Wylie,

et al., 1993, pp. 54).
Figure 5.13 The Section Pipes window automates the process of

obtaining appropriate sectioning criteria
2/24/2003
1b.doc
1:35 PM
Page 148

Figure 5.13 shows the Section Pipes window. The controlling pipe (#8)
is shown at the right. The sectioning criteria is 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 5.14.
Computing
stations
Pipe sections
5
5
Figure 5.14 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 junctions upstream pipe to the first computing
station of the junctions downstream pipe (see Figure 5.15).
J1
J2
P1
P1
1,0
1,1
1,2
J2
1,3
J3
P2
1,4 2,0
P2
2,1
2,2
2,3
Figure 5.15 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 5.16).
2/24/2003
1:35 PM
Page 149

J1
J3
P2
P1
J2
P3
J4
P2
1,0
2,0
P1
2,1
J2
1,1
1,3
2,2
2,3
2,4
1,2
3,0
P3
3,1
3,2
3,3
3,4
Figure 5.16 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).
2/24/2003
1b.doc
1:35 PM
Page 150
Transient Control
Figure 5.17 shows the Transient Control window. At the top of the
window is an area called the Transient Solver Control area. This area
offers numerous input fields detailed in the following descriptions.
Figure 5.17 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.
2/24/2003
1:35 PM
Page 151

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
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 5.17 example,
there are 20112 time steps that will be calculated in the simulation. The
first option, "Save Output to File Every Time Step", will write 20112
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 5.17) "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 5.17, results
for every tenth time step will be written, for a total of 2011 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
2/24/2003
1b.doc
1:35 PM
Page 152

flows that are not in the fully rough zone, the dependence of pipe
resistance is weak. Third, the waterhammer phenomenon is not affected
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 steadystate 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 increase 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 3.0 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 steadystate 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
2/24/2003
1:35 PM
Page 153

immediately stop if artificial transients are detected, then select the
option to have the simulation stop. If you choose to not stop the
simulation, the artificial transients will be displayed in the warnings
section when the simulation is complete.
Include 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 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
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 pipe stations. Frequently
the user is only interested in what happens at the inlet and outlet of the
pipe, and, if so, a good choice is to only save the inlet and outlet. 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.
2/24/2003
1b.doc
1:35 PM
Page 154

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

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
Junctions tab (Figure 5.17) and then selecting the junctions in the list.
Estimated file size and run time

At the bottom of the Transient Control window (see Figure 5.17) is an
estimate of the output file size and runtime.
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
2/24/2003
1:35 PM
Page 155

include systems which contain pressure control valves, flow control
valves, internal relief valves, or flow-specified 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
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 scenarios specifications
Pass changes from a scenario to its variants
All scenarios are saved within the model file.
The Scenario Manager window

Illustrated in Figure 5.18, 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
Impulses Status Bar). On the right, the Summary tab displays a
summary of the selected scenarios specifications and a place where one
may record notes related to that scenario. The Details tab displays the
selected scenarios object tree, illustrating the scenarios pipes and
junctions by number and name, the values for the various objects
attributes and the source of those values (see Figure 5.19).
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 (see Figure 5.20).
2/24/2003
1b.doc
1:35 PM
Page 156
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 5.18 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.
Current
Workspace
scenario
Scenario tree
Summary
information for
selected
scenario
Rename,
delete, clone,
promote & save
scenarios by
clicking here
Create a new
scenario by
clicking here
Figure 5.18 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
2/24/2003
1:35 PM
Page 157

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.
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

siblingactive 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 Impulses 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 undelete or un-promote the scenario.
2/24/2003
1b.doc
1:35 PM
Page 158

Object tree for the 3 in
PVC Pipe scenario
Attributes of object
selected in object tree
along with current
Value, Units and
Scenario providing the
source of the value
The attributes list may
be filtered based on
the source of the
attribute values
Figure 5.19 Object tree and object attributes in Scenario Manager
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 Impulses
Status Bar.) Once a scenario is the current scenario, changes may be
made to differentiate it from its parent using any of AFT Impulses
editing tools and functions. When an attribute is changed within a
scenario it breaks the link for this attribute. Subsequent changes in the
scenarios ancestors will no longer affect this attribute.
Properties of a scenario that may differ from its ancestors include:
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
2/24/2003
1:35 PM
Page 159
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 scenarios parent is
linked to the parent. In turn, if the parents attribute value is the same as
that of its parent, then the childs 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 scenarios ancestors, parent, grandparent,
etc., will no longer affect that attribute. The Details tab of the Scenario
Manager provides the information needed to determine attribute linking.
See Figure 5.20.
2/24/2003
1b.doc
1:35 PM
Page 160

Attributes for a pipe in the
3 in PVC Pipe scenario
These attributes will
inherit changes made in
the Base Scenario
This attribute has been
changed in the 3 in PVC
Pipe scenario and will not
inherit changes made in the
Base Scenario
Figure 5.20 Attribute values and sources for the 3 in PVC Pipe
scenario in Scenario Manager
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 be entering the value or
selecting the Same As Parent option from within a pipe or junction
specifications window, Steady Solution Control, Output Control, System
Properties, Transient Control, 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.
2/24/2003
1:35 PM
Page 161
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 5.21).
Figure 5.21 The Waterhammer Assistant offers advice on modeling

issues
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
2/24/2003
1b.doc
1:35 PM
Page 162

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.
Workspace Find
The Find window utility shown in Figure 5.22 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.
Figure 5.22 The Find feature for the Workspace is helpful for
locating a specific pipe or junction. It is opened from
the View menu or Toolbar.
2/24/2003
1:35 PM
Page 163
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 5.23) and allows you to quickly select a group
of objects.
Figure 5.23 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.
2/24/2003
1b.doc
1:35 PM
Page 164

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 to be selected or deselected respectively.
Several select/deselect operations may be performed in succession.
When finished, click on OK button.
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, clear all pipe and junction selections by clicking the
None button. Then click on Select Special and, with Selected on
Workspace showing in the Selection Type, click Select Pipes, then click
OK. Repeat this for junctions and, 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
2/24/2003
1:35 PM
Page 165

2. By selecting the junction(s) on the Workspace and clicking the
Special Condition icon on the Toolbar
3. By opening the junctions 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 5.24, 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.
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 5.24 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.
2/24/2003
1b.doc
1:35 PM
Page 166

The Special Conditions for Control Valves are dependent on the Action
If Control Is Lost setting for each valve. If the Action is to Fail Open
then the Special Condition for the control valve is to be failed open with
the loss equal to the Loss When Fully Open value. If, on the other hand,
the Action is to Always Control or Fail Closed then the Special
Condition for the control valve is to be closed.
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 steadystate 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.
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
2/24/2003
1:35 PM
Page 167

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
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.
2/24/2003
1b.doc
1:35 PM
Page 168
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.
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 valve. Thus not only is the data
ignored, the junction itself is ignored. This is convenient when locating
one of these devices by trail 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.
2/24/2003
1:35 PM
Page 169
Merging models
Two models can be merged together using the Merge command on the
File menu. Heres 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:
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 5.25 allows you to
access special print features such as Print Preview and Fit to One Page.
It is accessed from the File menu.
2/24/2003
1b.doc
1:35 PM
Page 170

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.
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.
Figure 5.25 The Print Preview/Special window is opened from the

File menu or Toolbar
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.
2/24/2003
1:35 PM
Page 171

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 Impulses 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
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 mode
AFT Impulse models can be run in batch mode, a feature that is
especially useful for running a group of models during lunch or
overnight.
To run models in batch mode, follow these steps:
1. Create a text file that lists the file names of all models to be run.
This is called the batch list file.
2. Either copy the input model files to the same directory as the batch
list file, or specify the complete path name for the input model files.
Include the disk drive letter in the full path name. If the models are
2/24/2003
1b.doc
1:35 PM
Page 172

copied to a common directory, then only the file names are required
in the batch list file.
3. Choose the Start Batch Run option from the AFT Impulse File menu.
4. Using the Print Format window, select the print content and format
you desire, including font style. All reports will use the specified
font.
5. After closing the Print Format window, choose the batch list file you
want to run.
6. Assign the destination of the output results and start the run.
AFT Impulse then opens each model in sequence, automatically runs
each model that has a completed Checklist, and sends results to the
specified destination. A special file called BATCH.LOG keeps track of
all pertinent information about the batch run. This file will automatically
print at the end of the batch run.
The BATCH.LOG file also logs any error messages that are displayed in
the process of running the models. The normal error messages you
receive when running AFT Impulse in interactive mode are not displayed
on the screen, because they would interrupt the batch run.
The entire batch run can be canceled at any time by clicking the Cancel
button in the Solution Progress window.
Running Scenarios in Batch Mode
If you save a model in any scenario that is not the base scenario, upon
opening the model AFT Impulse will prompt you whether you want to
open into the scenario which was active when the model was saved, or
into the base scenario.
However, if you are running in batch mode the scenario active when the
model was saved is automatically opened without prompting.
Thus to run scenarios in batch mode, just save the model with the
scenario active which you want to run. If you want to run multiple
scenarios from the same model, for each scenario you want to run copy
the model file as a different file name and then save the desired scenario
in each file.
2/24/2003
1:35 PM
Page 173
Print Format
With either the Model Data or Output window active, you can open Print
Format from the View menu (see Figure 5.26), 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.
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.
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.
Figure 5.26 Print Format window specifies the print content of the
Model Data or Output window. It is opened from the
View menu or Toolbar.
2/24/2003
1b.doc
1:35 PM
Page 174

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 3.0 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.
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
2/24/2003
1:35 PM
Page 175

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 NewtonRaphson 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 3.0, 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.
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 3.0, 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.
2/24/2003
1b.doc
1:35 PM
Page 176

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.
2/24/2003
1:35 PM
Page 177
2/24/2003
1b.doc
1:35 PM
Page 178
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
2/24/2003
1b.doc
1:35 PM
Page 179

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.
2/24/2003
1b.doc
1:35 PM
Page 180
Chapter 6 Pipe and Junction Specifications Windows 181
Figure 6.1a Pipe Specifications window Pipe Model tab.
Figure 6.1b Pipe Specifications window Additional Loss tab.
2/24/2003
1:35 PM
Page 181
Figure 6.1c Pipe Specifications window Design Alerts tab.
Figure 6.1d Pipe Specifications window Optional tab.
2/24/2003
1b.doc
1:35 PM
Page 182
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 less than 10,000. 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).
2/24/2003
1:35 PM
Page 183

Same As Pipe
The Same As 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 Same As Previous Pipe button functions the same way as the Same
As 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 Same As 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.
2/24/2003
1b.doc
1:35 PM
Page 184

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.
2/24/2003
1:35 PM
Page 185

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 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
2/24/2003
1b.doc
1:35 PM
Page 186

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.
2/24/2003
1:35 PM
Page 187

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. 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.
7. MIT Equation The MIT Equation is appropriate for crude oil. See
Chapter 8 for a discussion of how this method is implemented.
8. 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
methods are still accessible, but the Pipe Specifications window always
comes up with your preferred method selected.
2/24/2003
1b.doc
1:35 PM
Page 188

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 2000-5000 ft/sec (7001600 m/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.
Additional Losses
The additional loss factor is one way to enter loss factors in your model.
The additional loss 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 Additional
Losses tab (Figure 6.1b) and then click the Additional Losses button to
open the Additional Pipe Losses window (Figure 6.2). This window
allows you to specify losses from a table of standard losses.
Because many of the additional losses are dependent on pipe diameter,
you must specify a diameter before you can open the Additional Pipe
Losses window.
2/24/2003
1:35 PM
Page 189

The types of pipe losses are divided into several categories represented
by tabs. To add a loss to the pipe, click the tab which contains the type
of loss, then select the loss from the list at the top of the window. You
can add up to 99 of each loss by entering the quantity in the box below
the selection list. Press the Add to Pipe button and you will see the loss
added to the list of losses currently in the pipe. 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
Additional Pipe Losses window opened from the Pipe

If you want to change the quantity of a loss which is currently in the

pipe, select the loss you want to change from the bottom list and then
edit the quantity in the box below the list. To delete a loss, select it from
the bottom list and click the Delete button.
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.
2/24/2003
1b.doc
1:35 PM
Page 190

The abbreviations used in the Additional Loss window are listed at the
bottom of the window. You can double-click the image to see the
reference information for the loss.
You can set a specific loss as the default so when that particular type is
chosen the default loss will be already selected. To set a default, first
choose a loss then click the Set As Default button. You can have a
default for each loss type except Area Changes.
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 flow 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.
2/24/2003
1:35 PM
Page 191

Workspace display
You have the option of specifying what to display on the Workspace.
You can choose to display either the pipe ID number, the pipe name,
both, or neither. This is helpful in large models to 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 additional losses that you add to the pipe.
Design factors allow you to specify multipliers on the pipe friction
calculations and additional 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.
2/24/2003
1b.doc
1:35 PM
Page 192

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
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.
2/24/2003
1:35 PM
Page 193

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
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 pipes 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
2/24/2003
1b.doc
1:35 PM
Page 194

prepare a good conceptual model of the physical system of interest.
Table 6.1 compares the features of the junction types.
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.
2/24/2003
1:35 PM
Page 195

Table 6.1
Comparison of junction capabilities
Junction Type
Connecting
Pipes
Custom
Database
Window
Format
1/1-25
2-10
Check Valve
Control Valve
Dead End
Area Change
Assigned Flow
Assigned Pressure*
Branch
Gas Accumulator
1-25
Special
Conditions
General Component
Liquid Accumulator
1-25
Pump
1-2
Relief Valve
1-2
Reservoir
1-25
Spray Discharge
1-4
Surge Tank
1-25
Tee/Wye**
Vacuum Breaker Valve
1-2
Valve
1-2
Volume Balance
*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.
2/24/2003
1b.doc
1:35 PM
Page 196

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
pipes' 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.
2/24/2003
1:35 PM
Page 197

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
2/24/2003
1b.doc
1:35 PM
Page 198

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 less than 10,000.
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.
2/24/2003
1:35 PM
Page 199
Figure 6.4
Specifications window format #2 (pipes displayed in a

table).
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.
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
2/24/2003
1b.doc
1:35 PM
Page 200

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
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.
Same As Junction list

The Same As 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 Same As Jct list there is an
option to set the pipe to Parent Junction Data. This will change the
selected pipe 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).
2/24/2003
1:35 PM
Page 201
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.
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, 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).
2/24/2003
1b.doc
1:35 PM
Page 202

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.
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.
2/24/2003
1:35 PM
Page 203

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.
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)
2/24/2003
1b.doc
1:35 PM
Page 204

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.
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
2/24/2003
1:35 PM
Page 205

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.
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 either 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 steadystate 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 valve. Thus not only is the data
ignored, the junction itself is ignored. This is convenient when locating
one of these devices by trail 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
2/24/2003
1b.doc
1:35 PM
Page 206

design factor multiplies the pressure rise. Design factors are helpful for
adding safety margins to your design calculations.
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.
2/24/2003
1:35 PM
Page 207
Status
The Status tab shows which required input data has not been entered.
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.
2/24/2003
1b.doc
1:35 PM
Page 208

The critical parameter that influences the magnitude of the area change
loss is the area ratio between upstream and downstream pipes. Until you
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.
2/24/2003
1:35 PM
Page 209
Figure 6.8
The Assigned Flow Specifications window
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).
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.
2/24/2003
1b.doc
1:35 PM
Page 210

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.
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 steadystate.
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 steadystate.
2/24/2003
1:35 PM
Page 211
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 twentyfive 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.
2/24/2003
1b.doc
1:35 PM
Page 212
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.
2/24/2003
1:35 PM
Page 213

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.
2/24/2003
1b.doc
1:35 PM
Page 214
Figure 6.11a
The Branch Specifications window Loss Coefficients

table.
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.
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.
2/24/2003
1:35 PM
Page 215

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.
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.
2/24/2003
1b.doc
1:35 PM
Page 216
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
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
2/24/2003
1:35 PM
Page 217

the valve. The valve can close during the steady-state or transient portion
of the simulation.
The Special Condition for a check valve is to fail the valve open so that
flow will be allowed to pass through it even if the flow is reversed.
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)
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.
2/24/2003
1b.doc
1:35 PM
Page 218

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
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
2/24/2003
1:35 PM
Page 219
Control Valve types

You can model four types of control valves:
Pressure Reducing Valves (PRVs)
Pressure Sustaining Valves (PSVs)
Flow Control Valves (FCVs)
Pressure Drop Control Valves (PDCVs)
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.
2/24/2003
1b.doc
1:35 PM
Page 220
Action if control is lost

It is possible that the control valve cannot maintain the specified control
setting during steady-state and/or transient conditions. There are four
actions the valve can take. The default action is Always Control (Never
Fail). This option is physically unrealistic but useful for steady-state
modeling purposes. The reason this option is physically unrealistic is
that the valve must add pressure to maintain control, thus acting like a
pump. If this occurs, the user is clearly warned in the output. The
Always Control (Never Fail) option is useful because it gives a clear
indication to the user of the how far away the valve is from being able to
control.
For transient analyses, the Always Control (Never Fail) should not be
used. If it is, and the valve fails, the transient analysis will terminate.
Other options for valve failure are Fail Closed or Fail Open.
Additionally, if it fails open you can choose to have it close if the flow
reverses. Once the control valve fails, open or closed, Impulse will solve
the model again. These are the appropriate choices for transient
analyses.
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 cant open wide enough
(becoming less restrictive) to maintain 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
The Special Condition for a control valve is dependent on the Action If
Control Is Lost setting. If the setting is Always Control (Never Fail) or
Fail Closed, then the Special Condition for the control valve is to close.
If the setting is to Fail Open, then the Special Condition will be to open.
2/24/2003
1:35 PM
Page 221

In this case the valve will have a pressure drop equal to the Loss When
Fully Open value.
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 Impulses flow solution. However, engineers frequently desire to
know the valves 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 a failure, 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
2/24/2003
1b.doc
1:35 PM
Page 222

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.
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.
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
influence the transient response of the system. Unless you know
otherwise for a specific system, dead ends should always be included in
the model.
2/24/2003
1:35 PM
Page 223
Figure 6.14 The Dead End Specifications window
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.
2/24/2003
1b.doc
1:35 PM
Page 224

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,
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.
2/24/2003
1:35 PM
Page 225
Figure 6.15 Gas Accumulator Specifications window
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.
2/24/2003
1b.doc
1:35 PM
Page 226

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
LC
m& A
Connector
Pipe
Figure 6.16 Schematic of a gas accumulator
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.
2/24/2003
1:35 PM
Page 227

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.
2/24/2003
1b.doc
1:35 PM
Page 228
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
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
2/24/2003
1:35 PM
Page 229

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.
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.
2/24/2003
1b.doc
1:35 PM
Page 230
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.
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.
2/24/2003
1:35 PM
Page 231

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
assigning a flow, the total pressure/head developed by the pump is
calculated. This is useful for modeling positive displacement pumps.
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
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.
2/24/2003
1b.doc
1:35 PM
Page 232

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.
Figure 6.20 The Pump Specifications window.
2/24/2003
1:35 PM
Page 233

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.
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, then a particular combination can be
chosen.
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. Data for head, NPSH, and
efficiency or power can be entered. Also data for the pumps 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).
Controlled discharge 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 discharge pressure you set. The
resulting speed will be reported in the Pump Summary in the General
2/24/2003
1b.doc
1:35 PM
Page 234

Results section on the Output window. Alternatively, a discharge
pressure, it can be controlled only if it exceeds some level, otherwise the
pump will operate on its regular pump curve.
These parameters are set in the Auxiliary Input tab.
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.
2/24/2003
1:35 PM
Page 235
Figure 6.21 Pump Configuration window Pump Data tab above

where raw data is entered and curve fit, Composite
Graph tab below.
2/24/2003
1b.doc
1:35 PM
Page 236
Multiple configuration
selections displayed
Figure 6.22 Pump Specifications window with multiple

configurations available.
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.
2/24/2003
1:35 PM
Page 237
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.
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
2/24/2003
1b.doc
1:35 PM
Page 238

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 timebased 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.
Figure 6.23 A speed transient can be entered for a pump with a

curve
2/24/2003
1:35 PM
Page 239

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:
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
2/24/2003
1b.doc
1:35 PM
Page 240

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 initial 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.
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).
2/24/2003
1:35 PM
Page 241
Figure 6.24 Pump with four quadrant data for pump trip modeling
AFT Impulse provides data of FH and FB vs. 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
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).
2/24/2003
1b.doc
1:35 PM
Page 242

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 initial 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.
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.
2/24/2003
1:35 PM
Page 243

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).
The Figure 6.26 transient essentially increases the speed of the pump to
achieve 150 gpm 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.
2/24/2003
1b.doc
1:35 PM
Page 244
Figure 6.25 A pump with a variable speed controller can have a

transient on the control point.
2/24/2003
1:35 PM
Page 245
Figure 6.26 A control point transient can be entered for a pump.
Pumps as assigned flow

When a pump is modeled as an assigned 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.
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.
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".
2/24/2003
1b.doc
1:35 PM
Page 246

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
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 semilogarithmic diagram relating flow rate, head rise and viscosity. The
details of this diagram have been digitized into numerical relationships
inside AFT Impulse 3.0. The corrections to head rise, flow rate and
2/24/2003
1:35 PM
Page 247

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.
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
2/24/2003
1b.doc
1:35 PM
Page 248
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
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).
2/24/2003
1:35 PM
Page 249
Transient Data
AFT Impulse allows the user to specify an opening 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 opening will occur
as soon as the cracking pressure is exceeded. If no opening transient
profile is entered, the valve will open instantaneously.
You can also specify a closing profile of Cv or K vs. time, which is the
profile the valve will follow if it closes. If no closing transient profile is
entered, the valve will close 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.
More information on transient data is given earlier in this chapter and in
Chapter 10.
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
until it is fully open. Any further pressure increases to not increase the
Cv, and so the fully open Cv is used for all higher pressures.
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.
2/24/2003
1b.doc
1:35 PM
Page 250
Figure 6.28a
The Reservoir Specifications window Reservoir

Model tab.
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.
2/24/2003
1:35 PM
Page 251
Figure 6.28b Individual pipe depths (or elevations) and loss

coefficients may be specified in the Reservoir
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.
2/24/2003
1b.doc
1:35 PM
Page 252
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.
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.
2/24/2003
1:35 PM
Page 253
Figure 6.29 The Spray Specifications window
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 254

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 twentyfive connecting pipes. After you add a fifth pipe a scroll bar appears,
allowing you to review all the pipes in the table.
2/24/2003
1:35 PM
Page 255
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.
2/24/2003
1b.doc
1:35 PM
Page 256
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
LC
Orifice
m& T
Connector
Pipe
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.
2/24/2003
1:35 PM
Page 257
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 100
psig must have a liquid surface at least 230 feet 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.
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.
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
2/24/2003
1b.doc
1:35 PM
Page 258

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.
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,
2/24/2003
1:35 PM
Page 259

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 steadystate (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
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).
2/24/2003
1b.doc
1:35 PM
Page 260
LA
LL
LC
Orifice
m& T
Connector
Pipe
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
connecting pipes are arranged on the Arrangement tab 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
2/24/2003
1:35 PM
Page 261

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
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.
2/24/2003
1b.doc
1:35 PM
Page 262
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
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
2/24/2003
1:35 PM
Page 263

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.
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 264
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
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 junctions allows air to flow in and out.
Alternatively, it also model only air inflow or outflow.
2/24/2003
1:35 PM
Page 265
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.
Figure 6.37 The Volume Balance Specifications window
2/24/2003
1b.doc
1:35 PM
Page 266

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.
Global Pipe editing

The Global Pipe Edit window, shown in Figure 6.38, allows you to edit
most of the pipe data in the model at the same time. In the Global Pipe
Edit window, select the pipes you want to edit, then 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.
When you are done entering the data to be changed, click the OK button
in the Pipe Specifications window. The Global Pipe Edit window now
lists the parameters to change in a list on the right. Select the parameters
you want to change and then click the Apply Selections button.
Note that you cannot change some data, like the pipe ID number, using
this global method.
2/24/2003
1:35 PM
Page 267
Figure 6.38 The Global Pipe Edit window allows you to change data
for many pipes at the same time
Global Junction editing

The Global Junction Edit window allows you to edit elevation,
name/number display option, initial pressure, design factor, and
Workspace icon size for all junctions (see Figure 6.39a). This window
also offers the ability to select specific junction types for specific data
changes (see Figure 6.39b). For example, with a couple mouse clicks all
valve junctions can be turned off or the pump curves in your model can
be changed to a new pump curve.
To edit the common junction data, select Common Data option enter the
data in the Common Parameters area. Then check the boxes on the right
to select the data to be changed and click the Apply Selections button.
2/24/2003
1b.doc
1:35 PM
Page 268
Figure 6.39a
The Global Junction Edit window allows you to

change data that is common among different types
of junctions, as shown above.
Editing specific junction data

To edit specific junction data do the following:
1. Select Specific Data For option
2. Select the junction type in the dropdown list
3. In the Junction List, choose the junctions you want to edit
4. Click the Select Specific Junction Data button. The Specifications
window for the junction type you selected will be opened.
5. Enter the data you want to change for each of the selected junctions.
The Specifications window for the junction type you selected will be
opened. Enter the data you want to change for each of the selected
junctions.
When you are done entering the data to be changed, click the OK button
in the Junction Specifications window. The Global Junction Edit
2/24/2003
1:35 PM
Page 269

window now itemizes the parameters to change in a list on the right.
Select the parameters you want to change then click the Apply
Selections button.
Figure 6.39b
2/24/2003
1b.doc
The Global Junction Edit window allows you to

change data for many junctions at the same time.
You can change detailed data for junctions of the
same type, as shown above.
1:35 PM
Page 270
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 IMPULSE3.INI file in your Windows directory.
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_USER3.DAT file in your Windows directory.
If you pursue extensive customizing of AFT Impulse, you are strongly
advised to make regular backup copies of the two files (IMPULSE3.INI
and IMP_USER3.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
during assembly of your model, AFT Impulse offers your default
2/24/2003
1b.doc
1:35 PM
Page 271

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
are applied to your pipes and junctions, here is where you do it.
2/24/2003
1b.doc
1:35 PM
Page 272
Chapter 7 Customizing AFT Impulse 273

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.
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.
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.
2/24/2003
1:35 PM
Page 273
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 274

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.
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 Impulses 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.
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 well discuss how databases are configured and
administered through your local or wide area network.
2/24/2003
1:35 PM
Page 275
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 Impulses 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 Recall 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 companywide database, allowing standardized reporting.
If you have made changes that you dont 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.
2/24/2003
1b.doc
1:35 PM
Page 276
Figure 7.3
The Pipes and Junctions tab in the Workspace

Preferences window.
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.
2/24/2003
1:35 PM
Page 277

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.
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.
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 278

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.
Icon source
Limitations exist in Windows 95, 98 and Me that can affect AFT
Impulses 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).
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.
2/24/2003
1:35 PM
Page 279

The grid can be included in printouts by specifying its inclusion in the
Print Preview/Special window.
Displaying source/sink indicator
The Branch and Assigned Flow junctions specify flow into (source) or
out of (sink) the system. These junctions can display a small symbol
indicating whether the junction represents a flow source or sink. You
can change these symbols, for example, to a + and - by entering the
symbol in the appropriate field in the Flow Junction Symbol area. To not
show any symbol on any of these junctions, uncheck the Display
Source/Sink Indicator checkbox.
Figure 7.4
The Display Options tab in the Workspace Preferences

window.
2/24/2003
1b.doc
1:35 PM
Page 280

Special condition junction symbol
To better locate an object which has Special Conditions set (e.g., a pump
turned off), the object will have a Closed Symbol preceding the ID
number. By default this is an X, but can be customized by entering the
desired symbol. If no symbol is entered no symbol will be displayed.
This symbol is also used in the Junction Results tables on the Output
report. To not use any symbol on any object, uncheck the Display Closed
Indicator checkbox. Additionally, junctions which are closed in any way
are outlined using the Closed Pipe Line Style. To not outline the
junctions, uncheck the Display Outline Around Closed Junctions
checkbox.
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-bypipe 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.
2/24/2003
1:35 PM
Page 281

Junction transient symbols
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 282
Figure 7.5
The Colors and Fonts tab in the Workspace Preferences

window.
Sample Workspace
The tab for Sample Workspace gives you 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.
2/24/2003
1:35 PM
Page 283

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 well 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
have these used by default (instead of Impulses 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 Recall 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 companywide database, allowing standardized reporting.
If you have made changes which you dont 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
2/24/2003
1b.doc
1:35 PM
Page 284

As Default button. From then on, AFT Impulse will always load with
your default Toolbox settings.
Figure 7.6
Toolbox Preferences window allows you to customize

the Workspace Toolbox.
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.
2/24/2003
1:35 PM
Page 285

You can also reorder the icons on the Toolbox by dragging and dropping
them to the desired locations on the Current Toolbox.
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
Enable Tip of the Day feature
Require a check of the Waterhammer Assistant before running

models
Default locations for saving models
Common database locations
Backup folders where your IMP_USER3.DAT and IMPULSE3.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
2/24/2003
1b.doc
1:35 PM
Page 286

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.
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
2/24/2003
1:35 PM
Page 287

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.
In addition, an Auxiliary Graph Formatting window is available to make
it easier to change graph colors and curve thickness. This window also
allows one to specify default colors and curve thickness.
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.
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 Additional Pipe Losses

The Additional Pipe Losses window presents organized lists of pipe loss
fittings. This list is customizable. You can add your own fittings to the
2/24/2003
1b.doc
1:35 PM
Page 288

base list provided with AFT Impulse. The Additional Pipe Losses
window is accessed through the Pipe Specifications window.
The standard losses provided with AFT Impulse are located in the
IMPULSE3.DAT file in the same directory as the AFT Impulse
executable file. You should not change anything in this file.
The custom user information loaded into the Additional Pipe Losses
window is found in the ADDLOSS.DAT file in the directory where AFT
Impulse is installed. Each data line represents a loss and is divided into
the nine fields listed in Table 7.1. The structure of this loss data is
identical to that in the IMPULSE3.DAT file, which you can open and
use as an example.
Table 7.1
Pipe loss data fields for additional losses
Field #
Description
Required
Reference number
Yes
Name of component
Yes
Source
Yes
Diameter
No
Angle
No
Percent open
No
Area ratio percentage
No
K factor
Yes
Page reference
No
All lines in the ADDLOSS.DAT file must have a reference number

(field 1), name (field 2), source (field 3), and K factor (field 8). Any
custom information you add must have a letter U in the source field to
indicate it is a user-entered item. The C, I, and M characters in the
source field for other components are for references to Crane, Idelchik,
and Miller, respectively.
The reference number must be unique between both the
ADDLOSS.DAT file and the IMPULSE3.DAT file.
Table 7.2 lists the reference numbers for Bends, Valves, Check Valves,
Orifices, Entrances, Exits, and other components. Enter custom valves
using reference numbers consistent with the subcategories.
No area change data is included in the ADDLOSS.DAT file. Area
change loss factors are all calculated based on standard correlations.
2/24/2003
1:35 PM
Page 289

As it is infrequently used, no Additional Pipe Losses data customization
features have been built into the Impulse interface, relying instead on
directly editing text files to effect changes.
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 Dead End and Tee/Wye. AFT Impulse
incorporates these databases internally and offers them to you through
lists.
2/24/2003
1b.doc
1:35 PM
Page 290

Table 7.2
Additional loss component reference numbers
Component
Bends
Valves
Reference number
Description
100000-100999
Standard
101000-101999
Mitre
102000-102999
Smooth Flanged
200000-200999
Angle
201000-201999
Ball
202000-202999
Butterfly
203000-203999
Cylindrical
204000-204999
Gate
205000-205999
Globe
206000-206999
Plug
207000-207999
Poppet
208000-208999
Three-way
300000-300999
Swing
301000-301999
Lift
302000-302999
Tilting Disk
303000-303999
Stop-check
Orifices
400000-400999
Sharp-edged
401000-401999
Long
Entrances and Exits
500000-500999
Entrances
501000-501999
Exits
600000-600999
Differential Flow Meters
601000-601999
Honeycomb
602000-602999
Screen
603000-603999
Tee
Check Valves
Miscellaneous
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_USER3.DAT file in the Windows directory.
In AFT Impulse you can create multiple external or local custom
databases. The Database Manager window on the Database menu
simplifies this process.
2/24/2003
1:35 PM
Page 291

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 PCs. 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 7.8 shows the layout
of the Fluid Database window. This window allows you to review
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 7.8
Fluid Database window
2/24/2003
1b.doc
1:35 PM
Page 292

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 7.9). 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.
Figure 7.9
Add New Fluid window
Density, viscosity and bulk modulus are required for all fluids. Vapor
pressure is required if cavitation modeling is desired.
2/24/2003
1:35 PM
Page 293

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
might have used is displayed if you elected to save the data in a raw data
source file.
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.
Adding custom pipe materials to the Pipe Material

Database
You can open the Pipe Material Database window from the Database
menu. Figure 7.10 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 7.10 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.
2/24/2003
1b.doc
1:35 PM
Page 294

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.
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
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 0.005 inches. 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 of Steel, it will apply only to the pipe schedules that are 1
inch.
2/24/2003
1:35 PM
Page 295

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 schedule 40 Steel pipe and assign its roughness value as
0.006 inches. Then all Steel pipe would use 0.005 inches for 25 year
usage friction while the different value for 1 inch schedule 40 of 0.006
inches would supercede the value of 0.005 inches associated with all
Steel pipe.
Finally, among multiple friction data sets you can specify a default data
set.
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 date 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.
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.
2/24/2003
1b.doc
1:35 PM
Page 296
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
7.11). 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.
Figure 7.11 Component Database window
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
2/24/2003
1:35 PM
Page 297
Setting available databases

To work with a database you must first make it available to AFT
Impulse. Referring to Figure 7.12, clicking on the Add Engineering
Database button 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 button. Contents of a database may be reviewed by selecting a
database from the list and clicking on the Review Content button.
Figure 7.12 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
2/24/2003
1b.doc
1:35 PM
Page 298

Database tab. To connect a database, select it from the list of available
databases and click on Add to Connections (see Figure 7.13). 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
which the sections to be used in the model may be selected; i.e.
components, fluids or pipe materials.
Figure 7.13 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.
2/24/2003
1:35 PM
Page 299
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
7.14).
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_USER3.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
2/24/2003
1b.doc
1:35 PM
Page 300

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 7.14 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.
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 7.15).
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
2/24/2003
1:35 PM
Page 301
Figure 7.15 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.
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
2/24/2003
1b.doc
1:35 PM
Page 302

If the user connects to a database with these files and selects to be
connected to any of these control files, his 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 4.13 is specified as
coming from the New Mile High Stadium Project database, it would
appear as shown in Figure 7.16.
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.
2/24/2003
1:35 PM
Page 303
Database
connection
Figure 7.16 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
maintain the network databases. The discussion below will be given
from the viewpoint of this person.
2/24/2003
1b.doc
1:35 PM
Page 304

Figure 7.17 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 Windows 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 users 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.
2/24/2003
1:35 PM
Page 305

Database Groups and Filenames
Database Titles/Descriptions
Impulse Internal Databases

AFT Default Internal Database
IMPULSE3.DAT
IMP_USER3.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 7.17 An example of the relationship between External

Databases and External Shared Databases.
This discussion will focus on the first and second methods. Lets 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 networkwide 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_USER3.DAT) to a new file using the Database Editor in the
Database Manager (see Figure 7.14). 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 well use Vail Resort
Materials). Then click Add to copy the pipe materials from the Local
2/24/2003
1b.doc
1:35 PM
Page 306

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, lets 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, lets create the file using the
Database Editor in the Database Manager (see Figure 7.14). 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 7.14). 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 Add 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.
2/24/2003
1:35 PM
Page 307

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 Window 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 four 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 Add Engineering Database button on the Connect
to Database tab (see Figure 7.12). 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 Windows 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
Network Neighborhood, then the computer and then browse to the
2/24/2003
1b.doc
1:35 PM
Page 308

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 four 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
7.13). 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.
2/24/2003
1:35 PM
Page 309
2/24/2003
1b.doc
1:35 PM
Page 310
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 SteadyState 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
2/24/2003
1b.doc
1:35 PM
Page 311

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:
HGL =
P
+z
g
(8.1)
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
& ij represents the mass
along pipes connecting two junctions. Thus, m
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:
Pf = f
2/24/2003
1b.doc
L 1
2
V
D2
1:35 PM
Page 312
(8.3)
Chapter 8 Steady-State Hydraulic Theory and Loss Models 313

where Pf is the frictional pressure loss. The total pressure change
between junctions is given by the momentum equation in the form of the
Bernoulli equation:
P1 +
1
1
V12 + gz1 = P2 + V2 2 + gz2 + Pf
2
2
(8.4)
The static and stagnation pressure are related as follows:

Po = P +
1
V 2
2
(8.5)
Substituting Equation 8.5 into 8.4

Po,1 + gz1 = Po,2 + gz2 + Pf
Solving for the frictional pressure drop for a constant area pipes yields:
Po,i Po, j + g ( zi z j ) = Pf
(8.6)
where i and j denote upstream and downstream junction values.

The definition of mass flow rate is:
m& = AV
(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 + g ( zi z j )
ij
R
= m& ij
(8.8)
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
+ K ij
Rij =
Dij
2 A 2
ij
(8.9)
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:
H = RQ 2
2/24/2003
1:35 PM
Page 313

Substituting Equation 8.8 into Equation 8.2 results in the equation to be
applied to each junction i:
1/ 2
ij
R
j =1
=0
(8.10)
where n is the number of pipes connected to junction i.

To be completely general, Equation 8.10 should be written for junction i:
1/ 2
ij
R
j =1
= m& i, Applied
(8.11)
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:
1/ 2
Fi =
ij
R
j =1
m& i, Applied
(8.12)
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:
2/24/2003
1b.doc
1:35 PM
Page 314
F1
F1 F1
P P L P
o, n
o,1 o,2
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 F1F
(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.
2/24/2003
1:35 PM
Page 315
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
Ploss
Pressure
1
V2 2
2
Location 2
Po,4
1
V42
2
Location 3
P4
Location 4
Distance (x)
Figure 8.2
Steady-state pressure profile along horizontal pipe

shown in Figure 8.1.
2/24/2003
1b.doc
1:35 PM
Page 316

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)V22.
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 isnt 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.
2/24/2003
1:35 PM
Page 317

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., PRVs and PSVs) 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.
2/24/2003
1b.doc
1:35 PM
Page 318

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 Impulses
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
P1
J3
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.
2/24/2003
1:35 PM
Page 319

J8
P9
J1
J9
P8
P6
J5
P4
P11
J4
J2
P1
P9
J8
J3
P3
P10
J11
J6
P7
J10
P2
J9
P8
J6
P7
J12
P4
J1
P6
J5
P10
P11
J4
J2
P1
Figure 8.4
J3
P3
J10
P2
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
2/24/2003
1b.doc
1:35 PM
Page 320

flow will enter or leave because there isnt 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.
2/24/2003
1:35 PM
Page 321
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

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 nonNewtonian fluids viscosity can change.
2/24/2003
1b.doc
1:35 PM
Page 322

On the other hand, a Newtonian fluids 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.
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.
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:
f =
64
Re
Re < 2300 (default)
(8.14)
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).
2/24/2003
1:35 PM
Page 323

Turbulent flow
The Colebrook-White iterative friction factor equation is used to obtain
friction factors in the turbulent flow regime.
9.35
f = 1.14 2 log +
D Re f
Re > 4000 (default)
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.
Hydraulically smooth
The friction factor in the turbulent range is calculated using ColebrookWhite 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.
CHW =
17.25
(VD) 0.081
0.54
where V is in ft/s and D is in ft.
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):
2/24/2003
1b.doc
1:35 PM
Page 324
dP 139.212 fQ 2 ( water )
=
L
D5
f =
Re
4 0.86859 ln
1964
.
ln( Re) 38215
.
Re =
10,059Q
d 2 818.56
where:
dP
= pressure drop (psid)
= length (miles)
= friction factor, MIT
= volumetric flow rate (barrels/hour)
= density (lbm/ft3)
water
= density of water (62.3 lbm/ft3)
= diameter (inches)
Re
= Reynolds number, MIT
= 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):
Q=
d 3(
water )( dP L)
log
4
.
35
+
0.5
2
5.9113( water )
d 2.5 ( dP L)
where:
dP
= pressure drop (psid)
= length (miles)
2/24/2003
1:35 PM
Page 325

Q
= volumetric flow rate (barrels/hour)
= density (lbm/ft3)
water
= density of water (62.3 lbm/ft3)
= diameter (inches)
= 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
Vwater = 122
. C 1.4
If V < Vmax, then
dH
= FKV C D
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)
2/24/2003
1b.doc
1:35 PM
Page 326

Vwater
= water velocity in above equation (limit at which pulp

effects are negligible), (m/s)
= inner pipe diameter (mm)
= correction factor for temperature or pipe roughness

(dimensionless)
K, K, , , , = 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.
Re =
f =
D 0.205V
C1.157
3.97
Re1.636
dH =
fV 2 LK
D
where:
Re = pseudo Reynolds Number (dimensionless)
D = inner diameter (feet)
V
= average velocity (ft/s)
= stock density (lbm/ft3)
C = pulp consistency (% dryness)
2/24/2003
1:35 PM
Page 327

f
= pseudo friction factor (dimensionless)
K = constant depending on pulp type (dimensionless)

L
= length (feet)
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

The Power Law model of pipe friction is given by the following (Tilton
1997)
1n
P
Q =
2 KL
n (1+ 3n ) n
r
1 + 3n
(8.15)
where:
K
= constant with units of dynamic viscosity
= dimensionless constant
Power Law flow is laminar. Substituting Equation 8.15 into 8.14 yields:
A 3+ n V
f =
K
2
m& D
D
n 1
1 + 3n
This friction factor can then be used in Equation 8.3.
Bingham Plastic non-Newtonian

The Bingham Plastic model of pipe friction is given by the following
(Carthew et al.):
Sy
H
RV
= 32
+ c 2
L pipe
6gD gD
(8.16)
where:
Rc
= coefficient of rigidity
Sy
= yield stress
2/24/2003
1b.doc
1:35 PM
Page 328

The head loss and pressure loss are related as follows:
gH = p = f
L1
V 2
D2
(8.17)
Bingham Plastic flow is laminar. Substituting Equations 8.17 into 8.14

yields:
f =
64
V
S y Rc
6V + D
This friction factor can then be used in Equation 8.3.

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 additional 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
2/24/2003
1:35 PM
Page 329
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.
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 SteadyState 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.
2/24/2003
1b.doc
1:35 PM
Page 330

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.
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:
If (For All Junctions)
Pj , new Pj , old
Pj , new
< TOLrel Then
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.
2/24/2003
1:35 PM
Page 331

AFT Impulse 3.0 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.
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.
Iteration #
1
2
3
4
5
6
Pressure
(psia)
100.0
60.0
72.0
66.0
67.0
66.9
Relative
Change
0.6667
0.1667
0.0909
0.0149
0.0015
Abs olute
Change (psia)
40.00
12.00
6.00
1.00
0.10
Figure 8.6
Tolerance input on Steady Solution Control window.
2/24/2003
1b.doc
1:35 PM
Page 332

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.
Table 8.2
Iteration Mass Flow

#
Rate (lbm/s)
1
2
3
4
5
6
7
8
9
20.0000
12.0000
14.0000
13.5000
13.6000
13.5900
13.5913
13.5907
13.5909
Relative
Change
0.666667
0.142857
0.037037
0.007353
0.000736
0.000096
0.000044
0.000015
Example of four tolerance methods for a mass flow rate

solution. Absolute tolerance = 0.0005 lbm/s, Relative
tolerance = 0.0001.
Absolute
Change
(lbm/sec)
8.000000
2.000000
0.500000
0.100000
0.010000
0.001300
0.000600
0.000200
Meets Abs.
Tolerance ?
Meets Rel.
Tolerance ?
Meets Abs.
OR Rel.
Tolerance ?
Meets Abs.
AND Rel.
Tolerance ?
NO
NO
NO
NO
NO
NO
NO
YES
NO
NO
NO
NO
NO
YES
YES
YES
NO
NO
NO
NO
NO
YES
YES
YES
NO
NO
NO
NO
NO
NO
NO
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 Impulses 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. This of course assumes that there are no input errors.
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
2/24/2003
1:35 PM
Page 333

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:
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
Relaxation
Value
1
0.5
0.2
0.1
0.05
0
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.
Old Flow Rate

(lbm/s)
10
10
10
10
10
10
Ideal New Flow

Actual New Flow
Rate (lbm/s)
Rate (lbm/s)
20
20
20
15
20
12
20
11
20
10.5
20
10
There are two relaxation parameters in AFT Impulse 3.0 (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
2/24/2003
1b.doc
1:35 PM
Page 334

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
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.
2/24/2003
1:35 PM
Page 335
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 Impulses 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.
2/24/2003
1b.doc
1:35 PM
Page 336
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.
There are different warning levels depending on the importance of the
warning.
2/24/2003
1:35 PM
Page 337

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.
Junction types and the Steady-State Solver

The junction types fall into four broad categories: pressure specification
junctions, flow specification junctions, pressure modifying junctions,
and pipe joining junctions. With judicious junction type selection you
can model almost any pipe system.
Pressure specification junctions

The Reservoir junction, Assigned Pressure junction, and the Spray
Discharge junction always behave as pressure specification junctions.
The Control Valve junction, when set as a pressure reducing or
sustaining valve, also behaves as a pressure specifying valve.
In addition, the Valve junction and the Relief Valve junction (when
cracked open) both allow valves to be specified as exit valves. Exit
valves have an associated back pressure, and therefore behave as
pressure specification junctions as well. A submerged pump acts like a
pressure junction. These junctions offer several complementary ways of
inputting known pressure information to your model.
Three other junction types can function as pressure junctions as special
cases. The Assigned Flow junction, when it is closed directly or by
another junction, can act as a pressure junction. The Gas Accumulator
can act as a pressure junction if it is in a location where it acts as a
topping tank or pressurizer. In such a case it will have no flow. A Surge
Tank can also act as a pressure junction. In each of these three cases, the
2/24/2003
1b.doc
1:35 PM
Page 338

junctions do not act as pressure junctions unless specifically specified by
the user. This specification is made on the Optional tab of each junction
in the area referring to steady-state flow behavior.
Pressure specification junctions do not balance flow. Instead, they apply
a known pressure at a specific location in the model. By applying a
known pressure, the flow at these junction types adjusts to the flow
required to maintain the pressure you specify. The output for one of
these junction types will show flow exiting or entering the system
through the junction. This is in contrast to other junction types where the
flow is always balanced and no flow goes in or out of the system unless
explicitly defined.
Pressure junctions are an infinite source of flow (i.e., mass). See Chapter
12 for more discussion on how pressure junctions work.
Flow specification junctions

The Assigned Flow junction and the Dead End junction always act as
flow specification junctions. In the case of Dead End junctions, the flow
is specified as zero. The Control Valve junction behaves as a flow
specification junction when it is specified as a Flow Control valve;
similarly, the Pump junction behaves as a flow specification junction
when it is set to an assigned flow.
The Assigned Flow junction allows nonzero flow rate specification. This
flow is defined as outflow or inflow, and the pressure adjusts at the
junction to that required to maintain the specified flow. For FCVs and
pumps with assigned flows, the pressure change across the junction
adjusts to achieve the specified flow.
How does a flow junction compare to a pressure junction? Besides the
obvious difference of providing pressure vs. flow, the flow junction is an
infinite source of pressure. In contrast, the pressure junction is an
infinite source of flow (i.e., mass).
For example, lets say you used an Assigned Flow junction to specify an
inflow to your pipe system. In so doing you want a flow rate of a million
pounds per day. But you accidentally assign the units wrong and input a
million pounds per second. Because the assigned flow junction is an
infinite source of pressure, it can provide an infinitely high pressure
whatever is necessary to drive the specified flow.
2/24/2003
1:35 PM
Page 339
Pressure modifying junctions

Pressure modifying junctions comprise the following junction types:
Area Change, Valve, Check Valve, Relief Valve (when internal open),
Pump, and General Component junctions. These junction types have
special geometric or flow control features that reduce the pressure in the
system. In the case of Pump junction, pressure is added to the system.
Pipe joining junctions

AFT Impulse has six pipe joining junctions: the Branch and the
Tee/Wye are two. The Tee/Wye junction joins three pipes together. The
Branch junction can join from two to twenty-five pipes. The Gas
Accumulator, Liquid Accumulator and Surge Tank accept from two to
twenty-five pipes and by default act essentially as branches during
steady flow. The Vacuum Breaker Valve junction accepts only two pipes
and acts a branch junction under steady flow.
Special junction
The Volume Balance junction is a special case. It acts like a pseudotransient junction where the mass flow rate is abruptly changed to
maintain a constant volume flow of two fluids with different densities.
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.
2/24/2003
1b.doc
1:35 PM
Page 340
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
Area Change
Crane 1988, A-26 and

Idelchik 1994, 208, 216
Orifice
Idelchik 1994, 218, 220
Tee/Wye
Idelchik 1994, chapter 7

and Miller
Valve
Ingersoll-Rand 1970 Crane

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:
Ploss = K
1
V 2
2
(8.18)
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.
2/24/2003
1:35 PM
Page 341
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:
Aup

K up = 2.6 sin 1
Adown
2
Aup
K up = 1
Adown
( < 45 degrees)
( > 45 degrees)
The conical contraction correlation (Crane 1988, A-26) is:
K up =
K up =
0.8 sin 1 down

Aup
2
Adown
Aup
0.51 down sin
Aup
2
Adown
Aup
( < 45 degrees)
( > 45 degrees)
The abrupt expansion correlation (Idelchik 1994, 208) is:
2/24/2003
1b.doc
1:35 PM
Page 342
Aup
K up = 1
Adown
The abrupt contraction correlation (Idelchik 1994, 216) is:
K up =
A
0.51 down
Aup
Adown
Aup
0.75
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
r/d
20 f T
24 f T
1.5
14 f T
10
30 f T
12 f T
12
34 f T
12 f T
14
38 f T
14 f T
17 f T
16
42 f T
50 f T
2/24/2003
1:35 PM
20
Page 343

Table 8.6
Pipe friction factors used for Crane formulas
Nominal
Size
Friction
Factor f T
Nominal
Size
Friction
Factor fT
Nominal
Size
Friction
Factor fT
0.027
0.019
8-10
0.014
0.025
2 , 3
0.018
12-16
0.013
18-24
0.012
0.023
0.017
0.022
0.016
0.021
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:
K = (n 1) 0.25f 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.
2/24/2003
1b.doc
1:35 PM
Page 344

Table 8.7
(deg.)
K values for Mitre bends

K
2 fT
15
4 fT
30
8 fT
45
15 f T
60
25 f T
75
40 f T
60 f T
90
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):
Aup
K up =
Aorifice
0.7071 Aorifice
Aup
Figure 8.9
0.375
Aorifice
+ 1
Adown
Sharp-edged orifice
The round-edged orifice, shown in Figure 8.10, is given by the

following (Idelchik 1994, 220):
2/24/2003
1:35 PM
Page 345
Aup
K up =
Aorifice
K 1 Aorifice

Aup

0.75
Aorifice
K 1 orifice
+21
Adown
Aup
Aorifice
+ 1
Adown
0.375
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.
2/24/2003
1b.doc
1:35 PM
Page 346

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
Ac
Ast
As
Ast
s = side branch
Ac
As
Figure 8.11 Idelchiks 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 Impulses 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 Arrangement tab.
The angle itself is specified on the Tee/Wye Model tab, where there
are other options for how to model the tee.
2/24/2003
1:35 PM
Page 347
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):
V 2
Vs
s
K c, s =
= 1 + 2 cos
2
Vc
Vc / 2
Vc
Ps
where is found from Table 8.8.
2/24/2003
1b.doc
1:35 PM
Page 348

Table 8.8
Diverging case values
As / Ac
Qs / Qc
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):
K c, st =
Q
= st s
Vc2 / 2
Qc
Pst
where st is given in Table 8.9:

Table 8.9
Diverging case st values
As / Ac
Qs / Qc
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):
K c, s =
Ps
Vc2 / 2
Q A 2
A
= A1 + s c 2 c
Ast
Qc As
2
2
Qs
Ac Qs
1
2 cos
As Qc
Qc
Straight run, constant area (Idelchik 1994, 417):

2
Q
A Q
K c, st =
= 1 1 s 2 c s cos
As Qc
Vc2 / 2
Qc
Pst
2/24/2003
1:35 PM
Page 349

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
Constant area, sharp
Constant area, r/D = 0.1
Area sum, sharp
Wye
Direction
Reference
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*
Diverging tee
Branch
Idelchik 1994, 459-461
Diverging tee
Run
Idelchik 1994, 453
Converging tee
Branch
Idelchik 1994, 439-442
Converging tee
Run
Idelchik 1994, 439-442
Splitting tee
From branch
Miller 1990, 323*
Combining tee
To branch
Miller 1990, 315*
Diverging tee
Branch
Idelchik 1994, 452
Diverging tee
Run
Idelchik 1994, 454-455
Converging tee
Branch
Idelchik 1994, 433-437
Converging tee
Run
Idelchik 1994, 439-442
Splitting tee
From branch
Miller 1990, 323*
Combining tee
To branch
Miller 1990, 315*
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):
CV =
Pvalve ( 62.4 / )
where Q is in gpm, P is in psid, and is in lbm/ft3.
2/24/2003
1b.doc
1:35 PM
Page 350

It can be shown that K and Cv are related by the following approximate
expression (Lyons 1982):
Kvalve
A2
= 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
P
K sprinkler =
where Q and P 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 P 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 P.
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
CD A =
2/24/2003
water
1:35 PM
K sprinkler
Page 351
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:
H1 s1
=
H 2 s2
Q1 s1
=
Q2 s2
(8.19)
where s is the speed. If the pump data is input as a polynomial, the speed
will affect the curve as follows:
2
H1 = a + bQ1 + cQ1 + dQ1 + eQ1
4
2
H 2 = s 2 H1 = s 2 a + s 2bQ1 + s 2 cQ1 + s 2 dQ1 + s 2eQ1

2
H 2 = s 2 a + s 2b
Q2
Q
Q
Q
+ s 2 c 2 + s 2 d 2 + s 2 e 3
s
s
s
s
3
Q
Q
H 2 = s a + sbQ2 + cQ2 + d 2 + e 23
s
s
2
(8.20)
where H1 is the input curve that typically represents 100% speed and
H2 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.
2/24/2003
1b.doc
1:35 PM
Page 352
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.
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.
2/24/2003
1:35 PM
Page 353
2/24/2003
1b.doc
1:35 PM
Page 354
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:
P = a V
2/24/2003
1b.doc
1:35 PM
(9.1)
Page 355

where:
P =
pressure surge
density
wavespeed
V =
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, such as transient cavitation, the pressure can exceed the
instantaneous prediction. The engineer should therefore be cautious in
its application.
Heres 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.
ft
ft 1 ft 2
lbf s 2
Pmax = 62
* 4000 *10 *
*
+ 50 psig
s
s 144 in 2 32.2 lbm - ft
ft 3
Pmax = 535 + 50
lbm
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):
2/24/2003
1b.doc
1:35 PM
Page 356
Chapter 9 Theory of Waterhammer and Solution Methodology 357
a2 =
K
1+ c1[(K E )(D e )]
where:
a
= wavespeed
K = fluid bulk modulus of elasticity
= fluid density
= 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 2
2. Thin-walled pipe anchored throughout
c1 = 1 2
3. Thin-walled pipe free to expand throughout
c1 = 1
4. Thick-walled pipe anchored at upstream end only
c1 =
2e
(1 + ) + D (1 2)
D
D+e
5. Thick-walled pipe anchored throughout
c1 =
2/24/2003
2e
(1 + ) + D 1 2
D
D+e
1:35 PM
Page 357

6. Thick-walled pipe free to expand throughout
c1 =
2e
(1 + ) + D
D
D+e
7. Circular tunnel
c1 =
2e
(1 + )
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:
t = 2
L
a
(9.2)
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.
2/24/2003
1b.doc
1:35 PM
Page 358
Momentum equation
Following from Wylie, et al., 1993, page 22, the momentum equation
can be expressed as
fV V
1 P V
+
+ g sin( ) +
=0
2D
x t
(9.3)
where:
P
= pressure
= velocity
= density
= distance along pipe
= time
= acceleration due to gravity
D = diameter
f
= friction factor
= 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
g x t
(9.4)
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.
2/24/2003
1:35 PM
Page 359

The equations are identified as L1 and L2 .
fV V
1 P V
+
+ g sin( ) +
= L1
2D
x t
a 2
V P
+
= L2
x t
(9.5)
(9.6)
Combine these two equation linearly using a parameter, , (Wylie, et al.,

1993, pp. 27)
L1 + L2 = 0
or,
V
1 P P V
+
+ a 2
+
x
x t t
fV V
+ g sin( ) +
2D
=0
(9.7)
If
dx 1
= = a2
dt
(9.8)
then
1
a
and
dx
= a
dt
From calculus
dP P dx P
=
+
dt x dt t
(9.9)
dV V dx V
=
+
dt x dt t
(9.10)
Substituting Equation 9.8 into 9.7
2/24/2003
1b.doc
1:35 PM
Page 360
V dx
P dx P V
+
+
+
x dt
x dt t t
+ g sin( ) +
fV V
2D
=0
(9.11)
and substituting Equations 9.9 and 9.10 into 9.11
fV V
dP dV
=0
+
+ g sin( ) +
2D
dt dt
or
fV V
1 dP
dV
+
+ g sin( ) +
=0
dt
2D
a dt
(9.12)
Now multiply Equation 9.12 by (a dt), which also equals dx
dP + adV + g sin( )dx +
fV V
2D
dx = 0
(9.13)
The gravity term contains the term , and the term can be converted as
follows
dz
dx = gdz
dx
g sin( )dx = g
(9.14)
where dz is the vertical change in pipe elevation. Therefore, substituting

9.14 into 9.13
dP + adV + gdz +
fV V
2D
dx = 0
Further substitution of the mass flow rate for velocity yields
dP +
a
f
dm& + gdz +
dx m& m& = 0
A
2 DA2
Integrating Equation 9.15 with the positive sign along the C+

characteristic from point A to P (Figure 9.1) yields
2/24/2003
1:35 PM
Page 361
(9.15)

PP
a
dP + A
PA
m& P
zP
dm& + g dz +
m& A
zA
xP
m& m& dx = 0
2 DA2 x
A
(9.16)
where the subscript P denotes the current solution point at x = i at time t

= t, 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 Ccharacteristic from point B to P.
The term on the right before the equal sign in Equation 9.16 can be
& to
integrated assuming a second order accurate representation of m
obtain
(PP PA ) + a (m& P m& A ) + g (z P z A ) +

A
f x
2 DA2
(9.17)
m& P m& A = 0
t = 5t
t = 4t
t = 3t
t = 2t
P
t = t
C-
C+
t=0
x=0
A
x = i-1
x=i
x = i+1
x=L
Figure 9.1 Method of Characteristics grid
Similarly
2/24/2003
1b.doc
1:35 PM
Page 362
(PP PA ) +
+
f x
2 DA2
a
(m& P m& A ) + g (z P z A )
A
m& P m& A = 0
(9.18)
Introducing two convenient parameters:

Impedance:
B=
a
A
Resistance:
R=
f x
2 DA2
Equations 9.17 and 9.18 then become
(PP PA ) + B(m& P m& A ) + g (z P z A ) + Rm& P m& A
=0
(PP PA ) + B(m& P m& A ) + 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:
C P = Pi 1,old + Bm& i 1, old g ( zi zi 1 )

C M = Pi +1,old Bm& i +1, old + 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
2/24/2003
1:35 PM
Page 363
& i, new .
BM are known, the only unknowns in 9.19 and 9.20 are Pi,new and m
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
BP + BM
(9.21)
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
BP + BM
(9.22)
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
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
2/24/2003
1b.doc
1:35 PM
Page 364
Vvapor , new = Vvapor ,old
+ m& up, new m& up, old + m& down, new + m& down, old
) 2t
(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
n t
a=
(9.24)
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:
ni =
Li
ai t
(9.25)
In Equation 9.25 as applied to each pipe in the system, the time step t
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.
2/24/2003
1:35 PM
Page 365

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 t. 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
ni =
Li
(1 )ai t
where 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
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.
2/24/2003
1b.doc
1:35 PM
Page 366
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).
CM
P
m& i, new = known
BM
Conversely, if the pressure is known at the downstream end, the positive
m& i, new =
C P Pknown
BP
If the known pressure changes with time, then a current pressure is

obtained for the current time step and used in the equations.
2/24/2003
1:35 PM
Page 367
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 g + 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
CM
P
m& i, new = known
BM
m& i, new =
C P Pknown
BP
If the reservoir height changes with time, then a current height is

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.
2/24/2003
1b.doc
1:35 PM
Page 368

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
) 2t
& up terms are known because they are specified terms for inflow
The m
& down terms are known. Similar to
junctions. For outflow junctions the m
a pipe interior node, when the vapor volume is negative, the cavity
collapses and the fluid pressure then rises above the vapor pressure.
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.
2/24/2003
1:35 PM
Page 369
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
) 2t
& down terms are always zero. For dead ends at the beginning
since the m
& up terms are zero. Similar to a pipe interior node, when
of a pipe, the m
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.
The compatibility equation is written for each junction. For the pipes
flowing into the branch, the positive equation is used (Equation 9.19)
m& i, new =
C P Pj
BP
where Pj is the branch junction pressure, as yet unknown.
2/24/2003
1b.doc
1:35 PM
Page 370

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
CP + CM
SC =
BP
SB =
BM
1
1
+
BP
BM
(9.27)
(9.28)
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
Pj =
SC + m& Applied
SB
(9.29)
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.
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:
2/24/2003
1:35 PM
Page 371
m& in, new m& in, old
Vvapor , new = Vvapor , old + + m& out, new + m& out, old
m& B, new + m& B, old
t
2
& B is the branch source/sink term defined as positive flow into

where m
& in terms are obtained from the positive compatibility
the branch. The m
& out terms are obtained from the negative equation.
equation, while the m
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:
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 = Rvalve m& valve 2

where
2/24/2003
1b.doc
1:35 PM
Page 372

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 = Rvalve m& valve 2

Rearranging terms,
m& valve 2
+
1
Rvalve
1
Rvalve
(CP,up BP,up m& valve )

(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
& valve , the upstream and downstream

After solving for flow rate, m
pressures can be obtained from the compatibility equations.
If the valve closes then 1/Rvalve = 0 and
m& valve = 0
Exit valves
If the valve is an exit valve, then the downstream pressure is known and
2/24/2003
1:35 PM
Page 373
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
) 2t
Vvapor , down, new = Vvapor , down, old

+ ( m& valve, new m& valve, old + m& down, new + m& down, old )
t
2
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.
pressure.
2/24/2003
1b.doc
1:35 PM
Page 374
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, t 3Pi +1,2t + Pi + 2,3t

m& infinity,new = 3m& i, t 3m& i +1,2t + m& i + 2, 3t
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)
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 g ( zi zi +1 )
2/24/2003
1:35 PM
Page 375
(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.
m& spray
Pexit
Figure 9.2
Spray discharge schematic
The relationship for pressure drop across the nozzle is as follows
2/24/2003
1b.doc
1:35 PM
Page 376
P =
1
V 2
2
P = Pj , new Pexit = sgn(m& spray )
2 C D Aspray
m& spray 2
& spray is defined positive flowing out of the system.

where, for clarity, m
In addition, the spray junction is similar to a branch junction, so the
following equation applies with the flowrate defined positive out.
Pj , new =
SC m& spray
SB
where SC and SB are given by Equations 9.27 and 9.28.

Combining the preceding equations and eliminating flow rate,
sgn (m& )
1
C D 2 Aspray 2 2
m& 2 +
S
1
m& + Pexit C = 0
SB
SB
or
&2
sgn(m& )m +
+
2 C D 2 Aspray 2
SB
m&
2 C D 2 Aspray 2 (S B Pexit SC )
SB
=0
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.
2/24/2003
1:35 PM
Page 377
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& spray , new + m& spray , old
t
2
& spray is the spray outflow term. The m& in terms are obtained
where m
& out terms are
from the positive compatibility equation, while the m
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.
Modeling pump as a known speed

Assuming that the pump curve can be modeled by a quadratic, then
H = 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
2/24/2003
1b.doc
1:35 PM
Page 378
2
Q
= constant1
= constant2
where is the pump speed. Applying these relationships the pump curve
is transformed
= a + b
Q pump
+ c
Q pump 2
H = 2 a + bQ pump + cQ pump 2
Converting to pressure rise
b
P = g 2 a +
m& pump +
m& pump 2
P = 2 a + bQ pump + cQ pump 2
(9.33)
where the primes on the constants were dropped after they were divided
through by g.
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
and substituting these equations into Equation 9.33
(C M + BM m& pump ) (C P BP m& pump )

= 2 a + bm& pump + cm& pump 2
This is a quadratic equation for the pump flow rate which can be solved
by the quadratic formula
2/24/2003
1:35 PM
Page 379
BP,up + BM , down b
Q pump =
2
c

4b 2 a + C P ,up C M , down
1 1
BP,up + BM , down b 2

)1 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:
H = a + bQ pump + cQ pump 2 + dQ pump 3 + eQ pump 4

and with the speed included,
= a + b
Q pump
+ c
Q pump 2
+ d
Q pump 3
+ e
Q pump 4
Combine this with the compatibility equation as in the previous section

to obtain:
(CM + BM m& pump ) (CP BP m& pump ) = 2a + bm& pump

+ cm& pump 2 +
m& pump 3 +
m& pump 4
This equation is then solved by Newton-Raphson iteration to obtain the

& pump .
flow rate, m
Backflow through the pump
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. In the
case of backflow it is assumed that the head rise for all negative flow
2/24/2003
1b.doc
1:35 PM
Page 380

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.
Flow rates past the pump maximum
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,
AFT Impulse 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.
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 b
m& pump =
2c

4b 2 a + PR CM , down
1 1
(BM ,down b )2

)1 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.
2/24/2003
1:35 PM
Page 381
Vvapor ,up, new = Vvapor ,up,old
+ m& up, new m& up, old + m& pump, new + m& pump, old
) 2t
Vvapor , down, new = Vvapor , down, old
+ m& pump, new m& pump, old + m& down, new + m& down, old
) 2t
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.
BM , down b
m& pump =
2
2c

4b 2 a + Psat CM , down
1 1
(BM ,down b )2

)1 2
Similar to a pipe interior node, when the vapor volume is negative the
pressure.
Modeling pump as with unknown speed

There are four pump models in AFT Impulse 3.0 that account for pump
inertial effects and predict the speed over time. These are:
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
2/24/2003
1b.doc
1:35 PM
Page 382

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.

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 efficiency data is also required as a function of
flow.
The method is presented by Watters (1979, p. 169). The unbalanced
torque is given by:
T = I
d
dt
(9.34)
where T is the torque on the pump, I is the moment of inertia of the

pump and entrained liquid, and d/dt is the change in angular speed with
time.
The pump power is related to the pump torque by:
P = T
(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
d
dt =
I
T
Assuming an average value of torque over the time step one obtains:
t
1
=
I
T
With the speed known at the previous time step, one can solve for the
new speed as
new = old T
2/24/2003
1:35 PM
t
I
Page 383

Adding in the definition of torque from Equation 9.35
new = old
Tnew + Told t
2
I
(9.36)
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.

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).
Tmotor T fluid = I
d
dt
Integrating over a time step:
new = old +
Tmotor , new + Tmotor , old T fluid , new + T fluid , old
2
2
(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.
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.
2/24/2003
1b.doc
1:35 PM
Page 384
Figure 9.3
Motor torque vs. speed data is used for startup

modeling
2/24/2003
1:35 PM
Page 385
Figure 9.4 Example graph of motor torque vs. speed

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:
H
P
=
H R PR
T
=
TR
Q
m&
v=
=
QR m& R
h=
2/24/2003
1b.doc
(9.38)
N
NR
1:35 PM
Page 386

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, is the ratio of the pump torque to the rated torque, v is the ratio
of the flowrate to the rated flowrate, and 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:
FH ( ) =
FB ( ) =
h
+ v2
2
2 + v2
where
2/24/2003
1:35 PM
Page 387
(9.39a)
= + tan 1
(9.39b)
Some references choose to define slightly differently:
= tan 1
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).
Figure 9.6
Variation of FH and FB for a pump with specific speed

of 0.46 (see Wylie, et al., 1993).
2/24/2003
1b.doc
1:35 PM
Page 388

AFT Impulse provides data of FH and FB vs. for pumps of twenty-one
specific speeds, as given by Wylie, et al., which was derived from
Donsky (1961). See Table 9.1 for a list of all specific speeds available.
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
(-)
0.3
0.39
0.42
0.46
0.48
0.55
0.57
0.71
0.79
1.21
1.36
1.44
1.61
1.83
1.9
2.32
2.48
2.54
2.78
3.2
4.94
Four quadrant data available in AFT Impulse
-- Specific Speed -(gpm units) (metric units)

810
15.7
1060
20.5
1140
22.1
1270
24
1320
25.5
1490
28.8
1565
30.3
1935
37.4
2160
41.8
3300
64
3725
71.9
3940
76.1
4400
85.1
5000
97
5200
101
6340
123
6790
131
6940
134
7600
147
8760
170
13500
261
Reference
Brown & Rogers, Partial, From Nevada Test, 1980
Kittredge, 1956
Brown & Rogers, from Bureau of Reclamation, 1980
Donsky, 1961
Kittredge, 1956
Thorley, 1996
Thorley, 1996
Thorley, 1996
Thorley, 1996
Brown & Rogers, from Bureau of Reclamation, 1980
Thorley, 1996
Thorley, 1996
Thorley, 1996
Thorley, 1996
Thorley, 1996
Donsky, 1961
Thorley, 1996
Donsky, 1961
With data for FH and FB, the compatibility equations (9.19, 9.20) are
used as follows:
Ppump = C M C P + (BM + BP )m& pump
(9.40)
But from Equations 9.38 and 9.39,
Ppump = PR 2 + v 2 FH
2/24/2003
1:35 PM
Page 389
(9.41)
m& pump = vm& R

Substituting Equations 9.41 into 9.40 obtains:
C P CM vm& R (BP + BM ) + PR 2 + v 2 FH = 0
(9.42)
Since the value of FH depends on , which depends on 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,
= NR
0 =
=
60
T0
TR
T
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,
d = N R
2
2
d = N R
( 0 )
60
60
Equation 9.34 can be written
T=
T + T0
d
= I
2
dt
(9.43)
Substituting these terms into Equation 9.43 obtains
+ o I
NR
( 0 ) = 0
TR 15t
(9.44)
Using Equations 9.38 and 9.39 in Equation 9.44 obtains
2/24/2003
1b.doc
1:35 PM
Page 390
( 2 + v 2 )FB + o I NT R 15t ( 0 ) = 0
(9.45)
Equations 9.42 and 9.45 are solved simultaneously using a NewtonRaphson 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 T f =
Tm + Tm,0
T f + T f ,0
d
=I
dt
2
where the m subscript refers to the motor torque and the f subscript
refers to the fluid torque. Equation 9.44 becomes:
f + f , o m m, o I
NR
( 0 ) = 0
TR 15t
and Equation 9.45 becomes:
( 2 + v 2 )FB + f ,o m m,o I NT R 15t ( 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:
2/24/2003
1:35 PM
Page 391
( ) N
P
I p = 1.5 107
0.9556
(9.46)
where P is the power in kW and N is the rotational speed in rpm.

An estimate of the motor rotational inertia is
1.48
P
I m = 118
N
(9.47)
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 liquidfilled 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.
Pj , new =
SC + m& A, 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,
K =
P
V 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:
V = (m& A, new + m& A, old )
t
2
Therefore,
2/24/2003
1b.doc
1:35 PM
Page 392
Pj , new = Pj , old
K t m& A, new + m& A,old

2V
Substituting and eliminating flow rate
Pj , new =
2V
m& A, old
K t
2V
SB +
K t
SC + Pj , old
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& A, new + m& A, old
t
2
& A is the accumulator flow out of the junction. The m& in terms
where m
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.
2/24/2003
1:35 PM
Page 393
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.
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.
Pj , new =
SC m& exit , new

SB

The pressure drop across the exit is
P =
1
V 2
2
or,
P = Pj , new Pexit =
2 Aexit
m& exit , new2
Substituting
SC m& exit , new

SB
Pexit =
2 Aexit
m& exit , new2
& exit , new can be obtained from the

This is a quadratic equation in which m
quadratic formula.
m& exit , new2 +
S
1 2K
2 K
Pexit C = 0
m& exit , new +
2
2
SB
S B Aexit
Aexit
If using the variable Cv model, the K factor above itself becomes a

function of P, which requires iteration to solve for the mass flow rate.
2/24/2003
1b.doc
1:35 PM
Page 394
Inline Relief Valve

When this valve opens, it is similar to a Spray Discharge junction with
two pipes connected. If the Variable Cv model is used, the solution
requires iteration.
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.
If the valve cracks, it acts like a spray discharge junction.
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.
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.
2/24/2003
1:35 PM
Page 395
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.
C CM P
m& i, new = P
BP + BM
Surge Tank theory

A surge tank is an open tank to the atmosphere (Figure 9.7) and is used
as a surge suppression device (Wylie, et al., 1993, pp. 128-129).
2/24/2003
1b.doc
1:35 PM
Page 396

The flow balance on the surge tank junction is similar to a branch
&T .
junction, with the deficit equaling the flow out of the surge tank, m
Similar to a branch, it can be shown that for all connecting pipes to the
junction
Pj , new =
SC + m& T , new
SB
LL
Orifice
m& T
LC
Figure 9.7
Connector
Pipe
Surge tank schematic
The optional connector pipe and orifice effect is incorporated by the

lumped inertia relationship as follows:
Pj , new = Psurface + gLL, new C1 C2 m& T , new

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:
LL, new = LL, old
m& T , new + m& T , old
2 AT
Substitution yields
2/24/2003
1:35 PM
Page 397
SC + m& T , new
m& T ,old t
Psurface + g LL, old +

SB
2
A
T
+g
m& T , newt
2 AT
C1 C2 m& T , new = 0
& T , new
Solving for m
m& T , new =
m&
t
SC
Psurface g LL,old T , old + C1
SB
2 AT
gt
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.
Gas Accumulator theory

A gas accumulator (shown in Figure 9.8) 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
& A.
junction, with the deficit equaling the flow out of the accumulator, m
Similar to a branch, it can be shown that for all connecting pipes to the
junction,
Pj , new =
2/24/2003
1b.doc
SC + m& A, new
SB
1:35 PM
Page 398

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
Pipe
m& A
LC
Figure 9.8
Gas accumulator schematic
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 )
2/24/2003
1:35 PM
Page 399
t
2

By substitution
SC + m& A, new
= PA, new C1 C2 m& A, new
SB
or
PA, new =
SC + m& A, 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
C4 =
1
+ C2
SB
C5 = V A, old + m& A, old

C6 =
t
2
t
2
This is a non-linear equation that can be solved using Newton-Raphson

iteration. Newton-Raphson has the following form:
xi +1 = xi
F ( 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

& A, new has been found that causes F to equal zero,
When the value of m
the equation will be satisfied. To solve by Newton-Raphson the
derivative of F is given by,
2/24/2003
1b.doc
1:35 PM
Page 400
dF
C7
= F =
+ C4 C5 + C6 m& A, new n
dm& A, new
C5 + C6 m& A, new
where
C7 = nC AC6
& A, new is calculated. This value can then be backBy iteration, m
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
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
m& =
2/24/2003
C d Ain Po
RTo
1:35 PM
1
1 +
( +1) (2(1 ))
Page 401
(9.48a)
Subsonic inflow
( +1)
2
Pi
2 Pi

1 Po
P
C A P
m& = d in o
RTo
(9.48b)
Sonic outflow
C A P
1
m& = d out i 1 +
2
RTL
( +1) (2(1 ))
(9.48c)
Subsonic outflow
m& =
C d Aout Pi
RTL
2
( +1)
Pi
2 Pi

1 Po
Po
(9.48d)
where:
Pi =
pressure in the pipe at the junction node
Po =
atmospheric pressure
To =
atmospheric temperature
TL =
liquid temperature
CdAin
CdAout =
flow area for inflow

flow area for outflow
gas constant
= 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
2/24/2003
1b.doc
1:35 PM
Page 402
Pi
t
Qdown, new + Qdown,old Qup, new Qup, old =

Vold +
2
RTL
m& gas , new + m& gas, old

mold +
2
Eliminating the volumetric flowrate using the mass flowrate obtains
Pi 2Vold
1
m& down, new + m& down,old m& up, new m& up, old
+
RTL t
L
2mold
+ m& gas , new + m& gas, old
) =
Using Equations 9.19 and 9.20 to eliminate mass flowrate yields:
Pi 2Vold
C P
1 Pi C M
& down, old P i m& up,old =

m
+
+
RTL t
BP
L BM
2mold
+ m& gas , old + m& gas, new
t
(9.49)
& gas, new as follows:

Now we can use Equations 9.48 to eliminate m
Sonic Inflow
The pressure at the junction can be solved directly from Equation 9.49 as
follows:
1
Pi 2
+
L BM
C
+ Pi M
L BM
2m
RTL old
t
2/24/2003
L BP
m& d
m& u
2V
CP
+ old old + old
t
L BP
L
L
+ m& gasold + m& sonic = 0
1:35 PM
Page 403
& sonic is the sonic inflow which can be calculated from Equation
where m
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
+
L BM L BP
m& d
m& u
2V
C
C
+ Pr Po old M P + old old
L BM L BP
L
L
t
2m
RTL old + m& gasold

t
2R 2
Pr
+ Cd Ain PoTL
Pr ( +1) = 0
(
)
1 To
Defining terms for Newton-Raphson solution yields
2/24/2003
1b.doc
1:35 PM
Page 404
)0.5 + C2 Pr 2 + C3Pr + C4 = F
0.5
dF
0.5C1 (Pr a Pr b ) (aPr a 1 bPr b 1 )+ 2C2 Pr + C3 =
dP
C1 Pr a Pr b
a=
+1
b=
C1 = Cd Ain PoTL
2R
( 1)To
1
1
+
C2 = Po 2
B
B
L M
L P
m& d
m& u
2V
C
C
C3 = Po old M P + old old
L BM L BP
L
L
t
2m
C4 = RTL old + m& gasold

t
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:
2/24/2003
1:35 PM
Page 405
1
1
+
Pr 2 Po 2
L BM L BP
m& d
m& u
2V
C
CP
+ Pr Po old M
+ old old
L BM L BP
L
L
t
2m
RTL old + m& gasold

t
+ Pr (C D Aout Po )
2RTL 2
Pr
Pr ( +1) = 0
1
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
+ C 2 C4
a=
1
Pr
)0.5 ( aPr a 1 + bPr b1 )
dF
dPr
+1
b=
C1 = Cd Aout Po
2RTL
1
1
1
+
C2 = Po 2
L BM L BP
m& d
m& u
2V
C
CP
+ old old
C3 = Po old M
L BM L BP
L
L
t
2m
C4 = RTL old + m& gasold

t
where the function F is solved for Pr.
2/24/2003
1b.doc
1:35 PM
Page 406

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
L BM L BP
m& d
m& u
2V
C
CP
+ Pi old M
+ old old
L BM L BP
L
L
t
( +1) (2(1 ))
+ Pi C D Aout RTL 1 +
2m
RTL old + m& gasold = 0

t
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 9.9.
1
z
2
orifice
Figure 9.9
Lumped inertia with friction schematic
2/24/2003
1:35 PM
Page 407

Summing forces acting on the lumped section, 2,
F1 F3 F friction Forifice + Fg = A2 L2
dV2
dt
or,
F1 F3 F friction Forifice + Fg = L2
dm& 2
dt
Now,
F1 = P1 A2
F3 = P3 A2
F friction =
Forifice =
fL2 m& 2 2
A2
D2 2 A2 2
m& 2 2
2 C D 2 Ao 2
A2
Fg = A2 gz
Substituting terms and dividing by area A2
P1 P3
1
fL2 1
L dm& 2
& 2 2 gz = 2
m& 2 2
m
D2 2 A2 2
A2 dt
2 C D 2 Ao 2
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 A2 2
1
L (m& m& 2 )
& 2 m& 2 gz = 2 2
m
A2
t
2 C D 2 Ao 2
Collecting terms and simplifying
P1 P3 = C1 + C2 m& 2
where
2/24/2003
1b.doc
1:35 PM
Page 408
C1 = P3 P1 + 2 gz
C2 =
2 L2
m& 2
A2 t
1
2 L2
fL2 1
m& 2 +
m& 2 +
2
2
2
D2 A2
A2 t
C D Ao
(9.50)
(9.51)
If a connector pipe exists but no orifice then
C1 = P3 P1 + 2 gz
C2 =
2 L2
m& 2
A2 t
2 L2
fL2 1
m& 2 +
D2 A2 2
A2 t
If an orifice exists but no connector pipe
C1 = P3 P1
C2 =
C D 2 Ao 2
m& 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
2/24/2003
1:35 PM
Page 409

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.
2/24/2003
1b.doc
1:35 PM
Page 410

Wylie, et al., (1993, pp. 4) gives the instantaneous waterhammer
equation as:
P = aV
(9.52)
where:
P =
pressure surge
density
wavespeed
V =
velocity change
To apply this equation, one would typically determine the steady-state

velocity and assume the flow stops suddenly. The -V 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 V obtained based
on the pipe diameter at the valve. However, if smaller diameter pipes
exist upstream that carry the same flowrate, their V will be larger
because they have less area. This will result in a larger pressure surge in
the smaller pipe.
2/24/2003
1:35 PM
Page 411

To be safe, Equation 9.52 needs to be applied to the pipe with the largest
V 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 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.
2/24/2003
1b.doc
1:35 PM
Page 412
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 timebased 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.
2/24/2003
1b.doc
1:35 PM
Page 413
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
2/24/2003
1b.doc
1:35 PM
Page 414
Chapter 10 Modeling Time and Event Based Transients 415

criteria is such that the transient will be initiated when the outlet static
pressure in Pipe 2 exceeds 150 psig. If the pressure at this location never
reaches 150 psig, 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 150
psig 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.
2/24/2003
1:35 PM
Page 415

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
50 psig. 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.
2/24/2003
1b.doc
1:35 PM
Page 416
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.
2/24/2003
1:35 PM
Page 417

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.
2/24/2003
1b.doc
1:35 PM
Page 418
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 SteadyState 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.
2/24/2003
1:35 PM
Page 419
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 timebased 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.
2/24/2003
1b.doc
1:35 PM
Page 420
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.
2/24/2003
1:35 PM
Page 421
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.
2/24/2003
1b.doc
1:35 PM
Page 422
Figure 10.6 All junction transient data is shown in the Model Data
window on the Transient Data tab.
2/24/2003
1:35 PM
Page 423
2/24/2003
1b.doc
1:35 PM
Page 424
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).
2/24/2003
1b.doc
1:35 PM
Page 425
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.
2/24/2003
1b.doc
1:35 PM
Page 426
Chapter 11 Reducing Surge Pressures 427
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.
2/24/2003
1:35 PM
Page 427
2/24/2003
1b.doc
1:35 PM
Page 428
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
2/24/2003
1b.doc
1:35 PM
Page 429

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
2/24/2003
1b.doc
1:35 PM
Page 430
Chapter 12 Special Topics: Including Troubleshooting Models 431

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.
2/24/2003
1:35 PM
Page 431
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.
2/24/2003
1b.doc
1:35 PM
Page 432

12
Pressure (psid)
10
8
6
4
2
0
0
50
100
150
200
150
200
Flow (ft3/s ec)

12
Pressure (psid)
10
8
6
4
2
0
0
50
100
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.
2/24/2003
1:35 PM
Page 433
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.
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
2/24/2003
1b.doc
1:35 PM
Page 434

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
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 Impulses 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 lbm/sec, even though some operate at 0.1
lbm/sec and others at 10 lbm/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.
2/24/2003
1:35 PM
Page 435
Change boundary conditions

Changing the boundary pressures or flow rates can sometimes give you
insight into what is giving the model trouble.
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 submodels will not converge, it is doubtful your original model will
converge either.
Control valve failure issues in steady-state

AFT Impulse offers modeling of PRVs, PSVs and FCVs. 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
2/24/2003
1b.doc
1:35 PM
Page 436

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.
Altogether, there are four failure options for a control valve:
1. Maintain control by adding pressure (useful for diagnostics only)
2. Go to the valves full open state
3. Close the valve
4. Go to the valves full open state but close if the flow reverses
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 during the steady-state part of the simulation.
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.
2/24/2003
1:35 PM
Page 437
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 60 miles 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
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.
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.
One approach is to model pipe P1 as the full 60 mile length. Thus all the
pipe stations along 60 miles 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 60 mile
pipe except for two stations at the entrance.
2/24/2003
1b.doc
1:35 PM
Page 438
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.
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
2/24/2003
1:35 PM
Page 439

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.
Discharge
Tank (J2)
Transfer
Pipeline
Supply
Tank (J1)
Transfer
Pump (J3)
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
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.
2/24/2003
1b.doc
1:35 PM
Page 440

J2
P3>
J1
J3
P1>
J4
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.
Discharge
Tank (J2)
Transfer
Pipeline
Supply
Tank (J1)
Transfer
Pump (J3)
Shoreline
Elevation
Lake Michigan (J4)
Figure 12.7 Physical system represented by model in Figure 12.6.
2/24/2003
1:35 PM
Page 441

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.
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
2/24/2003
1b.doc
1:35 PM
Page 442

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.
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.
Inlet
Q = 1 ft3/s
El = 0 ft
L = 100 ft
D = 2 in.
f = 0.02
Outlet
Q = 1 ft3/s
El = 0 ft
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
2/24/2003
1:35 PM
Page 443

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
nine 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 steadystate 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
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.
2/24/2003
1b.doc
1:35 PM
Page 444
L = 100 ft
D = 2 in.
f = 0.02
Inlet
Q = 1 ft3/s
El = 0 ft
L = 100 ft
D = 2 in.
f = 0.02
Inlet
P = 10 psig
El = 5 ft
L = 100 ft
D = 2 in.
f = 0.02
Inlet
P = 10 psig
El = 5 ft
Outlet
P = 0 psig
El = 5 ft
Outlet
Q = 1 lbm/s
El = 0 ft
Outlet
P = 0 psig
El = 5 ft
Figure 12.0 a-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
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.
2/24/2003
1:35 PM
Page 445

Table 12.1 Four logical possibilities of boundary conditions for
single pipe systems
CASE
Inlet
Pressure
Fig. 12.9
Fig. 12.10a
Fig. 12.10b
Fig. 12.10c
Inlet
Flow
X
Outlet
Pressure
Outlet Flow
OK ?
No
Yes
X
X
Yes
Yes
Comments
No reference pressure, cannot
calculate pressures
1 known flow, can calculate
everything directly
everything directly
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.
Outlet
Q = 0.5 ft3/s
El = 0 ft
Inlet
Q = 1 ft3/s
El = 0 ft
Outlet
Q = 0.5 ft3/s
El = 0 ft
Figure 12.11
2/24/2003
1b.doc
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.
1:35 PM
Page 446

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 1 ft3/s, and the
downstream flow is demanding 1 ft3/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.
Inlet
P = 10 psig
El = 10 ft
J1
J2
J3
Outlet
Q = 1 ft3/s
El = 0 ft
FCV
Q = 1 ft3/s
Inlet
P = 10 psig
El = 5 ft
J1
J2
J3
Outlet
P = 0 psig
El = 5 ft
FCV
Q = 1 ft3/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.
2/24/2003
1:35 PM
Page 447
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.
2/24/2003
1b.doc
1:35 PM
Page 448

Outlet
Q = 1 ft3/s
El = 0 ft
Inlet
P = 10 psig
El = 5 ft
Pump with Known (assigned) Flow
Q = 1 ft3/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 FCVs 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, lets say we choose a pump that results in the pressure drop
across the FCV being 0.01 psid. 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.
2/24/2003
1:35 PM
Page 449
Inlet
P = 10 psig
El = 5 ft
Inlet
P = 10 psig
El = 5 ft
Outlet
P = 0 psig
El = 5 ft
FCV
Q = 1 ft3/s
Pump with
Assigned Flow
Q = 1 ft3/s
FCV
Q = 1 ft3/s
FCV
Q = 1 ft3/s
Outlet
P = 20 psig
El = 5 ft
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.
2/24/2003
1b.doc
1:35 PM
Page 450
Inlet
P = 10 psig
El = 5 ft
Outlet
P = 20 psig
El = 5 ft
Pump with
Known Flow
Q = 1 ft3/s
Inlet
P = 10 psig
El = 5 ft
Pump with
Known Flow
Q = 1 ft3/s
FCV
Q = 1 ft3/s
PDCV
dP = 5 psid
Inlet
P = 10 psig
El = 5 ft
Outlet
P = 20 psig
El = 5 ft
Outlet
P = 20 psig
El = 5 ft
Pump with Curve FCV
Q = 1 ft3/s
dP = f(Q)
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.
Lets 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 FCVs, 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.
2/24/2003
1:35 PM
Page 451

J6
P5^
^P8
J5
J8
FCV
Q = 100 gpm
FCV
Q = 100 gpm
P4^
J1
J3
J2
P1>
P2>
P7^
J7
J4
P3>
P6>
Pump known flow

Q = 200 gpm
Figure 12.17
A pump with an assigned flow in series with two

parallel FCVs. 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 FCVs are put in parallel, frequently the pipe design has one of
the FCVs 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 FCVs, 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 FCVs in the parallel system will have
greater than the minimum pressure drop thus satisfying the pressure
drop requirement for those FCVs as well.
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 FCVs 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
2/24/2003
1b.doc
1:35 PM
Page 452

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 FCVs
have identical pressure drops, any of the FCVs will serve as the PDCV.
The preceding process can be extended to cases where there are three or
more FCVs in parallel, and also cases with more than one pump in
parallel supplying the FCVs.
In Figure 12.17 the most remote valve is J8, the one on the right. If the
minimum pressure drop is 5 psid, change J8 to a PDCV with a 5 psi
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 5
psid because it is closer to the pump. Also notice how the flow through
J8 is still 100 gpm even though it is not controlling flow. Once again, the
reason for this is that the pump is controlling the flow to 200 gpm, and
with the J5 FCV controlling to 100 gpm, the excess must go through J8.
The pump is thus sized and its pump head rise is shown in Figure 12.18b
(113.5 feet).
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.
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
2/24/2003
1:35 PM
Page 453

PRV is used to control pressure downstream of the valve, while a PSV
controls pressure at the valve inlet. One thing that PRVs and PSVs
have in common with FCVs 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 PRVs pressure drop, the pressure upstream
of the PRV cannot be determined and thus no unique solution exists.
Inlet
Q = 1 ft3/s
El = 10 ft
Outlet
P = 0 psig
El = 10 ft
PRV
Pdown = 10 psig
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 PRVs and PSVs, 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.
2/24/2003
1b.doc
1:35 PM
Page 454

a
Inlet
Q = 1 ft3/s
El = 10 ft
Outlet
P = 0 psig
El = 10 ft
PRV
Pdown = 10 psig
Inlet
P = 30 psig
El = 10 ft
Outlet
Q = 1 ft3/s
El = 10 ft
PRV
Pdown = 10 psig
Inlet
P = 30 psig
El = 10 ft
Inlet
Q = 1 ft3/s
El = 10 ft
Inlet
P = 30 psig
El = 10 ft
Inlet
P = 30 psig
El = 10 ft
PRV
Pdown = 10 psig
PSV
Pup = 20 psig
PSV
Pup = 20 psig
PSV
Pup = 20 psig
Outlet
P = 0 psig
El = 10 ft
Outlet
P = 0 psig
El = 10 ft
Outlet
Q = 1 ft3/s
El = 10 ft
Outlet
P = 0 psig
El = 10 ft
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.
2/24/2003
1:35 PM
Page 455

Table 12.2 Summary of six possibilities for PRV and PSV
configurations.
CASE
Inlet
Pressure
Fig. 12.20a - PRV
Fig. 12.20b - PRV
Fig. 12.20c - PRV

Fig. 12.20d - PSV
Fig. 12.20e - PSV
Fig. 12.20f - PSV
Inlet
Flow
X
Outlet
Pressure
X
Outlet Flow
X
X
OK ?
Comments
No
No reference pressure upstream of

PRV, cannot calculate
Yes

everything directly
everything directly
No reference pressure downstream
of PSV, cannot calculate
Yes
Yes
X
No
Yes
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 FCVs.
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.
2/24/2003
1b.doc
1:35 PM
Page 456

J1
J5
a
P1v
J2
P4^
J3
J4
P2>
P3>
Open
J1
J7
P6>
J5
P1v
J2
P4^
XJ4
J3
P2>
J6
P5>
P3>
J7
P6>
J6
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.
2/24/2003
1:35 PM
Page 457

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!
2/24/2003
1b.doc
1:35 PM
Page 458
Figure 12.22
The Section Pipes window alerts you to pipes that

could be combined and possibly reduce model
runtime.
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.
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
2/24/2003
1:35 PM
Page 459

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
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.
2/24/2003
1b.doc
1:35 PM
Page 460
APPENDIX A
Keyboard Modifiers and Shortcuts
The keyboard modifiers and shortcuts for the Workspace are

summarized here for convenient reference
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.
2/24/2003
1b.doc
1:35 PM
Page 461

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.
2/24/2003
1b.doc
1:35 PM
Page 462
Appendix A
Keyboard Modifiers and Shortcuts 463
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 additional losses in the original pipe will be left in the original
pipe, and the new pipe will have no additional losses.
2/24/2003
1:35 PM
Page 463
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 by 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.
2/24/2003
1b.doc
1:35 PM
Page 464
Appendix A
Keyboard Modifiers and Shortcuts 465
The F5 function key also behaves this way for junction specifications
windows.
2/24/2003
1:35 PM
Page 465
2/24/2003
1b.doc
1:35 PM
Page 466
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
2/24/2003
1b.doc
1:35 PM
Page 467

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.
2/24/2003
1b.doc
1:35 PM
Page 468
APPENDIX C
Installation Issues
Customization files
As you customize AFT Impulse, two new files are created in your
WINDOWS directory:
IMPULSE3.INI
IMP_USER3.DAT
The IMPULSE3.INI file contains the information to customize the
features in AFT Impulse. The IMP_USER3.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 and copy protection tokens 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.
2/24/2003
1b.doc
1:35 PM
Page 469

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 both the data and the copy protection token are located.
These paths are specified in the control.aft file. Typically the token is in
the same location as the data files, however, this is not a requirement
(see Special Network Installations).
When Chempak is installed the setup procedure automatically updates
any control.aft file specified with the path to the data and the token.
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):
ChempakTokenPath = C:\AFT Products\Chempak Data
ChempakDataPath = C:\AFT Products\Chempak Data
This will instruct AFT Impulse to look in the C:\AFT Products\Chempak
Data directory for the data and also for the token.
Special network installations (remote tokens)

At times it may be more convenient to have AFT Impulse installed
locally on each client but have the token located on a network so
multiple users can access AFT Impulse concurrently. This is especially
true for dialup and remote login networks where access to the server is
relatively slow. With AFT Impulse installed locally on the client, it will
load quickly and the only network traffic is the checking for tokens.
The token normally resides in a hidden, system folder of the application
folder. The hidden folder for AFT Impulse is named 0930.092. The key
file, cimp.dll, is located in the application folder.
2/24/2003
1b.doc
1:35 PM
Page 470
Appendix C Installation Issues 471

However, to setup a remote token, where the token is placed in a
different location than the application, first create a folder on the remote
location or server named AFT Impulse 3.0 (for example,
R:\Engineering\AFT Products\AFT Impulse 3.0). Then copy the key file,
cimp.dll, from the Master Floppy Disk into it. Then move the token
using CCMOVE and specify the folder you created
(R:\Engineering\AFT Products\AFT Impulse 3.0) as the destination. The
hidden folder will automatically be created as the token is moved.
In order for this remote token method to work, AFT Impulse on the
client has to know where to look for the tokens. The path to the tokens
(actually the folder were the key file is located) must be specified for
each client. Manually edit the control.aft file to specify the folder where
the cimp.dll file is located. Control.aft should be located on the client in
the application folder or the Windows\System folder. Continuing with
the above example, if cimp.dll is located on R:\Engineering\AFT
Products\AFT Impulse 3.0 then add/edit the following line to the
[SEARCH PATHS] section:
ImpulseTokenPath = R:\Engineering\AFT
Products\AFT Impulse 3.0
The application can be installed both on the network and the client. In
this case the clients that reference the application on the server do not
need to have their control.aft file modified.
As an extension of the remote token method, the data files for the
Chempak database can also be installed on the local client and the
Chempak token can be placed on the server. You will need to modify the
information in control.aft as described in the Specifying Where
Chempak is Located section.
Network installations
AFT Impulse 3.0 requires a number of auxiliary files to run properly.
Because of how 32-bit Windows operating systems work (Windows 9x
and NT), certain files must be registered in the Windows registry. For
software installed directly on the client PC, this situation works well.
However, it creates a dilemma and an inconvenience for software
installed on a local or wide area network.
2/24/2003
1:35 PM
Page 471

The dilemma is where to locate the auxiliary files on the network in
the AFT Impulse folder or on the client PC local hard drive. The
implications of this dilemma are discussed at length below.
The inconvenience is that the installation becomes a two-step process.
The first step is the network administrators installation to the network,
and the second step is the registration of auxiliary files on each client
PC.
Following is a discussion of the this issue as related to AFT Impulse
installation and operation.
Locate auxiliary files on the client PC hard drive The installation
process is a two-step process. The first step is to install AFT Impulse on
the network using the Network option in the setup program. This
performs a typical network installation except that the auxiliary files are
copied into a subfolder below AFT Impulse. The second step is for each
client to run the AFT Impulse Client Setup program installed into the
AFT Impulse network folder. The Client Setup program automatically
detects that the Network option was chosen in the first step, and installs
the auxiliary files from the network to the clients SYSTEM folder and
registers them there. This process uses standard installation protocol
whereby the auxiliary files are copied to the SYSTEM folder only if they
do not exist, or are newer versions than those that already exist on the
client PC.
Advantages of this approach are: 1) Since the auxiliary files for each
user are installed and registered on the client PC, the network folder can
be more easily relocated on the server because the auxiliary files will not
need to be registered again. 2) Problems that occur with loading the
Graphics Server will be avoided.
Disadvantages of this approach are: 1) Later re-installs of AFT Impulse
may use more up to date files than the ones in the current installation,
and these files will also need to be copied to all client PCs. 2) All client
PCs will have these files on the local hard drive, thus using extra disk
space.
Although this approach 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).
2/24/2003
1b.doc
1:35 PM
Page 472
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 3.0.
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
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 3.0, 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 3.0, 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
2/24/2003
1:35 PM
Page 473

files, you can change the names back to the original names. This will
then cause AFT Impulse 3.0 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 3.0 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 3.0 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
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 3.0 at the same time as the other software which
needs these files.
2/24/2003
1b.doc
1:35 PM
Page 474

5. Reinstall AFT Impulse with the Network option to run auxiliary
files on the client (recommended) This option is the best longterm 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.
2/24/2003
1:35 PM
Page 475
2/24/2003
1b.doc
1:35 PM
Page 476
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 3.0 Help system. To learn more about
Chempak, open the Help system and search for Chempak.
2/24/2003
1b.doc
1:35 PM
Page 477
2/24/2003
1b.doc
1:35 PM
Page 478
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 180 Days
AFT provides free support for the first 180 days 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 the first 180 days, 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
2/24/2003
1b.doc
1:35 PM
Page 479

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
2/24/2003
1b.doc
1:35 PM
Page 480
References
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.
Karassik, I.J., J.P. Messina, P. Cooper and C.C. Heald, Pump Handbook,
3rd Edition, McGraw Hill, New York, NY, 2001.
2/24/2003
1b.doc
1:35 PM
Page 481

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.
2/24/2003
1b.doc
1:35 PM
Page 482
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.
2/24/2003
1b.doc
1:35 PM
Page 483

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.
2/24/2003
1b.doc
1:35 PM
Page 484
Glossary 485
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 but less than 10,000.
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.
2/24/2003
1:35 PM
Page 485

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.
2/24/2003
1b.doc
1:35 PM
Page 486
Glossary 487
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.
are entered.
The window in which an object's properties
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.
2/24/2003
1:35 PM
Page 487

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.
2/24/2003
1b.doc
1:35 PM
Page 488
Example models 7
Index
Getting started 6
installation 5
A
Absolute roughness 62, 186, 187, 295,
323
New features in version 3.0 11

Overview 7
Absolute tolerance See Steady

Solution Control window
qualifications for use 4
Additional Pipe Losses 63. See Losses
using online help 7
summary of capabilities 2
customizing 288
design factors 192
AFT IMPULSE LOCAL USER

DATABASE 298
in Solver 340
AFT Impulse window 16, 71
Model Data window 87
AFT Mercury
Converting models from 10
Non-Newtonian flow 329
AFT Standard fluid 29, 134
specifying 62
splitting pipes 463
Change Fluid Data window 294
when splitting pipes 124
Database 4, 141, 353
window 189
Database editing 146
ADDLOSS.DAT 289
Aligning objects 81
Adjust Pipes Automatically 81
Analysis type
steady-state vs. transient 175
Affinity laws 232
Animate Using Output File See

Animation
for pumps 352, 378, 382

AFT DEFAULT INTERNAL
DATABASE 298
Animate Using Solver See Animation

Animation 53, 107
AFT Fathom
Annotation Manager 78
Converting models from 10
Annotation tool 17, 78
AFT Impulse
converting custom databases from 1.0
9
Annotations 125
engineering assumptions 4
Area Change
2/24/2003
1:35 PM
Page 489
Annotion Manager 125

as Additional Pipe Loss 190
use in open and closed systems 318
Area Change junction 340
using static and stagnation pressure

318
abrupt transition 208, 342
waterhammer theory 367
conical transition 208, 342
Atmospheric pressure 144
loss factors 209, 342
Autosave 286
Specifications window 208
Auxiliary Graph Formatting window

111, 288
Artificial transients 46, 153, 459

manual detection 460
Stop run when detected 153
ASME Steam Tables 29
Base area 61, 208, 217, 230, 263, 347
Assigned Flow junction 339
Specifying 207
behavior in hydraulic solution 339
Batch mode 172
contrast with pressure-type junction

59, 339
BATCH.LOG 173
example 443
Bends
loss factors 343
loss factors 62
source/sink indicator on Workspace
280
Bernoulli equation 313

Bingham Plastic See non-Newtonian
Special features for steady flow 211
Black and white junction icons 172
Branch junction 18, 262, 340
Transient Data 211
flow source or sink 216
loss factors 62
Specifications window 34, 214
Assigned Pressure junction 338

How pressure junctions work 439
Transient Data 216
loss factors 62

Brecht & Heller method See nonNewtonian
periodic data 214

the role of pressure junctions 442
Transient Data 214
2/24/2003
1:35 PM
Page 490
Index 491
compatibility equations 363, 367, 368,
369, 370, 373, 375, 379
C
Calculator 175
Component Database 296
Cavitation 140, 293
computing stations 149, 409
transient 153
graphing 52
Check Valve
in output file 92
theory 395
Check Valve junction 340
Delta Pressure to Re-Open 218
Forward Velocity to Close Valve
218
max/min results in output 96

numbering convention 149
results in output window 99
Connectivity 126, 128, 129
inherent event logic 418
graphical connection 127
loss factors 218
in Junction Specification window

201
Special Condition check valve 218

Transient Data 218
in Pipe Specifications window 184

model connection 128
Checklist 26, 132
Constant Pressure Drop Valve See

Pressure Drop Control Valve
Chempak 29, 134, 141, 353
Continuity equation
accuracy options 353
steady-state 312
mixtures 29
transient 359
specifying Chempak file locations

470
chopped sine wave 211, 214
Clipboard 90, 111
Color Map See Visual Report window
column separation See vapor cavitation
communication time 43, 65, 66, 358,
375, 437
1:35 PM
action if control is lost 221

Control Transient 223
failure to control issues 436
Colebrook-White 324
2/24/2003
Control Valve junction 338
Page 491
Failure Transient 222

Open Percentage Table 222
sizing pumps with flow control valves
448
special condition options 167, 221

Special Conditions 221
Parameter Preferences 271
pipe materials 294
static and stagnation pressure 318
Toolbox Preferences 284
Transient Data 222
Unit Preferences 273
types 220
Visual Report window 288
Workspace window 82, 276
Control.aft file 469

controlling pipe 43, 44, 147, 366, 458
Conventions See also Base area
flow entering and exiting 60
Convergence 336
Converting models from AFT Impulse
1.0 9
Copy 79, 80, 122, 124, 126
Customizing AFT Impulse See

Databases. See Model Data Control
window. See Parameters and Unit
Preferences window. See Toolbox
Preferences window. See Workspace
Preferences window. See Output
Control window. See Visual Report
Control window
Cut 79, 126
cyclic dual events 416. See Event
transients
Create Mixture window 141

Crude oil 324
CTRL key See Keyboard modifiers.
See Keyboard modifiers
D
Darcy-Weisbach 187, 312, 324
Database Manager 160, 291, 297
Custom databases See Databases

Customize Graph window 110
Benefits of shared databases 302
Customizing AFT Impulse
Connecting and disconnecting

databases 298
additional pipe losses 288
creating network databases 305
databases 290
DATABASE.LIB 298, 304, 307
files created 469
Editing databases 300
fluids 292
Enterprise network databases 304
graph styles 288

junction (component) databases 296
IMP_DBUSER.LIB 305
Making databases available 298
Output window 92, 287
2/24/2003
1:35 PM
Page 492
Index 493
sharing databases using
DATABASE.LIB 307
DATABASE.LIB 298, 304, 307, 309
Databases 32, 290
Delete 79, 126

Density 140
Design Alerts See Pipe Specifications
window
Benefits of shared databases 302

Component database 201, 232, 296,
297
cross-plotting against results 106

Design factors
defaults 272
Connecting to external shared

databases 309
converting from AFT Impulse 9
Creating an enterprise-wide network
database system 304
junctions 206, 342

pipes 192, 329
Discrete Vapor Cavity Model 364, 409
Distributed loss 62, 340
custom junctions 296
Drag-and-drop 3
fluid See AFT Standard fluid or

Chempak
Duffy method See non-Newtonian
fluid database 141, 146, 292
Dynamic Pressure 317
fluid properties 146, 290
Dynamic viscosity 140
Duplicate 79, 122, 124, 126
Friction Data Set 295

IMP_USER3.DAT 271, 291, 300,
306, 469
pipe materials 185, 294
special control files 302
Specifying common locations 286
Dead End junction 339
Defining objects
Checklist - Define All Pipes and
Junctions 30, 134
E
Editing
on Workspace 79
EGL 106
Elbow
See Bend 343
Elevation 106, 200
default 273
Event messages 103, 420
Event transients 414
Example models 7
object status 80
Exit valve 373. See Valve junction
2/24/2003
1:35 PM
Page 493

Explicit friction factor 62, 188
MIT Equation 324
Extended Model Check 175
non-Newtonian 326
Relative roughness 323
Frictional Losses See Losses
Fanning friction factor 323
Frictionless pipe 62, 188, 326
FCV See Flow Control Valve

Find utility 163
Flip workspace See Workspace

window
Gas Accumulator junction 340

connector pipe 227
Flow Control Valve 220, 339, 436,

447, 450
Flow rate relaxation See Steady
Solution Control window
flow restrictor 227

Graphing transient data 229
in output file 155
Fluid database See AFT Standard fluid

or Chempak
Initial Gas Volume 225
Fluid Database window 292, 294
Maximum and minimum volumes

226
fluid transients
lumped inertia 407
orifice 227
See waterhammer
Polytropic Constant 225
fluidhammer
See waterhammer
Fouling 187
Friction Data Set 186, 295
surge reduction 427
Friction factor 145, 187
changes during transient flow 153

Friction models 62, 322
Absolute roughness 323
Gaussian Elimination 336

With pivoting 336
General Component junction 340
Frictionless pipe 326
loss factors 229
Hazen-Williams 324
Hydraulically smooth 324
General Preferences 286
Miller Turbulent method 325
Global Junction Edit window 268
2/24/2003
1:35 PM
Page 494
Index 495
Global Pipe Edit window 267
Glossary 483
Icon size 278
Graph Results window 8, 51, 72, 104,

288
ID numbers 22, 31, 122, 124, 146, 183,

281
Animation 53, 107
ID Reduction - Scaling 186
Import Graph Data 105
IMP_DBUSER.LIB 305, 308, 309
Save Graph Data 105

Select Data 52
IMP_USER3.DAT 271, 286, 291, 300,

306, 469
Select Graph Data 104
impedance 363
Impulse Examples.hlp 7
Graphics Server
problems loading 473
Gravitational acceleration 5, 145
IMPULSE3.INI 271, 286, 469

Infinite Pipe
example 437
Grid
On Workspace 24, 115, 279
Grouping objects 80, 125

theory 375
Initial guesses
flow guess 191
Hazen-Williams 188, 324
generation 205
H-Equation method 311
pressure guess 204
HGL 106, 312
using for troubleshooting models

435
Hide Empty Columns

Highlight feature 32, 129, 180
Homologous pumps See Affinity Laws
Inspection feature 33, 130, 197, 199

Showing only undefined items 464
Installation 469
Hydraulic Grade Line See HGL
Network installations 471
hydraulic transients
special network installations 470

Intermediate Pipe Elevations 193
See waterhammer
Hydraulically smooth 62, 188
Irrecoverable losses See Losses
Hydrocarbons - roughness 325
Iteration history 336, 435
2/24/2003
1:35 PM
Page 495
Notes 207
Jacobian matrix 314
Number 199
Jump 128, 179, 202
Pump 232
Junction icons
Relief Valve 248

Reservoir 250
printing black and white 172

Junction label position 281
Same As Junction 201
Junction Results table 90
Junction Specifications window 194
Spray Discharge 253
Area Change 208
Status 208
Assigned Flow 209
Surge Tank 255
Assigned Pressure 212
Tee/Wye junction 261
Branch 214
Vacuum Breaker Valve 264
Check Valve 217
Valve 263
connected pipes 201
Volume Balance 266

Junctions 59
Control Valve 219

Database list 201
and the Steady-State Solver 338
Dead End 223
Base area 207
design factor 206
comparison of types 195, 338
displaying name 206
connectivity 126
Elevation 200
creating 124
format 197
customizing icons 285
Gas Accumulator 224
defining 128
General Component 229
displaying data in Output window

96
Initial pressure guess 204

Jumping to next higher or lower
number junction 129
displaying in Model Data 88

displaying in Visual Report 115
Liquid Accumulator 231
displaying on Workspace 83
Name 199
displaying transient data in Graph

Results 106
2/24/2003
1:35 PM
Page 496
Index 497
displaying transient results in Output
window 103
editing 79, 126

Lock 79, 125
finding 163
Loss factor 60, 62, 189, 197, 198, 208,

214, 217, 221, 229, 251, 261, 289,
341
flow specifying 339

Global Editing 268
icons on Toolbox 79
ID numbers 124
keyboard shortcuts 463
loss factor 62
negative flow rates 197
pipe joining 340
loss models 36, 38, 341

Losses 60, 62, 313, 323, 340. See
Additional Pipe Losses
LU Decomposition 336
Lumped inertia pipe with orifice
theory 407
placing on Workspace 124
pressure losses 340
Matrix Method 336
pressure modifying 340
Merging models 170
pressure specifying 338

specifying data 128
with multiple scenarios 170

Method of Characteristics 43, 358,
359, 458
reliability with cavitation 410
Keyboard modifiers 461
Miller Turbulent Method 188, 325
CTRL 73, 74, 121, 126
minor losses See Losses
SHIFT 73, 74, 77, 80, 121, 125
MIT Equation 188, 324

Mixtures 4, 29, 141, 142, 353
Model Data Control window 88
Laminar flow 145, 323
database connections 89
Legend 114, 118
showing junction data 88
Liquid Accumulator junction 340
showing pipe data 88
Elasticity 231
showing selected pipes and junctions

89
Initial Volume 231
2/24/2003
1:35 PM
Page 497

Additional Pipe Losses 329
using saved format files in Database

Manager 302
Bingham Plastic 143, 328
Model Data window 8, 41, 71, 86
Brecht & Heller method 327
Additional Losses 87
Duffy method 326
customizing 88
junction losses 329
opening pipe and junction

specifications windows from 128,
179
Power Law 143, 328

pulp and paper 326, 327
NPSH 140, 234, 237
Pipe Detail Summary 87

Pipe Loss Data table 87
constant for variable speed 234
Pipe table 87
NPSHR 237
print format 174
NPSP 140
transient junction data 422
NPSPR 237
NPSP See NPSH

430
Model Status Light 132
modulus of elasticity 187
Object status 80, 130, 146, 194, 208
Momentum equation
One-dimensional flow 59
steady-state 312
Open Pipe/Jct Window 86, 128, 179
transient 359
Orifice
Moody friction factor 323
loss factors 345
morph 124, 463
round edged 345

sharp edged 345
N
Network databases 201. See Database
Manager
Newton-Raphson method 3, 312, 314,
321
transient solution for gas
accumulators 400
1:35 PM
Page 498
Format and Action 95
General Output Control 94
Pipe and Junction parameters 92, 94
non-Newtonian 143, 144, 322, 326
2/24/2003
Output Control window 27, 50, 92,

134, 136, 287, 321
scenarios 160
Index 499
Show Selected Pipes/Junctions 96
Manager 302

Manager 302
Paste 79, 80, 122, 124, 126
PDCV See Pressure Drop Control
Valve
Output file 47, 154

transient data 91
Output window 8, 48, 71, 90
Piezometric head 311

Pipe Detail Summary
customizing 287
Event messages 420

Pipe Drawing tool 17, 22, 74, 121, 462
exporting output 90
number formatting 95
Pipe flow direction 210

Pipe Loss Data table
print format 174
setting defaults 97
updates 104
Overview of AFT Impulse 7
Pipe Material Database window 185,

294
Pipe Results table 90
pipe sectioning 43, 365, 366. See
Section Pipes
P
Parallel Pipes
effect on run time 457
in Pipe Specifications window 192

Parameter and Unit Preferences window
271
Pipe Specifications window 38, 39,

180
Additional Losses 189
connected junctions 184
default roughness model 188
Design Alerts 191
Design Factors 272
displaying name 192
elevation 273
Fluid Properties 194
setting defaults 276
friction model 187
setting preferred units 275
ID Reduction - Scaling 186
unit preferences 273
initial flow 191
Use Most Recent Pipe Size Data 272
jumping to next higher or lower

number pipe 129
2/24/2003
1:35 PM
Page 499

material 185
displaying data in Output window

96
Notes 194
displaying in Model Data 88
optional input 191
displaying in Visual Report 115
Parallel Pipes 192
displaying on Workspace 83
pipe diameter 186
displaying transient data in Output

window 99
Pipe Length 185

Pipe Line Size and Color 192
editing 125
pipe modulus of elasticity 187
finding 163
Pipe Name 183

Pipe Number 183
flow direction 60, 122, 124, 164,

184, 197
pipe Poisson ratio 187
Global Editing 267
pipe support 189
handles 75, 125
pipe wall thickness 186
ID numbers 122
roughness model 187
length on Workspace 122, 185
Same as previous pipe 184
loss factor 62
Same Specs as Pipe 184
losses 60, 61, 62, 323, 324, 340
Status 194
number of connections 122
wavespeed 189
Reverse direction 164
pipe stations See computing stations
segmenting 75
Pipe Support 189
specifying data 128
Pipe table
Splitting 124, 463

Pipes 22, 59
stretching 125
Point loss 62
connectivity 126
Poisson Ratio 187
creating 121
Popup menu 80
defining 128
positive flow direction 23
Design factors 192
Power Law See non-Newtonian

Pre-mixtures 142
2/24/2003
1:35 PM
Page 500
Index 501
Pressure
Graphing pump data 248
defining system pressure 155, 442
in output file 155
Pressure Drop Control Valve 220, 450
Known speed transients 238, 239
Pressure Reducing Valve 220, 436,

453
NPSHR 237
static and stagnation pressure

discussion 318
Pressure relaxation See Solution
Control window
Pressure Sustaining Valve 220, 436,
453, 455
static and stagnation pressure
discussion 318
problems from poor curve fits 431

pump configurations 234
Pump Startup With Inertia 243
Pump trip with inertia four quadrant
241
Pump trip with inertia and no
backflow 240, 241
sizing with flow control valves 448
Primary windows 7, 71
Special Conditions 246, 248
Print Format 88, 90, 174
Print Preview 170
Submerged pump 233
Print Special 170
Transient Data 238
PRV See Pressure Reducing Valve
variable speed 232, 234, 352
PSV See Pressure Sustaining Valve
viscosity corrections 247
pulp and paper stock See nonNewtonian

pump speed 379
Pump junction 339, 340, 448
Pumps 140, 352
as assigned flow 246
efficiency 234, 238
check valve 234
power usage 234, 238
controlled discharge pressure 234,

352
Pump configurations 234
controlled flow 234, 352

Controller transient 244
efficiency and power usage 238
Estimating the pump inertia 243
2/24/2003
1:35 PM
Page 501
sizing method 440

sizing with flow control valves 448
submerged 381
viscosity corrections 247

Renumbering See Workspace window.
See Workspace window
R
README.TXT 5
Repeat transient 420
Reference Information 94, 174
Reservoir junction 17, 60, 338
Reference positive flow direction
Difference between a Reservoir and

Surge Tank 260
See positive flow direction

Reference Pressure 443
How pressure junctions work 439
References 481
loss factors 62
Relative location 18
Specifications window 31, 36, 250
Relative roughness 62, 186, 188, 323
the role of pressure junctions 442
Relative tolerance See Steady Solution

Control window
Transient Data 253
Relaxation See Steady Solution

Control window
use in open and closed systems 318

resistance 363, 367

434
variable during transient 63
Relief Valve junction 21, 338, 340
Resistance Curve 230
entering transient data 249
Reverse Direction 164
Exit 394
Reynolds number
effects during transient flow 153
Internal 394
transition 145, 324
Internal vs. exit relief valves 249

loss factors 62, 217
Scale/flip workspace See Workspace

window
surge reduction 426
Scaling See ID Reduction - Scaling
Transient Data 250

Transient valve cracking 249
Renumber Wizard 85
2/24/2003
1:35 PM
Scenario Manager 156

re-establishing broken links 161
Scenarios
merging models with multiple
scenarios 170
Page 502
Index 503
same as parent junction 201
same as parent pipe 184
Section Pipes window 43, 135, 147,
152, 176, 366, 457, 458
Recommendations on Steady Solver

176
Steady-State 134, 175, 311, 329
Transient 176, 358
Select Flow Path 80, 125
Transient Solver control 151
Select Graph Data window 105
verification of steady-state 321
Select Special 80, 97, 164
working with the steady and transient

solvers 175
Selection Drawing tool 16, 73, 80, 125,

461
selecting objects completely or
partially in box 278
sequential dual events 417. See Event
Transients
Setup program 6
Sort output feature 90

434
Sparger 254
Special Conditions 165, 436
Ignore 169
SHIFT key See Keyboard modifiers.

See Keyboard modifiers. See
Keyboard modifiers
Show list box
in Output window 90
Show Object Status 146
single event transients 414
Solution Balance Summary table 321
Solution Progress window 46, 90, 173,
336
iteration history 336
View Results button 90
junctions 206
No reflections - infinite pipe 169
Pipe/Junction Symbol on Workspace
281
Transient 169
Workspace appearance 278
Specifications windows 86, 179
accessing 179
inspecting in 131
Jump feature 180
Splitting pipes 124, 463
Spray Discharge junction 338
Solver
discussion of Steady-State and
Transient 46
Graphing Spray Discharge data 255
output file for transient data 47, 91
2/24/2003
1:35 PM
Page 503
in output file 155

Transient Data 255
Annual Support 479
e-mail 480
Spreadsheet
surge flow
exporting data to 90
See waterhammer
Stagnation pressure
surge suppression 425
at pressure control valves 318
Surge Tank junction 20, 340
explanation 315
as a branch during steady-state 259
Static pressure
as a finite size reservoir during

steady-state 260
at pressure control valves 318
as a pressurizer during steady-state

259
explanation 315
Status Bar 132
Steady Solution Control window 134,
135
absolute tolerance 331
convergence 330
false convergence 335
flow rate relaxation 334
Matrix Method 336
maximum iterations 336
parameters 329
pressure relaxation 335
relative tolerance 331
relaxation 333
relaxation example 334
tolerance 330
tolerance example 332
Submerged pumps 381
Support 479
2/24/2003
1:35 PM
connector pipe 257

Difference between a Reservoir and
Surge Tank 260
flow restrictor 257
Graphing Surge Tank data 261
in output file 155
lumped inertia 407
Model as dipping tube vessel 260
Only Allow Outflow 258
orifice 257
surge reduction 427
Tank area 256
Tank geometry 256
Tank height 256
Transient Data 258
Page 504
Index 505
balance verification 321
surge transients
See waterhammer
System pressure 155, 442
System Properties window 28, 134,
138
AFT Standard fluid database 141
atmospheric pressure 144
Chempak database 141
creating mixtures 141

435
Toolbars 119
Toolbox 72, 79, 121, 124
customizing See Toolbox
Preferences window
Toolbox Preferences window 284
Current Toolbox/Unused Tools 285
fluid list 293

gravitational acceleration 145
scenarios 160
Transition Reynolds number 145

Manager 302
ToolTips 79
Transfer Results to Initial Guess 90,

172
Technical Support See Support
Transient Control window 45, 135,

151, 176, 229, 244, 248, 255, 261,
266, 413, 457
Tee/Wye junction 198, 340

convergence problems and
recommendations 434
artificial transients 459
loss factors 62, 262, 346
Estimated file size and run time 155
Simple, no loss 262
Include in transient output file 154
Model transient cavitation 153
Save Output to File Every Time Step

152
Terminology 1
time step 365, 458, 459
Time-based transients 413
Tip of the Day 286
scenarios 160
Start and stop time 151
Tolerance
1:35 PM
Saving pipe output 154

specifying content for output file 91
Title 94, 114
2/24/2003
Saving junction output 155
Page 505

Stop run if artificial transient detected
153
variable resistance 64, 152

surge reduction 427
Valve coefficient 350
Transient Data
Valve junction 21, 60, 338, 340
Absolute vs. relative 419
Exit valve 263, 338
Repeat transient 420
loss factors 62, 343, 345
Transient output file See Output file

Transient special conditions See
Special Conditions

Transient vapor cavitation See vapor

cavitation
Transient Data 264

vapor cavitation 46, 150
Transition flow 145, 324
transient 364, 409
Transition Reynolds number 145
vapor cavitation waterhammer theory
Transport properties 353

Troubleshooting steady-state models
430
assigned flow 369
Turbulent flow 145, 324
branch 371
assigned pressure 368

check valve 395
dead end 370
Undo 79, 126

Unit Preferences See Parameter and
Unit Preference window
gas accumulator 401

infinite pipe 376
liquid accumulator 393
Unspecified fluid 28
pump 381
Relief Valve 395
Vacuum Breaker Valve junction 340
spray discharge 378
Graphing Vacuum Breaker Valve

data 266
surge tank 398

valve 374
in output file 155
Vapor pressure 140
Variable fluid properties 139, 194
2/24/2003
1:35 PM
Page 506
Index 507
variable pipe resistance 152
Verification
Verification against published results
7
Verifying steady-state network
solutions 321
View Results button See Solution
Progress window
Viscosity corrections 247
Visual Report Control window 112,
288
scenarios 160
Water Data from ASME Steam Tables

29
waterhammer 1, 64
description 64
instantaneous 65, 355
Waterhammer Assistant 162
wavespeed 39, 43, 64, 66, 147, 148,
189, 356, 358, 365, 366, 409
assumption that it remains constant
5
Calculated 185
User Specified 186

Manager 302
Visual Report window 8, 55, 71, 111,
288
Web site 480

Workspace
grid 279
Reverse Direction 23
Color Map 115

Display Parameters 113
General Display 113
input-only mode 115
legend 114, 118
Workspace Preferences window 125,

146, 166, 169, 206, 276
Action When Dragging Junctions
279
Allowable Workspace Label
Movements 281
Save Options 118

Show Selected Pipes and Junctions
115
Visual Report Control 112
Volume Balance junction 340
colors and fonts 282

display options 279
displaying name and/or ID numbers
281
icon size 278
pipe line options 277
Walk Through Example 15
Sample Workspace 283
wall thickness 186
2/24/2003
1:35 PM
Page 507

source/sink indicator 280
Zoom feature 83, 111
Special Condition Pipe/Junction

Symbol 281
Zoom Select tool 17, 77, 463
special conditions 278

Transient indicators on the
Workspace 421
Manager 302
Workspace window 8, 16, 71, 72
changing junction icons 207
colors and fonts 282
customizing 82, 276
Displaying name and/or ID numbers
281
editing 79
finding pipes and junctions 163
flipping 84
pipe lengths on 185
relative locations 18
renumbering 84
scaling 84
setting the junction label position
See Workspace Preferences
window
size 83
Special Condition Pipe/Junction
Symbol 281
zoom feature 83
2/24/2003
1:35 PM
Page 508

Impulse 3.0 Users Guide

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Impulse 3.0 Users Guide

Uploaded by

Copyright:

Available Formats

AFT Impulse

AFT Impulse version 3.0

Applied Flow Technology

c:\aft manuals\i3\impulse3 manual

2003 Applied Flow Technology Corporation. All rights reserved.

c:\aft manuals\i3\impulse3 manual

c:\aft manuals\i3\impulse3 manual

iv AFT Impulse 3.0 Users Guide

11. Reducing Surge Pressures ..................................... 425

c:\aft manuals\i3\impulse3 manual

c:\aft manuals\i3\impulse3 manual 1b.doc

vi AFT Impulse 3.0 Users Guide

2. A Walk Through AFT Impulse...................................... 15

c:\aft manuals\i3\impulse3 manual

Table of Contents vii

c:\aft manuals\i3\impulse3 manual 1b.doc

viii AFT Impulse 3.0 Users Guide

4. The Five Primary Windows.......................................... 71

c:\aft manuals\i3\impulse3 manual

c:\aft manuals\i3\impulse3 manual 1b.doc

x AFT Impulse 3.0 Users Guide

5. Building and Running Models................................... 121

c:\aft manuals\i3\impulse3 manual

c:\aft manuals\i3\impulse3 manual 1b.doc

xii AFT Impulse 3.0 Users Guide

c:\aft manuals\i3\impulse3 manual

Table of Contents xiii

6. Pipe and Junction Specifications Windows ............ 179

c:\aft manuals\i3\impulse3 manual 1b.doc

xiv AFT Impulse 3.0 Users Guide

c:\aft manuals\i3\impulse3 manual

c:\aft manuals\i3\impulse3 manual 1b.doc

xvi AFT Impulse 3.0 Users Guide

c:\aft manuals\i3\impulse3 manual

Table of Contents xvii

c:\aft manuals\i3\impulse3 manual 1b.doc

xviii AFT Impulse 3.0 Users Guide

7. Customizing AFT Impulse ......................................... 271

c:\aft manuals\i3\impulse3 manual

Table of Contents xix

8. Steady-State Hydraulic Theory and Loss Models.... 311

c:\aft manuals\i3\impulse3 manual 1b.doc

xx AFT Impulse 3.0 Users Guide

c:\aft manuals\i3\impulse3 manual

Table of Contents xxi

9. Theory of Waterhammer & Solution Methodology .. 355

c:\aft manuals\i3\impulse3 manual 1b.doc

xxii AFT Impulse 3.0 Users Guide

c:\aft manuals\i3\impulse3 manual

Table of Contents xxiii

c:\aft manuals\i3\impulse3 manual 1b.doc

xxiv AFT Impulse 3.0 Users Guide

10. Modeling Time and Event Based Transients ......... 413

11. Reducing Surge Pressures ..................................... 425

c:\aft manuals\i3\impulse3 manual

Table of Contents xxv

12. Special Topics & Troubleshooting Models ............ 429

c:\aft manuals\i3\impulse3 manual 1b.doc

xxvi AFT Impulse 3.0 Users Guide

Appendix A. Keyboard Modifiers and Shortcuts ......... 461

Appendix B. Limitations ................................................ 467

c:\aft manuals\i3\impulse3 manual

Table of Contents xxvii

Appendix D. Chempak Technical Information ............. 477