impulse 4.0

8/10/2007 4:10:28 PM c:\aft manuals\i4\impulse4 manual (si) g.doc

AFT ImpulseTM

User’s Guide

AFT Impulse version 4.0

Waterhammer Modeling in Piping Systems

Applied Flow Technology

8/10/2007 4:10:28 PM c:\aft manuals\i4\impulse4 manual (si) g.doc

CAUTION!

AFT Impulse is a sophisticated waterhammer and surge transient modeling program designed for qualified engineers with experience in waterhammer analysis and should not be used by untrained individuals. AFT Impulse is intended solely as an aide for pipe flow analysis engineers and not as a replacement for other design and analysis methods, including hand calculations and sound engineering judgment. All data generated by AFT Impulse should be independently verified with other engineering methods.

AFT Impulse is designed to be used only by persons who possess a level of knowledge consistent with that obtained in an undergraduate engineering course in the analysis of pipe system fluid mechanics and is familiar with standard industry practice in waterhammer analysis.

AFT Impulse is intended to be used only within the boundaries of its engineering assumptions. The user should consult the manual for a discussion of all engineering assumptions made by AFT Impulse.

Information in this document is subject to change without notice. No part of this User’s Guide may be reproduced or transmitted in any form or by any means, electronic or mechanical, for any purpose, without the express written permission of Applied Flow Technology.

© 2007 Applied Flow Technology Corporation. All rights reserved.

Printed in the United States of America.

"AFT Impulse", "AFT Fathom", "AFT Mercury", "Applied Flow Technology", and the AFT logo are trademarks of Applied Flow Technology Corporation

Chempak is a trademark of Madison Technical Software, Inc.

Microsoft, "Visual Basic", Excel and Windows are trademarks or registered trademarks of Microsoft Corporation. “CAESAR II” is a registered trademark of COADE, Inc.

8/10/2007 4:10:28 PM Page iii c:\aft manuals\i4\impulse4 manual (si) g.doc

Contents

Summary

Summary..................................................................................................iii Detailed Contents ..................................................................................... v Nomenclature....................................................................................... xxix

1. Introduction .................................................................... 1

2. A Walk Through AFT Impulse...................................... 13

3. Fundamental Concepts................................................ 59

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

5. Building and Running Models................................... 125

6. Pipe and Junction Specifications Windows ............ 193

7. Customizing AFT Impulse ......................................... 287

8. Steady-State Hydraulic Theory and Loss Models.... 325

9. Theory of Waterhammer & Solution Methodology .. 367

10. Modeling Time and Event Based Transients ......... 429

iv AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page iv c:\aft manuals\i4\impulse4 manual (si) g.doc

11. Reducing Surge Pressures ..................................... 441

12. Special Topics: Troubleshooting Models .............. 445

Appendix A. Keyboard Modifiers and Shortcuts ......... 477

Appendix B. Limitations ................................................ 483

Appendix C. Installation Issues .................................... 485

Appendix D. Chempak Technical Information ............. 491

Appendix E. Obtaining Technical Support................... 493

References...................................................................... 495

Glossary.......................................................................... 497

Table of Contents v

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

c:\aft manuals\i4\impulse4 manual (si) g.doc

Detailed Contents

Summary..................................................................................................iii Detailed Contents ..................................................................................... v Nomenclature....................................................................................... xxix

1. Introduction .................................................................... 1

Terminology ............................................................................................. 1 Modeling capabilities ............................................................................... 2 Interface features ...................................................................................... 3 Thermophysical property data .................................................................. 4 Who can use AFT Impulse ....................................................................... 4 Engineering assumptions in AFT Impulse ............................................... 4 Installing AFT Impulse............................................................................. 5

Check hardware and system requirements ......................................... 5 Read the README.TXT file............................................................. 5 Run the setup program ....................................................................... 6

Getting started with AFT Impulse ............................................................ 6 Example models ................................................................................. 7

Using online help...................................................................................... 7 Verification models .................................................................................. 7 AFT Impulse overview ............................................................................. 7

Input windows.................................................................................... 8 Output windows ................................................................................. 8

Converting models and databases from previous versions....................... 9 Pipe friction........................................................................................ 9 Infinite pipe junctions ........................................................................ 9 Transient junction data....................................................................... 9 Junction loss models .......................................................................... 9 Reservoir transients............................................................................ 9 Databases ......................................................................................... 10

vi AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page vi c:\aft manuals\i4\impulse4 manual (si) g.doc

Converting models and databases from AFT Fathom or Mercury ......... 10 Combining pipes .............................................................................. 10 Junctions with polynomial loss data or interpolated x-y data.......... 10 Databases ......................................................................................... 10

What’s new in version 4.0 ...................................................................... 11 General ............................................................................................. 11 Pipes ................................................................................................. 11 Junctions........................................................................................... 11 Transient Control ............................................................................. 11 Solver ............................................................................................... 11 Workspace........................................................................................ 12 Model Data....................................................................................... 12 Output............................................................................................... 12 Graph Results ................................................................................... 12

Cavitation pressure changes ...................................................... 12

2. A Walk Through AFT Impulse...................................... 13

Step 1. Start AFT Impulse ...................................................................... 13 The Workspace window ............................................................ 14

Step 2. Lay out the model ....................................................................... 15 A. Place the first reservoir ............................................................... 15

Objects and ID numbers ............................................................ 15 Editing on the Workspace.......................................................... 16

B. Place the first branch ................................................................... 16 C. Place surge tank........................................................................... 18 D. Place the second reservoir ........................................................... 18 E. Add the remaining junctions........................................................ 19 F. Draw a pipe between J1 and J2.................................................... 20

Reference positive flow direction.............................................. 21 G. Add the remaining pipes ............................................................. 22

Step 3. Set the unit preferences .............................................................. 23 Step 4. Complete the first three checklist requirements ......................... 24

A. Specify Steady Solution Control ................................................. 25

Table of Contents vii

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


B. Specify output control ................................................................. 25 C. Specify system properties............................................................ 26

Step 5. Define the model components (checklist item #4) ..................... 28 Object status .............................................................................. 28 Undefined Objects window ....................................................... 29

A. Define Reservoir J1..................................................................... 29 The Inspection feature ............................................................... 31

B. Define Branch J2 ......................................................................... 32 C. Define Surge Tank J3 .................................................................. 33 D. Define Reservoir J4 and Branches J5 and J6 .............................. 34 E. Define Valve J7 ........................................................................... 34 F. Define Valve J8 ........................................................................... 35 G. Define Pipe P1............................................................................. 37

The Pipe Specifications window ............................................... 39 H. Define Pipes P2 through P7 ........................................................ 40

Reviewing input in Model Data window................................... 40 Step 6. Complete the last two checklist requirements ............................ 41

A. Specify pipe sectioning ............................................................... 41 B. Specify transient control.............................................................. 44

Step 7. Run the Solver ............................................................................ 46 The two solvers ................................................................................ 46 The transient output file ................................................................... 47

Step 8. Review the output....................................................................... 48 A. Modify the output format ............................................................ 51 B. Graph the results.......................................................................... 52 C. Animate the Results..................................................................... 53 D. View the Visual Report............................................................... 55

Conclusion .............................................................................................. 57

3. Fundamental Concepts................................................ 59

Pipes and junctions ................................................................................. 59 Convention for flow entering and exiting............................................... 60 Features for modeling irrecoverable losses ............................................ 60

viii AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page viii c:\aft manuals\i4\impulse4 manual (si) g.doc

Incorporating equivalent length data ......................................... 61 Convention for specifying junction base area.................................. 61 Specifying losses.............................................................................. 62 Modeling valve closures with K factors .......................................... 62 Specifying frictional losses in pipes ................................................ 62 Comment on importance of hydraulic losses during transients ....... 63

Static vs. dynamic losses ........................................................... 63 Variable resistance .................................................................... 63

Introduction to wavespeed and waterhammer phenomenon................... 64 Description of waterhammer............................................................ 64 Instantaneous waterhammer............................................................. 65 Wavespeed ....................................................................................... 66 Communication time in pipes .......................................................... 66 Conceptual example ......................................................................... 66

Phase A: 0 < t < L/a................................................................... 67 Phase B: L/a < t < 2L/a.............................................................. 67 Phase C: 2L/a < t < 3L/a............................................................ 69 Phase D: 3L/a < t < 4L/a............................................................ 69

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

Overview................................................................................................. 71 The Workspace window ......................................................................... 72

The Shortcut Button ......................................................................... 73 The Selection Drawing Tool ............................................................ 73

Selecting objects completely or partially inside the box........... 73 The Pipe Drawing Tool.................................................................... 74

Pipe handles and segmenting a pipe .......................................... 76 The Zoom Select Tool...................................................................... 77 The Annotation Tool........................................................................ 78 Junction icons................................................................................... 79 Editing features ................................................................................ 80

Selecting groups of objects........................................................ 80 Selecting objects in a flow path................................................. 81

Table of Contents ix

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


Aligning objects......................................................................... 81 Creating and Managing Saved Groups ............................................ 81 Special group usage.......................................................................... 82 Bookmarks on the Workspace.......................................................... 82 Last View ......................................................................................... 82 Panning the Workspace.................................................................... 82 Customizing features........................................................................ 83

Workspace Preferences ............................................................. 83 Adjusting the Workspace size ................................................... 83 Selective display of pipes and junctions.................................... 83 Scale/flip workspace.................................................................. 84 Renumbering the Workspace objects ........................................ 85 Using the Renumber Wizard ..................................................... 86 Incrementing Object Numbers................................................... 86

Background Graphic ........................................................................ 86 Specifications windows.................................................................... 87 Printing and exporting the Workspace............................................. 88

The Model Data window ........................................................................ 88 Display and printing features ........................................................... 90 The Model Data Control window .................................................... 90

Pipe data display........................................................................ 90 Junction data display ................................................................. 90 Show selected pipes and junctions ............................................ 91 Scenario Format......................................................................... 91 Database connections ................................................................ 92

The Output window................................................................................ 93 Transient output files ....................................................................... 94 The Output Control window ............................................................ 95

Pipe and Junction parameters .................................................... 95 Pipe Transient parameters ......................................................... 97 General output ........................................................................... 97 Summaries ................................................................................. 97 Format and action ...................................................................... 98

x AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page x c:\aft manuals\i4\impulse4 manual (si) g.doc

Show selected pipes/junctions................................................... 99 Database connections .............................................................. 100 Command buttons.................................................................... 100 Output parameter descriptions................................................. 101

Steady flow results ......................................................................... 101 Transient flow results for pipes ..................................................... 101

Transient Max/Min results for pipes ....................................... 102 Summary Max/Min results ...................................................... 103 Transient event messages ........................................................ 104

Output window updates ................................................................. 104 The Graph Results window .................................................................. 106

The Select Graph Data window ..................................................... 107 Transient pipe results............................................................... 107 Transient junction results ........................................................ 108 Profile Along a Flow Path and EGL, HGL and Elevation Profile… .................................................................................. 108 Animating profile results ......................................................... 109 Transient force results ............................................................. 112 Other graph controls ................................................................ 112

The Customize Graph window....................................................... 112 The Auxiliary Graph Formatting window...................................... 114 Printing and exporting the Graph Results image and data............. 114

The Visual Report window................................................................... 114 The Visual Report Control window............................................... 116

Display parameters .................................................................. 117 General display........................................................................ 117 Show selected pipes and junctions .......................................... 118 Color map ................................................................................ 119 Command buttons.................................................................... 121

Visual Report annotations.............................................................. 122 Annotations for the Workspace ............................................... 122 Creating Visual Report annotations ........................................ 122

The Toolbars......................................................................................... 122

Table of Contents xi

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


5. Building and Running Models................................... 125

Creating objects .................................................................................... 125 Pipes ............................................................................................... 125 Junctions......................................................................................... 128

Morphing junctions ................................................................. 128 Splitting pipes .......................................................................... 128

Annotations .................................................................................... 129 Moving objects ..................................................................................... 129

Keyboard modifiers........................................................................ 130 Editing objects ...................................................................................... 130 Connecting objects ............................................................................... 130 Defining objects.................................................................................... 132

Specifying required property data.................................................. 132 Highlighting required information ................................................. 133 Use status feature ........................................................................... 133 Undefined objects window............................................................. 133 Satisfying connectivity requirements............................................. 134

Inspecting objects ................................................................................. 134 Using the Checklist............................................................................... 136

Analysis Type................................................................................. 138 Specify Steady Solution Control.................................................... 138 Specify Output Control .................................................................. 138 Specify System Properties.............................................................. 138 Define All Pipes and Junctions ...................................................... 139 Section Pipes .................................................................................. 139 Transient Control ........................................................................... 139

Steady Solution Control........................................................................ 139 Output Control ...................................................................................... 140 System Properties ................................................................................. 142

System fluid property variation...................................................... 143 Density and dynamic viscosity....................................................... 144 Bulk modulus of elasticity ............................................................. 144 Vapor pressure and cavitation........................................................ 144

xii AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xii c:\aft manuals\i4\impulse4 manual (si) g.doc

The fluid databases ........................................................................ 145 The AFT Standard Database ................................................... 145 The ASME Steam Tables Database ........................................ 145 The Chempak database............................................................ 146 Accuracy option....................................................................... 147 The ASME Steam Tables Database ........................................ 147

Viscosity models ............................................................................ 148 Non-Newtonian flow in non-pipe elements............................. 148

Atmospheric pressure..................................................................... 149 Gravitational acceleration .............................................................. 149 Transition Reynolds Numbers ....................................................... 150 Editing the AFT Standard fluid database....................................... 150

Finding object definition status ............................................................ 151 List Undefined Objects .................................................................. 151

Section Pipes......................................................................................... 152 Numbering convention................................................................... 154

Transient Control.................................................................................. 157 Start and stop time.......................................................................... 157 Save output to file… ...................................................................... 158 Use variable pipe resistance........................................................... 158 Model transient cavitation.............................................................. 159 Stop run if artificial transient detected........................................... 159 Including data in transient output file ............................................ 160

Saving pipe output ................................................................... 160 Disadvantages to saving all stations ........................................ 160 Advantages to saving all stations............................................. 161

Saving junction output ................................................................... 161 Defining force sets ......................................................................... 161 Estimated file size and run time..................................................... 162

Calculating and outputting unbalanced forces ..................................... 163 Including friction and momentum.................................................. 163

Area changes............................................................................ 165 Viewing force data ......................................................................... 165

Table of Contents xiii

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


Exporting transient force results .................................................... 165 Defining pressures for the system ........................................................ 166 Using scenario manager........................................................................ 166

The Scenario Manager window ..................................................... 167 Creating, organizing and editing scenarios .................................... 167 Viewing scenario differences......................................................... 169 Modifying individual scenarios ..................................................... 170 Passing changes to child scenarios................................................. 171

Scenario logic examples .......................................................... 172 Re-establishing broken links ................................................... 176

Fast Scenario Changes ................................................................... 176 Special modeling features..................................................................... 176

Waterhammer Assistant ................................................................. 176 Workspace Find ............................................................................. 177 Reverse Direction........................................................................... 178 Select Special ................................................................................. 178 Special Conditions ......................................................................... 180

Special conditions with no transient data ................................ 181 Special conditions with transient data ..................................... 181 Pump special conditions .......................................................... 182 Special conditions only applied to junctions........................... 183

Transient special conditions........................................................... 183 No reflections – infinite pipe................................................... 183

Special Condition Ignore................................................................ 184 Merging models ............................................................................. 184

Merging models with multiple scenarios................................. 184 Print Preview/Special ..................................................................... 185 Transfer Results to Initial Guesses ................................................ 186 Batch runs ...................................................................................... 187 Print Content .................................................................................. 188 Extended Model Check.................................................................. 189 Math calculator .............................................................................. 189

Working with the steady and transient solvers..................................... 189

xiv AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xiv c:\aft manuals\i4\impulse4 manual (si) g.doc

Steady-state solver ......................................................................... 189 Recommendations on Steady Solver ....................................... 190

Transient solver.............................................................................. 190 Artificial transients .................................................................. 191

6. Pipe and Junction Specifications Windows ............ 193

Highlight feature ............................................................................ 194 Jump feature ................................................................................... 194

Pipe Specifications window ................................................................. 194 Common input parameters ............................................................. 197

Pipe number ............................................................................. 197 Pipe name ................................................................................ 197 Copy Data From Pipe .............................................................. 198 Scenario same as parent........................................................... 198 Connected junctions ................................................................ 198

Pipe Model ..................................................................................... 198 Pipe length ............................................................................... 199 Pipe material ............................................................................ 199 Pipe diameter ........................................................................... 200 Pipe wall thickness .................................................................. 200 Pipe modulus of elasticity ....................................................... 201 Pipe Poisson ratio .................................................................... 201 Friction Model ......................................................................... 201 Pipe Support ............................................................................ 203 Pipe Wavespeed....................................................................... 203

Fittings & Losses ........................................................................... 203 Design Alerts.................................................................................. 205 Optional input ................................................................................ 205

Initial Flow Rate Guess ........................................................... 205 Workspace display................................................................... 205 Design Factor........................................................................... 206 Pipe Line Size and Color ......................................................... 206 Parallel Pipes ........................................................................... 206

Table of Contents xv

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


Support for Partially Full Pipe................................................. 206 Intermediate Pipe Elevations ................................................... 207

Fluid Properties .............................................................................. 208 Notes .............................................................................................. 208 Status .............................................................................................. 208

Junction Specifications windows ......................................................... 208 Format #1: Junctions with one or two connecting pipes............... 210 Format #2: Junctions with more than two connecting pipes.......... 211

Parameters common to all junction Specifications windows ............... 212 Junction Number ............................................................................ 212 Junction Name................................................................................ 212 Junction Elevation.......................................................................... 212 Database List.................................................................................. 214 Copy Data From Jct list ................................................................. 214

Scenario same as parent........................................................... 214 Pipe connectivity............................................................................ 214 Jumping to another junction........................................................... 214 Transient Data ................................................................................ 215

Repeat Transient...................................................................... 216 Transient Special Conditions................................................... 216 Time based and event based transients.................................... 217

Initial Pressure/Head Guess ........................................................... 217 Displaying junction names and numbers ....................................... 219 Special Conditions ......................................................................... 219 Design Factor ................................................................................. 219 Changing the icon graphic ............................................................. 220 Changing the icon size ................................................................... 220 Specifying Base Area..................................................................... 220 Notes .............................................................................................. 220 Status .............................................................................................. 220

Area Change Specifications window.................................................... 221 Assigned Flow Specifications window................................................. 222

Transient Data ................................................................................ 224

xvi AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xvi c:\aft manuals\i4\impulse4 manual (si) g.doc

Special Conditions ......................................................................... 224 Special features for steady flow..................................................... 224

Assigned Pressure Specifications window ........................................... 225 Transient Data ................................................................................ 227

Branch Specifications window ............................................................. 227 Transient Data ................................................................................ 229

Check Valve Specifications window.................................................... 229 Specifying losses............................................................................ 230 Forward Velocity to Close Valve................................................... 231 Delta Pressure to Re-Open............................................................. 231 Transient Data ................................................................................ 231

Control Valve Specifications window.................................................. 231 Control Valve types ....................................................................... 232

PRV/PSV static vs. stagnation pressure .................................. 233 Action if setpoint not achievable ................................................... 233 Special Conditions ......................................................................... 234 Open Percentage Table .................................................................. 235 Transient Data ................................................................................ 235

Failure Transient...................................................................... 235 Control Transient..................................................................... 236

Dead End Specifications window......................................................... 236 Gas Accumulator Specifications window............................................. 237

Initial conditions and polytropic process ....................................... 237 Maximum and minimum volumes.................................................. 239 Interface with pipe system ............................................................. 239 Special features for steady flow..................................................... 240 Graphing Gas Accumulator data.................................................... 241

General Component Specifications window ........................................ 241 Liquid Accumulator Specifications window ........................................ 243

Initial conditions and elastic process ............................................. 244 Pump Specifications window ............................................................... 244

Pump Model ................................................................................... 244 Submerged pump ..................................................................... 246

Table of Contents xvii

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


Do not allow backwards rotation............................................. 246 Check valve at pump discharge ............................................... 246

Pump configurations ...................................................................... 246 Controlled pressure or flow (variable speed)................................. 248 NPSHR and NPSPR (optional in pump configuration) ................. 249 Efficiency and power usage (optional in pump configuration)...... 249 Transient Data ................................................................................ 249

Pumps with curves................................................................... 250 No transient from pump, no back flow.................................... 250 No transient from pump, but back flow possible..................... 250 Known speed transients, no back flow.................................... 250 Pump startup transient with possible reverse flow.................. 251 Pump transients with inertia .................................................... 251 Trip with inertia and no back flow or reverse speed ............... 252 Startup with inertia and no back flow or reverse speed .......... 252 Trip with inertia – four quadrant ............................................. 253 Startup with inertia - four quadrant, known motor torque/speed…......................................................................... 254 Estimating the pump inertia..................................................... 254 Event modeling – including trips followed by restarts or vice versa......................................................................................... 254 Controller transient.................................................................. 255 Pumps as fixed flow ................................................................ 255

Special Conditions ......................................................................... 257 Representing Multiple Pumps With One Junction......................... 258 Viscosity corrections...................................................................... 258 Graphing pump data ....................................................................... 258

Relief Valve Specifications window .................................................... 259 Three relief valve configurations: internal, exit and inline............ 259 Specifying losses............................................................................ 260 Transient valve cracking ................................................................ 260 Transients with variable Cv ........................................................... 260

Reservoir Specifications window ......................................................... 261 Transient Data ................................................................................ 263

xviii AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xviii c:\aft manuals\i4\impulse4 manual (si) g.doc

Spray Discharge Specifications window .............................................. 263 Loss Model..................................................................................... 265 Transient Data ................................................................................ 265 Graphing Spray Discharge data ..................................................... 265

Surge Tank Specifications window ...................................................... 265 Tank geometry................................................................................ 266 Interface with pipe system ............................................................. 267 One-Way Surge Tank..................................................................... 268 Transient Data ................................................................................ 268 Special features for steady flow..................................................... 268

Surge tank as a branch during steady-state.............................. 269 Surge tank as a pressurizer during steady-state ....................... 270 Surge Tank as a finite size reservoir during steady-state ........ 270 Difference between a Reservoir and Surge Tank .................... 270

Model as dipping tube vessel ......................................................... 270 Graphing Surge Tank data ............................................................. 271

Tee/Wye Specifications window.......................................................... 271 Loss factors .................................................................................... 272

Valve Specifications window ............................................................... 273 Transient Data ................................................................................ 274 Special Conditions ......................................................................... 274

Vacuum Breaker Valve Specifications window................................... 274 Three-Stage Valve Modeling ......................................................... 276 Graphing Vacuum Breaker Valve data .......................................... 276

Volume Balance Specifications window.............................................. 276 Global Edit windows ............................................................................ 277

Global Pipe editing......................................................................... 278 Global Junction editing .................................................................. 279

Editing common junction data................................................. 280 Editing specific junction data .................................................. 282

7. Customizing AFT Impulse ......................................... 287

Parameter and Unit Preferences ........................................................... 287

Table of Contents xix

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


Parameter Preferences.................................................................... 288 Unit Preferences............................................................................. 289

Setting preferred units ............................................................. 290 Description of selected unit ........................................................... 291 Database connections..................................................................... 291 Command buttons .......................................................................... 291

Workspace Preferences......................................................................... 292 Pipes and Junctions ........................................................................ 292

Pipe Line Options .................................................................... 292 Special Conditions options ...................................................... 293 Action When Drawing Selection Right To Left...................... 294 Pipe endpoint adjustments....................................................... 294 Modifying icon size ................................................................. 294 Action when dragging junctions.............................................. 294 Auto increment labels.............................................................. 294 Icon source............................................................................... 295

Display Options.............................................................................. 295 Workspace grid........................................................................ 296 Workspace symbols ................................................................. 296 Pipe direction arrows............................................................... 297 Displaying name and/or ID numbers ....................................... 297 Setting the default junction label position............................... 297 Allowable Workspace label movements ................................. 297 Popup menu ............................................................................. 297 Special Conditions Graphics ................................................... 298 Background Picture Scaling .................................................... 298

Colors and Fonts ............................................................................ 298 Sample Workspace......................................................................... 298 Database connections..................................................................... 298 Command buttons .......................................................................... 299

Toolbox Preferences............................................................................. 300 Customizing Toolbox contents ...................................................... 300

Toolbox shortcuts .................................................................... 301

xx AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xx c:\aft manuals\i4\impulse4 manual (si) g.doc

Database connections..................................................................... 302 General Preferences.............................................................................. 302 Customizing the Output........................................................................ 303 Specifying graph style preferences....................................................... 304

Auxiliary Graph Formatting window ...................................... 304 Specifying Visual Report preferences .................................................. 304 Customizing Pipe Fittings & Losses..................................................... 305 Building custom databases ................................................................... 305

Adding custom fluid properties to the Fluid Database .................. 306 Adding custom pipe materials to the Pipe Material Database ....... 309

Friction Data Sets .................................................................... 310 Pipe Physical Properties .......................................................... 311

Adding custom junction data to the Component Database ............ 311 Editing the Component Database................................................... 311

Database Manager ................................................................................ 312 Setting available databases............................................................. 312 Connecting and disconnecting to a database.................................. 313 Editing databases............................................................................ 314 Benefits of shared databases .......................................................... 316

Creating an enterprise-wide, network database system........................ 318 Overview ........................................................................................ 318 Creating database files ................................................................... 319 Sharing database files using DATABASE.LIB ............................. 321 Connecting to the external shared databases ................................. 323

8. Steady-State Hydraulic Theory and Loss Models.... 325

The Steady-State Solver ....................................................................... 325 Stagnation vs. static pressure boundaries....................................... 329

When to use static pressure ..................................................... 331 Static pressure at pressure control valve ................................. 332

Open vs. closed systems ....................................................................... 332 Verifying network solutions ................................................................. 335 Pressure drop calculation methods ....................................................... 335

Table of Contents xxi

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


Pressure drop calculation methods for Newtonian fluids .............. 336 Roughness-based methods ............................................................. 337

Laminar flow ........................................................................... 337 Turbulent flow......................................................................... 337 Transition flow ........................................................................ 337

Hydraulically smooth ..................................................................... 338 Hazen-Williams method................................................................. 338 Resistance....................................................................................... 338 MIT Equation for crude oil ............................................................ 338 Miller Turbulent method................................................................ 339 Frictionless pipes ........................................................................... 340

Pressure drop calculation methods for non-Newtonian fluids ............. 340 Duffy method for pulp and paper stock ......................................... 340 Brecht & Heller method for pulp and paper stock ......................... 341 Power Law non-Newtonian............................................................ 342 Bingham Plastic non-Newtonian.................................................... 343

Non-Newtonian flow through non-pipes ................................. 344 Design factors ....................................................................................... 344 Steady Solution Control parameters ..................................................... 344

Solution tolerance specification..................................................... 345 How tolerances relate to solution accuracy............................. 348

Relaxation ...................................................................................... 348 Flow Rate Relaxation .............................................................. 349 Pressure Relaxation ................................................................. 350 Avoiding false convergence .................................................... 350

Maximum Iterations ....................................................................... 351 Matrix Method ............................................................................... 351

Solution Progress window and iteration history................................... 351 Warnings in the solution....................................................................... 352 Modeling irrecoverable losses .............................................................. 353 Loss models and reference material ..................................................... 354

Design factors ................................................................................ 355 Area change.................................................................................... 355

xxii AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xxii c:\aft manuals\i4\impulse4 manual (si) g.doc

Bend ............................................................................................... 356 90 degree bends ....................................................................... 356 Non-90 degree bends ............................................................... 357

Valve .............................................................................................. 358 Orifice ............................................................................................ 358 Tee/Wye ......................................................................................... 359

Diverging Case ........................................................................ 361 Converging Case...................................................................... 362

K and CV (valve coefficient) .......................................................... 363 K for fire sprinklers........................................................................ 364

Pumps ................................................................................................... 365 Chempak thermophysical property database ........................................ 366

Accuracy options ..................................................................... 366 ASME steam tables database................................................................ 366 AFT Standard fluids ............................................................................. 366

Viscosity......................................................................................... 366

9. Theory of Waterhammer & Solution Methodology .. 367

Description of waterhammer ................................................................ 367 Instantaneous waterhammer ................................................................. 367 Wavespeed............................................................................................ 368 Communication time in pipes............................................................... 370 Method of Characteristics..................................................................... 370

Momentum equation ...................................................................... 371 Mass continuity equation ............................................................... 371

Transient vapor cavitation .................................................................... 376 Pipe sectioning...................................................................................... 377 Effects of flow resistance ..................................................................... 379 Assigned Pressure theory...................................................................... 379

Assigned Pressure vapor cavitation theory .................................... 380 Reservoir theory ................................................................................... 380

Reservoir vapor cavitation theory .................................................. 381 Assigned Flow theory........................................................................... 381

Table of Contents xxiii

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


Assigned Flow vapor cavitation theory.......................................... 381 Dead End theory ................................................................................... 382

Dead End vapor cavitation theory.................................................. 382 Branch theory ....................................................................................... 382

Branch vapor cavitation theory ...................................................... 384 Tee/Wye theory .................................................................................... 384 Valve theory ......................................................................................... 384

Exit valves...................................................................................... 386 Valve vapor cavitation theory ........................................................ 386

Infinite Pipe theory ............................................................................... 387 Infinite Pipe vapor cavitation theory.............................................. 388

Spray Discharge theory......................................................................... 388 Spray Discharge vapor cavitation theory ....................................... 390

Pump theory.......................................................................................... 390 Modeling pump as a known speed ................................................. 391

Pumps with higher order curve fits.......................................... 392 Backflow through the pump – one quadrant ........................... 393 Backflow through the pump – four quadrants ......................... 393 Flow rates greater the pump maximum – one quadrant .......... 393 Flow rates greater the pump maximum – four quadrants ........ 393 Submerged pumps ................................................................... 394

Pump with known speed vapor cavitation theory .......................... 394 Modeling pump with unknown speed ............................................ 395 Trip With Inertia and No Back Flow or Reverse Speed ................ 395 Startup With Inertia and No Back Flow or Reverse Speed............ 397 Trip With Inertia – Four Quadrant................................................. 399 Startup With Inertia – Four Quadrant, w/Motor Torque/Speed..... 404 Estimating the pump inertia ........................................................... 404

Liquid Accumulator theory .................................................................. 405 Liquid Accumulator vapor cavitation theory ................................. 406

Relief Valve theory............................................................................... 406 Internal relief valve ........................................................................ 406 Exit relief valve.............................................................................. 407

xxiv AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xxiv c:\aft manuals\i4\impulse4 manual (si) g.doc

Inline Relief Valve ......................................................................... 407 Variable Cv model ......................................................................... 408 Relief Valve vapor cavitation theory ............................................. 409

Internal relief valves ................................................................ 409 Exit relief valves...................................................................... 409 Inline relief valves ................................................................... 409

Check Valve theory .............................................................................. 410 Check Valve vapor cavitation theory............................................. 410

Control Valve theory ............................................................................ 411 Flow control valves ........................................................................ 411 Pressure control valves................................................................... 411 Pressure drop control valves .......................................................... 411

Surge Tank theory................................................................................. 412 Surge Tank vapor cavitation theory ............................................... 413

Gas Accumulator theory....................................................................... 414 Gas Accumulator vapor cavitation theory...................................... 416

Vacuum Breaker Valve theory ............................................................. 416 Sonic inflow ................................................................................... 417 Subsonic inflow.............................................................................. 417 Sonic outflow ................................................................................. 417 Subsonic outflow............................................................................ 417 Equation of state............................................................................. 418

Sonic Inflow ............................................................................ 419 Subsonic Inflow....................................................................... 419 Subsonic Outflow .................................................................... 420 Sonic Outflow.......................................................................... 422

Lumped inertia pipe with orifice theory............................................... 422 Interpreting cavitation results ............................................................... 424 Instantaneous waterhammer equation exceptions ................................ 425

Cavitation ....................................................................................... 426 Pipe size changes............................................................................ 426 Multiple pressure surge sources..................................................... 427 Line pack due to friction ................................................................ 427

Table of Contents xxv

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


Elevation changes........................................................................... 427

10. Modeling Time and Event Based Transients ......... 429

Time-based transients ........................................................................... 429 Event-based transients .......................................................................... 430

Single event transients ................................................................... 430 Dual event transients: cyclic and sequential .................................. 431

Cyclic events............................................................................ 432 Sequential events ..................................................................... 433

Junctions with inherent event logic ...................................................... 434 Check valve .................................................................................... 434 Relief valve .................................................................................... 434 Vacuum breaker valve.................................................................... 434

Thought experiment to further clarify event transients ........................ 435 Other transient features......................................................................... 435

Absolute vs. relative transient data ................................................ 435 Repeat transient.............................................................................. 436 Add time offset............................................................................... 436 Graphing the transient data ............................................................ 436

Event messages..................................................................................... 436 Transient indicators on the Workspace ................................................ 437 Transient data in Model Data ............................................................... 438

11. Reducing Surge Pressures ..................................... 441

General considerations ......................................................................... 441 Larger pipe sizes ................................................................................... 442 Flexible hose......................................................................................... 442 Slower or modified valve closure profiles............................................ 442 Parallel valves....................................................................................... 442 Relief valves ......................................................................................... 442 Vacuum breaker valves for low pressures ............................................ 443 Surge tank ............................................................................................. 443 Gas accumulator ................................................................................... 443

xxvi AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xxvi c:\aft manuals\i4\impulse4 manual (si) g.doc

12. Special Topics: Troubleshooting Models .............. 445

The philosophy of computer modeling................................................. 445 Troubleshooting steady-state models ................................................... 446

Use the Model Data window.......................................................... 446 Poor pump curve fits ...................................................................... 447 Use the Output window Sort feature.............................................. 448 Tee/Wye junction complexity........................................................ 450 Lower the flow rate relaxation....................................................... 450 Try absolute tolerance.................................................................... 451 Make initial flow rate guesses for pipes ........................................ 451 Review the Iteration History .......................................................... 451 Change boundary conditions.......................................................... 451 Turn off parts of the model ............................................................ 452 Break a large model into submodels .............................................. 452

Control valve failure issues in steady-state .......................................... 452 Infinite pipe junctions........................................................................... 453

Application example ...................................................................... 453 How pressure junctions work in steady-state ....................................... 455 The role of pressure junctions in steady-state ...................................... 458

Examples ........................................................................................ 459 Sizing pumps with flow control valves .......................................... 464 Pressure control valves................................................................... 470 Closing parts of a system ............................................................... 472

Effect of pipe sectioning on transient run times................................... 473 Artificial transients ............................................................................... 474

Manually detecting artificial transients.......................................... 476

Appendix A. Keyboard Modifiers and Shortcuts ......... 477

Selection Drawing Tool........................................................................ 477 Double-clicking tool ................................................................ 477 Control key .............................................................................. 477 Shift key................................................................................... 477 Alt key ..................................................................................... 478

Table of Contents xxvii

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


Drag left to right ...................................................................... 478 Drag right to left ...................................................................... 478

Pipe Drawing Tool................................................................................ 478 Double-clicking tool ................................................................ 478 Control key .............................................................................. 478 Shift key................................................................................... 478

Zoom Select Tool ................................................................................. 479 Shift key................................................................................... 479

Panning the Workspace ........................................................................ 479 Dragging junctions from Toolbox ........................................................ 479

Control key .............................................................................. 479 Shift key................................................................................... 479

Inspecting pipes and junctions.............................................................. 480 Shift key................................................................................... 480 Control key .............................................................................. 480

Moving junctions with connected pipes ............................................... 480 Control key .............................................................................. 480

Pipe and junction specifications window ............................................. 480 F2 function key........................................................................ 480 F5 function key........................................................................ 480

Appendix B. Limitations ................................................ 483

Appendix C. Installation Issues .................................... 485

Customization files ............................................................................... 485 Special file control features .................................................................. 485

What is the Control.aft File?.......................................................... 485 Specifying where Chempak is located ........................................... 486

Network installations............................................................................ 486 Problems loading the Graphics Server ................................................. 487

Appendix D. Chempak Technical Information ............. 491

xxviii AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xxviii c:\aft manuals\i4\impulse4 manual (si) g.doc

Appendix E. Obtaining Technical Support................... 493

AFT's three levels of support................................................................ 493 1. Free Software Support for One Year ......................................... 493 2. Annual Support, Upgrade and Maintenance (SUM).................. 493 3. Waterhammer Modeling Consulting Support ............................ 493

Contacting AFT .................................................................................... 494 Telephone support.......................................................................... 494 Web site.......................................................................................... 494 E-Mail support ............................................................................... 494 Mail support ................................................................................... 494

References...................................................................... 495

Glossary.......................................................................... 497

8/10/2007 4:10:28 PM Page xxix c:\aft manuals\i4\impulse4 manual (si) g.doc

Nomenclature

a wavespeed

A cross-sectional flow area of a pipe

BM parameter as defined after Equation 9.20

BP parameter as defined after Equation 9.20

CD discharge coefficient

CM parameter as defined after Equation 9.20

CP parameter as defined after Equation 9.20

CV valve coefficient

CHW Hazen-Williams factor

D diameter of a pipe

E modulus of elasticity

e pipe wall thickness

f friction factor

fT turbulent friction factor (used by Crane, 1988)

F error value in mass balance

FB Dimensionless pump torque parameter (see Chapter 9)

FW Dimensionless pump head parameter (see Chapter 9)

g gravitational constant

h head ratio for pump

H head

HGL hydraulic gradeline (piezometric head)

I moment of inertia of pump and entrained liquid

J Jacobian matrix

K loss factor

K constant for Power Law fluid

KB liquid bulk modulus of elasticity

Ksprinkler fire sprinkler loss factor

L length of a pipe

xxx AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xxx c:\aft manuals\i4\impulse4 manual (si) g.doc

m& mass flow rate

n general constant

n constant for Power Law fluid

N pump synchronous speed (rpm)

P pressure, static

Po pressure, stagnation

Pw wetted perimeter

P power

Q volumetric flow rate

r radius

r relaxation

R, R’ resistance

Rc coefficient of rigidity

Re Reynolds Number

s pump speed

SB parameter as defined in Equation 9.27

SC parameter as defined in Equation 9.28

Sy yield stress

t time

T torque on pump

v pump flow ratio to reference

V velocity

x distance along pipe centerline

z elevation

α, μ, θ angle

α pump speed ratio

β diameter ratio β torque ratio

λ method of characteristics multiplier

ν kinematic viscosity

ε roughness

Nomenclature xxxi

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


ω pump speed

ρ density μ dynamic viscosity

Subscripts

f fluid

i junction or pipe station at which solution is sought

j junctions with pipes connecting to junction I

m motor

R rated conditions for a pump, usually BEP

o stagnation

∞ infinity, far away, ambient

new current time for computation

old most recent time for computation

xxxii AFT Impulse 4.0 User’s Guide

8/10/2007 4:10:28 PM Page xxxii c:\aft manuals\i4\impulse4 manual (si) g.doc

8/10/2007 4:10:28 PM Page 1 c:\aft manuals\i4\impulse4 manual (si) g.doc

C H A P T E R 1

Introduction

Welcome to AFT Impulse™ 4.0 for Windows. AFT Impulse is a graphical platform for modeling waterhammer and surge transients in pipe networks.

AFT Impulse’s advanced Windows graphical interface simplifies the complex process of building waterhammer models. The transient solution engine employs the proven and well understood Method of Characteristics to solve the fundamental equations of waterhammer.

An integral part of the interface is AFT Impulse's Workspace window. Here you place icons and draw lines to represent the components of a pipe system. You then enter data for the components in the associated Specifications windows.

AFT Impulse shows you both input data and analysis results in graphical form, allowing rapid analysis of the model's validity. Identifying poor assumptions, catching typographical errors, and rerunning models are all accelerated because of AFT Impulse's graphical environment. This reduces the possibility of modeling errors.

Whether your pipe system model will be used to evaluate and improve an existing system or to design a new one, AFT Impulse increases your productivity in the modeling process.

Terminology

The term “waterhammer” has been used for more than one-hundred years, and describes a transient phenomenon in piping systems caused by

2 AFT Impulse 4.0 User’s Guide


changing pressure or flow boundary conditions, or changing operating conditions of pipe system equipment. The result is often a pressure surge that builds much higher than the steady state pressure that preceded it.

There have been and are many terms employed to refer to the waterhammer phenomenon. The oldest and probably most recognized is the term waterhammer itself. One drawback of this term is that it suggests a phenomenon associated only with water. This is misleading, because the surge pressures resulting from waterhammer occur in all liquid piping systems.

To help avoid this confusion, the term “fluidhammer” has been advocated and/or used in some of the literature. Fluidhammer does better convey the breadth of fluids that can experience waterhammer, but has not gained wide acceptance.

Another alternate term for waterhammer is “hydraulic transients”. Again, a suggestion of water-related events are present in this term. Another term, “fluid transients” was therefore put forth. This term also has problems. Specifically, a fluid system can have many kinds of transient behavior, of which waterhammer is only one.

Finally, another commonly heard term is “surge”. This term is more specific than fluid transients, but is still somewhat ambiguous.

The unfortunate situation, therefore, is that there is no universally accepted or recognized term to refer to this important phenomenon.

Arguably, the term waterhammer remains the most recognized term, even with its deficiencies. Therefore, with the qualification that it is not limited solely to water systems, waterhammer is the preferred term and the one that will be used most commonly when referring to AFT Impulse.

Modeling capabilities

AFT Impulse can be used to model a wide variety of pipe system transients, including:

• Transients in open and closed (recirculating) systems

• Network systems that branch or loop, with no limit on the number of loops

• Systems with valve transients

Chapter 1 Introduction 3

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



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

Thermophysical property data

AFT Impulse derives physical properties from one of three sources. The first is the standard AFT Impulse set of incompressible fluids which contains data for about 10 common fluids (called AFT Standard). These are the same fluids as were available in previous versions of AFT Impulse.

The second are water properties from the ASME Steam Tables database.

The third source is Chempak™. Chempak has a database of approximately 700 fluids, and supports user specified fluid mixtures. AFT Impulse is restricted to non-reacting mixture calculations.

The user does have the option of manually entering physical properties. This is called an Unspecified Fluid.

In the remainder of this User’s Guide it will be assumed in general discussion that the user has access to the Chempak Database. If the user does not, then the portions describing Chempak will not be applicable.

Who can use AFT Impulse

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

Engineering assumptions in AFT Impulse

AFT Impulse is based on the following fundamental fluid mechanics assumptions:

• Liquid flow

• One-dimensional flow


8/10/2007 4:10:28 PM Page 5


• No chemical reactions

• Wavespeed remains constant during transients

• Non-condensable gas release is negligible

• Bubbles that form during transient cavitation do not move

AFT Impulse allows you to change the gravitational acceleration of the system, a feature useful for aerospace vehicle applications or pipe systems that are designed for extraterrestrial environments.

Installing AFT Impulse

You install AFT Impulse on your computer by running the SETUP.EXE program located on the installation CD. The setup program will install AFT Impulse, the Help system, example models, verification models, and AFT Impulse support files to your hard disk.

Note: You must use the setup program to install AFT Impulse. You cannot simply copy the files from the installation disks to your hard drive.

Before you run the setup program, make sure your computer platform meets the minimum requirements.

Check hardware and system requirements

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

Read the README.TXT file

The README.TXT file is included on the installation CD. README.TXT contains information about installation and support material that is more up to date than this User’s Guide.



Table 1.1 Computer hardware and software requirements to use AFT Impulse

Feature Required

Processor Pentium II or higher

Hard disk space 40 MB

CD-ROM drive Yes

Monitor resolution SVGA (800x600)

RAM 128 MB

MS-Windows Windows 98, NT 4.0, 2000, ME, XP or later versions

Run the setup program

When you run the setup program, you will be prompted for the directory where you want AFT Impulse installed. To run the setup program:

1. Insert the CD into your CD-ROM drive.

2. If Autorun is disabled, from the Start button choose Run. Assuming your CD-ROM is drive “d”, type d:\ setup, or click the Browse button and search for the setup program on your CD-ROM drive in the Impulse 4.0 directory.

3. Select Impulse 4.0 on the displayed menu.

4. Follow the instructions on the screen.

Getting started with AFT Impulse

If you are not familiar with basic techniques used in Windows applications (working with the mouse, selecting menus, using dialog boxes, and so on), first consult your Windows documentation.

To gain an understanding of the powerful features available in AFT Impulse, read Chapters 3-12 of this User’s Guide. For an overview of the model building process and AFT Impulse’s capabilities, read Chapter 2 and follow the instructions for building a simple model.


8/10/2007 4:10:28 PM Page 7


Example models

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

Using online help

To access AFT Impulse's online help, press the F1 function key or select Help from the menu bar. For convenient access, much of the content of this User’s Guide is included in the Help system. The Help button in each dialog window provides context-sensitive help on the features of that window.

In the Help system you have the option of searching for information on specific topics or searching through the hierarchical layout to find more general information.

Verification models

A large number of verification models have been built and compared to published results from the open literature. These are included in the Verification sub-folder below AFT Impulse. Along with the models you will find documented comparisons in the Verification help file also installed in the Verification folder.

AFT Impulse overview

The AFT Impulse window has five subordinate windows that work in an integrated fashion. You work exclusively from one of these windows at all times. For this reason they are referred to as primary windows.

Of the five primary windows, two are input windows, two are output windows, and one displays output and input information. Figure 1.1 shows the relationship between the primary windows.



Graph Results

Visual Report

Workspace Output

Model Data

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.


8/10/2007 4:10:28 PM Page 9


Converting models and databases from previous versions

AFT Impulse 1.0, 2.0 and 3.0 models can be opened directly by choosing Open from the File menu. Once saved, they are saved in the version 4.0 format. Importing issues are discussed in the following.

Pipe friction

AFT Impulse 1.0 only worked with constant friction factor models. When these models are imported into version 4.0, the friction model for the pipe is set to Explicit Friction Factor or Frictionless, depending on whether the version 1.0 pipe had a zero or non-zero friction factor.

Infinite pipe junctions

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

Transient junction data

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

Junction loss models

A loss model found in several junction types in AFT Impulse 1.0 was K/A^2. This is no longer supported in version 4.0. Such data, whether steady or transient, is converted to a loss model that is supported such as K factor or Cv for valves.

Reservoir transients

AFT Impulse 1.0 allowed reservoir surface level transients as a delta from the steady-state. These will all get converted automatically to the version 4.0 convention of specifying the actual liquid surface level.



Databases

AFT Impulse 1.0 databases can be converted using the AFT Convert utility provided with the software.

Converting models and databases from AFT Fathom or Mercury

AFT Fathom™ 4.0, 5.0 and 6.0 models, and AFT Mercury™ 5.0 models can be read directly by AFT Impulse 4.0 by choosing Open from the File menu and opening the file. Only the base scenario can be imported. If you want to import scenarios that are not the base, you will need to first use the Scenario Manager in AFT Fathom or AFT Mercury to save the scenario into a separate model file.

When imported, all pipe specification data should be maintained, including friction data.

Combining pipes

AFT Impulse 4.0 will offer to combine pipes for you when you import AFT Fathom models. This will allow models to run much more efficiently. When combined, AFT Impulse will create intermediate pipe elevations at the points where junctions are deleted. All pipe resistance and junction losses will be maintained in the model.

Junctions with polynomial loss data or interpolated x-y data

Except for General Components, AFT Impulse 4.0 does not support polynomial or interpolated x-y loss data. When AFT Fathom or AFT Mercury models are imported, all such data is automatically converted to K factors.

Databases

AFT Fathom and AFT Mercury databases can be converted using the AFT Convert utility.


8/10/2007 4:10:28 PM Page 11


What’s new in version 4.0

General

• Users can create transient force files for use in CAESAR II pipe stress software

• Select Special on pipe material

• Significantly enhanced global editing of pipes and junctions (data now shown in categories)

• Batch runs for scenarios

• All printing can go directly to a PDF file

• In grids, cells that have drop-down capability have a persistent drop down symbol

• Initialization and data files now saved to user folders

Pipes

• Enhanced fittings & losses entry for pipes

• New pipe friction model uses resistance in dH/Q2

Junctions

• Valves now support Cv vs. percent open (like Control Valves)

• More advanced control valve logic

Transient Control

• Users can define “force sets” for purposes of plotting force data

• Runtime estimates based on users previous

Solver

• Power Law and Bingham Plastic now cover both laminar and turbulent flow



Workspace

• Renumber Increment Added

• Shortcut button provides fast access to explanations of shortcut functions

• Selections can be rotated right or left

• New window added to show all undefined pipes and junctions and missing properties

Model Data

• Can see data for all scenario direct ancestors on Model Data

Output

• Users can print Model Data with output

Graph Results

• Can graph forces on pipes

Cavitation pressure changes In AFT Impulse 2.0 and 3.0 there was an attempt at increased numerical precision by calculating cavitation when the static pressure reached vapor pressure. Hence the stagnation pressure was always greater than vapor pressure.

Further review showed that using static pressure as the cavitation criteria could result in increased numerical noise. A change was therefore made in AFT Impulse 4.0 to use stagnation pressure as the cavitation criteria. This will result in static pressures below vapor pressure in certain situations.


C H A P T E R 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.



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

Minimizedprimary

windows

Status Bar

Figure 2.1 The Workspace window is where the model is built

The Workspace window The Workspace window is the primary vehicle for building your model. This window has two main areas: the Toolbox and the Workspace itself. The Toolbox is the bundle of tools on the far left. The Workspace takes up the rest of the window. See Chapter 4 for more detailed information on the Workspace.

You will build your pipe flow model on the Workspace using the Toolbox tools. At the top of the Toolbox are a Shortcut button and four drawing tools. The Selection Drawing tool, on the upper left below the Shortcut button, is useful for selecting groups of objects on the Workspace for editing or moving. The Pipe Drawing tool, on the upper

Chapter 2 A Walk Through AFT Impulse 15

8/10/2007 4:10:28 PM Page 15


right below the Shortcut button, is used to draw new pipes on the Workspace. Below these two tools are the Zoom Select tool and the Annotation tool. The Zoom Select tool allows you to draw a box on the Workspace after which AFT Impulse will zoom into that area. The Annotation tool allows you to create annotations and auxiliary graphics.

Below the four drawing tools are nineteen icons that represent the different types of junctions available in AFT Impulse. Junctions are components that connect pipes and also influence the pressure or flow behavior of the pipe system. The nineteen junction icons can be dragged from the Toolbox and dropped onto the Workspace.

When you pass your mouse pointer over any of the Toolbox tools, a Tool Tip identifies the tool's function.

Step 2. Lay out the model

To lay out the walk-through model, you will place the two reservoir junctions, three branch junctions, a surge tank junction, a relief valve junction, and a valve junction on the Workspace. Then you will connect the junctions with pipes.

A. Place the first reservoir

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.



The Reservoir you have created on the Workspace will take on the default ID number of 1. You can change this to any desired number greater than zero and up to 30,000.

Figure 2.2a Walk Through Model with one reservoir placed

Editing on the Workspace Once on the Workspace, junction objects can be moved to new locations and edited with the features on the Edit menu. Cutting, copying, and pasting are all supported. A single level of undo is available for all editing operations.

Note: The relative location of objects in AFT Impulse is not important. Distances and heights are defined through dialog boxes. The relative locations on the Workspace establish the connectivity of the objects, but have no bearing on the actual length or elevation relationships.

B. Place the first branch

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).


8/10/2007 4:10:28 PM Page 17


Figure 2.2b Walk Through Model with one reservoir and one branch placed

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).



Figure 2.2c Walk Through Model after placing the Surge Tank

C. Place surge tank

D. Place the second reservoir

The second reservoir can be created the same way as the first one or it can be derived from the existing reservoir.

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).


8/10/2007 4:10:28 PM Page 19


Figure 2.2d Walk Through Model after placing the second Reservoir

E. Add the remaining junctions

Next add the four remaining junctions - two branches, and two regular valves (Figure 2.2e).

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 valve junctions by selecting the junction icons from the Toolbox and placing them on the Workspace as shown in Figure 2.2e. These junctions will become J7 and J8, respectively.

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

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.


8/10/2007 4:10:28 PM Page 21


Figure 2.2f Walk Through Model after drawing first pipe

To place the pipe between J1 and J2, grab the pipe in the center with the mouse and drag it so that its left endpoint falls within the J1 Reservoir icon and drop it there. Next, grab the right endpoint of the pipe and stretch the pipe, dragging it until the endpoint terminates within the J2 Branch icon (see Figure 2.2g).

Reference positive flow direction Located on the pipe is an arrow that indicates the reference positive flow direction for the pipe. AFT Impulse assigns a flow direction corresponding to the direction in which the pipe is drawn. You can reverse the reference positive flow direction by choosing Reverse Direction from the Arrange menu or selecting the reverse direction button on the Toolbar.



Figure 2.2g Walk Through Model after placing first pipe

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

G. Add the remaining pipes

A quicker way to add a pipe is to draw it directly between the desired junctions.

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


8/10/2007 4:10:28 PM Page 23


At this point all the objects in the model are graphically connected. Save the model by selecting Save in the File menu or by clicking on the diskette button on the Toolbar.

Note: It is generally desirable to lock your objects to the Workspace once they have been placed. This prevents accidental movement and disruption of the connections. You can lock all the objects by choosing Select All from the Edit menu, then selecting Lock Object from the Arrange menu. The lock button on the Toolbar will appear depressed indicating it is in an enabled state, and will remain so as long as any selected object is locked. Alternatively, you can use the grid feature enabled on the Workspace Preferences window and specify that the pipes and junctions snap to grid.

Figure 2.2h Walk Through Model with all pipes and junctions placed

Step 3. Set the unit preferences

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.

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.


8/10/2007 4:10:28 PM Page 25


The Status Bar at the bottom of the AFT Impulse window also reflects the state of each checklist item (see Figure 2.1). Once the checklist is complete, the Model Status light in the lower left corner turns from red to green.

Figure 2.4 The Checklist tracks the model’s status

A. Specify Steady Solution Control

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

B. Specify output control

The second item on the checklist is Specify Output Control. Like Solution Control, this item is always checked when you start AFT Impulse. Default Output Control parameters and a default title are assigned.

You may want to add a descriptive title for the model. To do this, open the Output Control window (see Figure 2.5) and enter a title on the General tab. In addition, this window allows you to select the specific output parameters you want in your output. You also can choose the units for the output.



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

Figure 2.5 The Output Control window lets you customize the output

C. Specify system properties

The third item on the checklist is Specify System Properties. To complete this item, you must open the System Properties window (see Figure 2.6). This window allows you to specify your fluid properties (density, dynamic viscosity, bulk modulus and optional vapor pressure), viscosity model, gravitational acceleration and atmospheric pressure.

For models with variable fluid properties, the values for density, viscosity and bulk modulus are default fluid properties. You can then enter different property values, if desired, for any pipe in the Pipe Specifications window.


8/10/2007 4:10:28 PM Page 27


Figure 2.6 The System Properties window lets you enter physical properties of the fluid

You can model the fluid properties in one of five ways.

1. Unspecified fluid – This fluid model allows you to directly type in the density, viscosity, bulk modulus and vapor pressure.

2. AFT Standard fluid – This fluid model accesses fluid data from the AFT Standard database. These fluid properties are either temperature dependent or dependent on the solids concentration. You type in the desired condition (e.g., temperature), click the Calculate Properties button and the required properties are calculated. Users can add their own fluids to this database. Custom fluids are created by opening the Fluid Database window from the AFT Impulse Database menu or by clicking the Edit Fluid List button in the System Properties window.

3. Chempak Fluid – This fluid model allows you to select a single fluid from the Chempak database list. These fluid properties are pressure and temperature dependent, although some are temperature dependent only. Chempak is an optional add-on to AFT Impulse.



4. Chempak Mixture – This fluid model allows you to create a liquid mixture from among the Chempak database fluids. These fluid properties are pressure and temperature dependent. Upon entering a pressure, AFT Impulse will display a temperature range applicable to the liquid. Chempak mixture data can be exported to a data file, and then be shared with other AFT applications or users. Chempak is an optional add-on to AFT Impulse.

5. Water Data from ASME Steam Tables – As its name implies, this fluid model obtains water properties from ASME steam tables. This model is pressure and temperature dependent.

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.

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


8/10/2007 4:10:28 PM Page 29


colors are configurable through Workspace Preferences from the Options menu.)

Because you have not yet defined the pipes and junctions in this sample problem, all the objects' ID numbers will change to red when you turn on Show Object Status.

Undefined Objects window The Undefined Objects window lists all undefined pipes and junctions and further displays the items that are not yet defined.

A. Define Reservoir J1

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 junction’s name is the junction type. The name can be displayed on the Workspace, Visual Report or in the Output.

Most junction types can be entered into a custom database allowing the junction to be used multiple times or shared between users. To select a junction from the custom database, choose the desired junction from the Database list. The current junction will get the properties from the database component.



The Copy Data From Jct list will show all the junctions of the same type in the model. This will copy selected parameters from an existing junction in the model to the current junction.

The pipe table on the Pipe Depth and Loss Coefficients tab allows you to specify entrance and exit loss factors for each pipe connected to the reservoir (in this case there is one). You can enter standard losses by selecting the option buttons at the right. The default selection is the Custom option with loss factors specified as zero. To later change the loss factors, click within the pipe table and enter the loss. You can also specify a depth for the pipe.

The Optional tab allows you to enter different types of optional data. You can select whether the junction number, name, or both are displayed on the Workspace. Some junction types also allow you to specify an initial pressure as well as other junction specific data. The junction icon graphic can be changed, as can the size of the icon. Design factors can be entered for most junctions, which are applied to the pressure loss calculations for the junction in order to give additional safety margin to the model.

Each junction has a tab for notes, allowing you to enter text describing the junction or documenting any assumptions.

The highlight feature displays all the required information in the Specifications window in light blue. The highlight is on by default. You can toggle the highlight off and on by double-clicking anywhere in the window or by pressing the F2 key. The highlight feature can also be turned on or off by selecting it on the Options menu.

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.


8/10/2007 4:10:28 PM Page 31


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.



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.


8/10/2007 4:10:28 PM Page 33


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



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.

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):


8/10/2007 4:10:28 PM Page 35


Figure 2.10 Valve Specifications for J7 initiates the transient for the system

Time Cv 0 2392.6

10 797.4

100 797.4

F. Define Valve J8

Open the J8 Valve Specifications window. For this example, a regular valve junction will be used to model the behavior of a relief valve. This relief valve will vent to an external source, so it should be modeled as an Exit Valve, with an exit pressure of 1 atm. Enter an elevation of 50 meters.

Choose the loss model as Cv and enter a Cv value of 651.2. This represents the valve's Cv after it fully opens. The valve remains closed



until the cracking pressure is exceeded (defined on the Transient Data tab), at which point the valve opens to relieve pressure.

Click the Transient Data tab and select the Single Event option for the Initiation of Transient. Select Pressure Stagnation at Pipe as the Event Type, Greater Than or Equal To as the Condition, and a Value of 1.688 MPa at the outlet of Pipe 8. This Event represents the cracking pressure required initiate the opening transient of the relief valve.

Enter the valve transient Cv data below in the Transient Data table (see Figure 2.11a):

Time Cv 0 0

3 651.2

63 0

100 0

The data represents the valve initially closed at time zero. Time zero is the moment of cracking. The valve then opens and gradually closes again over a period of 60 seconds.

Because the valve is closed initially, set the valve Special Condition on the Optional Data tab to Closed (see Figure 2.11b).


8/10/2007 4:10:28 PM Page 37


Figure 2.11a Valve Specifications for J8 initiates the transient for the valve

G. Define Pipe P1

The next step is to specify all the pipes. To open the Pipe Specifications window, double-click the pipe object on the Workspace.

First open the Pipe Specifications window for Pipe P1 (Figure 2.12). Choose the Pipe Material as Unspecified, choose the User Specified Wavespeed option, and select the Friction Model Data Set as Unspecified with the Friction Model as "Explicit Friction Factor". Enter a length of 1001.2 meters, a diameter of 1.5 meters, and a friction factor of 0.012.



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

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

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


8/10/2007 4:10:28 PM Page 39


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

The Pipe Specifications window The Pipe Specifications window offers control over all important flow system parameters related to pipes.

The Inspect feature also works within the Pipe Specifications window. To Inspect a connected junction, position the mouse pointer on the connected junction's ID number and hold down the right mouse button. This is helpful when you want to quickly check the properties of connecting objects. (You can also use this feature in junction Specifications windows for checking connected pipe properties.)

By double-clicking the connected junction number, you can jump directly to the junction's Specifications window. Or you can click the Jump button to jump to any other part of your model.

Figure 2.12 Pipe Specifications window for P1



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 Length (meters)

Diameter (meters)

Friction Factor

Wavespeed (meters/sec)

1 1001.2 1.5 0.012 996.3

2 2000 1 0.013 995.3

3 2000 0.75 0.014 995

4 502.5 0.5 0.015 1000

5 502.5 0.5 0.015 1000

6 1001.2 1 0.014 996.3

7 2000.2 0.75 0.013 995.1

8 100.12 1 0.014 996.3

After entering the data for all the pipes, the fourth checklist item should be completed. If it is not, see if the Show Object Status is on. If not, select Show Object Status from the View menu or toolbar. If the fourth checklist item is not completed at this point, see if any of the pipes or junctions have their number displayed in red. If so, you did not enter all the data for that item.

Before continuing the model, save it to file one more time. It is also a good idea to review the input using the Model Data window.

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


8/10/2007 4:10:28 PM Page 41


The Model Data window allows access to all Specifications windows by double-clicking the appropriate ID number in the far left column of the table. You may want to try this right now.

Customize view

General data

Pipe data

Junction data

Figure 2.13 The Model Data window shows the input data in text form

Step 6. Complete the last two checklist requirements

A. Specify pipe sectioning

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



window helps divide the pipes into computation sections in a manner which is consistent with the Method of Characteristics (MOC).

Figure 2.14 The Transient Data tab in the Model Data Junction data area shows all transient data entered

Select Section Pipes on the Analysis menu to display the Section Pipes Window (Figure 2.15). For this model the controlling pipe is P8. This is the pipe with the shortest end-to-end communication time (i.e., L/a – the length divided by the wavespeed). To satisfy the MOC, the following equation must be applied:

nL

a tii

i

= Δ

where n is the number of sections in pipe i, L is the length, and a is the wavespeed. The Δt is the time step. Since all pipes in the network must be solved together, the same time step must be used for each pipe. With a given length and wavespeed for each pipe, it can be seen from the above equation that it is unlikely that the number of required sections, n, for each pipe will be a whole number.


8/10/2007 4:10:28 PM Page 43


To address this situation, it is helpful to recognize that the wavespeed, a, is the least certain input parameter. It is therefore acceptable to allow up to a 15% uncertainty in wavespeed. By adjusting the wavespeed for each pipe within this tolerance the sectioning can be made to come out as whole numbers for each pipe. The Section Pipes window automates this process by searching for sectioning which satisfies the required tolerance.

You can specify the tolerance on wavespeed by entering it into the Max. Percentage Error. The minimum and maximum allowable sections in the controlling pipe narrows the search space. Also, the Percentage Increment directs the routine in how fine to search the search space.

This model has been specified by Karney in such a way that the pipe sectioning falls closely along whole numbers.

Figure 2.15 The Section Pipes window automates the sectioning process and calculates the time step.

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.

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.

Select Transient Control on the Analysis menu to display the Transient Control window (Figure 2.16 top). Enter zero for Start Time and 65 for Stop Time.

The Transient Control window lets you enable or disable transient cavitation modeling. It also offers control over how AFT Impulse should respond to artificial transients. Artificial transients are a problem that can sometimes occur when steady-state and transient conditions are inconsistent.

Data can be saved for all pipe stations or only selected stations. The selected stations are shown on the Pipe Station Output tab (see Figure 2.16 bottom). With the list next to the “Change All Pipes To” button set to “All Stations”, click the “Change All Pipes To” button. This will save all pipe station data for all pipes, which will be useful later for animation purposes.


8/10/2007 4:10:28 PM Page 45


Figure 2.16 The Transient Control window offers features to specify the time span for the transient and what output data is written



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 2.17). You can use this window to pause or cancel the Solver's activity.

The two solvers

AFT Impulse has two Solvers. The first is called the Steady-State Solver, which as its name suggests obtains a steady-state solution to the pipe network. The second Solver is called the Transient Solver. This solves the waterhammer equations.

Before a transient simulation can be initiated, the initial conditions are required. These initial conditions are the steady-state solution to the system. After the steady-state solution is obtained by the Steady-State Solver, AFT Impulse uses the results to automatically initialize the Transient Solver and then run it.


8/10/2007 4:10:28 PM Page 47


Figure 2.17 The Solution Progress window displays the state of the simulation

When the solution is complete, click the View Output button to display the text-based Output window. The information in the Output window can be reviewed visually on the screen, saved to file, exported to a spreadsheet-ready format, copied to the clipboard, printed to an Adobe PDF file, and printed out on the printer.

The transient output file

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



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

Step 8. Review the output

The Output window (Figure 2.18) is similar in structure to the Model Data window. Three areas are shown, and you can enlarge each area by selecting the options from the Show list box on the Toolbar or from the View menu. The items displayed in the tables are those items you chose in the Output Control window (checklist item # 2).

The Output window allows you to review both the steady-state and transient results. A summary of the maximum and minimum transient results for each pipe is given on the Transient Max/Min tab in the pipe area (Figure 2.19). You can also review the solutions for each time step (i.e., a time history) for which data was written to file. These two data sets are located on the Transient Output tab and Transient Max/Min tab in the pipe area of the Output window (see Figures 2.17 and 2.18).


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


Customize view

General results

Pipe results

Junction results

Figure 2.18 The Output window displays output results in text form



Figure 2.19 The Transient Output and Transient Max/Min results tabs are in the Output window Pipe section area. These show transient results for pipes.


8/10/2007 4:10:28 PM Page 51


A. Modify the output format

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

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 it has been restored or, if minimized, clicking on the minimized window at the bottom of the Workspace and restoring it. The Graph Results window offers full-featured Windows plot preparation.

Choose Select Graph Data from the View menu or the Toolbar to open the Select Graph Data window (see Figure 2.20). From the All Pipe Stations… list choose pipe 7. Select the Outlet of pipe 7. This is the station connected to the exit valve. A graph of this station shows the pressure and flow histories at the valve.

Figure 2.20 The Select Graph Data window controls the Graph Results content

Click the Add >> button to add this station to the list on the right.


8/10/2007 4:10:28 PM Page 53


From Graph Parameters select Hydraulic Gradeline (HGL) and choose meters as the Y-Axis units. Then click the Show button to display the graph (Figure 2.21).

This pressure/head history can be compared to that published in Karney (1992). This is done in the Verifications documentation provided with AFT Impulse.

Figure 2.21 The Graph Results window offers full-featured plot generation.

C. Animate the Results

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



Figure 2.22 The Select Graph Data window allows selection of animation parameters.

Additional animation control features will be displayed on the Graph Results window (see Figure 2.23). Push the Play button and AFT Impulse will run through all of the results in pipes 1, 6 and 7 for the entire transient which animates the results. The animation can be paused or stopped, and the time can be reset to any desired time using the time slider. If the animation is too fast, slow it down using the animation speed control at the right.

You can use the other buttons in the Graph Results window to change the graph appearance and to save and import data for cross-plotting. The Graph Results window can be printed, saved to file, copied to the clipboard, or printed to an Adobe PDF file. The graph’s x-y data can be exported to file or copied to the clipboard.


8/10/2007 4:10:28 PM Page 55


Figure 2.23 The Graph Results window allows animation of output to be displayed.

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 it has been restored or, if minimized, clicking on the minimized window at the bottom of the Workspace and restoring it. This window allows you to integrate your text results with the graphic layout of your pipe network.



Figure 2.24 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 2.24. Default parameters are already selected, but you can modify these as desired. For now, select Max Pressure Stagnation and Min Pressure Stagnation in the Pipe Results area. Click the Show button. The Visual Report window graphic is generated (see Figure 2.25).

It is common for the text in the Visual Report window to overlap when first generated. You can change this by selecting smaller fonts or by dragging the text to a new area to increase clarity (this has already been done in Figure 2.25, as has the selection to show units in a legend). This window can be printed, copied to the clipboard for import into other Windows graphics programs, saved to file, or printed to an Adobe PDF file.


8/10/2007 4:10:28 PM Page 57


Figure 2.25 The Visual Report window integrates results with the model layout

Conclusion

You have now used AFT Impulse's five primary windows to build a simple model. Review the rest of this User’s Guide for more detailed information on each of the windows and functions.


C H A P T E R 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.



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

Chapter 3 Fundamental Concepts 61

8/10/2007 4:10:28 PM Page 61


specifying loss information are provided where they are consistent with the loss factor method. For example, you will find Cv for valves.

Incorporating equivalent length data The equivalent length method is a popular method for modeling losses in fittings and equipment. There are two ways you can incorporate equivalent loss data into AFT Impulse – one way which is acceptable and one which isn't.

First we will discuss the unacceptable way. You can account for losses by adding to the actual pipe length, thus creating a fictitious length. For steady flow calculations this is acceptable. However, in modeling waterhammer transients the true length is a critical aspect of reflection of pressure waves. If the length is artificially increased, this will change the dynamic characteristics of the pipe and yield incorrect results. Any approximations which result in artificially changing pipe length are unacceptable.

The second way to incorporate equivalent length data is to apply a design factor to the pipe. The design factor is specified on the Pipe Specifications window Optional tab, and acts as a multiplier on the standard pipe friction. Thus if the pipe is physically 30 meters long, and has fittings with 3 meters of equivalent length, a design factor of 1.1 can be used to account for this. This maintains the true physical length and thus the dynamic behavior.

Convention for specifying junction base area

Loss factors are area dependent. When specifying a loss factor at a junction, the base area is of critical importance. By default AFT Impulse adheres to the following convention: whenever it makes sense physically, the base area is always the upstream pipe flow area. However, this can be modified by the user. In most junction Specifications windows, the base area can be specified as the upstream pipe, downstream pipe, user specified area or diameter.

However, some junction types, especially those that allow three or more connecting pipes, specify loss factors based on each connecting pipe flow area, whether upstream or downstream. An example of this is a wye in which each of the connecting pipes has a different diameter. The idea of specifying a loss factor based on upstream flow area loses meaning because there may be more than one upstream pipe. In these cases the



loss factor is referenced to the base area of the actual downstream pipe(s) where it connects to the junction. The types of junctions that do not always follow the upstream pipe area convention are Tee/Wye, Branch, Reservoir, Assigned Pressure, and Assigned Flow. Other exceptions are Valves and Relief Valves specified as exit valves.

Specifying losses

Local losses can be specified in two ways. The most flexible way of including a loss factor in an AFT Impulse model is to associate it with a junction specifically defined for that type of loss.

Secondly, you can include loss factors with pipes. When you specify Fittings & Losses in the Pipe Specifications window, the loss is assumed to be distributed evenly along the pipe length, much like friction losses. For this reason a pipe-associated loss factor is referred to as a distributed loss. The losses at junctions, on the other hand, are point losses.

When modeling for waterhammer, it is preferred that losses for all static components be included in pipes as Pipe Fittings & Losses. This reduces the number of pipes and model run time.

Modeling valve closures with K factors

You can close a valve by specifying the transient K factor as a –1 (negative one).

Specifying frictional losses in pipes

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

There are also two special loss models. The MIT equation is applicable for petroleum. And the Miller turbulent is applicable for light hydrocarbons.


8/10/2007 4:10:28 PM Page 63


Comment on importance of hydraulic losses during transients

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

Static vs. dynamic losses Pipe system fittings and equipment can be static or dynamic. Whenever possible, static elements should be lumped together with other elements or, better, into pipes. Examples of static elements that should be lumped are elbows, orifices, filters, and valves whose position remains constant over time. To discourage users from using static elements as junctions in their model, such junctions are not offered in AFT Impulse. There are no junction types for elbows, orifices or filters. However, valves can be either static (i.e., no position changes), or dynamic (i.e., they can open or close during transient modeling). The user can thus model a static valve if he/she chooses. However, it is better to only use valves for dynamic cases, as this will improve model run time.

Static losses (such as elbows, orifices, filters, and valves) can be included as Fittings & Losses in pipes.

Variable resistance When experiencing waterhammer transients, the flowrate in a pipe will vary for each pipe computing station at an instant in time, and all pipe stations will vary over time. As the flowrate varies, so does the Reynolds number. This means that, strictly speaking, the pipe friction factor will change as the flowrate changes, and will also change from station to station.

This change in friction factor is typically neglected for two reasons. First, it complicates the computation method. And second, it is usually not important.

AFT Impulse allows the user to model either constant friction factor or variable. The variable friction factor model results in model runtimes



roughly three times as long. The choice is available in the Transient Control window in the Use Variable Pipe Resistance field.

Introduction to wavespeed and waterhammer phenomenon

Description of waterhammer

Waterhammer is a phenomenon that occurs in all liquid piping systems whenever some event disturbs the steady state. Rapid disturbances can cause large transient pressures that in extreme cases can cause a catastrophic failure of the piping system and/or damage to associated equipment.

For more slowly occurring disturbances, the transient pressures may be small or even unnoticeable. In such cases it is sometimes said that waterhammer doesn’t exist. What is meant by such statements is that large, potentially destructive waterhammer doesn’t exist. Regardless of the magnitude, waterhammer always exists when the liquid velocity in a piping system is changed. In cases of small velocity changes waterhammer transients are small and are typically and justifiably ignored by engineers.

The disturbances consist of coupled pressure and velocity waves that propagate throughout a pipe system. Across these waves a discontinuity in magnitude exists. An example is given later in this section to clarify.

To picture what happens during a waterhammer event, consider a 1220 meter long pipeline flowing water in a steady state condition. Any disturbance in the fluid will propagate near the acoustic speed of the liquid in both the upstream and downstream directions.

It turns out that the propagation speed is influenced by the piping material, wall thickness, diameter and structural support method. These other factors combine into a parameter called the wavespeed, which is traditionally referred to as a. The wavespeed is thus analogous to the acoustic speed, but combines the effect of pipe interaction. Although the wavespeed varies from case to case, a nominal value of 1220 m/sec is reasonable for water.

Now picture a transient event in this pipe, specifically that the valve is suddenly closed at the downstream end (see Wylie, et al., 1993, pp. 7).


8/10/2007 4:10:28 PM Page 65


Assume the upstream supply is a reservoir, which acts as a fixed pressure.

When the valve closes this fact is communicated up the pipe to the supply, telling it to stop supplying fluid because fluid is no longer needed. This communication takes a finite time, and this time is in fact related to the wavespeed. In our system it will take one full second for the disturbance downstream to be communicated to the upstream end because the wavespeed is 1220 m/sec and the pipe is 1220 meters long. During this second, the flow continues because it has not yet received the message to stop.

Once the message reaches the upstream supply, however, it takes another full second for the supply’s response to reach the valve. There is thus a full two second period in which the pipe disturbance communicates upstream and the response comes back.

Instantaneous waterhammer

With few exceptions, it is safe to calculate the maximum possible waterhammer pressure surge by using the instantaneous waterhammer equation. The instantaneous waterhammer equation assumes that the transient event occurs either instantaneously or rapidly enough such that it is in effect instantaneous.

In such a case, it can be shown (Wylie, et al., 1993, pp. 4) by use of the mass and momentum equation that the pressure transient is given by the following equation:

VaP Δ−=Δ ρ (3.1)

where:

ΔP = pressure surge

ρ = density

a = wavespeed

ΔV = velocity change

By adding the pressure surge to the existing static pressure, one can obtain the maximum theoretical pressure in the pipe. However, in some cases the pressure can exceed the instantaneous prediction. These are discussed in Chapter 9.



Here’s an example. Assume flow in a pipe filled with water is instantaneously stopped by a valve closure. Assume the density is 1000 kg/m3, the wavespeed is 1220 meters per second, and the initial velocity is 3 meters per second. Also assume that before the transient the static pressure is 350 kPa.

kPa 4010

3503660

kPa 350s

m3*

s

m1220*

m

kg1000

max

max

3max

=+=

+=

P

P

P

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:

a

Lt 2=Δ (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


8/10/2007 4:10:28 PM Page 67


flowing at a steady state. At time zero the valve is instantaneously closed. We will look at four phases of the transient response.

Phase A: 0 < t < L/a The first phase occurs after the valve has closed but before the resulting wave has traveled the pipe length and reached the upstream reservoir (see Figure 3.1). The wave, moving at speed a, travels from the valve to the reservoir (see Figure 3.1, Phase A – the wave initially travels from right to left). The fluid upstream of the wave front (on the left) does not yet know of the closed valve, and thus remains at the steady-state pressure and velocity. Thus before the wave reaches the inlet, fluid continues to flow into the pipe at velocity Vsteady even though the valve is closed.

The portion of the pipe to the right, where the wave has already passed, knows of the closed valve and thus the motion has stopped (the velocity is zero). The kinetic energy of the fluid has been converted into potential energy by increasing the fluid pressure and expanding the pipe circumference. Figure 3.1, Phase A, shows the pipe expanded to the right (exaggerated for effect), but at the original diameter on the left. It further shows a plot of the pressure and velocity in the pipe.

Since the valve was closed instantaneously, the magnitude of the pressure wave can be obtained from the instantaneous waterhammer equation (Equation 3.1).

Phase B: L/a < t < 2L/a The second phase occurs after the wave has reflected from the inlet reservoir but before it has come back to the valve. The wave, still moving at speed a, travels from the reservoir to the valve (see Figure 3.1, Phase B). The fluid upstream of the wave front (on the left) now is aware of the closed valve. Further, the high pressure that built up during Phase A means there is excess fluid in the pipe. During Phase B this fluid is expelled to the only place that can receive it – the upstream reservoir. Thus backflow occurs as the pressure returns to the steady-state level.

The fluid downstream of the wave front (on the right) does not yet know the reservoir is available to relieve the pressure, so it remains at the high pressure and zero velocity.



a

V = Vsteady

V = 0

P

ΔPinstantaneousPsteady

xV

Vsteady

x

a

V = Vsteady

V = 0

P

ΔPinstantaneousPsteady

xV

-Vsteady

x

a

V = Vsteady

V = 0

V

-Vsteadyx

x

P

ΔPinstantaneous

Psteady

VVsteady

x

a

V = Vsteady

V = 0

x

P

ΔPinstantaneous

Psteady

Phase A: 0 < t < L / a Phase B: L / a < t < 2L / a

Phase D: 3L / a < t < 4L / aPhase C: 2L / a < t < 3L / a

Figure 3.1 Pressure and velocity in a pipe after instantaneous valve closure. Pipe diameter changes with pressure exaggerated for effect


8/10/2007 4:10:28 PM Page 69


Phase C: 2L/a < t < 3L/a The third phase occurs after the wave has reflected from the inlet reservoir, traveled back to the valve and reflected from the valve, and is now moving back towards the reservoir (see Figure 3.1, Phase C). The wave is still moving at speed a.

Due to the pipe's attempt to establish equilibrium by flowing back towards the reservoir in Phase B to relieve the high pressure, the system has overcompensated. During Phase C the pressure drops below the steady value. The pressure drop is the same magnitude as the pressure rise in Phase A, but opposite in sign.

As the wave passes each location during Phase C, the fluid motion stops and the pressure is reduced.

Phase D: 3L/a < t < 4L/a The fourth phase occurs after the wave has reflected from the inlet reservoir for the second time and is moving back towards the valve (see Figure 3.1, Phase D). The fluid upstream of the wave front (on the left) has returned to the original steady values of pressure and velocity. The fluid downstream of the wave front (on the right) is still at a reduced pressure.

The fourth phase occurs after the wave has reflected from the inlet reservoir for the second time , traveled back to the valve and reflected from the valve, and is now moving back towards the reservoir (see Figure 3.1, Phase C). The wave is still moving at speed a.

Once the wave reaches the valve for the second time, the four phases are repeated. If the pipe is frictionless, the wave motion will continue indefinitely. If friction effects are included the wave will decay over time and motion would eventually stop altogether.


C H A P T E R 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



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.

Graph Results

Visual Report

Workspace Output

Model Data

Figure 4.1 Primary window workflow in AFT Impulse

The Workspace window

The Workspace window is the sole area where you specify the model layout and connectivity. The Workspace window has two separate functional areas: the Toolbox on the far left and the Workspace itself encompassing the rest of the screen (see Figure 4.2).

At the top of the Toolbox (Figure 4.3), a Shortcut button and four tool icons are slightly offset from the rest of the junction icons. These are the Selection Drawing Tool, Pipe Drawing Tool, Zoom Select Tool, and the Annotation Tool.

Chapter 4 The Five Primary Windows 73

8/10/2007 4:10:28 PM Page 73


The Shortcut Button

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

The Selection Drawing Tool

The drawing tool in the upper left corner (and below the Shortcut button) of the Toolbox is the Selection Drawing Tool (see Figure 4.3). This tool is one means of selecting groups of objects on the Workspace for group operations such as dragging or duplication.

To select a group of objects, click the Selection Drawing Tool and draw a box around the desired objects on the Workspace.

If you Double-click the Selection Drawing Tool it remains active until you click it again a single time. This allows you to make a series of selections.

Three keyboard modifiers extend the Selection Drawing Tool's functionality. If you hold down the CTRL key when completing the selection drawing (just before releasing the mouse button), the Selection Drawing Tool remains active. The CTRL key thus performs the same function as double-clicking the tool.

If you first double-click the Selection Drawing Tool or hold down the CTRL key, you can also hold down the SHIFT key while drawing the selection box and the enclosed pipes and junctions will be deselected.

Finally, if you first double-click the Selection Drawing Tool or hold down the CTRL key, you can also hold down the ALT key while drawing the selection box and the enclosed pipes and junctions will toggle. That is, any pipes or junctions that were selected will become deselected, and vice versa.

Selecting objects completely or partially inside the box If you select the objects by dragging left to right, you will select all objects completely inside the box.



If you select the objects by dragging right to left, you will select all objects completely or partially inside the box (if this feature is enabled in the Workspace Preferences window).

Toolbars

Workspace

Toolbox

Minimizedprimary

windows

Status Bar

Figure 4.2 The Workspace window is where you assemble your model

The Pipe Drawing Tool

The tool in the upper right corner (and below the Shortcut button) of the Toolbox is the Pipe Drawing Tool. With this tool you can draw new pipes on the Workspace, starting and ending at any desired location.

To draw a pipe, click the Pipe Drawing Tool. The mouse pointer will change to a crosshair when you move it over the Workspace. Position the cursor at the pipe’s starting point, hold down the left mouse button while dragging the cursor to the pipe endpoint, then release the mouse button.


8/10/2007 4:10:28 PM Page 75


Two keyboard modifiers extend the Pipe Drawing Tool's functionality. If you hold down the SHIFT key down while drawing on the Workspace, you can draw pipes that are perfectly horizontal or vertical. This may improve the aesthetics of your model.

If you hold down the CTRL key when completing the pipe drawing (just before releasing the mouse button), the Pipe Drawing Tool remains active, and you can draw a series of pipes without returning to the Toolbox each time.

Selection Drawing Tool

Branch

Assigned Flow

Area Change

Tee/Wye

Control Valve

Dead End

Surge Tank

Vacuum Breaker Valve

Pump

General Component

Zoom Select Tool

Pipe Drawing Tool

Reservoir

Assigned Pressure

Relief Valve

Valve

Check Valve

Gas Accumulator

Liquid Accumulator

Spray Discharge

Volume Balance

Annotation Tool

Shortcut Button

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 representation. It does not add an “elbow” to the model or the resulting pressure drop effects of an elbow.

Figure 4.4 Pipe with five segments

This feature is useful in graphically bending the pipe around another part of the model or graphically showing that there are embedded elbows as Fittings & Losses.

You can add a segment to a pipe in two ways. The easiest way to add a single segment is to select Pipe Segments and Add from the Arrange menu. If there are no segments, then a single segment can also be added by selecting the Add/Remove Segment from the Toolbar. If the pipe


8/10/2007 4:10:28 PM Page 77


already has one or more segments, this toolbar selection will remove all segments.

A more flexible way of adding or removing segments is to use the Pipe Segment Manager. To open the Pipe Segment Manager, choose Pipe Segments and Manager from the Arrange menu. A window will display (Figure 4.5) where you can add, remove and merge segments.

Figure 4.5 The Pipe Segments window allows you to create, merge and remove pipe segments.

The Zoom Select Tool

The Toolbox tool below the Selection Drawing Tool is the Zoom Select Tool. With this tool you can draw a box that will cause the Workspace to zoom to that box. The maximum zoom size is 200%.



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

The Annotation Tool

The Toolbox tool below the Pipe Drawing Tool is the Annotation Tool. With this tool you can draw a box that will contain an Annotation displayed on the Workspace.

As shown in Figure 4.6, the Annotation Tool allows you to create a text message and, by clicking on the appropriate tab, show an outline around the message and draw lines with pointers to certain areas of the model. Different style, color and font options are provided.

Figure 4.6 The Workspace Annotation tool dialog box.

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


8/10/2007 4:10:28 PM Page 79


Annotations print out with the model. Figure 4.7 shows the Workspace with an Annotation displayed.

The Annotation Manager, available on the View menu, allows you to specify whether an annotation will be shown or hidden, and in which scenarios it should appear.

Figure 4.7 Annotation added to the workspace

Junction icons

Below the four drawing tools are nineteen junction icons. The junctions allow you to model a large variety of pipe system components important in waterhammer. To add a junction to a model, drag the desired icon from the Toolbox and drop it anywhere on the Workspace.

As you move the mouse pointer over the Toolbox, a Tool Tip identifies the type of junction under the mouse pointer.

The types and order of junction icons on the Toolbox can be modified through the Toolbox Preferences window. For more details on how to customize the Toolbox, refer to Chapter 7.



Editing features

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

Selecting groups of objects A group of objects can be selected for manipulation in five ways:

1. By clicking the Selection Drawing Tool and drawing a box around the desired objects

2. By holding down the SHIFT key while clicking each object in the group

3. By selecting a pipe or junction, then choosing Select Flow Path from the Edit menu

4. By choosing Select All from the Edit menu

5. By choosing Select Special from the Edit menu

6. By choosing a group from the Edit menu (Groups -> Select)

When the SHIFT key is depressed, you can click repeatedly on the same object to select and deselect it. Selected junctions are outlined in red; selected pipes are also shown in red. The selection color is configurable in the Workspace Preferences.

The Copy and Paste operations allow you to duplicate a group on the Workspace. When you copy an item, it is placed on the Windows clipboard; from there you can paste it into another Windows graphics program as a picture.

New objects added to the Workspace from the Toolbox are undefined (refer to Chapter 5 for details on defining objects). Objects derived through duplication or copying retain all relevant Specifications window information. If a group of fully defined objects is duplicated or copied, AFT Impulse retains the object status and connectivity within the new group.


8/10/2007 4:10:28 PM Page 81


Selecting objects in a flow path The pipes and/or junctions in a flow path may all be easily selected using the Select Flow Path item on the Edit menu, Toolbar or the Popup menu (shown when the right mouse button is pressed while the cursor is over an empty area on the Workspace). A flow path is a sequence of pipes and junctions where there is no branching or flow sources or sinks. The endpoint junctions can be included in the flow path if the All Objects Including Endpoints option is chosen.

If a junction is selected which has multiple flow paths attached to it (e.g., a branch with 3 or more pipes) then objects in each of the connected flow paths will be selected. This feature is very useful in duplicating or deleting sequences of pipes and junctions. It can also be used to quickly define groups for graphing or for selected Output or Model Data listings.

Aligning objects Tools to align components in your model are located on the Arrange menu.

Creating and Managing Saved Groups

Groups of pipes and junctions can be named for later recall. This is done using the Group Tools on the Edit menu. The easiest way to create a named Group is as follows:

1. First select on the Workspace the pipes and junctions you want in the group

2. Choose Groups -> Create from the Edit menu or toolbar.

3. You will be prompted for a name for the group. Enter a meaningful name and click OK.

4. The Group Manager then opens. The selection of pipes and junctions in the Group will match those currently selected on the Workspace. If you want to modify the selections once more, you can do it here.

You can include Subgroups inside of your Groups. Subgroups are existing Groups. If, for example, you included a Subgroup in your Group, and later change the contents of the Subgroup, the Group which contains the Subgroup would also be changed.



After you have created a Group, you can then quickly reselect that Group of pipes or Junctions on the Workspace by choosing Group -> Select from the Edit menu or Toolbar and choosing the name of the Group.

The Group Manager can be opened by choosing Group -> Manager from the Edit menu or Toolbar. The Group Manager allows you to modify, create, delete or rename Groups.

Another place Groups are available is in the Select Special window.

Special group usage

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

Bookmarks on the Workspace

Bookmarks can be added on the Workspace in relation to individual pipes or junctions. You can quickly navigate around the Workspace by moving from Bookmark to Bookmark.

Use the Bookmark tools on the View menu to set Bookmarks and to display different bookmarked pipes or junctions.

Last View

The last view of the Workspace is always saved in memory and can be displayed by selecting Last View from the View Menu or Toolbar.

Panning the Workspace

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


8/10/2007 4:10:28 PM Page 83


Customizing features

Workspace Preferences The Workspace Preferences on the Options menu lets you customize the appearance and behavior of the Workspace. New defaults can be saved so AFT Impulse will always appear and behave according to your preferences.

See Chapter 7 for more information on Workspace Preferences.

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

Selective display of pipes and junctions Pipes and junctions can be shown or not shown on the Workspace by choosing Display Pipes and Junctions on the View Menu (see Figure 4.8).



Figure 4.8 Pipes and junctions can be selectively displayed on the Workspace using the Display Pipes and Junctions window on the View menu.

Additionally, junctions may be alternatively displayed as an outline or a solid gray box. By selecting the pipe(s) or junction(s) in the lists and pressing the << Shift or Shift >> buttons, an “x” will move left or right indicating how the objects will be displayed on the Workspace. A junction which is shown as an outline or a solid box may still be opened for editing.

AFT Impulse does not limit the number of junction icons you can show at a time.

Scale/flip workspace The Workspace can be scaled horizontally and/or vertically by choosing Scale/Flip Workspace on the Arrange menu (see Figure 4.9). Scaling will move the junctions and will stretch or shrink the pipes so that the area the model covers is changed by the percentage specified.

Flipping the Workspace horizontally and/or vertically will move the pipes and junctions to their mirror image locations. This is useful when


8/10/2007 4:10:28 PM Page 85


you are building a system which has symmetrical or mirrored segments. For example, you could select the segment, duplicate it, and then flip it horizontally, vertically, or both, with Scale/Flip Selected Objects Only checked.

Figure 4.9 The Workspace can be scaled or flipped

Renumbering the Workspace objects Objects on the Workspace can be renumbered in one of three ways: Renumber Automatic, Renumber Wizard and Renumber Increment. By choosing Renumber Automatic from the Edit menu (see Figure 4.10). Selected pipes and/or junctions can be renumbered starting at the lowest selected number or at a specified number. If a pipe or junction which is not selected to be renumbered already has the next number in sequence, it will be changed only if the Force Sequential checkbox is checked.

Impulse will start at the beginning of the flow path containing the lowest numbered selected pipe and will attempt to march down the pipes in the direction of flow. It is recommended that renumbering be performed after the model has been completed so that the junctions and pipes are completely connected.

If you want certain parts of the model to be numbered in sequence, then use the Renumber Automatic feature on the selected parts only. It is also preferable in the case of large or complicated models to renumber small sections at a time.



Figure 4.10 Pipes and junctions can be renumbered using the Renumber Automatic window opened from the Edit menu

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

Incrementing Object Numbers The Renumber Increment feature, initiated from the Edit menu, allows you to add or subtract an increment from select pipes and junctions.

Background Graphic

You can load a graphic into the background of the Workspace and build your model on top of that graphic. The graphic could be, for instance, a topographic map or a facility drawing. Such graphics are loaded from the File menu by choosing Load Background Picture. At your option the background graphic can be included in Workspace printouts.


8/10/2007 4:10:28 PM Page 87


Figure 4.11 The Renumber Wizard allows you to renumber objects by simply clicking on them.

Specifications windows

Each object on the Workspace has an associated Specifications window. The Specifications window lists all the input data for the selected object. There are three ways to open an object’s Specifications window from the Workspace:

1. Double-click the object

2. Select the object and click the Open Pipe/Jct Window icon on the Toolbar

3. Select the object and press ENTER

The Specifications windows can also be accessed through the Model Data window.



Printing and exporting the Workspace

The Workspace image can be:

• Copied to the clipboard

• Saved to a picture file

• Saved directly to an Adobe PDF file

• Sent to the printer

The Print Preview/Special window allows the above outputs to be applied to all or selected Workspace objects.

The Model Data window

The Model Data window offers a text-based perspective on all engineering information in the pipe flow model (see Figure 4.12). This window is immensely useful for obtaining a global view of the model and rapidly checking the input.

The Model Data window does not offer any tools to build or add to a model; all model assembly must occur in the Workspace.

The Model Data window allows you to manipulate existing elements from the Workspace. Therefore, a complete model can be assembled in the Workspace window without ever opening the Model Data window. In fact, for small models this may be the preferred approach.

The Pipes table allows display of all input data for all pipes in the model. Details of fittings and losses are displayed in the Pipe Loss Data table, accessed with an adjacent tab. The Pipe Loss Data table displays all fittings with K factors grouped into columns.


8/10/2007 4:10:28 PM Page 89


Open Model Data Control

General data section

Customize View

Pipe data section

Junction data section

Figure 4.12 The Model Data window provides a convenient summary of model element specifications

The other display tab in the pipe area is the Pipe Detail Summary, which assembles a list of details about each pipe in another format useful to some engineers.

The junction data area is below the pipe data area. Here the junction data is separated into tables for each junction type. Click the folder tabs to see the data for any type of junction.

The transient data for each junction is shown on the Transient Data tab (see Figure 4.12).

To edit information in the Model Data window, you can open the Specifications window for the pipe or junction. This can be done by double-clicking the far left column where the pipe or junction ID numbers are shown. Alternatively, you can use the Global Edit windows opened from the Edit menu.



Display and printing features

The Show Data options on the View menu (and Toolbar) lets you specify the combination of pipe, junction, and general information you want to see in the Model Data window.

Frequently, there are empty columns in the Model Data window tables. These can be hidden by selecting Hide Empty Columns from the Options menu.

The Print Content window lets you specify the content to include in the printed Model Data report and select the font to be used for printing. This window is opened from the View menu (or Toolbar).

The Model Data window contents can be:


• Saved to a formatted text file

• Exported to a delimited file suitable for direct importing into spreadsheet software



The Model Data Control window

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

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

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


8/10/2007 4:10:28 PM Page 91


junction type. Each type also includes the general data which may be selected/deselected. You can select/deselect the general data for all junction types simultaneously by clicking Set Common.

Figure 4.13 The Model Data Control window allows selection of the parameters to show in the Model Data window

Show selected pipes and junctions Specific pipes and junctions can be shown in Model Data. Click the Show Selected Pipes/Junctions tab, then click the items in the lists to select the pipe(s) and junction(s) to be shown.

Scenario Format A powerful Model Data feature when using scenarios is to show all direct ancestor data in the Model Data along with data from the current scenario. The Scenario Format features allow you to do this, and to highlight where data changes occur. See Figure 4.14. Note that the scenario name is appended to the pipe and junction number in the left columns.



Figure 4.14 Model Data display of scenario ancestor data. This functionality is enabled in Model Data Control.

Database connections In the lower left of the Model Data Control window there is a Database checkbox. If this box is checked, your Model Data Control parameters are set up as determined by the database to which you are connected. This is referred to as an active database. To make it inactive, uncheck the box or change one of the Model Data Control settings controlled by the database.

If the checkbox is unchecked, but enabled, you are connected to a database but the settings are not being passed to the Model Data Control window. The database is thus inactive. To make it active, check the box then click the OK button.

If the checkbox is disabled, there is no connected database.


8/10/2007 4:10:28 PM Page 93


See Chapter 7 for an in-depth discussion of how databases are configured and administered through your local or wide area network.

The Output window

The Output window is the primary vehicle for communicating the results of an analysis in text form (see Figure 4.15). The Output window follows the same general window organization as the Model Data window.

When a model is submitted to the Solver, the Solution Progress window appears. After processing is complete, you can view the results in the Output window by clicking the View Output button.

If the information in the Output window is not in the desired format, you can choose Output Control from the Analysis menu (or Toolbar) to modify the format. The Output Control window (see Figure 4.16) lets you specify the parameters, their units, and the order of their display in the Output window (Figure 4.15).

Data in the Output window can be:


• Saved to a formatted text file




The printed and Adobe PDF output follows the same flexible, organized, and accessible format as the Output window itself.

The Show options on the View menu (or Toolbar) lets you enlarge each of the output sections for greater clarity. The Print Content window, available on the View menu (or Toolbar), lets you specify the content to include in the printed Output report and select the font to be used. The Sort window, opened from the Arrange menu (or Toolbar), offers the ability to sort the Pipe Results or Junction Results tables according to the values in any of the columns. The Transfer Results to Initial Guess feature on the Edit menu (and Toolbar) sets all pipe and junction initial



flows and pressures equal to the steady-state results, which causes future steady-state runs to converge much faster.

You can view the input data for individual pipes or junctions directly from the Output window in one of two ways. First, if you press down the right mouse button on the pipe or junction number in the far left column of the table you will see the Inspection window. Second, if you double-click the pipe or junction number you will open the Specifications window in read-only form where you can review input in detail.

Figure 4.15 The Output window displays output in text form

Transient output files

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

Open Output Control

Customize view

General section

Pipe output section

Junction output section


8/10/2007 4:10:28 PM Page 95


specific output file is used for a scenario, the name is given in the General section of the Output window.

The content of the Output file is specified in the Transient Control window. Details are given in Chapter 5, and summarized here. As the Transient Solver progresses, solutions are obtained for all pipe stations.

Each pipe is divided into multiple computing stations. In order to perform the simulation, solutions must be obtained for each station for each time step. When all pipe stations and time steps are considered, the amount of output data can be enormous. Saving all of this data can thus create a huge output file. The Transient Control window allows you to only save selected pipe stations at selected time intervals to the output file. This allows more reasonable output file sizes, while still retaining the data of most interest.

The output file saves the fundamental pressure and flow data that allows all important transient parameters to be determined. However, the user can only review output data that exists in the output file. If data for a pipe station is not saved, the user cannot review data in the Output window or Graph Results window at that location because the data was not saved.

The Output Control window

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

Pipe and Junction parameters The first two tabs control the parameters to be shown for the pipe and junction steady-state results. The list on the left side indicates the parameters not shown but available. This list can be viewed in alphabetized or categorized mode, which makes it easier to find a parameter. Choose a parameter to be shown by selecting it on the left-hand list and clicking the Add >> button. The selected item will move to the list on the right which shows the order of the parameters to be displayed in the output and the units (if any) in which to display the results.

You can change the order of the output by selecting a parameter on the right-hand list, then clicking the Reorder scroll bar on the far right either



up or down. The list will be reordered accordingly. The output is presented in the order defined in this list.

Figure 4.16 The Output Control window selects parameters for the Output window. Here the Pipe parameters (for steady- flow results) are shown.

The units used for a parameter can be changed by selecting the parameter then choosing the desired units from the dropdown list below the list of parameters. By selecting the units of interest, you can obtain your results in whatever units you find most convenient and meaningful. This means that you can enter all your input parameters in one set of units (or a variety of units, for that matter), and have all your output parameters in a completely different set of units. This feature has obvious benefits for international applications.

To set selected units for both pipes, junctions and summary reports to the preferred units (set in Parameter and Unit Preferences), click the Use Preferred Units button.

You will note that some output parameters do not offer the ability to specify output units. These parameters either do not have units associated with them (such as Reynolds Number or friction factor), or are in fact echoes of input parameters (such as Pipe Nominal Size).

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


8/10/2007 4:10:28 PM Page 97


Pipe Transient parameters The third and fourth tabs control the parameters to be shown for the pipe and junction transient results. The list on the left side indicates the parameters not shown but available. Choose a parameter to be shown by selecting it on the left-hand list and clicking the Add >> button. The selected item will move to the list on the right which shows the order of the parameters to be displayed in the output and the units (if any) in which to display the results.

Reordering the parameters and changing the units works similarly to the steady-state methods described previously.

General output The General section is the text area located at the top of the Output window. The project title, reference information and other settings affecting the General Output are defined here.

You can enter a descriptive title which will be used in the Model Data, Output and Visual Report windows. A title is required and can have up to 100 characters.

You can keep a lengthy explanation or any other documentation about your model in the Reference Information section. Names of projects, individuals, and assumptions can all be kept with the model. This information can then be included in your reports to improve tracking.

The checkboxes allow you to enable certain special reports to be shown in the General Section.

Summaries Several Special Summary Reports can be displayed in the Output window. These include a Pump Summary and Valve Summary. The contents of these summary reports and the engineering units can be specified in the Summaries list at the right.



Format and action The results can be automatically sent to the printer or saved to a file when the model has reached convergence. Additionally, the results can be automatically saved as initial conditions by checking the Transfer Results to Initial When Done option. The model can also be automatically saved after the initial conditions are transferred. The valve state (open, closed or failed) can also be saved along with the initial conditions.

When enabled, the Wrap Table Column headers option makes the output tables wrap the column headers an extra line if the header is wider than usual. There are four formatting methods for the pipe and junction output data:

1. Align on the decimal point

2. Show only the specified number of digits

3. Align only when the differences in magnitudes are less than a user-set limit

4. Use exponential notation with the specified number of digits when the differences in the magnitudes are greater than a user-set limit

The first method allows the magnitude of the data to be easily compared as the data is read down the column. However, a number whose magnitude is small may force larger numbers to show many digits. For example, assume that the minimum number of digits is set to four, and there are two output values: 1234.5 and .09876. These two values will be displayed in the output as:

1234.50000

0.09876

The second method will force the data to use exactly the number of digits specified, which may result in the decimal point location changing from value to value. Using this method with the above example and specifying four digits, the output will be displayed as:

1235.

0.09876

The third method will align the data on the decimal point only if the maximum value in the column divided by the minimum value is less than a limiting order of magnitude that you set. If this ratio is greater than the limit then each value in the column will be exactly the number of digits


8/10/2007 4:10:28 PM Page 99


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



3. Click the Workspace button for both pipes and junctions

4. Click the Output Control window OK button

Database connections 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.

If the checkbox is unchecked, but enabled, you are connected to a database but the settings are not being passed to the Output Control window. The database is thus inactive. To make it active, check the box then click the OK button.


See Chapter 7 for an in-depth discussion of how databases are configured and administered through your local or wide area network.

Command buttons There are eight buttons at the bottom of the Output Control window. Impulse has built-in default parameters, units and settings which you can choose by clicking the Impulse Default button. You can also develop your own settings, tailored to your project or industry, and have these used by default (instead of Impulse’s defaults). To make your own default, first select the output parameters, units and settings you would like to use then click the Set As Default button. Your settings will be saved and will be used each time any new project is initiated. If you make changes to the settings, and want to get back to your defaults, click the User Default button. The saved default settings are updated only when you click Set As Default.

You can save the output settings (except for the selected pipes/junctions and junction deltas) to a file by pressing the Save Control Format button and entering a file name. These settings are loaded again by pressing the Load Control Format button and choosing the file name. For example, you may have a final report format that is always desired. You can load this format before generating final results. If you have another format you use for reviewing model accuracy, you may want a larger number of parameters in the output.


8/10/2007 4:10:28 PM Page 101


The format files you create can be placed on a network for sharing among a group or company, or incorporated into a company-wide database, allowing standardized reporting.

If you have made changes you don’t want to keep, click the Cancel button. Click OK to use the settings you have specified.

Output parameter descriptions At the bottom of the Pipe, Junction, Pipe Transient and Junction Transient tabs a description is given for the selected parameter. There are three lines shown.

First is the Output Window table header. This is the text that is shown in the headers of the pipe and junction tables on the Output window.

Second is the Visual Report Abbreviation. This is the text shown on the Visual Report window when you display an output parameter. This text is very short so as to save space.

Third is the description of the term.

Steady flow results

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

Transient flow results for pipes

You can review the transient results for any pipe station for any time which was saved to file as specified in the Transient Control window. These results are shown in the Output window Transient Output table (Figure 4.18)

Figure 4.18 shows the transient flow results at the initial time step (which is equal to zero here). The columns of data shown in Figure 4.18 are specified in the Output Control window.

Note that the pipe stations shown are those saved to the output file.



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

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


8/10/2007 4:10:28 PM Page 103


Time step for currently viewed

results

Time step selection tool slides up and

down

Buttons to increment or

decrement currently

displayed time

Figure 4.18 The Transient Output table in the Output window shows output results for pipe stations and at specified times.

Summary Max/Min results There is a summary version of the Transient Max/Min table. The version shown in Figure 4.19 is what is called the Detailed version. You can switch between detailed and summary views by selecting the option on the Output Control "Format & Action" tab (see Figure 4.20).

The summary version shows only maximum and minimum values for the entire pipe, as opposed to each pipe station. The summary also shows the pipe station that is the overall maximum or minimum for the pipe. Therefore, for each Output Control transient parameter there are six



columns in the summary version of the Transient Max/Min table. See Figure 4.21 for an example.

Figure 4.19 The Transient Max/Min table details the maximum and minimum values for the entire simulation.

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

Output window updates

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


8/10/2007 4:10:28 PM Page 105


forgetting to rerun it, and mistaking the previous Output window for the current model's results.

Figure 4.20 Detailed and Summary versions are available for the Transient Max/Min table

Figure 4.21 Summary version of the Transient Max/Min table



The Graph Results window

The Graph Results window, shown in Figure 4.22, allows you to generate high quality printed graphs.

The Select Graph Data window, opened from the View menu (and Toolbar), lets you control the content of the graph and offers several types of standard graphs. The Customize Graph window, opened from the Options menu (and Toolbar), offers you control of the graph's general style. Fonts and colors can be selected.

The graphical data can be exported to a file for later import as a cross plot against results for a different model. For example, let's say you would like to see the effect on the system pipe pressure drop when you change a valve loss factor. You could run the Solver for one value of K and export the results to a file. Then you could rerun the Solver for a new value of K and import the results from the previous case for cross plotting. These features are accessed by choosing Save Graph Data and Import Graph Data from the File menu.

Figure 4.22 The Graph Results window offers full-featured plot generation


8/10/2007 4:10:28 PM Page 107


Graph “Sets” can be created and saved by the user using the tools in the Select Graph Data window.

The Graph Results window is always erased when you make changes to model input. Rerun the model to graph the new results.

The Select Graph Data window

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

Figure 4.23 The Select Graph Data window controls the Graph Results window

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



pipe is expanded, the pipe stations for which data was saved are shown. These pipe stations can be added to the list on the right.

The user can plot the data for the entire simulation, or only for some specified time frame. Further, if Design Alerts are specified for the pipe (e.g., max/min operating pressures), these can be cross-plotted against the transient data.

Transient junction results Similar to transient pipe results, transient junction results can be plotted vs. time for any junction that is saved to the output file. Junction data includes items such as pump speed and accumulator volume. Similar to pipe output, these are specified in the Transient Control window.

Profile Along a Flow Path and EGL, HGL and Elevation Profile These two types of graphs will plot the selected parameter along a flow path. The pipes that comprise the path must form a single, continuous path. The independent variable is the length along the pipe(s).

A path of pipes can be specified in one of two ways. First, the user can select the pipes in the displayed list. This permits only a single path to be specified.

Second, one or more flow paths can be specified using groups (see Figure 4.24). Only groups which consist of a pipe sequence will be displayed. Groups are set up on the Workspace using the Groups tool on the Edit menu.

The results can be the values at a particular point in time, or the overall maximum and/or minimum values. The Profile Along a Flow Path graph type also allows cross-plotting of Design Alerts such as max/min allowed pressures.

Both of these graph types support animation, discussed in the next section.


8/10/2007 4:10:28 PM Page 109


Figure 4.24 Multiple paths can be cross-plotted using predefined groups.

Animating profile results There are two animation options: Animate Using Output File, and Animate Using Solver.

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



As its name implies, the Animate Using Output File option obtains the data for animation by reading in data saved to the transient output file and then displaying it in the graph.

The option to Animate Using Solver does not use the transient output file at all. Instead, it actually reruns the Transient Solver in the background to regenerate the transient data. This rerun does not write any output to file. Because the model is being rerun, all data for all pipe stations is available for each time step.

Figure 4.25 shows the Graph Results window when using animation. As shown, additional controls become visible to allow the user to adjust animation. At the left is the Play button, which starts the animation. Next to it are the Pause and Stop buttons. If the Pause button is pushed, the animation stops at that time step. One can print the graph or copy the data if desired or, if using the Animate Using Output File option, use the time slider controls (to the right of the Stop button) to set the time forward or backwards. If using the Animate Using Solver, the solver can be made to march one step at a time when one clicks the “+” button. The Stop button stops the transient and resets the time to zero.

The current time (in seconds) and time step are shown to the right of the time controls. At the far right are the speed controls. Slowing the animation involves adding time delays to the data display. Increasing the speed involves skipping time steps. The default speed, right in the middle, has no delays or skips.

If plotting a single path using the list of pipes in Select Graph Data, one can cross-plot on the animation design alerts (in Profile Along a Flow Path) or steady-state results (in EGL, HGL and Elevation Profile).

Each of these options has advantages and disadvantages. The Animate Using Output File option is usually a little faster. It also allows you to use the time slider to reset the time to any time step forwards or backwards. The reason being that the transient output file has all of the required data over time, and thus obtaining data for any other time is just a matter of reading the data from the proper place in the output file. The Animate Using Solver does not have this flexibility, since it can only obtain the data by marching through all of the time steps.


8/10/2007 4:10:28 PM Page 111


Figure 4.25 The animation options display additional controls

Another advantage of the Animate Using Output File option is that it can be made to run much faster. When the animation speed control is set to increase speed, intermediate time steps are skipped and not plotted. If the speed is increased too much, the quality of the animation deteriorates because too much missing time data makes it difficult to follow what is happening. When the speed is increased, the Animate Using Output File option just skips time steps and reads in the data for each required time step.

On the other hand, the Animate Using Solver option cannot skip calculation steps because it is running the Transient Solver. To obtain the results for a future time step, it must calculate all of the intermediate time steps.

The biggest advantage of the Animate Using Solver option is that it does not require one to save all of the required pipe station data to disk in order run an animation. Because it is obtaining data from the Transient Solver, it can always run any animation desired. If one uses the Animate Using Output File option and has not saved all of the required data, the



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

Transient force results If transient force sets have been defined in the Transient Control window (see Transient Control section in Chapter 5), the resulting force data can be plotted vs. time. Figure 4.26 shows the Select Graph Data window when plotting transient force data.

The differential pressure forces for each force set selected from the list of defined force sets will be plotted. If desired, the user may select to include momentum and friction effects in the force balance calculation, as well.

The differential force data can be plotted for the entire transient simulation, or the user has the option of selecting a specific time frame to display.

The options to include pipe friction and momentum in the force balance calculation are discussed in Chapter 5. Force data can be exported to a file for use with CAESAR II pipe stress software, which is also discussed in Chapter 5.

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

The Customize Graph window

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


8/10/2007 4:10:28 PM Page 113


1. Fonts for titles, numbers, and legend

2. Colors for the graph background, foreground, and data sets

3. Line types and thicknesses

4. Symbol shapes

5. Normal, semi-log, or log-log plots

6. Borders and gridlines

7. Printing preferences (color or monochrome)

You can save the settings you set up by clicking the Save button on the System tab. If you make changes to the settings, and want to get back to previous settings, click the Load button on the System tab and load the file.

Figure 4.26 Transient Force data can be plotted using force sets defined in the Transient Control window



The Auxiliary Graph Formatting window

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

Printing and exporting the Graph Results image and data

The Graph Results image can be:





The x-y data in the Graph Results graph can be:



The Visual Report window

The Visual Report will help you visualize the model data and steady-state and transient solutions in relationship to the model layout. The Visual Report window displays the pipe system layout from the Workspace window with information in text form (see Figure 4.27). Additionally, magnitudes of steady-state and transient parameters may also be shown with color variations.

The values to be shown may come from the input data (when in input mode), Steady-State Output results, or Transient Output results. You can change between these by selecting the option at the top of the Visual Report Control window.

In the Visual Report window, the positions of the junctions and pipes are fixed based on their locations on the Workspace. While the pipe system layout cannot be edited in the Visual Report window, the integrated text


8/10/2007 4:10:28 PM Page 115


results for each pipe or junction can be rearranged by dragging and dropping to improve the final appearance of the window.

The contents of the Visual Report window can be:





The View menu allows you to zoom in and out at ratios of 25%, 50%, and 75% of the normal size, as well as choose custom sizes. This feature compresses more results onto a single page. The printed size is consistent with the zoom state. That is, if you zoom out to 50%, the Visual Report screen prints at 50% of its normal size. The Custom Zoom feature allows you to specify the exact zoom value you desire. There also is a Zoom to Fit selection, which zooms the model so that one can see the entire model.

The Visual Report window is always erased when you make changes to model input. If viewing output results, rerun the model to view the new results.

Figure 4.27 Visual Report window integrates input data and output results with model layout



The Visual Report Control window

The Visual Report Control lets you control the content and appearance of the Visual Report window (see Figure 4.28). You can also save custom Visual Report window layouts to disk for later recall, so you don't have to reformat this window repeatedly.

When the Visual Report window is opened and no output results are currently displayed, Visual Report Control will be in the Input Display Mode. If output results are displayed, it will be opened in the Steady Output or Transient Output Display Mode. Parameters available for display vary depending upon the mode, i.e. input, steady output or transient output. If output results exist, the display mode may be changed at any time.

Figure 4.28 The Visual Report Control window allows you to configure the Visual Report display


8/10/2007 4:10:28 PM Page 117


Display parameters The parameters to be shown in the Visual Report window are selected from the Display Parameters tab. The left list selects the pipe parameters and the right list selects the junction parameters. In these lists you can show only the parameters you have selected in the Output Control window, or all output parameters. You can select more than one parameter; however, the Visual Report window may get crowded if too many are selected.

General display The General Display tab allows you to configure several other items.

The title, defined in the Output Control window, and filename can be shown or not shown depending on your preference. The scenario name can also be shown.

Figure 4.29 The Visual Report with the units and Color Map in the Legend

The units for all parameters can be shown in a legend instead of next to the parameter value (see Figure 4.27 or 4.28). This will reduce the amount of text and crowding shown in the Visual Report window. The legend can be moved to any location in the Visual Report window. If this



option is chosen while in input-only mode, all values will be converted to a consistent set of units.

You can change the font for pipe and junction text by clicking the Fonts button and choosing the font desired or by using the font size change buttons on the Toolbar.

The number of digits to show in the numerical display can be configured using the provided drop-down list.

Normally on the Workspace, closed objects are displayed using special graphics (dashed lines, X’s, etc.). The number of a junction which has special conditions set, by default, has an X preceding the J (e.g. XJ56). If you want these symbols to be used in the Visual Report window also, choose the Show Closed Symbols checkbox. Normally, pipes in closed sections are shown as dashed lines, and junctions outlined with a dashed line. If you want these symbols to be used in the Visual Report window also, choose the Use Closed Object Style checkbox.

If a background picture exists on the Workspace, it can be shown in the Visual Report by selecting the checkbox provided. Similarly, if a grid is shown on the Workspace it can be shown on the Visual Report.

Show selected pipes and junctions You can choose, on an individual pipe and junction basis, what you would like to show in the Visual Report Control window (see Figure 4.30). This display customization feature is useful for large models where only some of the data points are of interest. For simplicity, this discussion will focus on just the pipes; however, the junctions behave in the same fashion.

Each pipe is listed with a set of x’s to the left of the name. The x’s indicate what will be shown. To change the settings, first choose the pipe(s) to change. Then use the checkboxes on top to set or clear the x’s. If more than one pipe is selected and the current options are different, the checkbox will look grayed and the first click will clear the checkbox and the second click will set it. You can have any combination of settings, including no settings at all. To restore the settings to what is shown in the Workspace and the data, click the Workspace Settings button.


8/10/2007 4:10:28 PM Page 119


Color map The Color Map feature allows you to visualize magnitudes of various parameters by coloring the pipes (see Figures 4.28 and 4.30). There is one Color Map for the steady output mode, one for transient output mode, and another for the input-only mode. A Color Map is useful to show graphically where, for example, high velocities or reverse flows are located or to visualize the pipe diameters or lengths in a model.

To create a Color Map, first choose a parameter and units for which all pipes will be categorized. Then choose a color, select the operator (i.e. >, >=, =, <=, <) and enter a value, then press Add to Map. Values can be categorized on an absolute value basis by selecting this option.

Figure 4.30 The Visual Report Control window allows you to individually select the object, name, number and data to be shown in the Visual Report window.

The order of entry is not important. Impulse will sort the entries first by value and then by the operator, with the > being above the <. The pipes



will be colored using the first match found in the map, starting from the top of the map. If no match is found, the pipe will be the Workspace pipe color. Be careful to avoid overlapping operators and values. If, for example, the map had the following four colors,

1) >=60

2) <60

3) >=20

4) <20

then there is no way colors 3 and 4 will be used because any pipe less than 20 is also less than 60. Since <60 is before <20 in the map, the pipe will be colored using 2.

Figure 4.31 The Visual Report Control Color Map allows you to specify the display colors for pipes according to their output values.


8/10/2007 4:10:28 PM Page 121


To avoid overlapping colors, only use less than (<) as the last entry. A better map would be to use

1) >=60

2) >=20

3) <20.

You can choose to categorize the pipes based on their absolute values by choosing the Display Absolute Value checkbox. To delete a map value, click the color you want to delete then click the Remove Setting button.

Instead of coloring the pipes, you could choose to color just the pipe numbers. This is done by choosing the Color Pipe Labels and clearing the Color Pipes checkboxes in the Color Options area. You could also apply the colorization to both. Choose the Print Black & White checkbox if you don’t have a color printer. The pipes will be printed in black.

The color map values can be displayed in a legend by checking the Display in Legend checkbox (see Figure 4.31). The color map will appear in the legend after the pipe and junction units, if that option was selected. The legend may be dragged and dropped to any location in the Visual Report window.

Command buttons You can save the output settings (except for the selected pipes/junctions) to a file by clicking the Save Options button and entering a file name. These setting are loaded again by pressing the Load Options and choosing the file name. This file may be shared among engineers or incorporated into a company-wide database.

Impulse will initially try to place the text for pipes and junctions based on their positions on the Workspace. You can drag the text to new locations which will give the best visual presentation. These positions are saved with the model. You can reset the locations back to where Impulse initially placed the text by clicking the Reset Locations button and then confirming the operation.

If you have made changes which you don’t want to keep, click the Cancel button. Click Show to apply your settings.



Visual Report annotations

Annotations for the Workspace Annotations created on the Workspace can be displayed on the Visual Report by opening the Workspace Annotation window (double-click the annotation) and selecting the Also Show on Visual Report on the Scenario Options tab.

Workspace annotations displayed on the Visual Report can be moved on the Visual Report but not edited.

Creating Visual Report annotations An annotation can be created specifically for the Visual Report by choosing Create Annotation from the Edit menu or Visual Report Toolbar. When the Create Annotation tool is activated, you can draw an annotation on the Visual Report just like on the Workspace.

Visual Report annotations have all the properties of a Workspace annotation, except they can be displayed only on the Visual Report.

The Toolbars

The Toolbars offer quick access to the features used most frequently in AFT Impulse. Figure 4.32 shows the Toolbars. There are actually six toolbars in all. One is the Common Toolbar, and it is always shown. Its features are common to all five Primary windows. In addition, there is one Toolbar for each Primary window, offering features appropriate for work in that window. Each of the functions on the Toolbar is also available through the menu system.


8/10/2007 4:10:28 PM Page 123


Figure 4.32 The Toolbars allow you to quickly access important features. There is one common Toolbar and a special Toolbar for each of the five Primary windows. You see only two toolbars at a time in AFT Impulse. The function of each tool is indicated in the ToolTips that appear when you hold the cursor over the button.


C H A P T E R 5

Building and Running Models

This chapter discusses the basics of building and running a waterhammer model with AFT Impulse. AFT Impulse assumes that the user has a good general understanding of engineering pipe hydraulics and a basic idea of what waterhammer is.

Creating objects

The pipe flow model is assembled on the Workspace by arranging graphical objects (pipes and junctions) that represent parts of a physical pipe flow system. In addition to pipe and junction objects, annotation objects can also be added.

Pipes

To create a pipe, select the Pipe Drawing Tool from the Toolbox and draw a line on the Workspace to represent the pipe. To draw multiple pipes more quickly, hold down the CTRL key while drawing each pipe. This keeps the Pipe Drawing Tool active, so you don't have to select it for each pipe. Alternatively, you can double-click the Pipe Drawing Tool to “lock” it down. You can then continue to draw pipes until you click the Pipe Drawing Tool a second time.

To draw a vertical or horizontal pipe, hold down the SHIFT key while drawing the pipe.



The length of the line on the Workspace has no relationship to the physical length of the pipe. The Workspace pipe is merely an abstract representation of a pipe.

If a pipe is drawn as an extremely short line, AFT Impulse eliminates the line immediately after it is drawn.

Another way to create new pipes is to duplicate existing pipes with the Duplicate or the Copy and Paste features found on the Edit menu. When a new pipe is derived from a previous pipe, the new pipe retains the engineering information associated with the original pipe.

When a new pipe is created it is assigned a default ID number. When the pipe is displayed on the Workspace, the ID number is preceded by the letter “P” to indicate that it is for a pipe. An arrow on the pipe indicates the reference positive flow direction for the pipe. From the Pipe Specifications window you can change the ID number to any value greater than zero and up to 30,000. The reference positive flow direction can be reversed by choosing Reverse Direction from the Edit menu or by selecting the reverse direction button on the Toolbar.

Tip: The pipe or junction ID number text can be dragged to a new location to improve visibility. The new location is saved with the model. If you want the text returned to its default location, double-click it. If you want all pipe and junction text returned to the default locations, choose Reset Label Locations from the Arrange menu.

Each pipe must be connected to two junctions. There are no exceptions to this rule.

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

Chapter 5 Building and Running Models 127

8/10/2007 4:10:28 PM Page 127


Figure 5.1 Pipe with five segments

This feature is useful in graphically bending the pipe around another part of the model or graphically showing that there are embedded elbows as Fittings & Losses.

You can add a segment to a pipe in two ways. The easiest way to add a single segment is to select Pipe Segments and Add from the Arrange menu. If there are no segments, then a single segment can also be added by selecting the Add/Remove Segment from the Toolbar. If the pipe already has one or more segments, this toolbar selection will remove all segments.

The other way to segment a pipe is to select the pipe on the Workspace then choose the Pipe Segments and Manager from the Arrange menu. A window will display where you can add, remove and merge segments. A “handle” will appear in the middle of the pipe for each new segment. As the cursor passes over the handle, the cursor will change to a crosshair. Press and hold down the left mouse button and drag the middle of the pipe to the desired location. By holding the SHIFT key the pipe segment will move only at right angles.



Junctions

To place a junction, drag the icon from the Toolbox onto the Workspace. The nineteen available junctions represent various pipe system components.

When you duplicate a junction using the Duplicate or Copy and Paste features on the Edit menu, the new junction retains the engineering information associated with the original junction.

Junctions are shown on the Workspace with default ID numbers, which you can change in the Junction Specifications window. The letter “J” precedes the ID number to signify that it is for a junction. From the junction’s Specifications window you can change the ID number to any desired value greater than zero and up to 30,000. Identical ID numbers can be assigned to both a pipe and a junction because the “P” or “J” will distinguish the two.

Unlike pipes, junctions do not have reference positive flow directions. Junctions for which the flow direction is important (such as pumps and control valves) derive the flow direction from the flow direction of the connecting pipes.

Morphing junctions A junction can be “morphed” from one type to another. To morph a junction, hold down the CTRL key and select a junction from the Toolbox and drop it onto an existing Workspace junction. The junction type will change to the new junction type, and data that can be kept will be copied into the new junction.

Splitting pipes If you hold down the SHIFT key while selecting a junction from the Toolbox and then drop the junction onto an existing pipe, the pipe will split into two pipes. The physical length of the original pipe will be automatically halved, and the new pipe will be assigned the balance. Thus the sum of the two pipe lengths will equal the length of the original pipe. Any fittings/losses in the original pipe will be left in the original pipe, and the new pipe will have no fittings/losses.


8/10/2007 4:10:28 PM Page 129


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



• 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.


8/10/2007 4:10:28 PM Page 131


AFT Impulse will attempt to automatically connect pipes and junctions when they are moved. This eliminates unnecessary opening of the pipe and junction Specifications windows. If Impulse cannot make the connection automatically, then the following three steps may be required to establish a connection between a junction and a pipe:

1. Graphically connect the objects on the Workspace. The pipe endpoint must terminate within the boundaries of a junction icon. Figure 5.2 shows a graphical connection between a pipe and a junction.

2. After the pipe is graphically connected to two junctions, Double-click the pipe to open its Specifications window, then click OK to accept the connected junctions as determined by AFT Impulse. It is most efficient to do this when you are entering the pipe's data in its Specifications window.

3. Open the Specifications window of the corresponding junction and accept the pipe connectivity determination. You can do this as you enter the junction's data in its Specifications window.

Figure 5.2 A pipe and junction graphical connection



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


8/10/2007 4:10:28 PM Page 133


6. Double-click the connected pipe or junction number in a Specifications window

When editing in a Pipe Specifications window, pressing the F5 function key jumps automatically to the next higher numbered pipe. Pressing F5 while holding the CTRL key jumps to the next lower number. Junction windows behave similarly.

Parameters that functionally describe the pipe or junction are entered in the Specifications window. Some parameters in the Specifications windows are required while some are optional. The optional parameters offer you added control over your model and the ability to obtain special information.

Highlighting required information

You can toggle the highlight feature on and off in any of the following ways:

• Double-click anywhere in the specifications window

• Press the F2 function key when a specifications window is open

• Choose “Highlight in Pipe and Jct Windows” from the Options menu

The highlighting feature may be especially useful when you are first learning to use AFT Impulse or when you are having difficulty obtaining a defined object status.

A more detailed discussion of the contents of pipe and junction Specifications windows is given in Chapter 6.

Use status feature

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

Undefined objects window

You can open the Undefined Objects window to view all undefined objects and the undefined properties. This is available on the View menu.



Satisfying connectivity requirements

In addition to entering the required information for each object in the model, you must satisfy these connectivity requirements:

• You must accept the connectivity information shown in each object's Specifications window (refer to “Connecting objects” earlier in this chapter).

• Each junction must have the appropriate number of pipes connected to it. For example, an area change junction must have exactly two connecting pipes, whereas a reservoir may have one to twenty-five connecting pipes. (Details on each of the junctions are given in Chapter 6 and summarized in Table 6.1.)

• Pipes must always be connected to two junctions, one at each pipe endpoint.

Should you ever delete any of the required information in the Specifications window or move a connected object from its connection, the object's status reverts to undefined.

For example, say you place a pump on the Workspace. A pump requires inlet and outlet piping (unless it is a submerged pump). AFT Impulse reflects this by requiring each pump to be assigned an upstream pipe and a downstream pipe. Entering only the pump constants and setting an elevation is necessary but not sufficient information to define the pump according to AFT Impulse's requirements. Once you have entered the pump data and properly connected it, the pump is defined.

Inspecting objects

The Inspection feature is a time saving feature for reviewing the data associated with a Workspace pipe or junction object. Inspection displays the data for an object in read-only format (see Figure 5.3).


8/10/2007 4:10:28 PM Page 135


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.

Using the Checklist

Assembling a model that is ready to submit to the Solver is not, at first glance, an obvious process. The new user will wonder where to begin, and even the experienced user will at times lose track of what information has already been entered and what information remains to be entered. To help you in this process, the Checklist window tracks the status of your model and communicates to you where you lack complete information.

The Checklist shows six items (see Figure 5.4). The first four are always required, and the last two are required if it is a transient model. When a requirement has been met, you will see a check mark next to the item in the Checklist window. You can view the Checklist by selecting Checklist from the View menu or clicking the check button on the Toolbar. As a guide to the user, the Checklist items are also shown on the Status Bar at the bottom of the AFT Impulse window (Figure 5.5).

When you have entered the proper information for all six Checklist requirements your model is complete and ready to submit to the Solver. As long as the Checklist remains uncompleted the Model Status Light in the lower left corner is red. If you try to run a model with an incomplete Checklist, you are informed what Checklist items are not completed. The Model Status Light will turn green when the checklist is complete.

Tip: If you double-click the checkbox on the Checklist window or Status Bar, the appropriate window will be opened. For example, double-clicking Specify Output Control will open the Output Control window.


8/10/2007 4:10:28 PM Page 137


Figure 5.4 The six Checklist items help you track the status of your model

Status bar

Status light

Figure 5.5 The Status Bar shows the status of each Checklist item. The Model Status Light turns green when the model is complete.



Analysis Type

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

Specify Steady Solution Control

For a new model, the Checklist shows two items checked and four unchecked. The first item is Steady Solution Control; this is checked because the solution control parameters have been set to default values which do not generally require modification.

AFT Impulse attempts to shield less experienced analysts from the details of numerical convergence by providing robust defaults. More experienced users of numerical modeling software, however, will desire access to the parameters that control the behavior of the Steady-State Solver. The Steady Solution Control parameters can be modified by opening the Steady Solution Control window from the Analysis menu. More information on Steady Solution Control is given later in this chapter and in Chapter 8.

Specify Output Control

The second Checklist item, Output Control, is also checked initially because Output Control default parameters have been specified. However, you will probably want to change the title of your analysis from the default (on the General tab).

The Output Control window allows you to choose the parameters you want included in your output and the units for those parameters. You can also rearrange the output parameters to appear in the order you find most productive. These features are discussed in more detail later in this chapter.

Specify System Properties

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


8/10/2007 4:10:28 PM Page 139


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

Define All Pipes and Junctions

The fourth Checklist item, Define All Pipes and Junctions, is more abstract and cannot be addressed like the first three items by simply opening a window and entering the proper information. This item requires that the proper specifications have been entered and connectivity requirements have been met for all objects in the model. When you have properly defined all Workspace objects, the fourth Checklist item will be completed.

Section Pipes

The fifth Checklist item, Section Pipes, is only used when the Analysis Type is Transient. The details of pipe sectioning are related to the transient solution method, the Method of Characteristics (MOC) discussed in Chapter 9. In summary, the pipe must be broken down into computing sections so the MOC can be used. There are strict criteria on how this is done, and the Section Pipes window automates this process.

The Section Pipes window can only be opened after required pipe input and fluid properties has been entered, as these data impact how the sectioning is performed.

Transient Control

The sixth Checklist item, Transient Control, is only used when the Analysis Type is Transient. Here the model start and stop times are defined, as well as what data to save to the output file and estimates of model run time and output file size.

Steady Solution Control

The solution control parameters you can specify are tolerances for pressure and flow rate balance, two numerical relaxation parameters, and the number of iterations to perform before stopping (see Figure 5.6).



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


8/10/2007 4:10:28 PM Page 141


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



Figure 5.7 The Output Control window selects parameters for Output. The list of available parameters (on the left) can be viewed alphabetically or by category.

System Properties

The System Properties window gives you control over the fluid and environmental properties that influence your pipe system analysis. Figure 5.8 shows the System Properties window, which is accessed through the Analysis menu.

The three fluid properties that are required for all analyses are density, dynamic viscosity and liquid bulk modulus of elasticity. If only a steady-state analysis is being performed, the liquid bulk modulus of elasticity will not be used although input is still required. Fluid vapor pressure is an optional specification.

These properties can be entered by hand when the fluid is selected as Unspecified, or they can be obtained from the database list of AFT Standard or Chempak fluids. When selected from the list, density,

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


8/10/2007 4:10:28 PM Page 143


viscosity and bulk modulus cannot be edited as they depend on the specified temperature (and pressure if Chempak or ASME Steam Tables).

Figure 5.8 The System Properties window is where you choose the fluid(s) for your model. The case above uses an AFT Impulse Standard fluid.

System fluid property variation

The System Properties window allows you to choose from one of two fluid property variation models. The first fluid property variation model is for constant fluid properties. When this model is chosen, all pipes in the system are assigned the fluid properties you specify in the System Properties window.

The second model is for variable properties. In this model, the fluid properties assigned in the System Properties window are used as the default fluid properties for each pipe. You can then assign different fluid properties within each Pipe Specifications window. This allows you to model systems that may be subject to large temperature variations or systems that contain more than one fluid.



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

Density and dynamic viscosity

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

Bulk modulus of elasticity

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

ρρ

ΔΔ= P

K B

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


8/10/2007 4:10:28 PM Page 145


even destroy the pump. In addition to pump cavitation, cavitation may occur downstream of restrictions such as valves and orifices.

The fluid databases

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

The AFT Standard Database The AFT Standard Database offers property data based on temperature or, if entered by the user, solids concentration.

To add a fluid into your model, select it from the list at the top and choose the Add to Model button. The fluid will appear in the lower list of Fluids in Current Model. Enter a temperature and click the Calculate Properties button to display the density, viscosity, bulk modulus and vapor pressure.

The AFT Standard Database supports two different independent variables. Most commonly, the independent variable will be temperature. All fluids provided by AFT are temperature dependent. User created fluids can depend on either temperature or solids concentration. Fluid properties dependent on solids concentration are useful for modeling sludge in pipe systems. The choice of which independent variable to use is offered in the Add New Fluid window opened from the Fluid Database window.

More information on the custom fluids is given in Chapter 7.

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



The Chempak database The optional Chempak database offers the user approximately 700 fluids. In addition, it offers non-reacting mixture calculations.

A Chempak fluid is added the same way as a Standard fluid, by selecting it in the list at the top and choosing the Add to Model button (see Figure 5.9). You also have the option of creating a mixture of Chempak fluids by choosing the Create New Mixture and Add button. This will open the Create Mixture window shown in Figure 5.10. Here you can specify a mixture of fluids in any composition ratio you desire.

Figure 5.9 The System Properties window is where you choose the fluid for your model. The case above uses methanol, a Chempak fluid.

The basis of all mixtures is either mass or mole, and is specified on the System Properties window itself. There are no limits to the number of components in a particular mixture.

The mixtures you create here are referred to as pre-mixtures because you are specifying the composition before the model is run.


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


Once a Chempak mixture has been defined, the mixture data can be exported to a data file which can be imported by other AFT applications or users.

Accuracy option There is one accuracy option for Chempak fluids and mixtures. The State Property Accuracy option influences the number of elements used in curve fits for calculating property data. Typically you should use the high accuracy option.

Figure 5.10 The Create Mixture window (opened from the System Properties window) allows you to create predefined mixtures for the model.

The ASME Steam Tables Database The ASME Steam Tables Database offers users access to water properties obtained from the IAPWS Industrial Formulation 1997 for the



Thermodynamic Properties of Water and Steam (ISPWS-IF97), (see ASME Press in References). This database is selected by choosing the option at the top, at which point “ASME ’97 Water” is automatically displayed in the Fluids in Current Model area.

Viscosity models

The default viscosity model is Newtonian, which applies to a wide variety of important liquids including water. Some liquids exhibit a dependence of viscosity on the fluid dynamics, and these fluids are called non-Newtonian. When one of these model is selected, additional input parameters are required (see Figure 5.11).

AFT Impulse offers four non-Newtonian viscosity models. Two of the models apply to pulp and paper. These are the Duffy model and the Brecht & Heller model. Also available are models for Power Law fluids and Bingham Plastics. See Tilton, 1997 for more information on these viscosity models.

Also available are models for Power Law fluids and Bingham Plastics. See Darby, 2001 for more information on these viscosity models.

More extensive discussion of these viscosity models is given in Chapter 8.

Non-Newtonian flow in non-pipe elements The non-Newtonian viscosity models are designed for calculating pressure loss in pipes. But what about other elements such as valves?

There does not seem to be any standard calculation methods for such elements. AFT Impulse therefore offers an approximate method. At your option, standard K values and/or loss data will be modified by ratioing the provided loss values (assumed to be Newtonian) by the non-Newtonian and Newtonian friction factors. See Chapter 8 for more discussion.

If you wish to use these correction factors, select the appropriate option in the provided fields.


8/10/2007 4:10:28 PM Page 149


Figure 5.11 Non-Newtonian fluid data on System Properties.

Atmospheric pressure

AFT Impulse uses a default atmospheric pressure of 1 standard earth atmosphere, or 101.325 kPa. You are free to change this value. This pressure is also used to convert absolute pressures to gauge pressures. Keep in mind that atmospheric pressure varies with altitude. If you are doing one project at Sea Level and another at 2500 meters you should adjust the Atmospheric Pressure. See Figure 5.12 for the System Data tab.

Gravitational acceleration

The default for gravitational acceleration is 1 standard earth gravitational acceleration, or 9.81 m/s2. You can change this to a multiple of standard earth accelerations (that is, number of g's) according to the design environment for your system. In principle, the body force on a fluid system does not have to be due to gravity. See Figure 5.12 for the System Data tab.



Figure 5.12 System Properties window System Data tab

Transition Reynolds Numbers

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

You can modify these transition Reynolds Numbers if you choose.

Editing the AFT Standard fluid database

AFT Impulse provides a fluid database for numerous common fluids. If the working fluid is not in the database, you may elect to enter the fluid properties in a custom database for later recall.

To enter or modify fluid properties in the fluid list database, click the Edit Fluid List button or select Fluid Database from the Database menu.

The fluids you define are incorporated into the fluid list as if they were native to AFT Impulse. The fluid data is saved to disk and read in when AFT Impulse is loaded. More information on building custom fluid databases into AFT Impulse is given in Chapter 7.


8/10/2007 4:10:28 PM Page 151


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

Finding object definition status

In order to complete the fourth checklist item you must properly define all objects on the Workspace.

The Show Object Status feature helps you identify which objects on the Workspace are defined and which are not. This feature can be toggled on and off by selecting it from the View menu or by clicking the flood light button on the Toolbar.

Turning Show Object Status on displays the undefined junction and pipe ID numbers in the Undefined Color, which is red by default. The color can be changed through the Workspace Preferences window. Show Object Status helps you rapidly assess where your model requires additional information.

Leaving Show Object Status always toggled on is not recommended for large models, as it will slow the Workspace graphics response. When this feature is on, every Workspace graphical operation you perform must be accompanied by a check of all pipe and junction connections. This check takes time, and could degrade your graphics performance.

The most efficient way to use Show Object Status is to turn it on only when you have finished adding and arranging the Workspace objects. If you need to graphically edit your model, it is best to turn Show Object Status off, perform your graphical editing, then turn it back on when you need it.

List Undefined Objects

The Undefined Objects window can be opened from the View menu or Workspace Toolbar. It lists all undefined pipes and junctions and displays the undefined properties as well. See Figure 5.13.



Figure 5.13 The Undefined Objects window displays undefined pipe and junctions and the undefined properties of each.

Section Pipes

In a transient simulation, a common time step must be used for all pipes. The Method of Characteristics (MOC) requires that each pipe section satisfies the following:

t

xa

ΔΔ= (5.1)

where a is the wavespeed in the pipe, Δx is the length of the section and Δt is the time step. The length of a pipe section is just the length of the pipe divided by the number of sections in that pipe:

n

Lx =Δ (5.2)

The time step can be obtained by combining Equations 5.1 and 5.2:


8/10/2007 4:10:28 PM Page 153


na

Lt =Δ (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:

min

max a

Lt =Δ (5.4)

In any pipe system there will be one pipe that is the controlling pipe. The controlling pipe is that pipe which has the least number of sections, frequently only one. Once the controlling pipe is identified, the time step is determined by solving Equation 5.4 for Δt. Then the number of sections in the remaining pipes is obtained from:

nL

a tii

i

= Δ

(5.5)

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:

( )nL

a tii

i

=±1 ψ Δ

where ψ is the accepted uncertainty in wavespeed up to ±15% (Wylie, et al., 1993, pp. 54).



Figure 5.14 The Section Pipes window automates the process of obtaining appropriate sectioning criteria

Figure 5.14 shows the Section Pipes window. The controlling pipe (#8) is shown at the right. The sectioning criteria are shown at the upper left, where the average and maximum sectioning errors are shown. The table at the bottom shows the selected sectioning strategy and resulting modified wavespeed. The resulting time step size is shown at the right.

Numbering convention

Once a pipe is broken into sections, computation takes place where the sections join together. These are called pipe stations. The total number of stations in a pipe is equal to the number of sections plus 1. The station at the beginning of the pipe is numbered zero. See Figure 5.15.


8/10/2007 4:10:28 PM Page 155


Computingstations

1 2 3 4 5 6

0 1 2 3 4 5

Pipe sections

6

Figure 5.15 The computing stations are at the pipe section boundaries. The number of stations is one more than the pipe sections.

To obtain the pressure difference across a junction, compare the outlet computing station of the junction’s upstream pipe to the first computing station of the junction’s downstream pipe (see Figure 5.16).

J1 J2 J3P1 P2

J2P1

1,0 1,1 1,2 1,3 1,4

P2

2,0 2,1 2,2 2,3

Figure 5.16 The pressure drop across the valve at J2 in the AFT Impulse model at the top is given by the difference between the solutions at Pipe 1, station 4 and Pipe 2, station 0.

For junctions with more than one connecting pipe, the principle is the same (see Figure 5.17).



J1

J2

J3

P1 P2

J4

P3

P11,01,1

1,21,3

P2

P3

2,02,1

2,22,3

2,4

3,03,1

3,2

3,33,4

J2

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


8/10/2007 4:10:28 PM Page 157


Transient Control

Figure 5.18 shows the Transient Control window. The first tab is an area called the Transient Solver Control area. This area offers numerous input fields detailed in the following descriptions.

Figure 5.18 The Transient Control window configures the behavior of the transient solver and output file data

Start and stop time

The Start Time is the time at which you want the simulation to begin. Usually this will be zero, but you can define the starting time as any time that is convenient for you.

Note the junction transient data that is time-based is referenced to the start time. That is, if you want the simulation to begin at -10 seconds, the initial transient data point should be at -10 seconds. The Stop Time is of course when you want the simulation to end.

Shown right below is the Time Step Size. This is selected in the Section Pipes window, and is statically shown here for reference. Along with the



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

Save output to file…

When AFT Impulse runs a simulation, the transient output data that is generated is written to an output file. This is a file with a .out extension.

There are two options in the lower left that allow you to specify how frequently data is written to the output file. In the Figure 5.18 example, there are 21468 time steps that will be calculated in the simulation. The first option, "Save Output to File Every Time Step", will write 21468 time steps of data to file if chosen.

Output files can become extremely large if the data is saved for every time step, and in such cases the second option (in Figure 5.18) "Save Intermediate Output to File…" is a better choice. With this option AFT Impulse will skip over writing output to file at certain time steps. It will write it on increments of time step as specified in the input box. It is recommended this value never be set above 25. In Figure 5.18, results for every tenth time step will be written, for a total of 2146 time steps.

When output data is skipped, it will not appear in the output results or graphs. The justification for skipping over data is that over small time increments the transient behavior is sufficiently linear that the graphed results, for example, will appear the same as when all data is graphed.

It should be noted that AFT Impulse tracks the maximum and minimum values for pressure and flowrate, and does this for all computations whether saved to the output file or not. Thus a review of the maximum and minimum results in the Output window will alert you to whether any significant data was skipped by using the intermediate output to file feature.

Use variable pipe resistance

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


8/10/2007 4:10:28 PM Page 159


very much by relatively small changes in pipe resistance. And finally, introducing varying pipe resistance complicates the calculation process.

Therefore, accounting for varying pipe resistance is not important in many waterhammer applications. One application where it is important is when the initial steady conditions have no flow in some pipes. For instance, when the transient being modeled is a pipe with a closed valve which is being opened. In such cases, it is difficult to obtain the steady-state pipe resistance because the flow is zero and thus no Reynolds number exists. Thus it is probably best to use variable resistance.

If you choose to not use variable resistance, the pipe frictional resistance obtained in the steady-state part of the simulation will be used as a constant in the transient part of the simulation. On the other hand, if you allow the resistance to vary, the pipe friction will differ for each pipe section because the flowrate (and hence Reynolds number) is varying along the pipe due to the waterhammer transient. Further, the resistance will vary with time as the flowrate changes.

If variable resistance is modeled, the required run time increases. Experience suggests the run time increases by about a factor of three.

Model transient cavitation

When the transient pressure drops below the vapor pressure, transient cavitation will occur. You can model this using the cavitation models provided. In AFT Impulse 4.0 there is only one model and it is called the Discrete Vapor Cavity Model. If you do not model cavitation, then the fluid pressure can go below the vapor pressure and, additionally, below absolute zero pressure.

If the steady-state results indicate cavitation is occurring, this will cause a problem for the transient simulation and an artificial transient may be generated. If so, a warning will be given.

Stop run if artificial transient detected

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



simulation, the artificial transients will be displayed in the warnings section when the simulation is complete.

Including data in transient output file

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

Saving pipe output The Pipe Station Output tab allows you to specify which pipe station data is saved for each pipe. For pipes you have six choices for each pipe:

1. All Stations

2. Inlet and Outlet

3. Inlet Only

4. Outlet Only

5. None

6. User Specified (Up to Five)

As mentioned, the default is to save data for all inlet and outlet pipe stations. If the user is interested in Animation, the first choice to save All Stations provides the most animation flexibility. The last choice, User Specified, allows one to specify up to five stations that will be saved to the output file.

By selecting in the table your choices, you can save all stations for some pipes and no stations for the others. If you want to change settings for all pipes, use the "Change All Pipes To" button.

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


8/10/2007 4:10:28 PM Page 161


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

Saving junction output

The Junction Output tab allows you to save transient junction data. As mentioned, by default no junction output is saved. In general, most data for junctions can be obtained from the pipes. For instance, the flowrate through a valve over time can be obtained by looking at the upstream pipe flow rate. However, certain types of junction data is not obtainable in this way. In such cases, the data can be saved separately to file and reviewed in the output or graphed.

Here are some examples of junction data that can be saved:

• Surge tank liquid surface height, inflow and total liquid volume

• Gas accumulator volume and pressure

• Spray discharge outflow

• Vacuum breaker valve inflow and total volume

• Pump speed

You can select which junction data you want to save by selecting the Junction Output tab and then selecting the junctions in the list.

If you want to save junction data by default, select the check box on the Junction Output tab and junction data will always be saved. The only exception is branch junctions, which in most cases have no transient data.

Defining force sets

In order to process force data, the locations where the forces are to be calculated must be defined. The forces may be defined as differential forces, or as point forces.



Forces are defined on the Force Output tab in the Transient Control window, as shown in Figure 5.19. New force sets are added by clicking the New button. Force sets can be duplicated and deleted using the appropriate buttons on the tab, as well as moved up and down the list of force sets by using the reorder buttons. By right-clicking on the start pipe or end pipe in a force set, a popup window displaying input information about the pipe is displayed.

Force sets are defined by a Start Node and an End Node. The node locations are selected from the pipe stations calculated during the Pipe Sectioning process. For each node, the user must select the pipe where the force is to be calculated, and specify the distance from the start of the pipe (Station 0) to the location of the node.

In the example shown in Figure 5.19, the first force set (Node 1) is defined with the Start Node located on pipe P1, at a distance of 0 meters from the start of the pipe. The End Node is located along pipe P3 at a distance of 6 meters from the start of the pipe. The closest pipe station to the specified distance along the pipe will be selected automatically. In this example, the nearest pipe station to the Start Node is Station 0, and the nearest pipe station to the End Node is Station 2 (at 6 meters).

Once a force set is defined, AFT Impulse can calculate the differential force between the two defined points, and the resulting forces can be displayed in the Graph Results window, or they can be exported to a Force/Time file which can be imported into the CAESAR II pipe stress analysis program. A typical application of force set data would be to select the pipe endpoints between a pair of pipe elbows.

A point force can be defined by selecting the Point option in the Force Set column. When a user defines a point force, the cells for the end node are disabled as they have no meaning. For point forces, an ambient pressure can be specified against which this force will act. The ambient pressure can be different for each force set if desired.

Estimated file size and run time

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


8/10/2007 4:10:28 PM Page 163


Figure 5.19 Force set data is defined on the Forces tab in the Transient Control window

Calculating and outputting unbalanced forces

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

Including friction and momentum

When calculating the resulting force unbalance between two pipe stations, multiple forces come into play. It is traditional to calculate the force imbalance based solely on the difference in pressure between points. Since the pressure can differ based on elevation, the hydropressure is included in the force balance in order to adjust the pressure difference so this important effect does not bias the pressure difference.

What is typically not taken into account is the effect of fluid momentum change and friction forces. Each of these will bias the pressure difference and hence force calculation. If these effects are included in



the force balance, then the pressure difference can be adjusted to account for them and hence a force calculation based purely on imbalanced forces can be obtained.

To understand this, consider the steady-state pressure difference between elbows as shown in Figure 5.20. It is clear there is a pressure difference between nodes A and B. The difference in this case is due solely to frictional pressure drop in the pipe and across the valve. However, as the flow is steady-state, all forces at Node A are balanced by Node B and hence there is no force unbalance. Thus even though there is a pressure difference, the pressure difference is caused by the friction which in fact acts as another force in the opposite direction and of equal magnitude such that all forces in the x-direction balance to zero.

Pre

ssur

e

Distance

x

y

Node A Node B

PA

PB

Flow

Pre

ssur

e

Distance

x

y

x

y

Node A Node B

PA

PB

Flow

Figure 5.20 Steady-state pressure difference between elbows

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


8/10/2007 4:10:28 PM Page 165


Another component of the force balance is the unbalanced momentum at Nodes A and B in Figure 5.20. In steady state these will be in balance, but during a transient they usually will not. Figure 5.21 shows the option to include momentum in the force balance calculation.

When including friction in the force calculation, AFT Impulse can only output force balances for pipe stations within a single pipe object, or two pipes which are directly connected with each other (with a single junction in between).

Should one desire to calculate forces for pipes which are not directly connected, a sequence of force sets can be created between connected pipes and summed together. AFT Impulse does not perform this summation, and hence the user must perform it as a side calculation – typically with spreadsheet software.

Area changes In cases where force sets span across two pipes, and the pipes have a different diameter, the area change itself cannot physically create an unbalanced force even though the static pressure changes. AFT Impulse will automatically account for this in the force balance calculation.

See the Examples Help for additional explanation of these issues.

Viewing force data

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

Exporting transient force results

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



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

Defining pressures for the system

AFT Impulse needs a pressure specified in the system, to be used as the reference point upon which all other pressures are based. In certain cases, however, more than one pressure must be specified. These cases include systems which contain pressure control valves, flow control valves, internal relief valves, or fixed flow pumps.

Chapter 12 contains a comprehensive discussion on this topic. See the Role of pressure junctions.

Using scenario manager

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

• Create, name and organize scenarios


8/10/2007 4:10:28 PM Page 167


• Select the scenario to appear in the Workspace (the ‘current’ scenario)

• Delete, copy and rename scenarios

• Duplicate scenarios and save them as separate models

• Review the source of a scenario’s specifications

• Pass changes from a scenario to its variants

All scenarios are saved within the model file.

The Scenario Manager window

Illustrated in Figure 5.22, the left side of the Scenario Manager window contains the scenario tree, displaying the names of the currently defined scenarios, their relationship to other scenarios and which is the current Workspace scenario (the current scenario is also identified in AFT Impulse’s Status Bar). On the right there is a place where one may record notes related to that scenario.

Those attributes whose value comes from a different scenario are linked, while those values coming from the selected scenario are not linked. Linked attributes will inherit changes made in the same object and attribute of the linked ancestor. This inheritance will occur over as many generations as the link exists.

Creating, organizing and editing scenarios

The “All Scenarios Defined in Model” portion of the Scenario Manager window shows the name and relationship of all currently defined scenarios. Scenarios are referred to as either a parent or child, plus the Base Scenario, which may be thought of as the root scenario. Upon opening a new model file, only a Base Scenario will be present. To create a new scenario below the Base Scenario, select the Base Scenario by clicking on it, then click the Create Child button and enter a name for the new scenario.

Figure 5.22 illustrates a model with a Base Scenario with three children, ‘3 in PVC Pipe’, ‘3.5 in PVC Pipe’ and ‘4 in PVC Pipe’, representing three different size pipe runs. Each of these, in turn, has two children, ‘North hydrant open’ and ‘South hydrant open’, for the two operating cases.



Current Workspace

scenario

Scenario Tree

Rename, delete, clone,

promote & save

scenarios by clicking here

Create a new scenario by

clicking here

Figure 5.22 Scenario Manager Window

‘3 in PVC Pipe’, ‘3.5 in PVC Pipe’ and ‘4 in PVC Pipe’ were created by first selecting the Base Scenario then clicking on Create Child. ‘North hydrant open’ and ‘South hydrant open’ were, in turn, created by selecting each of the above children scenarios for the three different pipe run sizes, then clicking on Create Child (Note: alternatively, one could have made the Base Scenario current, created the ‘3 in PVC Pipe’ child, then selected it and created the ‘North hydrant open’ and ‘South hydrant open’ scenarios below it, then, with ‘3 in PVC Pipe’ still selected, click on Other Actions / Clone With Children twice, then rename the new clones ‘3.5 in PVC Pipe’ and ‘4 in PVC Pipe’). The ‘North hydrant’ and ‘South hydrant’ scenarios may be thought of as grand children of the Base Scenario. This concept of scenario ancestry is helpful in keeping track of the differences and similarities between scenarios discussed later.

A box with a plus sign next to a scenario indicates it may be expanded to reveal children scenarios while a box with a minus sign indicates the level is fully expanded with all children scenarios shown, if any. No box indicates there are no children. Levels may be expanded by clicking on the plus sign and condensed by clicking on the minus sign.


8/10/2007 4:10:28 PM Page 169


Clicking on the Other Actions button reveals a menu providing the following choices:

• Rename – renames the selected scenario

• Delete – deletes the selected scenario

• Clone Without Children – creates a copy of the selected scenario at the same level (a sibling)

• Clone With Children – creates a copy of the selected scenario along with its children

• Promote – raises a scenario one level, i.e. from a child to a sibling…active only for scenarios more than one level below the base scenario

• Save Scenario to File Without Children – saves the selected scenario to a new file without its children

• Save Scenario to File With Children – saves the selected scenario and its children to a new file

It is important to remember that AFT Impulse’s Edit menu Undo does not apply to Scenario Manager operations. To undo a Rename scenario operation, you can rename the scenario again to the old name. To undo a Clone or Save Scenario operation, you can delete the clone or file created. Once a scenario is deleted or promoted, there is no way to un-delete or un-promote the scenario.

Viewing scenario differences

A scenario’s data and that for all of its direct ancestors can be viewed in the Model Data window (Figure 5.23). This functionality can be enabled in the Model Data Control window. Output from different scenarios can also be displayed in the Output window using Output Control features.



Figure 5.23 Model Data display of scenario ancestor data. This functionality is enabled in Model Data Control.

Modifying individual scenarios

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

Properties of a scenario that may differ from its ancestors include:


8/10/2007 4:10:28 PM Page 171


• Presence and location of objects – one may add or delete pipes and junctions within a scenario and the workspace locations may vary from scenario to scenario

• Pipe and junction specifications – all values set in the specification windows

• System properties – fluid selected, temperature of fluid, viscosity model, system data

• Solution control – solution method, tolerances, relaxation

• Output Control

• Transient Control

• Database Manager

• Visual Report Control

Workspace Preferences, Parameter and Unit Preferences, General Preferences, Toolbox Preferences and Model Data Control apply to all scenarios.

Passing changes to child scenarios

Scenario Manager will pass changes made in a scenario to linked attributes of its descendants, allowing multiple scenarios to be modified in one operation and maintain similarity between scenarios where desired.

Any attribute whose value is the same as that of the scenario’s parent is linked to the parent. In turn, if the parent’s attribute value is the same as that of its parent, then the child’s attribute is linked to its grandparent. This ancestral linking of attributes can exist across any number of generations. A change made to a scenario will be passed downward to its children, grandchildren, etc., as far as the link exists. Changes are NOT passed upward from child to parent.

Once an attribute is changed within a scenario, the link is broken. Subsequent changes in the scenario’s ancestors, parent, grandparent, etc., will no longer affect that attribute.



Scenario logic examples For many users, it is easiest to grasp Scenario Manager when it is explained how the coding logic is actually implemented. Blank fields for children, grandchildren, etc., mean to look to the parent for the data. The Base Scenario never has blank fields (Figure 5.24a). Data only passes downwards, never upwards.

Base

Child #1

GrandChild #1

Diameter

3

__

__

Length

25

__

__

Figure 5.24a Scenario Manager logic – Blank fields for Child #1 and GrandChild #1 mean that the data is to come from the parent. If the Base scenario data is changed, all descendants are changed.

If a child scenario does not have a blank field, then data for that property is initiated at that scenario level (see Figure 5.24b).

Base

Child #1

GrandChild #1

Diameter

3

2

__

Length

25

__

__

Figure 5.24b Scenario Manager logic – Child #1 does not have a blank field, so its Diameter would be 2, not 3, as would GrandChild #1.

If a scenario is changed, and its child has different data, then the change will not pass downwards (see Figure 5.24c).


8/10/2007 4:10:28 PM Page 173


Base

Child #1

GrandChild #1

Diameter

3

2

__

Length

25

__

__

Base

Child #1

GrandChild #1

Diameter

6

2

__

Length

40

__

__

Figure 5.24c Scenario Manager logic – Changing the Base Scenario Diameter from 3 to 6 would not impact Child #1 or any descendents in that line. Changing the Length from 25 to 40 would also change the length in Child #1, GrandChild #1, and any descendents of GrandChild #1.

If a child scenario has data that is different than the parent, its children cannot relink to the parent (see Figure 5.24d).

If a child scenario’s data, which was previously changed and is thus different from the parent, is changed back to the same vale as the parent, the inheritance link is re-established (see Figure 5.24e). Its descendant’s link is also re-established.



Base

Child #1

GrandChild #1

Diameter

3

2

3

Length

25

__

__

Figure 5.24d Scenario Manager logic – Even if the GrandChild #1 has the same Diameter as the Base, it is not linked to the Base because it and its parent are not blank.

Base

Child #1

GrandChild #1

Diameter

3

3

3

Length

25

__

__

Base

Child #1

GrandChild #1

Diameter

3

__

__

Length

25

__

__

Figure 5.24e Scenario Manager logic – If the Diameter in Child #1 is changed to be the same as the Base, it will be “blanked out” the next time the scenario is loaded and the link re-established. And so will GrandChild #1, if its Diameter is also the same.


8/10/2007 4:10:28 PM Page 175


Figure 5.24f shows data for two pipe properties across three scenarios. The data that would be used when each scenario is loaded is as follows:

1. Base scenario

a. Diameter = 3

b. Length = 25

c. Changes to Diameter will not pass downwards

d. Changes to Length will affect only Child #1.

2. Child #1 scenario

a. Diameter = 2

b. Length = 25

c. Changes to Base Diameter will not affect Diameter

d. Changes to Base Length will affect Length

3. GrandChild #1 scenario

a. Diameter = 2

b. Length = 15

c. Changes to Base Diameter will not affect Diameter

d. Changes to Child #1 Diameter will affect Diameter

e. Changes to Base Length or Child #1 Length will not affect Length

Base

Child #1

GrandChild #1

Diameter

3

2

__

Length

25

__

15

Figure 5.24f Scenario Manager logic – Input data for these properties explained in the text.



Re-establishing broken links A link may be re-established by returning the attribute to the same value as that of its parent. This can be done manually by entering the value or selecting the Copy Data From Pipe list and selecting the Parent Pipe Data option. Within other data windows it is done by selecting the Same As Parent option in Solution Control, Output Control, System Properties, Cost Settings, Database Manager or Visual Report Control.

Since links are identified by comparing attribute values of pipes or junctions with the same Workspace ID number, renumbering a scenario will break the links of all pipes and junctions renumbered. Since numbers must be unique, once a link has been broken by renumbering, it may not be re-established.

Fast Scenario Changes

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

Special modeling features

The modeling utilities described in this section help you build your model more rapidly and keep track of what you are doing and what assumptions you have made.

Waterhammer Assistant

The Waterhammer Assistant can be opened from the Help menu or Toolbar. The Waterhammer Assistant reviews your model and offers helpful advice (see Figure 5.25).

By default, whenever a model is run AFT Impulse checks with the Waterhammer Assistant to see if it has any advice. If it does, the Waterhammer Assistant is displayed. The user then has the option of canceling the run or ignoring the advice and continuing. If you do not want the Waterhammer Assistant to be consulted before each run, you can specify this choice in the General Preferences window opened from the Options menu.


8/10/2007 4:10:28 PM Page 177


Figure 5.25 The Waterhammer Assistant offers advice on modeling issues

Workspace Find

The Find window utility shown in Figure 5.26 helps you find a particular pipe or junction by its ID number and name. The Find utility becomes more useful for models that expand over multiple Workspace pages.

You can access the Find utility by selecting Find on the View menu or by clicking the button on the Toolbar. The Workspace is moved so that the pipe or junction is located near the center of the screen.



Figure 5.26 The Find feature for the Workspace is helpful for locating a specific pipe or junction. It is opened from the View menu or Toolbar.

Reverse Direction

Each pipe you create has a reference positive flow direction as indicated by the arrow on the pipe. If you want to reverse the reference positive flow direction, you can do so by manually moving the pipe endpoints. However, this procedure is tedious. The Reverse Direction utility toggles the reference positive flow direction for the selected pipe or pipes.

The Reverse Direction utility is accessed from the Arrange menu or from the reverse direction button on the Toolbar.

Select Special

Select Special is a tool for selecting or deselecting objects based on certain criteria (see Figure 5.27) and allows you to quickly select a group of objects.


8/10/2007 4:10:28 PM Page 179


Figure 5.27 You can customize your selections using Select Special. It is accessed from multiple locations within AFT Impulse to assist with selecting groups of pipes and/or junctions.

When first opened, Select Special will select all the objects currently selected on the Workspace. You may alternatively select objects based on several other criteria available from the Selection Type drop-down menu. Objects in the Pipe Selections or Junction Selections list may be selected or deselected based on the criteria by clicking the Select Pipes or Select Junctions button or the Deselect Pipes or Deselect Junctions button.

Invert will deselect all selected objects and select all unselected objects. Clicking on the Select All or None buttons will result in all of the objects being selected or deselected respectively.

Several select/deselect operations may be performed in succession. When finished, click on OK button.



This tool is available from several windows and will select the items on the window from which it is opened. For example, it could be used to display only the closed valves in the Output window when it is accessed from the Show Selected Pipes/Junctions tab in the Output Control window.

Tip: To quickly show only a part of the model in the Output window, first select the objects on the Workspace for which you want to see output. Then, on the Show Selected Pipes/Junctions tab in the Output Control window, select the Workspace buttons in the pipe and junction areas. Finally, click OK in the Output Control window and only the pipes and junctions you selected on the Workspace will be shown in the Output.

Special Conditions

Some junctions have Special Conditions that alter the normal state of the junction. For example, a valve can be closed, a pump turned off, or a relief valve opened.

The special conditions are set in one of three ways:

1. By selecting the junction(s) on the Workspace and choosing Special Conditions from the Edit Menu

2. By selecting the junction(s) on the Workspace and clicking the Special Condition icon on the Toolbar

3. By opening the junction’s Specifications window and clicking the appropriate condition on the Optional tab

The junctions that have special conditions set are shown using a special symbol before the ID number (an “X” by default). This symbol can be customized in the Workspace Preferences window.

As shown in Figure 5.28, when the special condition results in a section of the model being closed or “turned off”, the pipes which will have zero steady-state flow are displayed as dotted lines. This indicates visually that they are cut off from flow; the junctions are outlined with a dotted line. The default dotted line can be changed in the Workspace Preferences window.


8/10/2007 4:10:28 PM Page 181


An X next to the junction

number indicates a

special condition has

been set

Dashed lines for pipes and

around junctions

indicate closed model sections

Figure 5.28 Workspace display when a valve is closed using the Special Condition. Notice the “X” next to the valve number (J14), which is shown in red by default. Also notice the connecting pipes shown with dashed lines and junctions with dashed outlines which indicate nonflowing sections of the model.

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

Special conditions with transient data If a Special Condition is applied to a junction that has transient data, then the transient solution will ignore the Special Condition. It is then incumbent upon the engineer to relate the Special Condition to the initial transient data.



For instance, assume the user wants to simulate the transient that occurs during a valve opening. The steady-state solution will have the valve closed, and this is modeled by using a Special Condition for the valve junction. In such a case, the Cv data entered on the valve junction's Loss Model tab will be ignored.

On the Transient tab, a transient can be entered. The first data in the table should then be Cv = 0, which corresponds to a closed valve. From there the Cv can be increased above zero as desired to open the valve.

If the user were to keep the Cv equal to zero, then that would be equivalent to the previous case where the user set the Special Condition but did not enter any transient data.

Pump special conditions Pump Special Conditions are slightly more complicated than for valves. In fact, pump junctions have two types of Special Conditions. The first is to turn the pump off and have no flow through it. This is called "Pump Off No Flow". The second type of Special Condition turns the pump off but allows flow to go through the pump. This is called "Pump Off With Flow Through".

Why would one want to use one Special Condition rather than the other? The simple answer is that the first Special Condition is more appropriate for positive displacement type pumps and the second for centrifugal pumps.

Here's why. When a positive displacement pump is turned off and has a pressure difference across it, it will usually act like a closed valve and not allow flow to go through it. Thus the first Special Condition would be most appropriate. For instance, assume one wants to model the transient that occurs during the startup of a positive displacement pump. The best way to do this would be to use the first Special Condition with no flow, and then input a flowrate transient in which the first data point is zero flow.

On the other hand, when a centrifugal pump is turned off and has a pressure difference across it, in the absence of other valving which prevents flow the pump will usually allow flow to go through it. Thus the Special Condition that allows through flow would be most appropriate. For example, assume one wants to model the transient during the startup of a centrifugal pump. One would use the second


8/10/2007 4:10:28 PM Page 183


Special Condition, and, in conjunction with a pump curve entered at 100% speed, input a speed transient with the initial speed as zero.

Special conditions only applied to junctions Special Conditions in AFT Impulse can only be used for junctions. Unlike AFT Fathom, Special Conditions cannot be applied to pipes. The reason is fairly simple. In AFT Fathom a pipe Special Condition closes the pipe. But in reality one does not close a pipe. One instead closes a valve that is in the pipe. For user convenience AFT Fathom overlooks this fact.

In contrast, AFT Impulse does not have as much freedom in such cases. When transients occur, there are reflections off system components such as valves. If one were to "close a pipe" in AFT Impulse, it would not be known exactly where the closure exists and thus the transient reflections could not be properly modeled.

Transient special conditions

Junctions which accept transient data also support Transient Special Conditions. The default for all junctions is None. The other choice is Ignore Transient Data. This allows one to have no transient initiation at that junction without having to delete the transient data.

When a Transient Special Condition is set to Ignore, a # symbol is displayed adjacent to the junction number on the Workspace. This symbol can be customized in the Workspace Preferences window.

No reflections – infinite pipe The Assigned Pressure and Assigned Flow junctions support a third type of Transient Special Condition. It is called No Reflections (Infinite Pipe). This condition is appropriate for the entrance to a long pipe line from which no reflections will occur during the simulation period.

When this Transient Special Condition is specified, a * symbol is displayed adjacent to the junction number on the Workspace. This symbol can be customized in the Workspace Preferences window. A discussion of infinite pipe behavior is given in Chapter 12.



Special Condition Ignore

The Vacuum Breaker Valve, Gas Accumulator, Liquid Accumulator, Surge Tank, and Relief Valve junctions have a unique kind of Special Condition called Ignore. This makes the steady-state and transient solvers ignore the presence of the junction. Thus not only is the data ignored, the junction itself is ignored. This is convenient when locating one of these devices by trial and error. Rather than having to delete the junction, it can be ignored.

Here is an example of when this feature might be useful. Assume that one is trying to find the best location for a gas accumulator. If one does this with a single gas accumulator junction, it will be required to run a case, view the results, then relocated the junction by changing the pipe lengths and rerunning. If one uses the Special Condition Ignore, one can place multiple gas accumulator junctions in the model and then set all of them except one to Ignore so that only one of the junctions is used at a time.

Merging models

Two models can be merged together using the Merge command on the File menu. Here’s how:

• Open the first model

• Choose Merge on the File menu

• From the file list, select the model to be merged

If there are duplicate numbers, the objects in the second model will be changed. The second model will be selected (highlighted) so that you can move it into position (be careful not to click off the selected portions or the second model may be deselected).

Tip: To move a selected group of objects, use a junction icon, instead of a pipe, to drag the group. This will avoid missing the pipe and clicking on an empty area of the Workspace which will deselect your group.

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


8/10/2007 4:10:28 PM Page 185


1. Open the model which you want to merge.

2. Open the Scenario Manager.

3. Select the child scenario you want to merge.

4. On the Other Actions button, choose the Save Scenario to File Without Children.

5. Enter a name and click OK. This process creates a new Impulse model with the desired scenario as the base scenario.

6. Open the original model and merge in the new model file created in Step 5.

Print Preview/Special

The Print Preview/Special window shown in Figure 5.29 allows you to access special print features such as Print Preview and Fit to One Page. It is accessed from the File menu.

Figure 5.29 The Print Preview/Special window is opened from the File menu or Toolbar

The Print Preview/Special window also offers you a preview of the print content and format for all primary windows. By clicking the Print Preview button you can review your printed material at several zoom states. Individual pages can be printed.



The Fit On One Page feature causes the graphic image generated from the Workspace or Visual Report windows to be compressed onto a single page. This is a convenient feature for creating reports of your model. AFT Impulse supports whatever printer page sizes are supported by your printer, so use of legal paper or larger may allow you to get more visible details. The graphics can also be centered vertically and/or horizontally. You also can fit your content onto multiple pages.

The Print Preview/Special window allows you to set the page orientation, which can also be changed once you are in the Print Preview window itself.

Several other special printing features affect the Workspace and Visual Report. The Only Print Selected Pipes and Junctions works with Workspace printing by selectively printing the pipes and junctions currently selected on the Workspace.

The Include Hidden Pipes and Junctions feature includes hidden pipes and/or junctions on the Workspace or Visual Report in the printed content. By default this box is not selected, and hidden pipes and junctions are not included in Workspace and Visual Report printouts.

By default, AFT Impulse prints the junction icons for the Workspace and Visual Report with the same colors as displayed on the screen. It also prints the background color as the same color as your Workspace or Visual Report background. You can also specify that the junction icons be printed in black and white (still with the color background) or that the entire printout be black and white.

If you added a Workspace Picture into the background, you can include this picture in the printout.

Finally, if you have a grid displayed on the Workspace or Visual Report, you can include the grid in the printout.

Transfer Results to Initial Guesses

AFT Impulse’s iterative Steady-State Solver discussed in Chapter 8 takes initial guesses (also called initial conditions) and applies the principles of mass and momentum balance to improve the guesses. This is called iteration. It continues to refine the solutions until agreement with the governing equations is obtained.

After a converged answer is obtained, you can choose to take these results and use them as the initial guesses for a subsequent run. This


8/10/2007 4:10:28 PM Page 187


causes much faster convergence in subsequent runs because you are starting the Steady-State Solver at or very close to the actual solution.

This feature is called Transfer Results to Initial Guess, and it is on the Output window Edit menu and Toolbar.

Batch runs

AFT Impulse scenarios or models can be run in batch mode, a feature that is especially useful for running a group of scenarios or models during lunch or overnight.

To run models in batch mode, follow these steps:

1. Click “Start Batch Run” on the File menu.

2. Select the Batch Run Type. This is whether you want to run scenarios from the current model or different model files. If the current model does not have any scenarios then only the second option will be available.

3. If the Batch Run Type is “Scenarios in Current Model”, then choose the Add Scenarios button and select the scenarios you wish to run.

4. If the Batch Run Type is “Models from Different Files”, then the models can be selected individually using the “Add Model Files” button, or loaded from a Batch File (by Choosing the “Load From Batch File” button). A Batch File is a text file listing all models you would like to run. At any time a selected model file can be removed by clicking the “Remove Selections”. The current list of models can be saved to a Batch File for use in the future by clicking the “Save List to File” button.

5. Use the Output Options in the lower left to modify style and content of output, and whether to save the output data to a file and/or send it to a printer or Adobe PDF file. All reports will use the specified font.

6. Click the Start Run button to begin the batch run.

AFT Impulse then opens each scenario or model file in sequence, automatically runs each scenario/model that has a completed Checklist, and sends results to the specified destination. All pertinent information that is generated during the batch run (such as error messages) are automatically displayed when the final scenario/model is finished.



The entire batch run can be canceled at any time by clicking the Cancel button in the Solution Progress window.

Print Content

With either the Model Data or Output window active, you can open Print Content from the View menu (see Figure 5.30), allowing you to specify the content and format of Model Data and Output reports.

You can include Model Reference Information (entered in the Output Control window) to document your model or make important comments you want kept with the model.

Figure 5.30 Print Content window specifies the print content of the Model Data or Output window. It is opened from the View menu or Toolbar.

The various tables in the General section, Pipe section, and Junction section can be selectively printed. All Pipe and Junction Notes are shown in the Model Data General section and can be printed at your option.


8/10/2007 4:10:28 PM Page 189


Empty columns in the three Model Data tables can be ignored in the printed report. You can also have each of the Print Content sections start on a new page. Both of these are formatting options in the Print Format window.

The printer font and style of your choice can be selected for the printed report. Large reports can be condensed by selecting small fonts. A good choice for large reports is Arial True Type Regular Style at a 6 or 8-point size.

Extended Model Check

The connections and settings of each pipe and junction are exhaustively checked using the Extended Model Check item on the View menu. This is also helpful if the model seems to not be connected correctly.

Math calculator

For convenience, AFT Impulse offers you quick access to the Windows Calculator; simply select Calculator from the View menu or press CTRL+K in any specifications window.

Working with the steady and transient solvers

The term Solver conveys a mathematical equation solution method whose purpose is to model some physical phenomenon. The implementation is in the form of algorithms and subroutines which solve the equations. AFT Impulse 4.0 has two Solvers because it is solving two problems.

Steady-state solver

The Steady-State Solver obtains a balanced steady-state solution to the system being modeled. It employs a Newton-Raphson matrix method that is iterative nature. An initial guess to the steady-state solution is made, and the Newton-Raphson method progresses the guess towards a final solution that agrees with the governing equations within some tolerance. Details are given in Chapter 8.



Control of the Steady-State Solver is given in the Steady Solution Control window opened from the Analysis menu. Here the user can specify items such as tolerances, relaxation and matrix methods.

The Steady-State Solver is based on that in AFT Fathom™, a commercial steady-state modeling software developed by Applied Flow Technology. AFT Impulse contains most of the capability found in AFT Fathom for solving systems with known density, and can optionally be used for steady-state modeling without reference to transient modeling.

If you want to run a model in steady-state mode only, on the Analysis menu specify the Analysis Type as "Steady Only".

When you specify the model as Steady Only, certain input data normally required for transient models are no longer required. Some examples are wavespeed for pipes, liquid bulk modulus, and input for the Section Pipes window and Transient Control on the Analysis menu. In fact, when modeling Steady Only the Checklist adjusts to only require four items to be completed.

Recommendations on Steady Solver It is recommended that before running a transient simulation, the user first run the model in Steady Only mode to help ensure the model is assembled properly. Once confidence is established in the steady-state results, the Transient feature can be enabled and run.

Transient solver

The Transient Solver models the surge transients in a pipe system using the Method of Characteristics (MOC). Unlike the iterative Newton-Raphson matrix method used for steady-state solutions, the MOC is a direct solution method that marches in time. Details are given in Chapter 9.

To function properly, the MOC requires an accurate steady-state solution to provide the initial conditions. Thus when one runs a transient simulation in AFT Impulse 4.0, the Steady-State Solver is always run first. The steady-state solutions are then automatically passed on to the Transient Solver, which uses them as the initial conditions for the MOC. This will reduce artificial transient problems to a bare minimum.


8/10/2007 4:10:28 PM Page 191


Artificial transients The fundamental cause of artificial transients is mismatched steady-state conditions. One would hope that artificial transients would never occur in AFT Impulse 4.0, but that is not the case. If they do occur, there is one of two causes. First, it may be an indication of a problem with the steady solution. This could be a problem in the Steady-State Solver, or an input problem that could not be identified when the model was built.

Secondly, a user could potentially specify mismatched steady-state and transient behavior of a junction. For example, for a valve junction the user inputs a steady-state valve Cv on the Loss Model tab. In addition, the user can enter a transient valve Cv on the Transient tab. The initial transient Cv should match the steady-state Cv on the loss model tab. If they do not, then an artificial transient would occur.

AFT Impulse attempts to identify such inconsistencies when the data is first input and then again before the model is run. If for some reason AFT Impulse does not catch this problem and bring it to the user's attention, then an artificial transient will occur.


C H A P T E R 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



6. Double-clicking the connected pipe or junction number in another Specifications window

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.

Chapter 6 Pipe and Junction Specifications Windows 195

8/10/2007 4:10:28 PM Page 195


Figure 6.1a Pipe Specifications window Pipe Model tab.

Figure 6.1b Pipe Specifications window Fittings & Losses tab.



Figure 6.1c Pipe Specifications window Design Alerts tab.

Figure 6.1d Pipe Specifications window Optional tab.


8/10/2007 4:10:28 PM Page 197


Common input parameters

The common input parameters are located above the tabbed folders (Figure 6.1a-e) and are discussed below.

Pipe number Every pipe must have an ID number. When you create a new pipe, AFT Impulse assigns a default pipe number. The pipe number can be changed; however, duplicate pipe ID numbers are not accepted, and ID numbers must be greater than zero and up to 30,000. The pipe numbers you choose have no bearing on the model connectivity, direction, or layout. They are merely convenient identifiers.

Pipe name A name can be assigned to each pipe for reference purposes. The default name is simply “Pipe.” Names do not need to be unique. The name can be shown on the Workspace, in the Visual Report and in the Output window. By using names, attention can be called to important or critical pipes.

Figure 6.1e Pipe Specifications window Fluid Properties tab (available only when Variable Fluid Properties are selected in System Properties).



Copy Data From Pipe The Copy Data From Pipe dropdown list allows you to conveniently set the pipe properties to be the same as any existing pipe. You can copy all or part of the parameters from an existing pipe in the model to the current pipe by choosing one of the pipes in the list and then specifying which parameters to copy. Only the parameters you choose will be copied to the current pipe.

The Copy Previous button functions the same way as the Copy Data From Pipe except that the last pipe edited is used. This is convenient when there are several pipes that have the same or similar parameters.

Scenario same as parent If the current model is a child scenario, in the Copy Data From Pipe list there is an option to set the pipe to Parent Pipe Data. This will change the selected pipe to be the same as its parent.

Connected junctions The connected junctions area shows you AFT Impulse's interpretation of the junctions that are connected to the particular pipe. Every pipe must be connected to two junctions in order to be completely defined. When a junction does not yet exist at one of the pipe endpoints, “None” appears in the area where the junction ID number would normally be displayed.

All pipes on the Workspace have a reference positive flow direction. The reference positive direction is indicated on the Workspace by an arrow on the pipe. Based on the reference direction, AFT Impulse identifies an upstream junction and a downstream junction.

To determine the properties of the connected junctions, you can use the Inspection utility. To inspect, position the mouse pointer over the connected junction ID number and hold down the right mouse button.

Pipe Model

The Pipe Model tab (Figure 6.1a) allows you to input geometric data for the pipe.


8/10/2007 4:10:28 PM Page 199


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.



If you select the Pipe Material as Unspecified, these three fields are editable and you will need to enter data in each for wavespeed calculations.

If User Specified Wavespeed is selected, these three fields will be disabled no matter what Pipe Material is specified, as they are not needed when the user supplies the wavespeed. Finally, if the model is for steady-flow only, the three fields will always be disabled as they have no bearing on steady-state models.

In the Friction Model area, the value shown is not to be construed as the correct pipe roughness value for your application. The pipe roughness depends on many factors, most importantly the age of the pipe. AFT Impulse offers the pipe material roughness as a guide. You will want to enter the roughness value that is most representative of your application.

Multiple friction data sets can be compiled for each pipe. For example, friction data for pipes with different ages can be set up. These different data sets are then made available in the Data Set dropdown list. More information on creating multiple friction data sets is given in Chapter 7 in the section on custom pipe materials.

Pipe diameter The inner diameter of the pipe is entered in the Inner Diameter field in the Pipe Specifications window. For pipes whose material is unspecified, the diameter and units may be entered directly.

If the material is specified from the Pipe Material list, the diameter cannot be edited; the value is derived from the database for the selected material, size, and type.

The reduction in diameter due to scaling can be accounted for by entering a percent reduction in the ID Reduction - Scaling field. Zero percent would represent no reduction.

Pipe wall thickness The wall thickness is entered in the Size area of the Pipe Specifications window. Wall thickness is required for pipes in which the wavespeed is calculated.

If the material is specified from the Pipe Material list, the wall thickness cannot be edited; the value is derived from the database for the selected


8/10/2007 4:10:28 PM Page 201


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 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.

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. Relative roughness – Some pipe roughness specifications are given as a relative roughness. In this case, the roughness height is divided by the pipe diameter. Relative roughness has no units, so when you enter a relative roughness, the Units dropdown list is disabled.

3. Hazen-Williams – The Hazen-Williams method uses an empirical factor (CHW) to relate the flow rate to the pressure drop in the pipe. This method is still in common use in the field of water distribution. You can specify the Hazen-Williams factor for any pipe.

4. Explicit Friction factor – If you already know the friction factor for the pipe, you can enter the value explicitly.

5. Hydraulically smooth – You can also specify a pipe as hydraulically smooth. Modeling a pipe as hydraulically smooth implies that its roughness is negligible. However, having a small roughness is not the same as being frictionless. Rather, the pipe friction factor follows the hydraulically smooth curve in the turbulent region of a standard Moody diagram. You are encouraged to consult a standard fluid mechanics textbook for additional discussion of this subject.

6. Resistance – You can specify a pipe resistance in terms of head loss and volumetric flow rate. The pipe head loss will then be as follows: dH = RQ2.

7. Frictionless – For modeling purposes, it is occasionally useful to model a pipe as having no friction. There are limitations to where such a pipe can be located in your model. Basically, a frictionless pipe cannot be the sole connection between two branching junctions.

8. MIT Equation – The MIT Equation is appropriate for crude oil. See Chapter 8 for a discussion of how this method is implemented.

9. Miller Turbulent – The Miller Turbulent method is appropriate for light hydrocarbons. See Chapter 8 for a discussion of how this method is implemented.

Because you are free to specify the friction model for each pipe individually, you can mix and match the friction models throughout your pipe system if you choose.

If you find it more convenient to use a certain type of friction specification, like the Relative Roughness method for example, you may want to change the defaults in the Parameter and Unit Preferences window to reflect this. When you change the default method, the other


8/10/2007 4:10:28 PM Page 203


methods are still accessible, but the Pipe Specifications window always comes up with your preferred method selected.

The preceding friction models all apply to Newtonian viscosity models. The viscosity model is specified in System Properties window. If the user chooses one of the non-Newtonian models, the preceding friction data may not be used in the calculation.

A more detailed theoretical discussion of the different friction models is given in Chapter 8.

Pipe Support The Pipe Support selection is made from the provided drop down list to obtain the constant c1. This constant is required for pipes in which the wavespeed is calculated. AFT Impulse provides seven common pipe support types.

The meaning of the c1 constant and its relationship to the seven support types is given in Chapter 9.

Pipe Wavespeed The wavespeed is the speed at which transient events propagate through the pipe. Typical values for the wavespeed are 700-1600 m/sec (2000-5000 ft/sec).

This parameter depends on the liquid acoustic velocity and the pipe material and support. AFT Impulse will calculate the wavespeed for you, or you can provide your own user specified wavespeed.

Fittings & Losses

The Fittings & Losses K factor is one way to enter loss factors in your model. The Fittings & Losses K factor is added to the friction loss calculated by the Steady-State Solver. This is useful when a pipe section contains elbows and other fittings.

You can use any of the losses supplied in the Impulse database or you can enter a specific loss factor. To enter losses, click the Fittings & Losses tab (Figure 6.1b) and then click the Specify Fittings & Losses button to open the Fittings & Losses window (Figure 6.2). This window allows you to specify losses from a table of standard losses.



Because many of the fittings & losses are dependent on pipe diameter, you must specify a diameter before you can open the Pipe Fittings & Losses window.

The types of pipe losses are divided into several categories represented by tabs. To add a loss to the pipe, click the table cell which describes the fitting or loss and type in the quantity. When you are done entering all the losses in each of the categories, press OK and you will see a summary list in the Pipe Specifications window.

Figure 6.2 Fittings & Losses window opened from the Pipe Specifications window

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.


8/10/2007 4:10:28 PM Page 205


The abbreviations used in the Fittings & Loss window are listed at the bottom of the window. You can double-click the image to see the reference information for the loss.

Design Alerts

A design alert (Figure 6.1c) allows you to specify certain maximum and/or minimum parameters for the pipe. After running the Steady-State or Transient Solvers, any exceeded value will be identified in the Warnings section of the Output window and the specific cell containing the value in the Pipes section will be highlighted. Design Alerts can be cross-plotted in the Graph Results window vs. the actual results. One good example where this is useful is for cross-plots of maximum and minimum allowed operating pressures.

Optional input

There are several optional input parameters (Figure 6.1d). These are described in the following sections.

Initial Flow Rate Guess The Initial Flow Rate Guess is the initial guess at the mass flowrate, volumetric flowrate, or velocity solution through the pipe. Specifying a good initial guess of the flow solution decreases the number of iterations the Steady-State Solver must execute in order to arrive at a converged solution. However, specifying initial guesses takes time and good judgment.

If you do not specify an initial flow guess, AFT Impulse implements a proprietary method of generating first guesses. This method is usually sufficient to get the Steady-State Solver going in the right direction so a converged solution results. However, the method of generating first guesses is not 100% effective, and there may be times when you will have to specify the first guess in order to obtain a solution. More discussion on this subject is given in Chapter 8.

Workspace display You have the option of specifying what to display on the Workspace. You can choose to display the pipe ID number, the pipe name, the pipe nominal size, and the pipe schedule. This is helpful in large models to



reduce the text on the Workspace and to focus on what you think is important. The default setting for this can be changed in the Workspace Preferences window.

Design Factor You can specify two design factors for each pipe – one for the pipe friction and one for fittings & losses that you add to the pipe.

Design factors allow you to specify multipliers on the pipe friction calculations and fittings & losses that are applied by the Steady-State and Transient Solvers during the solution process. Design factors are helpful for adding safety margins to your design calculations.

Pipe Line Size and Color The default thickness of the line shown on the Workspace and default pipe color is defined in the Workspace Preferences window. You can change the displayed line thickness and color for each pipe individually if you desire.

Parallel Pipes Here you can specify that the pipe is one of a number of parallel pipes. The pipe properties thus represent a single pipe. This is useful, for example, to represent a bundle of tubes in a heat exchanger.

Support for Partially Full Pipe There is checkbox which indicates the pipe supports partially full pipe calculations. This means partially full along the axial direction, as opposed to partially full along the radial direction. This function only works for pipes that have a slope, and when the pipe is partially full it drains from the end with the higher elevation.

If this feature is not selected, the pipe will always be assumed to be liquid full. Selecting this feature implies that one end of the pipe is connected to a junction at its higher elevation endpoint which allows gas to flow in, thus allowing the pipe to drain.

If the option is selected, and the entry is 100%, it means the pipe is initially 100% full, but can drain if backflow occurs at the high elevation endpoint. The percent full can also be set to something less than 100%. In this case, the pipe is assumed to be partially full initially. This is only


8/10/2007 4:10:28 PM Page 207


valid when the steady-state condition for the pipe is such that there is no flow. Because if there were flow in that pipe during steady-state, the pipe would not be partially full.

This feature only works with pipes which connect to a vacuum breaker valve, spray discharge, assigned pressure or exit valve junction at the high elevation end.

When selected, the predicted level of the liquid surface and other related parameters can be graphed in the Graph Results window. When pipes can be partially full, more accurate predictions result when the pipe is modeled with more sections. A single section is not recommended. At least two sections should be used, and more if possible.

Intermediate Pipe Elevations Pipe inlet and exit elevations are obtained from the upstream and downstream junction elevation data. The pipe elevations in between are assumed to vary linearly.

In general, the pipe elevation profile will have no effect on the hydraulic results, and the linear assumption is suitable. However, even though the hydraulics are predicted accurately, the intermediate pipe pressures are not (unless the elevation is in fact linear). As long as these pressures are within the allowables for the pipe, and do not drop below vapor pressure, then no problems arise. However, if the pressures do exceed these conditions, it is important to find this out.

One way to account for this is to break the pipe up into smaller individual pipes, with additional junctions (usually branch junctions) connecting them. These junctions can be set at the proper elevations, and the actual elevation profile can thus be modeled. However, this too has ramifications, especially for transient models. With the increased number of pipes, this can force a shorter time step selection, which can drastically increase model runtime.

The solution to the just described dilemmas is to use intermediate pipe elevations. Here an elevation profile can be entered directly for the pipe, and only a single pipe is required as long as the diameter is constant.

To use intermediate pipe elevations, tick the check box for Use Intermediate Pipe Elevations, and enter the elevation profile. To do this, the pipe length entered on the Pipe Model tab (Figure 6.1a) is required. AFT Impulse will ensure the total length in the intermediate elevation



area matches the length on the Pipe Model tab. It does so by fixing the length of the final section to the difference between the sum of the lengths in the table and the actual length entered by the user. The end elevation is also fixed to that of the downstream junction. If the user later modifies the downstream junction elevation, the pipe’s final elevation section will be automatically updated.

Fluid Properties

This tab is only visible when variable properties are being modeled (Figure 6.1e). Variable property modeling is enabled in the System Properties window.

For models with variable fluid properties, you can assign different fluid properties to each pipe in the model. This is convenient for modeling systems that are subject to temperature variations and systems that have multiple fluids. The pipe fluid properties are displayed in the Pipe Specifications window, and can be modified by selecting the Fluid Properties button. Besides varying the fluid properties, the viscosity model can varied from pipe to pipe. Thus some pipes in the model can be modeled as Newtonian, while others as non-Newtonian.

Notes

Each pipe can have notes associated with it. This is useful for listing assumptions, reference documents, drawings, measurements, etc.

Status

The Status tab shows which required input data has not been entered.

Junction Specifications windows

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


8/10/2007 4:10:28 PM Page 209


Table 6.1 Comparison of junction capabilities

Junction Type Connecting Pipes

Special Conditions

Custom Database

Window Format

Area Change 2 √ 1

Assigned Flow 1 √ √ 1

Assigned Pressure* 1/1-25 √ 2

Branch 2-10 √ 2

Check Valve 2 √ 1

Control Valve 2 √ √ 1

Dead End 1 √ 1

Gas Accumulator 1-25 √ √ 2

General Component 2 √ 1

Liquid Accumulator 1-25 √ √ 2

Pump 1-2 √ √ 1

Relief Valve 1-2 √ √ 1

Reservoir 1-25 √ 2

Spray Discharge 1-4 √ √ 2

Surge Tank 1-25 √ √ 2

Tee/Wye** 3 √ 2

Vacuum Breaker Valve 1-2 √ √ 1

Valve 1-2 √ √ 1

Volume Balance 2 1

*Assigned Pressure junction connects to only 1 pipe for static pressure,

and up to 25 for stagnation

**Tee/Wye junction has a unique format closer to Format #2.

There are nineteen junction Specifications windows, one for each junction type. Each junction Specifications window falls into one of two basic window formats, depending on the number of connecting pipes allowed for that junction type.



Note: AFT Impulse does not prevent negative flow through junctions. Should this occur, AFT Impulse will use the same loss factor or pressure drop data referenced to the defined upstream pipe, just as in forward flow. Because it is possible that the loss factor or pressure drop data would not be the same for both flow directions, the flow solution could be misleading. For this reason it is important that you either properly define the pipe flow directions for junctions or pay close attention to the flow directions that result. Should negative flow occur in any junction, AFT Impulse will give a warning in the output.

Format #1: Junctions with one or two connecting pipes

The first basic window format is for junction types that allow only one or two connecting pipes (see Figure 6.3). While junctions do not have an explicitly defined flow direction like pipes do, typically those with two pipes or less adopt a direction from the connecting pipes. For these junction types, upstream and downstream pipes are recognized by AFT Impulse based on the reference positive flow direction of the connecting pipes. The upstream and downstream pipes are displayed separately near the top of the window.

In this first basic window type, it is generally important to have the pipe’s reference positive flow directions specified in the physically correct directions. A good example of this is a Pump junction which will add pressure to the system in the direction of positive flow through the connecting pipes. The Pump junction interprets where to add the pressure based on the directions of the connecting pipes.

To inspect the connected pipe information in this window format, position the mouse pointer over the connected pipe ID and hold down the right mouse button.


8/10/2007 4:10:28 PM Page 211


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



direction of the pipe in reference to the junction is also shown in the pipe table.

To inspect connected pipe information in this window format, position the mouse pointer over the pipe ID number in the left column of the pipe table and hold down the right mouse button.

Parameters common to all junction Specifications windows

There are several junction parameters common to most junctions described in this section.

Junction Number

Each junction has an ID number (see Figures 6.3 and 6.4) that you are free to change. However, no two junction numbers can be the same, and all junction numbers must be greater than zero and up to 30,000. Duplicate junction numbers are not accepted. The junction numbers you choose are arbitrary and have no bearing on model connectivity, direction, or layout. They are merely convenient identifiers.

Junction Name

Each junction can be assigned a name for reference purposes (see Figures 6.3 and 6.4). The default name is the junction type. The name does not need to be unique. The name can be displayed on the Workspace, Visual Report, Model Data and Output windows.

Junction Elevation

Each junction has an entry for the elevation of the junction (see Figures 6.3 and 6.4). Next to each elevation entry is a set of units. For junctions that use Format #1, the default inlet elevation is entered by the user, and the outlet elevation is assumed to the same. A different outlet elevation can be entered by clearing the “Same as Inlet” checkbox. Different inlet and outlet elevations will have no effect on the flow solution, but will offset the local static pressure at the connecting pipe by the hydrostatic pressure difference.


8/10/2007 4:10:28 PM Page 213


Figure 6.4 Specifications window format #2 (pipes displayed in a table).

Junctions that follow Format #2 have a single elevation and offsets from that elevation (if they exist) for each individual connecting pipe entered in the connecting pipe table.

Junction elevations are used to account for pressure changes due to gravity and thus to allow calculation of absolute pressures in pipes. In a sense, junction elevations are arbitrarily defined. The magnitude of the elevation is not important from a solution standpoint, but the relative elevation difference between junctions is of critical importance because of its effect on the pressure results.

Elevation is required input for every junction. AFT Impulse uses this input to determine the relative height changes between one end of a pipe and the other.

In the Parameter and Unit Preferences window you can set a default elevation for your system. When a default elevation is specified, each time you open a new junction Specifications window the default



elevation is automatically entered. For systems that reside primarily at the same elevation, this can save time in entering data.

Database List

The Database List (see Figures 6.3 and 6.4) is a list of all custom equipment you have entered into the Component database. It displays custom equipment from both the local and network database sources. Selection of an item automatically retrieves the data and enters it into the junction. Changes to the custom input data are not accepted. To stop using a database selection, choose (None) from the database list.

Copy Data From Jct list

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

Scenario same as parent If the current model is a child scenario, in the Copy Data From Jct list there is an option to set the pipe to Parent Junction Data. This will change the selected junction to be the same as its parent.

Pipe connectivity

In each junction Specifications window (see Figures 6.3 and 6.4), you can see AFT Impulse's interpretation of your system's connectivity. All junctions must have the proper number of connecting pipes in order to be completely defined (refer to Table 6.1).

Jumping to another junction

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


8/10/2007 4:10:28 PM Page 215


Transient Data

Many junction types have an area where active transient data can be specified. Table 6.2 lists the junction types.

Table 6.2 Junctions that support transient data

Junction Type Type of Transient

Assigned Flow Flowrate

Assigned Pressure Pressure

Branch Flowrate source or sink

Check Valve Opening/Closing profiles

Control Valve Opening/Closing profiles, pressure or flow

Pump Speed, flowrate, control, trip or startup parameters

Relief Valve Opening/Closing profile

Reservoir Surface level changes

Spray Discharge Discharge opening/closing profiles

Surge Tank Surface Pressure

Valve K or Cv profile for opening and closing in part or full

The active transient data is entered on the Transient Data tab (Figure 6.5). Table 6.2 lists the type of data that can be entered for each junction type. If a junction does not have any active transient behavior, there is no need to enter any data on the Transient Data tab.

In most cases, transient data can be entered as absolute values or as a percentage of steady-state. The distinction is made when choosing between Absolute Values and Relative to Steady-State Value (Figure 6.5).

The first data point always needs to match the steady-state value which is usually entered on the tab at the far left. If, for example in Figure 6.5, the steady-state data for Cv of 1000 is changed to 750, the transient data will also need to be changed. However, if in Figure 6.5 the Relative option is specified, the first data point would be 100%, and when the Cv is steady-state is changed from 1000 to 750 there is no need to change the transient data since the first data point is still 100%. But it is now relative to a different steady-state value.



Figure 6.5 Example of junction transient data. A here a valve is being closed over five seconds.

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

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


8/10/2007 4:10:28 PM Page 217


A third type of Transient Special Condition is available for Assigned Flow and Assigned Pressure junctions. This is the No Reflections (Infinite Pipe) option. In such a case, the flow or pressure is used only for steady-state calculations. During the transient the flow or pressure is not fixed, but varies in accordance with the concept that no transient reflections occur. All transient data at this junction will be ignored.

The reasons for using this feature are important, but somewhat difficult to grasp. A lengthy discussion is given in Chapter 12.

Time based and event based transients In Figure 6.5, there is an area called Initiation of Transient, and the selection is Time. This means that the specified transient is linked to the time of the simulation. Specifically, the valve transient data is activated as soon as the simulation is started. Because the transient data is constant for one second, the valve does not start to close until the simulation time reaches one second.

Transients can also be initiated based on events. For instance, we can change the valve in Figure 6.5 to start closing when the pressure at some other system location reaches a certain level. Or maybe when a tank has filled up and thus reached a certain level. In such a case, the time zero data in Figure 6.5 does not refer to the start of the simulation, but is relative to when the event occurs. If, for instance, the pressure never reaches the specified level, the valve never closes.

Chapter 10 discusses in much more detail how event transient modeling can be used.

Initial Pressure/Head Guess

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



Figure 6.6 The Optional tab offers several optional data inputs and is very similar for each junction type

Specifying a good initial guess decreases the number of iterations the Steady-State Solver must execute in order to arrive at a converged solution. However, similar to specifying Initial Flow guesses in pipes, specifying initial guesses takes time and good judgment.

The Reservoir and Assigned Pressure junctions do not allow you the option of specifying an initial guess. Because the pressure is a known quantity, no guess is needed. The pressure solution is already available.

If you do not specify an initial guess, AFT Impulse implements a proprietary method of generating first guesses. This method is usually sufficient to get the Steady-State Solver going in the right direction so a converged solution results. However, AFT Impulse's method of generating first guesses is not 100% effective, and there may be times when you will have to specify first guesses in order to obtain a solution.


8/10/2007 4:10:28 PM Page 219


Displaying junction names and numbers

Also on the Optional tab of each junction Specifications window (see example in Figure 6.6) is the option to display the junction ID number, the junction name, both, or neither. This is helpful in large models for reducing the text on the Workspace and focusing on what you think is important. The default setting for this can be changed in the Workspace Preferences window.

Special Conditions

Some junctions allow you to set Special Conditions (see example in Figure 6.6). This feature allows you to override the default behavior of the junction. For example, the special condition for a Valve junction is to close the valve. The special condition for a Relief Valve is for it to be open no matter what the pressure. Not all junction types support special condition settings.

The junction for which the Special Condition is set will have an ‘X’ placed before the junction ID number. If the special condition is to close the junction, by default it will be outlined with a dashed line. These settings may be changed in the Workspace Preferences window.

Chapter 5 discusses how to apply transient data to junctions with steady-state Special Conditions.

The Vacuum Breaker Valve, Gas Accumulator, Liquid Accumulator, Surge Tank, and Relief Valve junctions have a unique kind of Special Condition called Ignore. This makes the steady-state and transient solvers ignore the presence of the junction. Thus not only is the data ignored, the junction itself is ignored. This is convenient when locating one of these devices by trial and error. Rather than having to delete the junction, it can be ignored.

Design Factor

You can specify a design factor for each junction (see example in Figure 6.6). Design factors allow you to specify multipliers on the junction pressure loss. These multipliers are applied by the Steady-State and Transient Solvers during the solution process. In the case of a pump, the design factor multiplies the pressure rise. Design factors are helpful for adding safety margins to your design calculations.



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

Changing the icon graphic

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

Changing the icon size

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

Specifying Base Area

Some junction windows that follow Format #1 allow selection of the base area for loss models. By default the base area will be the upstream pipe area. The user has the option of using the downstream pipe area or directly specifying an area or diameter.

Notes

Each junction can have associated notes. This is useful for listing assumptions, reference documents, drawings, measurements, etc. Notes are entered in the provided area on the Notes tab.

Status

The Status tab shows which required input data has not been entered.


8/10/2007 4:10:28 PM Page 221


Area Change Specifications window

The Area Change Specifications window is shown in Figure 6.7. The Area Change junction type must always have two connecting pipes. This junction type allows you to model the irrecoverable loss that occurs at the area transition between two pipes, whether expansion or contraction.

The Area Change Specifications window follows the first of the two basic Specifications window formats, displaying the connecting pipes in a fixed format. A flow direction through the junction is adopted from the defined directions of the two connecting pipes. Consistent with AFT Impulse's convention, the loss factor base area is referenced to the upstream flow area shown as the Base Area.

Figure 6.7 The Area Change Specifications window

The Area Change Specifications window offers two standard types of area change losses: the Conical Transition and the Abrupt Transition. The window shows a schematic of the selected geometry.

The critical parameter that influences the magnitude of the area change loss is the area ratio between upstream and downstream pipes. Until you



have input the information for the upstream and downstream pipes that allow for flow area determination, the standard Area Change junction loss factor is not calculated. For conical transition geometries you also need to specify the conical angle.

If the standard loss factors provided are inadequate for your application, you may specify your own custom loss factor referenced to the upstream pipe flow area.

More information on the standard loss factor types for Area Change junctions is given in Chapter 8.

Assigned Flow Specifications window

The Assigned Flow Specifications window is shown in Figure 6.8. The Assigned Flow junction type allows you to connect one pipe and to input a known steady-state and/or transient flowrate entering or leaving at a particular location in your system. You also can include an irrecoverable loss with this junction type.

The Assigned Flow Specifications window follows the first of the two basic Specifications window formats, displaying the connecting pipe information in a fixed format. Depending on whether you specify the junction as an inflow or outflow type, the required connecting pipe (upstream/inlet or downstream/outlet) will be enabled while the other pipe is disabled. Note that if the flow is specified as Inflow (to the system) then the reference positive flow direction of the connected pipe must be away from the junction (i.e. the junction is upstream of the pipe).


8/10/2007 4:10:28 PM Page 223


Figure 6.8 The Assigned Flow Specifications window

Tip: If you see that the required pipe in the Specifications window states “None” but a pipe is connected to the junction graphically, then check the Inflow/Outflow setting on the junction or the reference positive flow direction of the pipe.

The Assigned Flow junction type allows you to specify positive flow rates as mass flow rates. Negative flow rates are not accepted for steady flow, although are for transient flow. If, for example, you have an outflow type junction and your system is physically flowing in, you cannot assign a negative flow rate to the junction. You must reverse the connecting pipe flow direction to be consistent with the actual direction and change from an outflow type to an inflow type.



Transient Data

On the transient data tab a flowrate transient can be entered. A special feature is the ability to model a sinusoidal flowrate transient. Input for the amplitude and frequency is required. This transient is summed to that specified in the transient data table. A chopped sine wave takes the absolute value of the sine function.

More information on transient data, including event transients, is given earlier in this chapter and in Chapter 10.

Special Conditions

You can set a Special Condition for an Assigned Flow junction, which will turn the flow off and make it act like a Dead End during steady-state.

Special features for steady flow

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

Usually this feature would be used in conjunction with a Special Condition on the junction where its flow was turned off during steady-state.


8/10/2007 4:10:28 PM Page 225


Figure 6.9 Optional tab allows junction to act like a fixed pressure during steady-state if initial flow is zero.

Assigned Pressure Specifications window

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



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.


8/10/2007 4:10:28 PM Page 227


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.

More information on transient data, including event transients, is given earlier in this chapter and in Chapter 10.

Branch Specifications window

The Branch Specifications window is shown in Figure 6.11a-b. The Branch junction allows up to twenty-five connecting pipes. The Branch junction is the most flexible junction type for building network models that have flow splits. Other significant features are the ability to impose a steady-state and/or transient flow source or sink at the junction and the ability to specify separate loss factors for each connecting pipe. The Branch junction must have at least two connecting pipes.

The Branch Specifications window follows the second of the two basic Specifications window formats. A table on the Loss Coefficients tab displays the connecting pipe information. This table grows in size to accommodate up to twenty-five connecting pipes. After you add a fifth pipe a scroll bar appears, allowing you to review and enter loss factors for all pipes in the table.



Figure 6.11a The Branch Specifications window Loss Coefficients table.

For each connecting pipe, the pipe table lists the reference flow direction and up to two loss factors. The first loss factor is for physical flow out of the pipe and into the branch. The second is for physical flow into the pipe and out of the branch. The loss factors can be specified independently or left as zero. To edit within the pipe table, simply click in the appropriate column and row.

The distance of each connecting pipe (if any difference exists) from the elevation can be entered in the pipe table.

The Branch junction also offers you the option of specifying a flow source or sink at the junction. According to AFT Impulse's convention, a source of flow into the system is defined as positive, while a consumption of flow with flow out of the system (a sink) is defined as negative.


8/10/2007 4:10:28 PM Page 229


Figure 6.11b The Branch Specifications window Optional tab allows you to specify a steady-flow source or sink.

Transient Data

On the transient data tab a transient source or sink flowrate can be entered. More information on transient data, including event transients, is given earlier in this chapter and in Chapter 10.

Check Valve Specifications window

The Check Valve Specifications window is shown in Figure 6.12. The Check Valve junction type requires two connecting pipes. This junction type allows you to model the effect of a check valve which closes to prevent back flow.

The Check Valve Specifications window follows the first of the two basic Specifications window formats, displaying the connecting pipes in



a fixed format. The Check Valve junction does not have an explicit flow direction, but adopts a flow direction from the connecting pipes. Consistent with AFT Impulse's convention, the loss factor base area is referenced to the upstream flow area.

Figure 6.12 The Check Valve Specifications window

A check valve is a device that allows flow in only one direction. AFT Impulse assumes that the Check Valve is initially open. If the flow solution indicates that forward flow will not occur, AFT Impulse closes the valve. The valve can close during the steady-state or transient portion of the simulation.

A check valve Special Condition can be either open or closed.

Specifying losses

You specify the losses for the Check Valve on the Valve Model tab in the Specifications window. For convenience, you can specify the constant loss characteristics of a Check Valve as a valve coefficient (CV)


8/10/2007 4:10:28 PM Page 231


or as a loss factor (K). Chapter 8 details the relationship between K and CV.

Forward Velocity to Close Valve

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

Delta Pressure to Re-Open

Frequently after a check valve closes, there is pressure across the valve at which it will open again. You can specify that delta pressure here.

Transient Data

The closing of a check valve is a complicated transient process because the mechanical movement is coupled to the hydraulics of the pipe. Sophisticated models exist in the literature which attempt to model this complicated process, but these require extensive data about the valve and typically more simplified models are employed.

AFT Impulse offers a more simplified model, whereby the user can specify a closing profile transient of Cv or K vs. time. Remember that when using the K loss model, a value of –1 (negative one) means the valve is closed. The closing will occur as soon as the closing velocity criteria is met (discussed above). If no closing transient profile is entered, the valve will close instantaneously.

You can also specify an opening profile of Cv or K vs. time, which is the profile the valve will follow if it reopens. If no opening transient profile is entered, the valve will open instantaneously.

The zero time for the closing transient is relative to when the valve starts to close because of backflow. For the opening transient, it is relative to when the valve starts to open.

Control Valve Specifications window

The Control Valve Specifications window is shown in Figure 6.13. Control Valve junctions are always internal to the system, with two connecting pipes. This junction type allows modeling of valves that offer



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 basic Specifications window formats, displaying the connecting pipes in a fixed format. The Control Valve junction does not have an explicit flow direction, but adopts a flow direction from the connecting pipes.

Figure 6.13 The Control Valve Specifications window

Control Valve types

You can model four types of control valves:

• Pressure Reducing Valves (PRV’s)

• Pressure Sustaining Valves (PSV’s)

• Flow Control Valves (FCV’s)

• Pressure Drop Control Valves (PDCV’s)


8/10/2007 4:10:28 PM Page 233


Loss information for a control valve is not required, because control valves are dynamic devices that change their geometry in response to the pipe system behavior. The loss that results is that required to maintain the control parameter. You can, however, specify the full-open loss. This is the loss that will occur should the valve fail to a full-open state.

A PRV is a device that controls the pressure in a pipe system. The PRV maintains a constant control pressure downstream of the junction as long as the upstream pressure exceeds the control pressure. If the upstream pressure is lower than the control pressure, the ability to control pressure is lost.

A PSV is similar to a PRV in that it controls pressure in a pipe system. While the PRV maintains a constant downstream pressure, the PSV maintains a constant upstream pressure. If the downstream pressure rises higher than the control pressure, the ability to control pressure is lost.

An FCV is a device that maintains a constant flow rate in a pipe system. By setting the junction to an FCV type and entering a flow rate, the junction will limit the flow through the connecting pipes to be equal to the control flow rate. The FCV can lose its ability to control flow when the pressure drop across it becomes zero.

A PDCV is a device that maintains a constant (stagnation) pressure drop. For this option, valve failure cannot be modeled. An indicator that an unrealistic pressure drop has been demanded is a failure to obtain a converged solution.

PRV/PSV static vs. stagnation pressure The control pressure for a PRV or PSV can be either static or stagnation. The default selection is static, as this is the most frequent application in industry. See Chapter 8 for more information.

Action if setpoint not achievable

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



pressure to maintain control, thus acting like a pump. If this occurs, the user is clearly warned in the output. The Always Control option is useful because it gives a clear indication to the user of how far away the valve is from being able to control.

For transient analyses, if the Always Control is used and the valve fails, the transient analysis will terminate.

For flow control valves, other options for valves which cannot achieve control are to have the valve Open Fully or Close. By default the valve fully opens. Once the control valve loses control during steady-state, whether open or closed, Impulse will solve the model again.

Pressure control valves are more complicated. These valves have two potential actions when control cannot be achieved. The first action is when the control cannot be achieved due to insufficient upstream pressure. The second is due to excessive downstream pressure. The default behavior for pressure reducing valves is to Open Fully due to insufficient upstream pressure, and close due to excessive downstream pressure. The defaults for pressure sustaining valves are the opposite. You can override the default behavior by clearing the Use Default Actions selection and specifying the desired actions.

AFT Impulse will determine if the valve can maintain control based on the Loss When Fully Open value. If this fully-open pressure drop is greater than the pressure drop which balances the system, the control valve fails. In other words, the control valve can’t open wide enough (becoming less restrictive) to achieve control, so it fails. If you enter data in the Open Percentage table on the Optional tab, you can select to have the data taken directly from the table for the 100% open case. You can also choose to not have any losses associated with the valve in the full-open state by selecting None. Note that when a control valve fails open this fully-open loss will be used for the valve.

It is possible for the valve to control properly during the steady-state, and then lose control during the transient part of the run. In fact, it can lose control and regain it repeatedly.

Special Conditions

Control valves have two special conditions. To open fully and not control, or to close.


8/10/2007 4:10:28 PM Page 235


Open Percentage Table

The Optional tab allows data to be entered for special control valve characteristics. Specifically, the valve Cv and Flow Area can be Specified vs. the Open Percentage of the valve. This data does not affect AFT Impulse’s flow solution. However, engineers frequently desire to know the valve’s open percentage during operation to ensure it meets design requirements. The Valve Summary (in the Output window) always displays the valve Cv, and if data is specified for open percentage and flow area it will also display open percentage and flow area at the operating point.

Transient Data

The Control Valve junction is the only one with two transient tabs. The first is the valve reaction to inability to achieve control, and the second is to the control point. More information on transient data, including event transients, is given earlier in this chapter and in Chapter 10.

Failure Transient Control valves have what might be called inherent event logic. That is, rather than respond to events at some other location, the failure transients respond to events of the valve itself. Specifically, the events are the valve's loss of control and possibly regaining of control.

The failure transient is the valve action when it loses control. If the valve is set to close upon failure, the closure transient will be used. This is a K or Cv table vs. time. If the valve were to reopen, the opening transient would be used. The time scale for the closing or opening of a valve is relative to when the failure occurs. That is called time zero in a relative sense.

You can elect to not enter data for the closing or opening transient. If the valve loses control, the valve action is assumed to be instantaneous. If the failure action is to fully open, then the closing transient data will have no meaning and not be used. In a sense the opening transient will have meaning, but in reality when the valve loses control it is already fully open by that point, so no transient will occur. The valve will just stay fully open.

The closing and opening transients are thus intended for use when the valve is set to close upon failure to control.



Control Transient The other type of transient data supported by the Control Valve junction is the Control Transient. This is an optional transient you can apply to the control point. For instance, a change in control pressure for a regulator valve can be modeled.

Control transients support events, although only a single event can be modeled. Thus the regulator pressure could be modified when some event occurred in the system.

Dead End Specifications window

The Dead End Specifications window is shown in Figure 6.14. The Dead End junction must always have one connecting pipe. This junction type is one way to define flow in a pipe as truly zero.

Figure 6.14 The Dead End Specifications window

In steady flow, dead ends have no flow and thus play no role in the hydraulic solution to the system. However, in transient flow pipes that terminate in dead ends can have flow. Thus dead ended lines can and do


8/10/2007 4:10:28 PM Page 237


influence the transient response of the system. Unless you know otherwise for a specific system, dead ends should always be included in the model.

The Dead End Specifications window follows the first of the two basic Specifications window formats, displaying the connecting pipes in a fixed format. The only required information is the connecting pipe and the elevation.

Gas Accumulator Specifications window

The Gas Accumulator Specifications window is shown in Figure 6.15. The Gas Accumulator junction allows up to twenty-five connecting pipes. This junction type is useful for modeling a gas volume designed as a surge protection device (Figure 6.16). The Gas Accumulator may be integral with the pipe system, separated by an orifice, or by a short connector pipe. The waterhammer theory for a gas accumulator is given in Chapter 9. During steady flow the gas accumulator typically acts a lossless branch junction, although it can be made to act like a pressure junction. This is discussed later in the gas accumulator section.

The Gas Accumulator Specifications window follows the second of the two basic Specifications window formats. A table displays the connecting pipe information. The table grows in size to accommodate up to twenty-five connecting pipes. After you add a fifth pipe a scroll bar appears, allowing you to review all the pipes in the table.

Initial conditions and polytropic process

AFT Impulse needs to know the initial volume of gas. This can be determined in one of two ways. First, the Actual Initial Volume can be entered. If the accumulator is passive, the initial volume depends on the initial pressure at the accumulator. Thus the initial pressure must be known in order specify the Actual Initial Volume. One way to do this is to first run the model in Steady Only mode to determine the pressure.

Some gas accumulators have an active design (with compressors) which maintain a specified volume. In such a case, the initial volume is known and independent of the gas pressure, and can thus be entered directly.

The second way is to enter the gas volume at some user specified pressure. This is called “Volume at Specified Pressure”. When selected,



the user must specify a Reference Gas Volume and Reference Gas Pressure. Frequently this will be the gas volume at atmospheric pressure. When the steady-state is run, the gas volume will be determined from the steady-state pressure, assuming the gas volume ratio is inversely proportional to the gas pressure ratio.

The gas volume will change during the transient according to the thermodynamic law. The law requires a polytropic constant to relate the pressure to the gas volume. An isentropic process should use a polytropic constant equal to the specific heat ratio of the gas (for air it is 1.4). An isothermal process should use a polytropic constant of 1. Typically the process falls in between the two and an average polytropic constant can be assumed (for air the average is 1.2).

The gas in the accumulator is assumed to behave as an ideal gas.

Figure 6.15 Gas Accumulator Specifications window


8/10/2007 4:10:28 PM Page 239


Maximum and minimum volumes

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

Interface with pipe system

The Gas Accumulator can connect into the pipe system in three ways: it can be integral with the pipe, it can be separated by an orifice, or it can be connected by a short pipe.

These options are provided in the Flow Restrictor and Connector Pipe areas. If neither option is used, the accumulator is assumed to be attached directly to the pipe with no hydraulic loss.

If the Flow Restrictor is used, the accumulator is assumed to be separated from the pipe system by an orifice, which causes a hydraulic loss as the liquid flows in and out of the accumulator. The loss is specified by entering a flow area and discharge coefficient. Further, this loss can be different for inflow vs. outflow.

LC

ConnectorPipe

Orifice

PA

Am&

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.


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.


8/10/2007 4:10:28 PM Page 241


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 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.


8/10/2007 4:10:28 PM Page 243


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

Liquid Accumulator Specifications window

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

Figure 6.19 Liquid Accumulator Specifications window

The Liquid Accumulator Specifications window follows the second of the two basic Specifications window formats. A table displays the connecting pipe information. The table grows in size to accommodate up to twenty-five connecting pipes. After you add a fifth pipe a scroll bar appears, allowing you to review all the pipes in the table.



Initial conditions and elastic process

The relationship between liquid accumulator volume and pressure is assumed to be linear. The amount of volume change for a given amount of pressure change is given by the elasticity of the accumulator. This input will need to be obtained from test data or approximated based on mechanical correlations.

The Initial Volume of the tank is a required input. This volume will change during the transient according to the linear law.

Pump Specifications window

The Pump Specifications window is shown in Figure 6.20. The Pump junction type requires one or two connecting pipes. This junction type allows you to model the pressure (head) that is added to the system by a pump, and the pump response to a trip or startup. The pump can be of either the centrifugal or positive displacement kind. It also provides access to curve fitting tools. Data can also be interpolated between the provided data.

The Pump Specifications window follows the first of the two basic 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 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


8/10/2007 4:10:28 PM Page 245


using a fixed flow, the total pressure/head developed by the pump is calculated. This is useful for modeling positive displacement pumps.

Figure 6.20 The Pump Specifications window. Top shows the Pump Model tab, and bottom shows Variable Speed.

The pump operating speed can be entered as a percent of full speed. It is assumed that the pump curve is entered at full speed. You can use a



speed that is lower or higher than full speed. Impulse will adjust the full speed pump curve using standard affinity laws relating speed ratios to pressure rise ratios and flow rate ratios.

Submerged pump If a pump is submerged, there is no upstream pipe. Thus only one connecting pipe is needed for the discharge. This is modeled by ticking the check box and entering the pump suction conditions.

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

Check valve at pump discharge AFT Impulse offers a dedicated check valve junction, but frequently the check valve is close to the pump discharge and the pipe length between them is very small. This can result in models with unacceptable run times if a check valve junction is used for transient models.

You can model a check valve at the pump discharge assuming negligible pipe length between them. The valve model is relatively simple, with instantaneous closure once backflow occurs. There is also an input area to account for valve closure starting at some other velocity than zero. You also can enter re-opening criteria based on pressure.

Pump configurations

Pump data can be entered for multiple configurations. This is available only for pumps modeled as pump curves. The default is a single configuration.

A pump configuration is a pump with a specific impeller trim and operating speed. Multiple impeller trims and operating speeds can be specified as part of the pump, and then a particular combination can be chosen.

These parameters are set in the Auxiliary Input tab.


8/10/2007 4:10:28 PM Page 247


During transients, the pump speed will vary as the transient conditions affect the pump. The resulting pump speed can be graphed in the Graph Results. The response of the controller during the transient is assumed to be instantaneous.

Figure 6.21 Pump Configuration window Pump Data tab above where raw data is entered and curve fit or interpolated, Composite Graph tab below.



Figure 6.21 shows the Pump Configuration window. Raw data can be entered, imported from file, or pasted from the clipboard. Then a polynomial curve can be fitted to the data or the data can be interpolated. Data for head, NPSH, and efficiency or power can be entered. Also data for the pump’s end of curve and NPSH constant for variable speed can be entered.

If multiple configurations are entered, they are displayed on the Pump Specifications window in dropdown lists for selection (see Figure 6.22).

Multiple configuration selections displayed

Figure 6.22 Pump Specifications window with multiple configurations available.

Controlled pressure or flow (variable speed)

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


8/10/2007 4:10:28 PM Page 249


NPSHR and NPSPR (optional in pump configuration)

The net positive suction head required (NPSHR) or net positive suction pressure required (NPSPR) can be expressed as a function of mass or volumetric flow rate. This data is optional and is not required to completely define the pump. Both the required and available values are given in the Pump Summary in the General Output section.

For variable speed pumps, you can supply an affinity exponent to adjust the full speed NPSHR/NPSPR to that at the alternate speed. According to the affinity laws, the head ratio and speed ratio are related by the square law. However, the NPSHR ratio does not necessarily obey this affinity law. This exponent allows you to account for this.

Note that in the Elevation area a data field is provided for the reference elevation for NPSH calculations. This may be the pump suction elevation or pump centerline.

Efficiency and power usage (optional in pump configuration)

The pump efficiency or power usage can be specified. This does not affect the pump performance during steady-flow, but allows for quick assessment of power used at the operating point and the proximity to BEP (Best Efficiency Point). If efficiency or power data is supplied, the power, efficiency and percent of BEP at the operating point is given in the Output Pump Summary.

During transient flow, the efficiency/power usage is not optional if four quadrant pump modeling is used. In this case, the efficiency/power data will affect the results.

Transient Data

As discussed previously, pumps can be modeled as a pump curve or as an assigned flow. Using assigned flow pumps in steady flow modeling is useful for sizing a centrifugal pump. In AFT Impulse, however, that is not appropriate during transient modeling. When modeling a centrifugal pump transient, a pump with a pump curve is essential. Positive displacement pumps act as constant flow devices, and the assigned flowrate feature allows one to adequately model such pumps.



Pumps with curves Once a pump curve is entered, several options are available for transient modeling. Modeling a pump with a curve is usually associated with a centrifugal pump.

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

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

Known speed transients, no back flow An estimate of the pump speed change with time can be entered in the transient data area (see Figure 6.23). The speed change can be a time-based or event-based transient. The appropriate Transient Model is “Without Inertia”.

This model uses the provided speed data and the pump homologous laws to determine the pump performance. Since the only pump performance data is for forward flow, only forward flow or zero flow can be modeled. As before, if there is a check valve after the pump which prevents backwards flow, this is the best choice when the pump speed is known.


8/10/2007 4:10:28 PM Page 251


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

Pump startup transient with possible reverse flow If the speed for startup is known or can be approximated, and reverse flow is possible (no check valve exists), then a four quadrant model can be used. With the speed known, we use the four quadrant data FH only. Other data for moment of inertia and initial synchronous speed in rpm is not required.

Chapter 9 offers more information on the parameters and equations, and the user is encouraged to consult one of the excellent texts for more background on this method (Chaudhry, 1987, Chapter 4, Swaffield, et al., 1993, pp. 154-166, Wylie et al., 1993, Chapter 7).

Pump transients with inertia There are four pump models in AFT Impulse that account for pump inertial effects and predict the speed over time for either. Two allow reverse flow and/or speed, while two only work with forward or zero flow. These are:



• Trip With Inertia and No Back Flow or Reverse Speed

• Startup With Inertia and No Back Flow or Reverse Speed

• Trip With Inertia – Four Quadrant

• Startup With Inertia – Four Quadrant, Known Motor Torque/Speed

Predicting the response of a pump during a trip or startup requires additional data for the pump. Use of the homologous pump laws is essential. If backflow is possible, "four quadrant" pump curve data is required. The first model mentioned above assumes the presence of a check valve which prevents backflow. The second model also assumes there is no backflow, although no check valve is assumed. In such cases, the pump will operate in the first quadrant. Thus four quadrant data is not required.

The third and fourth models allow backflow and possibly reverse speed, and thus require four quadrant data. All four models will be discussed in the following sections.

Trip with inertia and no back flow or reverse speed When no backflow or reverse speed occurs through the pump, the normal pump curve data for flow vs. head can be used to model the pump response during a trip. Pump power data is also required as a function of flow. Data for the pump rotating moment of inertia and the initial synchronous speed in rpm is required.

This model will use the pump head curve and power curve to determine the pump torque, and use the moment of inertia to predict the pump speed decay over time.

See Chapter 9 for more information on the methodology.

Startup with inertia and no back flow or reverse speed When no backflow occurs through the pump, and the pump motor torque vs. speed is known, the normal pump curve data for head vs. flow and power vs. flow can be used to model the pump response during a startup. Data for the pump rotating moment of inertia and the 100% operation synchronous speed in rpm is required.

This model will use the pump head curve and power curve to determine the torque from the fluid, balance this with the torque from the motor, and use the moment of inertia to predict the pump speed rise over time.


8/10/2007 4:10:28 PM Page 253



Trip with inertia – four quadrant This model allows the pump to flow and/or spin backwards, and requires pump data on the head and torque characteristics for reverse flow and speed. Data for the pump rotating moment of inertia and the initial synchronous speed in rpm is required.

In addition, dimensionless curves of FH and FB are required. The parameters FH and FB are properties of the pump, and it is typically assumed that for pumps of similar specific speed they are invariant. Figure 6.24 shows data for a pump with a specific speed of 0.46 as given by Wylie, et al., (1993, pp. 147-148).

Figure 6.24 Pump with four quadrant data for pump trip modeling

AFT Impulse provides data of FH and FB vs. θ for pumps of twenty-one specific speeds. The references come from Donsky (1961) as given by Wylie, et al. (1993), Brown & Rogers (1980), and Thorley (1996). See Chapter 9 for a list of the available selections. The conventional assumption is that you should use four quadrant data for a pump with the



closest specific speed. At lower specific speeds, Brown & Rogers (1980) question this assumption.

If you should need to construct data for other sets, please consult the three textbooks referenced below for additional information. Note that the textbooks do not give a method for constructing these data sets, but they contain references that you can pursue.

Chapter 9 offers more information on the parameters and equations, and the user is encouraged to consult one of the excellent texts for more background on this method (Chaudhry, 1987, Chapter 4, Swaffield, et al., 1993, pp. 154-166, Wylie et al., 1993, Chapter 7).

Startup with inertia - four quadrant, known motor torque/speed When backflow can occur through the pump, and the pump motor torque vs. speed is known, four quadrant modeling is required to model a startup. The options for four quadrant data are the same as just discussed for pump trip cases. Data for the pump rotating moment of inertia and the 100% operation synchronous speed in rpm is required.

This model will use the four quadrant data to determine the torque from the fluid, balance this with the torque from the motor, and use the moment of inertia to predict the pump speed rise over time.


Estimating the pump inertia Chapter 9 offers a method described by Wylie et al. (1993) to estimate the moment of inertia, I, for a pump and the entrained liquid.

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


8/10/2007 4:10:28 PM Page 255


Pumps also support dual events, both cyclic and sequential. When the first event is a pump trip, the second event will always be a pump start, and vice versa. This allows a pump to be tripped and then restarted at some later time, for example. The restart can occur while the pump is still spinning down.

If the first event is a pump trip using "Trip With Inertia and No Back Flow or Reverse Speed", then the restart model will be "Startup With Inertia and No Back Flow or Reverse Speed", and vice versa. This means that motor torque vs. speed data is required to model the restart.

On the other hand, if the first event is a pump trip using "Trip With Inertia - Four Quadrant", then the restart model will be "Startup With Inertia - Four Quadrant, Known Motor Torque/Speed", and vice versa. As before, this means that motor torque vs. speed data is required to model the restart.

Controller transient If the pump has a variable speed controller (see Figure 6.25), a control flow or pressure transient can be specified. Again, this transient can be time-based or event-based (see Figure 6.26).

Figure 6.26 transient essentially increases the speed of the pump to achieve 300 m3/hr of flow.

Note that the resulting pump speed transient with time can be saved to the output file (see Transient Control window in Chapter 5). If saved, the speed can be plotted vs. time.

Pumps as fixed flow When a pump is modeled as a fixed flow, then the flowrate can be varied over time. This flow transient can be time-based or event-based. The transient data might look like that in Figure 6.26.



Figure 6.25 A pump with a variable speed controller can have a transient on the control point.

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


8/10/2007 4:10:28 PM Page 257


Figure 6.26 A control point transient can be entered for a pump.

Special Conditions

Pump junctions in AFT Impulse have two types of Special Conditions. The first is to turn the pump off and have no flow through it. This is called "Pump Off No Flow". The second type of Special Condition turns the pump off but allows flow to go through the pump. This is called "Pump Off With Flow Through".

Why would one want to use one Special Condition rather than the other? The simple answer is that the first Special Condition is more appropriate for positive displacement type pumps and the second for centrifugal pumps.

Here's why. When a positive displacement pump is turned off and has a pressure difference across it, it will usually act like a closed valve and not allow flow to go through it. Thus the first Special Condition would be most appropriate. For instance, assume one wants to model the transient that occurs during the startup of a positive displacement pump. The best way to do this would be to use the first Special Condition with



no flow, and then input a flowrate transient in which the first data point is zero flow.

On the other hand, when a centrifugal pump is turned off and has a pressure difference across it, in the absence of other valving which prevents flow the pump will usually allow flow to go through it. Thus the Special Condition that allows through flow would be most appropriate. For example, assume one wants to model the transient during the startup of a centrifugal pump. One would use the second Special Condition, and, in conjunction with a pump curve entered at 100% speed, input a speed transient with the initial speed as zero.

Representing Multiple Pumps With One Junction

If multiple identical pumps exist in parallel or closely spaced in series, rather than modeling each as a separate junction they can all be collapsed into a single junction. When doing so, intermediate piping is ignored. This is often advantageous in that short pipes around the pump which can cause long run times can be ignored.

To use this feature, specify the number of pumps at the location and whether they are in parallel or in series.

Viscosity corrections

Viscosity corrections to standard pump curves have been published in the Hydraulic Institute Standards. The method, based on the 14th edition, 1983, is employed. The method itself consists of a semi-logarithmic diagram relating flow rate, head rise and viscosity. The details of this diagram have been digitized into numerical relationships inside AFT Impulse 4.0. The corrections to head rise, flow rate and efficiency, referred to as CH, CQ and CE, respectively, are given in the Pump Summary output.

The viscosity corrections are assumed to be valid during transients.

Graphing pump data

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


8/10/2007 4:10:28 PM Page 259


Relief Valve Specifications window

The Relief Valve junction type can have one or two connecting pipes. This valve stays closed until it cracks open, at which time it relieves to an external pressure or to another pipe system.

The Relief Valve Specifications window, shown in Figure 6.27, follows the first of the two basic Specifications window formats, displaying the connecting pipes in a fixed format.

Figure 6.27 Relief Valve Specifications window

Three relief valve configurations: internal, exit and inline

The Relief Valve junction can be an internal relief valve, an exit relief valve, or an inline relief valve. If internal, the valve relieves into another pipe which was previously isolated. In this case the valve can only have two connecting pipes. Before cracking open the internal relief valve acts like a regular valve which is closed.

A Relief Valve junction that is at an exit relieves to an external ambient pressure such as the atmosphere. An exit relief valve only has one



connecting pipe. Relief valves that are located at an exit must also be supplied with an associated back pressure.

An inline Relief Valve junction discharges to an external ambient pressure, and is located between two pipes. When closed, the valve is a lossless connection and allows flow between the pipes.

Specifying losses

You specify the losses for the Relief Valve on the Valve Model tab in the Specifications window. For convenience, you can specify the constant loss characteristics of a Relief Valve as a valve coefficient (CV), as a loss factor (K), or as a variable Cv where Cv depends on pressure. Chapter 8 details the relationship between K and CV.

Transient valve cracking

Transient data for a relief valve is only used for the constant Cv and constant K models. The variable Cv model does not use pre-defined transient data.

The transient data for a relief valve is somewhat different from other junctions. The transient data does not refer to the absolute time in the model. Time zero is the moment of cracking, not the start of the model.

You specify the transient relief valve data by clicking the Transient Data button. A dialog window will open where you can enter the data. Once entered, the data will display in the Transient Input Data table in the window.

A zero value of Cv indicates a closed valve. For a lossless valve, enter a very large Cv thus approximating infinity.

To simulate a closed valve with one of the K factor methods, use a K factor of –1 (negative one).

Transients with variable Cv

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


8/10/2007 4:10:28 PM Page 261


until it is fully open. Any further pressure increases do not increase the Cv, and so the fully open Cv is used for all higher pressures.

In general the cracking pressure can be specified as either an absolute pressure/head or a pressure/head difference. When working with the variable Cv model, the cracking pressure must be entered in pressure/head difference. In addition, the first Cv data must equal zero, and the corresponding pressure/head to zero Cv must be the same as the cracking pressure/head difference. See the Relief Valve Theory section in Chapter 9 for more information.

Reservoir Specifications window

The Reservoir Specifications window is shown in Figure 6.28a-b. The Reservoir junction type allows you to connect up to twenty-five pipes. One connecting pipe is required. The Reservoir junction is convenient for specifying a fixed pressure in your system. During a transient, you can vary the liquid surface elevation in a prescribed way.

A Reservoir junction applies a defined pressure at the junction location in the model. When solving a pipe flow system, a Reservoir causes the rest of the system to distribute the flow in a manner consistent with the defined pressure.

The Reservoir Specifications window follows the second of the two basic Specifications window formats. A table on the Loss Coefficients tab displays the connecting pipe information. This table grows to accommodate up to twenty-five pipes. After you add a fifth connecting pipe, a scroll bar appears, allowing you to review and enter loss factors for all pipes in the table.

The pipe table displays the reference positive flow direction of each connecting pipe. To enter loss factors, select the cell in the table and edit the value. Each pipe can use a different loss model or custom value.

The depth or elevation for each connecting pipe can be entered in the table. The depth is the distance below the liquid surface level.



Figure 6.28a The Reservoir Specifications window Reservoir Model tab.

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


8/10/2007 4:10:28 PM Page 263


Figure 6.28b Individual pipe depths (or elevations) and loss coefficients may be specified in the Reservoir Specifications window

Transient Data

On the transient data tab the surface elevation can be changed with time. More information on transient data, including event transients, is given earlier in this chapter and in Chapter 10.

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.

The exit pressure to which the Spray Discharge junction discharges must be entered.

Two geometries can be modeled. The first is a normal, one hole spray, where the discharge flow area of the hole is entered. The second is a sparger which, in principle, works in the same way as a spray. For the sparger, the flow area of a single hole and the number of holes are entered. This assumes that the holes are hydraulically close in proximity and are the same area.

Figure 6.29 The Spray Specifications window


8/10/2007 4:10:28 PM Page 265


Loss Model

The discharge coefficient for the junction is entered in the appropriate box. The physical area of the exit flow can also be entered. These two parameters are used to calculate the discharge flow rate, which depends on the difference between the internal pressure in the system and the specified exit pressure.

The K Fire Sprinkler model uses the K value as defined by the fire sprinkler industry. Whereas the standard K loss value is dimensionless, the K Fire Sprinkler has units associated with it. See Chapter 8 for more details.

Transient Data

On the transient data tab the discharge flow area or K data can be varied with time to simulate the opening or closing of the junction. More information on transient data, including event transients, is given earlier in this chapter and in Chapter 10.

Graphing Spray Discharge data

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

Surge Tank Specifications window

The Surge Tank Specifications window is shown in Figure 6.30. The Surge Tank junction allows up to twenty-five connecting pipes. This junction type is useful for modeling a small vessel designed as a surge protection device (see Figure 6.31). The Surge Tank may be integral with the pipe system, separated by an orifice, or by a short connector pipe.

The Surge Tank Specifications window follows the second of the two basic Specifications window formats. A table displays the connecting pipe information. The table grows in size to accommodate up to twenty-five connecting pipes. After you add a fifth pipe a scroll bar appears, allowing you to review all the pipes in the table.



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.


8/10/2007 4:10:28 PM Page 267


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

ConnectorPipe

Orifice

Tm&

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.



One-Way Surge Tank

One drawback of surge tanks is that they must be located in low pressure parts of the system because high pressure locations require a very tall surge tank (a water surge tank located where the pipe system is at 690 kPa must have a liquid surface at least 70 meters higher than the pipe). Unless there is a check valve connecting the surge tank to the pipe. Since the tank can only flow out, it is only useful to limit low transient pressures by flowing liquid into the system when the pressure drops. However, when the pressure rises, the check valve closes and the tank cannot accept any fluid and thus decrease the system pressure.

Since the surge tank liquid is isolated from the pipe system during the initial steady-state calculation, it is required to specify the liquid surface level in an otherwise optional input field. The Optional tab offers an area to specify the initial liquid height. This is discussed below in “Special features for steady flow”.

Such a design is called a One-Way Surge Tank, and it can be modeled by defining the check valve characteristics.

Transient Data

On the transient data tab the surface pressure can be varied with time to simulate a tank pressurization. More information on transient data, including event transients, is given earlier in this chapter and in Chapter 10.


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


8/10/2007 4:10:28 PM Page 269


On other occasions, one may want to model a fixed steady-state liquid level at a location, but is interested in the liquid level changes in a finite size tank.

These two possibilities have different implications which will be explored here, along with the standard use where the surge tank acts like a branch. The both cases the input is on the Optional tab (see Figure 6.32).

Figure 6.32 Surge Tank junctions allow the user to specify an liquid level for special steady-state behavior.

Surge tank as a branch during steady-state When a Surge Tank is used as a surge suppression device as in pumped systems, it will act like a branch during steady flow. There will be no flow in or out of the tank, and the surface level will equal the local gradeline of the system. Thus the surface level is not an input, but is calculated. This is the default behavior of the Surge tank junction.

During the transient, flow will go in and out of the tank, and the liquid level will change.



Surge tank as a pressurizer during steady-state When a pressurizer is modeled as a surge tank, there will typically be no other pressure type junctions in the system. In fact, the pressurizer is the component which controls the static pressure on the system during steady-state operations by virtue of its liquid surface level. However, during transients the liquid surface level will change. Thus there is a need to allow a surge tank to act as a pressure junction during the steady-state (in which the liquid level is fixed), and then like a normal surge tank during the transient (in which the liquid level varies). This is the function of the Initial Liquid Height for Steady-State Flow on the Optional tab (see Figure 6.32).

When the surge tank is a pressurizer during steady-state, it has zero net flow. The liquid level entered will be the steady-state static system pressure. A flow balance exists at the junction because the net flow is zero.

During the transient, flow will go in and out of the tank, and the liquid level will change.

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

Difference between a Reservoir and Surge Tank A Reservoir junction is assumed to be sufficiently massive in volume that surface level changes do not occur during the simulation. A good physical example of this is when the liquid source is a large lake.

However, sometimes the liquid source comes from a finite size tank, and the liquid surface level will in fact change during the simulation. To track the change, the geometry of the source tank must be known. This can be entered for a Surge Tank, but not a Reservoir.

Model as dipping tube vessel

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


8/10/2007 4:10:28 PM Page 271


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).

LL

LC

ConnectorPipe

Orifice

Tm&

LA

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 in the Arrangement area using provided lists and images.

Loss factors

There are two Loss Models available for Tee/Wye junctions: Simple and Detailed. If you choose the Simple type of Tee/Wye, no loss factors will be calculated. This is equivalent to a Branch with no loss factors entered for the connecting pipes. In some systems the losses associated with a Tee/Wye are small compared to the overall loss in the system. To avoid unnecessary calculations you should choose the Simple type of Tee/Wye unless you need the actual losses.

Figure 6.34 The Tee/Wye Specifications window

The Detailed type of Tee/Wye will calculate the loss factors as they depend on the flow split as well as the geometry. Because of the dependence on the flow split, loss factors usually cannot be calculated before performing an analysis. AFT Impulse incorporates the best available models to calculate tee and wye loss factors during the actual solution. These calculations account for the percentage of the flow split


8/10/2007 4:10:28 PM Page 273


and the angle of connection. This offers a significant advantage over hand calculated losses.

If you have specific loss factors that you want to use for a tee/wye junction, you should use a Branch junction instead.

The Tee/Wye junction functions the same during the steady-state and transient.

Valve Specifications window

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

Figure 6.35 The Valve Specifications window



The Valve 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 connecting pipes. Consistent with AFT Impulse's convention, the loss factor base area is referenced to the upstream flow area shown as the Base Area. Valve junctions are typically internal to the system, with two connecting pipes.

Valves that are located as exit flow control mechanisms are specified as exit valves. These valves require an associated back pressure definition. Exit valves can only be connected to a single upstream pipe. To specify an exit valve, tick the Exit Valve checkbox. Then specify the exit pressure.

Numerous handbook loss factors for generic valve configurations are given in the Valve Data Source area by selecting Handbook Data.

If you input Cv vs. open percentage on the Optional tab, you can enter the actual open percentage and AFT Impulse will calculate the Cv.

For more information on the loss models for these valves, refer to Chapter 8.

Transient Data

On the transient data tab a transient Cv or K table can be entered to simulate the opening or closing of the valve. For a closed valve, the K factor is infinite. To account for this, enter a K factor of –1 (negative one) in the transient data table. More information on transient data, including event transients, is given earlier in this chapter and in Chapter 10.

Special Conditions

The Special Condition for a valve always closes it.

Vacuum Breaker Valve Specifications window

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


8/10/2007 4:10:28 PM Page 275


The Vacuum Breaker Valve Specifications window follows the first of the two basic Specifications window formats, displaying the connecting pipes in a fixed format. The flow direction through the junction is determined by the defined directions of the connecting pipes.

Figure 6.36 Vacuum Breaker Valve Specifications window

The Vacuum Breaker Valve offers protection for low pressure conditions. When the internal pressure drops below the cracking pressure (usually atmospheric), it opens to admit air (or whatever gas you specify). This keeps the pressure from dropping much below the atmospheric pressure. When the internal pressure rises above the atmospheric pressure, the air is discharged. To prevent large pressure spikes, the discharge flowrate is frequently kept lower than the incoming flowrate by using a smaller discharge flow area for outflow than inflow.

A vacuum breaker junction can connect with one pipe or two. When connected to one pipe, the valve is located at the end of the pipe. When



connected to two pipes, the valve is “inline” and allows air to flow in and out at that location.

During steady-state a vacuum breaker valve with one connecting pipe acts as a dead end. If connected to two pipes, it acts as a lossless valve allowing liquid to flow between the two pipes.

By default, the vacuum breaker junction allows air to flow in and out. Alternatively, it also model only air inflow or outflow.

Three-Stage Valve Modeling

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

Graphing Vacuum Breaker Valve data

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

Volume Balance Specifications window

The Volume Balance Specifications window is shown in Figure 6.37. The Volume Balance junction type is internal to the system and requires two connecting pipes.

The Volume Balance Specifications window follows the first of the two 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.


8/10/2007 4:10:28 PM Page 277


Figure 6.37 The Volume Balance Specifications window

The Volume Balance junction is useful for modeling pipelines with a moving interface between fluids of different density. Such a scenario is in reality a transient application, because the mass in the system is not steady with time. AFT Impulse assumes the interface does not move during the transient.

Volume Balance junctions should be used in conjunction with the variable physical property model (see System Properties window).

Global Edit windows

The Global Pipe Edit and Global Junction Edit windows, accessible through the Edit menu, assist you in making large scale changes to your model.



Global Pipe editing

The Global Pipe Edit window, shown in Figures 6.38 and 6.39, allows you to edit data for multiple pipes at the same time. Here is how you use it:

1. Open the Global Pipe Edit window from the Edit menu.

2. Select the pipes you want to edit from the Pipe list on the left (see Figure 6.38).

3. Click the Select Pipe Data button. This will open the Pipe Specifications window (see Figure 6.1) in which you can enter the data you want to change on each of the selected pipes.

4. Enter the pipe data you want to change and click the OK button in the Pipe Specifications window.

5. The Global Pipe Edit window now lists the parameters to change in a list on the right (see Figure 6.39). Select the parameters you want to change.

6. Click the Apply Selections button.

Note that you cannot change some data, like the pipe ID number, using this global method.


8/10/2007 4:10:28 PM Page 279


Figure 6.38 The Global Pipe Edit window allows you to change data for many pipes at the same time. This is how the window appears when first opened.

Global Junction editing

The Global Junction Edit window, shown in Figures 6.40-6.44, allows you to edit data for multiple junctions at the same time.

Junction data falls into two broad categories. Common data such as elevation and initial pressure which apply to all junctions, and junction specific data, such as a pump curve, which are specific to certain types of junctions. You can switch between the two categories by selecting the option in the upper left (see Figure 6.40).



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

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


8/10/2007 4:10:28 PM Page 281


Choice of editing

common or specific data

Figure 6.40 Global Junction Edit window. This is how the window appears when first opened to edit Common data.

Here is how you globally edit junction common data:

1. Open the Global Junction Edit window.

2. Select the Common Data option in the upper left (see Figure 6.40).

3. Select the junctions you want to edit from the Junction list on the left (see Figure 6.40). Note that all junctions in the model are shown, no matter what the junction type.

4. Click the Select Common Junction Data button. This will open the Common Junction Data Edit window (see Figure 6.41) in which you can enter the data you want to change on each of the selected junctions.

5. Enter the junction data you want to change and click the OK button in the Pipe Specifications window.



6. The Global Junction Edit window now lists the parameters to change in a list on the right (see Figure 6.42). Select the parameters you want to change.


Note that you cannot change some data, like the junction ID number, using this global method.

Figure 6.41 Common Data Edit window for global junction editing.

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


8/10/2007 4:10:28 PM Page 283


Figure 6.42 The Global Junction Edit window after selecting the “Select Common Junction Data” button. Data can now be applied.

Here is how you globally edit specific types of junctions:

1. Open the Global Junction Edit window.

2. Select the Specific Data For option in the upper left (see Figure 6.43).

3. Select the type of junction you want to edit in the dropdown list in the upper left (see Figure 6.43). Here Spray Discharge is selected.

4. Select the junctions you want to edit from the Junction list on the left (see Figure 6.43). Note that only junctions of type selected in the upper left are shown Spray Discharge junctions in this case).

5. Click the Select Spray Discharge Data button (the button name will change based on the type of junction selected). This will open the appropriate junction Specifications window, in this case the Spray Discharge Specifications window.



Option for specific types of

junctions. Junction type

selected in list.

Figure 6.43 The Global Junction Edit window allows you to change data for specific types of junctions (Spray Discharge shown here)

6. Enter the junction data you want to change and click the OK button in the Spray Discharge Specifications window (or whatever window type is displayed).

7. The Global Junction Edit window now lists the parameters to change in a list on the right (see Figure 6.44). Select the parameters you want to change.



8/10/2007 4:10:28 PM Page 285


Figure 6.44 The Global Junction Edit window allows you to change data for specific junction types (Spray Discharge junctions shown here)


C H A P T E R 7

Customizing AFT Impulse

This chapter explains how to customize AFT Impulse in ways that increase your productivity. To accelerate repetitive operations you can customize the way the Workspace looks and behaves and the way output is formatted. Changes to the program's appearance and functionality are saved to the IMPULSE4.INI file in your application data folder m (e.g., C:\Documents and Settings\Your Account\Application Data\Applied Flow Technology\AFT Impulse\4.0).

You can also build custom databases that are read in during startup and used as if they were native to the program. These databases are saved to the IMP_USER4.DAT file in your application data folder.

If you pursue extensive customizing of AFT Impulse, you are strongly advised to make regular backup copies of the two files (IMPULSE4.INI and IMP_USER4.DAT). If the files ever become corrupted you could lose all the custom data.

Tools are also provided to allow creation of network databases to share frequently used data and report formats across a local or wide area network.

Parameter and Unit Preferences

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



during assembly of your model, AFT Impulse offers your default parameters as a first choice. This way you do not have to repeatedly change the parameter in the Specifications windows.

Parameter Preferences

The first parameter you can set is the Use Most Recent Pipe Size Data feature. When this is selected, AFT Impulse remembers your most recent pipe material size setting and sets new or unspecified pipes to this pipe material and size.

You can also set a default pipe material and default friction model. You can change the material and loss model of each individual pipe at any time, but this will be the first choice.

Figure 7.1 Parameter and Unit Preferences window sets default values to use.

There are Default Design Factors. The function of design factors is discussed in more depth in Chapter 8. If you want to set up defaults that

Chapter 7 Customizing AFT Impulse 289

8/10/2007 4:10:28 PM Page 289


are applied to your pipes and junctions, here is where you do it. If you want to change the design factors for existing pipes or junctions, see the Global Pipe Edit and Global Junction Edit topics in Chapter 6.

In the upper right you can specify default parameters for flow rate and pressure head.

Another parameter on this tab is the Default Elevation. The Default Elevation is especially useful when the majority of your junctions all reside at the same elevation. When you specify a default elevation, all the junctions you create will show the default elevation, saving you from having to enter the same elevation repeatedly. AFT Impulse defaults to working without a default elevation specified.

Note: The inputs here merely specify defaults for pipes and junctions not yet placed in the model. If you want to change data for existing pipes and junctions, use the Global Pipe Edit and Global Junction Edit windows.

Unit Preferences

You can set the units you prefer to use in AFT Impulse on the Unit Preferences tab (see Figure 7.2). These units are selected as a first choice in the unit selection boxes. You can choose any of the units available. To set a new default, choose the unit type from the list on the right and click the Set As Preferred Unit button.

You also can customize the units used by AFT Impulse. For each parameter, AFT Impulse can use many different units. However, if only a few are ever typically used, then the others can be removed. This will decrease the number of items in the unit selection boxes, making it easier to choose the one desired. If a preferred unit is removed from those available to Impulse, a new one will have to be chosen before leaving this area. You also must have one set of units available for each parameter.

To customize the units, first choose the parameter to modify on the list on the left. Then, among the unit types on the right, check or uncheck the boxes for the units you want or do not want to see.

You can select or unselect all displayed units by clicking the appropriate button. Also, English and SI units can be selected as a group for parameters by clicking the appropriate button.



Figure 7.2 Using the Unit Preference tab in the Parameter and Unit Preferences window, you can set the default units you want to work with and choose to show only the units you want to work with.

Tip: To choose only English or SI units for all parameters, go to the Unit Preferences tab, choose the Default Unit System then click the Apply Default Units button. This will use AFT Impulse’s default units. This will also change the preferred units.

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


8/10/2007 4:10:28 PM Page 291


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.

Database connections

In the lower left of the Parameter and Unit Preferences window there are two Database checkboxes. If these boxes are checked, your Parameter and Unit Preferences parameters are set up as determined by the database to which you are connected. This is referred to as an active database. To make them inactive, uncheck the boxes or change one of the Parameter and Unit Preferences settings controlled by the database. You can specify database connections separately for parameters and units.

If the checkboxes are unchecked, but enabled, you are connected to a database but the settings are not being passed to the Parameter and Unit Preferences window. The databases are thus inactive. To make them active, check the boxes then click the OK button.

If the checkboxes are disabled, there are no connected databases.

Later in this chapter we’ll discuss how databases are configured and administered through your local or wide area network.

Command buttons

There are eight buttons at the bottom of the Parameter and Unit Preferences window. Impulse has built-in default parameters, units and settings which you can choose by clicking the Impulse Default button. You can also develop your own settings, tailored to your project or industry, and have these used by default (instead of Impulse’s defaults). To make your own default, first select the parameters, units and settings you would like to use, then click the Set As Default button. Your settings will be saved and will be used each time any new project is initiated. If you make changes to the settings, and want to get back to your defaults, click the User Default button. The default settings are updated only when you click Set As Default.

You can save the parameter and unit settings to a file by pressing the Save Preferences button and entering a file name. These setting are loaded again by pressing the Load Preferences button and choosing the file name. The format files you create can be placed on a network for



sharing among a group or company, or incorporated into a company-wide database, allowing standardized reporting.

If you have made changes that you don’t want to keep, click the Cancel button. Click OK to use the settings you have defined.

Workspace Preferences

The Workspace Preferences window is accessed from the Options menu. The features offered in this window allow you to customize the appearance and behavior of the Workspace.

Pipes and Junctions

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

Pipe Line Options Normally, all pipes on the Workspace are displayed using the same line thickness. This thickness value is set using the dropdown list at the top of the Pipe Line Options area. A sample of how the pipe will be displayed is shown immediately below the dropdown list.

There are two additional options that will vary the line thickness according to the pipe diameter. If you choose the Vary Thickness With Pipe Diameter option, the pipe line thickness will be varied according to the pipe physical diameter. The smallest diameter pipe in the model will be displayed as a line 1 pixel thick and the largest diameter pipe will be displayed as a line the thickness you specify in the Pipe Line Thickness dropdown list. All other pipes will be shown using a thickness somewhere in between, according to their diameter.

Similarly, the Vary Thickness Using option will vary the pipe line thickness from the smallest diameter pipe in the model to the largest. However, in this option you can specify the range of pipe diameters and line thicknesses to use. If a pipe is smaller than the minimum diameter specified it will be shown using the minimum thickness. Accordingly, if a pipe is larger than the maximum diameter, the maximum thickness will be used. This option is useful to keep a few very large or very small pipes in the model from skewing the display.


8/10/2007 4:10:28 PM Page 293


Figure 7.3 The Pipes and Junctions tab in the Workspace Preferences window.

By varying the pipe thicknesses you can easily see the feeding pipes, the large flow area pipes, or the small supporting flow pipes.

Each individual pipe allows you to modify its thickness, which overrides the default size set here.

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



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

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

Pipe endpoint adjustments By default, pipes snap to the center of junction icons. You can disable this here if you desire.

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

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

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


8/10/2007 4:10:28 PM Page 295


Icon source Limitations exist in Windows 95, 98 and Me that can affect AFT Impulse’s ability to display large models on the Workspace. The limitations relate to what are called GDI resources. The practical result is that models that display more than 150-250 junctions may have trouble. In such cases, it is recommended the Icon Source be changed to the Large File Icon Method. This allows you to use all icons that are displayed on the Toolbox, plus it offers the ability to rotate icons. The Toolbox Icons Only option only allows display of icons that are shown on the Toolbox. When either of these options are chosen, the Workspace Icon area on each junction Specifications window is disabled. Using different color junctions is therefore not available.

Windows NT, 2000 and XP do not experience these limitations and the user should thus always use the Complete Icon Set.

Display Options

From the Display Options tab you can change the symbols used, text content, and location of the text on the Workspace (see Figure 7.4).



Figure 7.4 The Display Options tab in the Workspace Preferences window.

Workspace grid A grid can be displayed on the background for modeling convenience. The grid can be fine, medium or coarse. The grid lines can be solid or dashed. And finally, the pipes and junctions can be snapped to the grid you specify.

The grid can be included in printouts by specifying its inclusion in the Print Preview/Special window.

Workspace symbols Special symbols can be displayed on the Workspace adjacent to pipes and junctions to denote special behavior. For example, an X (or symbol of your choice) can represent a Special Condition. If you want to use


8/10/2007 4:10:28 PM Page 297


these symbols, click the “Use” checkbox and specify the desired symbol. By default all symbols are selected for use.

There are special symbols displayed on the Workspace for transient related data. These symbols can be modified here. By default, a "T" symbol is displayed if a junction has transient data. A "#" symbol means that a Transient Special Condition has been specified. And a "*" symbol means that the junction is to have no reflections.

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

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

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

Allowable Workspace label movements By default, pipe and junction labels Workspace can be moved to new locations. You can change this functionality to only allow movements when holding the shift key.

Popup menu The popup menu can be disabled (it is enabled by default) here. The popup menu is a context sensitive menu available on the Workspace.



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

Background Picture Scaling Background pictures loaded on the workspace can be scaled by specifying a scale multiplier here.

Colors and Fonts

The color of pipes, junction labels, etc., can be set on the Colors and Fonts tab (see Figure 7.5). To change the color of an item such as the Workspace background, select the Workspace Background from the list and click the Set Color button. You also can just double-click the Workspace Color itself. The color of individual pipes can be assigned in each pipe window. This will override the default pipe color specified here.

In addition to the color area, the junction and pipe font is displayed. To change fonts, click the Change Fonts button to open the Fonts window then, choose a new font or font style. A sample is displayed above the Change Font button.

Sample Workspace

The tab for Sample Workspace shows you a sample of how the Workspace pipes and junctions will appear if you click the OK button.


In the lower left of the Workspace Preferences window there is a Database checkbox. If this box is checked, your Workspace Preferences parameters are set up as determined by the database to which you are connected. This is referred to as an active database. To make it inactive, uncheck the box or change one of the Workspace Preferences settings controlled by the database.


8/10/2007 4:10:28 PM Page 299


Figure 7.5 The Colors and Fonts tab in the Workspace Preferences window.

If the checkbox is unchecked, but enabled, you are connected to a database but the settings are not being passed to the Workspace Preferences window. The database is thus inactive. To make it active, check the box then click the OK button.


Later in this chapter we’ll present a discussion of how databases are configured and administered through your local or wide area network.

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 Impulse’s defaults). To make your own default, first select the settings you would like to use, then click the Set As Default button. Your settings will be saved and will be used each time any new project is initiated. If you make changes to the settings, and want to get back to your defaults, click the User Default button. The settings are updated only when you click Set As Default.

You can save the parameter and unit settings to a file by pressing the Save Preferences button and entering a file name. These setting are loaded again by pressing the Load Preferences button and choosing the file name. The format files you create can be placed on a network for sharing among a group or company, or incorporated into a company-wide database, allowing standardized reporting.

If you have made changes which you don’t want to keep, click the Cancel button. Click OK to use the settings you have defined.

Toolbox Preferences

The Toolbox Preferences window is accessed through the Options menu. Figure 7.6 shows the layout of the Toolbox Preferences window. The features in this window allow you to customize the Toolbox in the manner that is most productive for you.

Once you have changed the appearance and functionality of the Toolbox, you can set your changes as the user default values by clicking the Set As Default button. From then on, AFT Impulse will always load with your default Toolbox settings.

Customizing Toolbox contents

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


8/10/2007 4:10:28 PM Page 301


Figure 7.6 Toolbox Preferences window allows you to customize the Workspace Toolbox.

You can also reorder the icons on the Toolbox by dragging and dropping them to the desired locations on the Current Toolbox.

Toolbox shortcuts You can show or hide Toolbox shortcut information.




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.

If the checkbox is unchecked, but enabled, you are connected to a database but the settings are not being passed to the Toolbox Preferences window. The database is thus inactive. To make it active, check the box then click the OK button.


Later in Chapter 7 a lengthy discussion is given of how databases are configured and administered through your local or wide area network.

General Preferences

The General Preferences window (Figure 7.7) allows you to:

• Specify automatic model saving

• Options for opening models with scenarios

• What content to display on the Status Bar

• Require a check of the Waterhammer Assistant before running models

• Enable Tip of the Day feature

• Number of data points to show in input tables

• Whether or not to confirm before clearing output results

• Setting parameters for “button flashing” for curve fit buttons on junction Specifications windows

• Default locations for saving models

• Common database locations


8/10/2007 4:10:28 PM Page 303


• Backup folders where your IMP_USER4.DAT and IMPULSE4.INI files are copied (this is useful for keeping backups should you change computers)

• Specify how AFT Impulse handles transient junction data that does not cover the entire time range of the simulation (either maintain the data as constant at the last data point or extrapolate). Maintain is the default option.

Figure 7.7 General Preference window

Customizing the Output

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



If you make changes to the Output Control window and you want to return to your user default setup, click User Default. Clicking Impulse Default activates the original defaults used by AFT Impulse. No updates to the defaults are made unless you click Set As Default.

You can save frequently-used output formats to file for later reuse. For example, for checking modeling errors you may wish to create an Output Control file named “ALL.DAT” that has all parameters, which you can quickly access when needed. You may have another named “FINAL.DAT” for final reports. The Output Control format files can be located on a PC network for access by all engineers to allow for standard reporting between different projects and groups.

More extensive discussion on using Output Control to customize your output is given Chapter 4.

Specifying graph style preferences

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

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

Specifying Visual Report preferences

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


8/10/2007 4:10:28 PM Page 305


The parameters and text locations are automatically saved with the model. However, you may want to save the settings for use with a different model. To save your preferred layouts to an AFT Impulse visual layout file, click the Save Options button and assign a file name. The default file extension is .ivs. Later you can recall your customized layouts for rapid preparation of the Visual Report window.

The many features in the Visual Report Control window are described further in Chapter 4.

Customizing Pipe Fittings & Losses

The Pipe Fittings & Losses window presents organized lists of pipe loss fittings. This list is customizable. You can add your own fittings to the base list provided with AFT Impulse.

The standard losses provided with AFT Impulse are located in the IMPULSE4.DAT file in the same directory as the AFT Impulse executable file. You should not change anything in this file.

The Fittings & Losses Edit window (Figure 7.8) can be opened from the Database menu.

Building custom databases

One of the more powerful features of AFT Impulse is that it provides extensive customization capabilities. AFT Impulse provides the tools to build custom databases for fluid properties, pipe materials, and all junction types except Volume Balance. AFT Impulse incorporates these databases internally and offers them to you through lists.

There are two types of databases that may be customized: local and external (Note: the AFT Default Internal Database is not included here since it may not be altered). The local database is located in the IMP_USER4.DAT file in the application data directory.



Figure 7.8 Fittings & Losses Edit window allows you to and edit fitting and loss data that is used in pipes.

In AFT Impulse you can create multiple external or local custom databases. The Database Manager window on the Database menu simplifies this process.

When the databases are placed on a network, the data management features described below can be implemented.

Specific, quality-checked data can be entered into databases located on a local or wide area network of PC’s. This data is not accessible to the end users through AFT Impulse itself, so it is guarded data.

Once placed on a network, users can connect to any database found in the Database Manager. Users can connect to as many databases as they want, and get access to all data in databases using Database Manager.

Adding custom fluid properties to the Fluid Database

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


8/10/2007 4:10:28 PM Page 307


currently defined custom fluids, to change the parameters that define the fluids, and to enter new fluids. All fluid properties are assumed to be a function of temperature only.

Figure 7.9 Fluid Database window

Once you have entered a custom fluid, it will appear in the Fluid List area of the System Properties window whenever you use AFT Impulse.

You can add new fluids to the database by clicking the Add New Fluid button and opening a new window (see Figure 7.10). The constants provided allow you to describe the variation of density, dynamic viscosity, bulk modulus of elasticity, and vapor pressure as they depend on temperature or solids concentration for newtonian fluids. In addition, there is an option to describe the variation of density, viscosity, rigidity coefficient and yield stress for a Bingham Plastic non-newtonian fluid as it depends on solids concentration. This window provides the tools to perform polynomial curve fits of fluid data; or, if you already have appropriate constants, you can input them directly. In addition, the raw data can be interpolated



Figure 7.10 Add New Fluid window

Density, viscosity and bulk modulus are required for all fluids. Vapor pressure is required if cavitation modeling is desired.

The Fluid Database window also allows you to open the Change Fluid Data window, which is essentially the same window as the Add New Fluid window. For easy editing, data for the fluid you wish to change is entered for you in the appropriate areas. The original curve fit data you entered is also displayed.

The Add New Fluid/Change Fluid Data windows displays cross-plotting of your custom curves against the curve fit data points. You can also import data from text files into these windows.


8/10/2007 4:10:28 PM Page 309


Adding custom pipe materials to the Pipe Material Database

You can open the Pipe Material Database window from the Database menu. Figure 7.11 shows the Pipe Material Database window. Using this window, you can easily add to the user database the geometric information for any pipe material in the length units of your choice.

Figure 7.11 Pipe Material Database window

After you have entered a custom material, each time you open a Pipe Specifications window the material will be available to you for quick selection. This minimizes the need to work with handbooks.

Pipe materials typically have a range of nominal sizes and a specific type, class, or schedule that describes the actual dimension. AFT Impulse works with the inner diameter of the pipe, since that is the parameter that directly affects the fluid mechanics.

You can also add multiple Friction Data Sets for each pipe material. This allows you to keep, for example, different friction values that may apply to the same pipe under different conditions. These conditions could be related to the age of the pipe or the type of fluid being carried.

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 0.127 mm. This data set is then available to all sizes and types of steel. However, if “25 year usage” is associated only with the 1 inch (2.66 cm ID) of Steel, it will apply only to the pipe schedules that are 1 inch (2.66 cm ID).

You can also associate friction data sets with a specific type (i.e., schedule). For instance, you can make a “25 year usage” and associate it only with 1 inch (2.66 cm ID) schedule 40 Steel pipe and assign its roughness value as 0.152 mm. Then all Steel pipe would use 0.127 mm for “25 year usage” friction while the different value for 1 inch (2.66 cm ID) schedule 40 of 0.152 mm would supercede the value of 0.127 mm associated with all Steel pipe.

Finally, among multiple friction data sets you can specify a default data set.


8/10/2007 4:10:28 PM Page 311


Pipe Physical Properties The Pipe Material Database window allows you to assign physical properties of the pipe. The density and Poisson ratio of the pipe can be entered as constants, and the modulus of elasticity as a function of temperature.

Adding custom junction data to the Component Database

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

To save a Custom junction to the database, open the junction Specifications window and enter all required data. Close the window and choose Add Component to Database from the Database menu. You will be prompted to enter a name for the custom component. Enter a descriptive name and close the window. If you open the junction again, the custom name will be selected in the Database List.

AFT Impulse accepts custom data for all junction types except Volume Balance. Volume Balances are not included because there is no custom information for Volume Balance and thus nothing to store.

Editing the Component Database

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



Figure 7.12 Component Database window shows junction items in database and allows editing

Database Manager

Accessed from the Database menu, Database Manager allows you to –

• Connect to and disconnect from databases

• Create and delete databases

• Review the content of databases

• Move contents between databases

Setting available databases

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


8/10/2007 4:10:28 PM Page 313


button. Contents of a database may be reviewed by selecting a database from the list and clicking on the Review Content button.

Figure 7.13 Available Databases in the Database Manager

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

Connecting and disconnecting to a database

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



which the sections to be used in the model may be selected; i.e. components, fluids or pipe materials.

Figure 7.14 Connect To Database tab of the Database Manager allows you to connect to and disconnect from databases, review their contents and select the sections of a database to be used.

The contents of a connected database may be reviewed by selecting it and clicking on Review Content.

Clicking on Set As Default will establish the list of connected databases as the default list, while clicking on User Default will set the connections to the default list.

Editing databases

The Edit Database tab of Database Manager allows you to create and delete databases, and edit their contents by copying sections or items from one database to another. The tab is divided into a Source Database area on the left and a Changing Database area on the right (see Figure 7.15).


8/10/2007 4:10:28 PM Page 315


To edit a database, first choose a source database by selecting one of the four options:

• Currently available external

• Data From Current Model – components, fluids and materials made in the current model using the Component Database, Fluid Database or Pipe Material Database

• Local User Database – IMP_USER4.DAT located in your Windows directory, where component, fluid and materials data created by the user is stored by default

• Other External – by selecting this option and clicking on the Select Database, you are able to browse and select a database file from any available drive and directory

Select the database tobe 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 bemoved or copied from

the source to thedatabase being

changed

Figure 7.15 Edit Database tab allows you to select sections from a specified database and add them to an existing or new database.

The file and descriptive name of the selected database are displayed as well as the sections available in that database, i.e. fluids, components or pipe materials.



The database to be changed and what changes are to be made are specified in the Changing Database portion of the Edit Database tab.

• Select Database to Edit… – browse to and select an existing database file

• Create New Database… - specify a location, file name and descriptive name for a new database. The file and descriptive name of the selected or created database appears below (Figure 7.16).

• Delete database – deletes the database displayed

• Review content – displays the content of the database listed

• Delete section – deletes from the listed database the section selected from the Database Sections list

Figure 7.16 After specifying a database file name, you will be prompted for a description which will then appear in the various lists of the Database Manager when this database is selected.

Benefits of shared databases

When the Impulse databases are placed on a network, powerful data management procedures can be implemented. Specific, quality-checked data can be entered into databases located on a local or wide area network of PC's. This data is not accessible to the end users through AFT Impulse itself, so it is guarded data.

Once the data is placed on a network, users can connect to any database listed in DATABASE.LIB by selecting the database in the Connect Database window on the Database menu. Users can connect to as many databases as they want, and get access to all data in databases listed in DATABASE.LIB.


8/10/2007 4:10:28 PM Page 317


In addition, special modeling control files can also be included in your network databases. These include:

• Output Control format files

• Model Data Control format files

• Workspace Preference format files

• Parameter Preference format files

• Unit Preference format files

• Visual Report Control format files

• Toolbox Preference format files

If the user connects to a database with these files and selects to be connected to any of these control files, the model will inherit the control formatting specified in the control files. This means that special report layouts, special graphical representations and colors, and other special controls can be decided upon by the database constructor, and those control features can be selected into local models by end users. This is highly useful for creating customized reports for certain customers or project requirements.

By way of example, if the Output Control of Figure 4.15 is specified as coming from the “New Mile High Stadium Project” database, it would appear as shown in Figure 7.17.

If you uncheck the Database checkbox or change certain data in the Output Control window, the database will remain connected but change to an inactive state. When the database is inactive, none of its data flows through to the model. To bring the Output Control data back into sync with the connected database, recheck the Database checkbox in Output Control.

This entire custom database concept lends itself naturally to the idea of specific project databases. Each project has differing requirements that depend on the nature of the project and the end customer. Data for each project can be assembled into separate databases, allowing faster and more accurate transmission of the controlled data and desired reporting formats to engineers.

Maintaining a network database allows commonly used data to be controlled, which may be helpful in instances where quality control is strict.



Figure 7.17 Output Control window with database connected and active.

Creating an enterprise-wide, network database system

Overview

Users of AFT Impulse on a network can access common databases to simplify model verification, improve analysis and reporting consistency, and eliminate redundant work.

To accomplish this there is some initial setup that must be done. First, the data for components, materials, formats, etc., needs to be created and saved to various data files. Secondly, these files must be made available to the users of AFT Impulse by creating a file named DATABASE.LIB in the IMPULSE folder. Thirdly, the users need to connect to the new databases listed and select the sections to use. Each of these broad steps is discussed below. To maintain control and integrity of the data, it would be advised that one person is given the responsibility to create and

Database connection


8/10/2007 4:10:28 PM Page 319


maintain the network databases. The discussion below will be given from the viewpoint of this person.

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

Creating database files

The first thing to do is create the files that contain the information to be shared. These files will be referred to as database files and can contain information about pipe materials, components (e.g. pumps, valves, etc.), fluids, Output Control formats, Visual Report settings, etc. The following are several ways to create database files:

• Use the Edit Database feature in the Database Manager, as described previously. Data can be taken from the current model or other databases.

• Save the control formats to separate files. This can be done on the Output Control window, Visual Report control window, etc.

• Copy and paste sections from one file into another file using a text editor.



Figure 7.18 An example of the relationship between External Databases and External Shared Databases.

This discussion will focus on the first and second methods. Let’s assume that the person responsible for the network database has already created a new pipe material. Also, assume that the goal is to create a network-wide database to contain this material, as well as two Output Control settings and a Visual Report format.

The information for the material is copied from the Local User Database (IMP_USER4.DAT) to a new file using the Database Editor in the Database Manager (see Figure 7.15). For the source database choose Local User Database and then select Pipe Materials. Then click Create New Database and enter a file name, VailMaterials.dat, for this example. Then enter a description of the database. This description is how AFT Impulse users will refer to the database and should concisely describe what is in the database (for this example we’ll use “Vail Resort Materials”). Then click the Copy button to copy the pipe materials from

Vail Resort Materials

Development Output Control Settings

Final Report Output Format

Visual Report Format

AFT Default Internal Database

AFT Impulse Local User Database

My Custom Visual Report Format

My Custom Output Settings

External Shared Databases (DATABASE.LIB)

VailMaterials.dat Development Output.dat

Final Report Output.dat

Visual Report Format.dat

Impulse Internal Databases IMPULSE4.DAT IMP_USER4.DAT

Database

Database Groups and

External Databases (IMP_DBUSER.LIB)

My Report Output.dat

My VR Format.dat


8/10/2007 4:10:28 PM Page 321


the Local User Database to this new database file. The pipe material database file has now been created and is ready to be used.

Now we need to create two Output Control settings, one with a wide variety of parameters for use during the project development phase and another, more concise set, with the pertinent data for a final report. Again, there are many ways of creating these database files. For the development report format, let’s create a database file directly. Open the Output Control window and select the pipe and junction output parameters needed. Then click Save Control Format and enter the file name (Development Output.dat in this case). You may want to open the file in Notepad or other text editor and change the file heading (first line of the file) from the default to something more descriptive, “Development Output Control Settings”, for example.

Now, for the settings for a final report, let’s create the file using the Database Editor in the Database Manager (see Figure 7.15). Open the Output Control window again, select the parameters needed for the final reports and click OK. Then open the Database Manager and go to the Edit Database tab (see Figure 7.15). Click Data From Current Model and then choose the Output Control section. Then click Create New Database and enter a file name, Final Report Output.dat, for this example. Then enter a description of the database, something descriptive like “Final Report Output Format”. Then click the Copy button to copy the Output Control format from the local model to the new database file.

To complete the database files, we need to create a Visual Report format. To do this a model must first be run; any model can be used in this case. After the model is run and a converged answer is obtained, go to the Visual Report Control window and select the parameters to show on the Visual Report. Then click Save Options, then enter the name of the file, for example, Visual Report Format.ivs. Again, you may want to open the file in Notepad or other text editor and change the file heading (first line of the file) from the default to something more descriptive, “Visual Report Format”, for example.

Sharing database files using DATABASE.LIB

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



On the other hand, if the database is to be common among many users and shared over a network it is put into the External Shared Database group.

Continuing the above example, the goal is to share these database files, so ultimately they need to be External Shared Databases and listed in the DATABASE.LIB file in the Impulse directory on the network. However, since Impulse does not write directly to the network files for security reasons, it is easiest to temporarily treat these files as local External Database files which will be listed in the IMP_DBUSER.LIB file located in the application data folder of the local machine.

It is recommended that External Shared Databases files be copied to the network server and placed in a subfolder of the AFT Impulse folder (e.g. AFT Products\AFT Impulse\Databases). The files may, however, be located anywhere on the network that is accessible to the users. For our example, we are going to put all the database files on the network in a subfolder called Databases. It is easier to copy the files to the Databases folder before they are made available in AFT Impulse. This will help establish the correct paths to the files as the files are made available using the Database Manager.

Once the database files are in their final locations, open the Database Manager and click the Edit Available List button on the Connect to Database tab (see Figure 7.13), and select Add Engineering Database. Each of the new database files we have created needs to be added to the Available Databases. Once added, the description of the database you just added appears in the Available Databases list. When you have added all the new database files, click Close and the External Database listing file (IMP_DBUSER.LIB) in your application data folder is automatically updated.

Now we need to discuss how the paths should be setup. When you add a database, AFT Impulse will write the full path of the database to the External Database listing (IMP_DBUSER.LIB). That means that if you browsed to the database file using a mapped drive letter then the path will include that drive letter. However, if other users try to access the databases but have the AFT Impulse folder mapped to a different drive letter they will not be able to locate the database files. Other options include using either relative paths or the computer name in the path (e.g. \\Network Server\Eng Apps\AFT Products\AFT Impulse\Databases\filename.dat). You can automatically have AFT Impulse use the computer name when you add the file by choosing


8/10/2007 4:10:28 PM Page 323


Network Neighborhood, then the computer and then browse to the database file. If you would rather use relative paths, you can manually edit them in the database listing file. All database files must be relative to the AFT Impulse folder where the executable is located.

When the database files were added above, they became External Databases and were listed in IMP_DBUSER.LIB. Now we need to move this listing from the local machine to the AFT Impulse folder on the network so that all users can access the database files. Once the listing file is moved to the network, it must be renamed DATABASE.LIB. Once this is done, all users can now access the database file we have created.

Connecting to the external shared databases

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


C H A P T E R 8

Steady-State Hydraulic Theory and Loss Models

AFT Impulse employs standard matrix solution techniques for solving steady-state pipe network problems. The solution techniques use the basic engineering concepts of mass and momentum balance to determine the steady-state flow distribution and pressure losses in the network.

This chapter presents an overview of the steady-state solution technique and the steady-state solution control parameters available to you. This chapter also describes each of the junction types as it relates to the solution scheme. Many junction types have standard, handbook loss models. The nature of the loss models that AFT Impulse uses and references for those loss models are also given.

The Steady-State Solver

AFT Impulse's Steady-State Solver is designed to achieve a balanced flow and pressure solution in a pipe network. Before using the Steady-State Solver in an application, you must have at least a basic understanding of pipe fluid mechanics. A review of basic pipe fluid mechanics can be obtained from standard undergraduate fluid mechanics textbooks (Fox and McDonald 1985; White 1979).

AFT Impulse makes use of standard matrix solution techniques (Jeppson 1976). The method is known as the H-Equation method, where H, the piezometric head, is solved for at each junction by forcing continuity of flow through each connecting pipe. Simultaneously, the head loss across



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:

zg

PHGL +=

ρ (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 along pipes connecting two junctions. Thus, ijm& represents the mass

flow rate in the pipe connecting junctions i and j.

Application of the law of mass conservation to each junction yields:

01

=∑=

n

jijm& (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:

⎟⎠⎞

⎜⎝⎛=Δ 2

2

1V

D

LfPf ρ (8.3)

Chapter 8 Steady-State Hydraulic Theory and Loss Models 327

8/10/2007 4:10:28 PM Page 327


where fPΔ is the frictional pressure loss. The total pressure change

between junctions is given by the momentum equation in the form of the Bernoulli equation:

fPgzVPgzVP Δ+++=++ 22

2212

11 2

1

2

1 ρρρρ (8.4)

The static and stagnation pressure are related as follows:

2

2

1VPPo ρ+= (8.5)

Substituting Equation 8.5 into 8.4

foo PgzPgzP Δ++=+ 22,11, ρρ

Solving for the frictional pressure drop for a constant area pipes yields:

fjijoio PzzgPP Δ=−+− )(,, ρ (8.6)

where i and j denote upstream and downstream junction values.

The definition of mass flow rate is:

AVm ρ=& (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:

ijij

jijoiom

R

zzgPP&=⎟

⎟⎠

⎞⎜⎜⎝

⎛

′−+−

2/1,, )(ρ

(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.

22

1

ijij

ij

ijijij

AK

D

LfR

ρ⎟⎟⎠

⎞⎜⎜⎝

⎛+=′ (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:

2RQH=Δ



Substituting Equation 8.8 into Equation 8.2 results in the equation to be applied to each junction i:

0)(

1

2/1,, =⎟

⎟⎠

⎞⎜⎜⎝

⎛

′−+−∑

=

n

j ij

jijoio

R

zzgPP ρ (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:

Appliedi

n

j ij

jijoiom

R

zzgPP,

1

2/1,, )(

&=⎟⎟⎠

⎞⎜⎜⎝

⎛

′−+−∑

=

ρ (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:

Appliedi

n

j ij

jijoioi m

R

zzgPPF ,

1

2/1,, )(

&−⎟⎟⎠

⎞⎜⎜⎝

⎛

′−+−

=∑=

ρ (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:


8/10/2007 4:10:28 PM Page 329


⎥⎥⎥⎥⎥⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢⎢⎢⎢⎢⎢

⎣

⎡

=

no

n

o

n

o

n

nooo

nooo

F

P

F

P

F

P

F

P

F

P

F

P

F

P

F

P

F

P

F

J

,2,1,

,

2

2,

2

1,

2

,

1

2,

1

1,

1

∂∂

∂∂

∂∂

∂∂

∂∂

∂∂

∂∂

∂∂

∂∂

L

M

L

L

The column matrix oPv

contains the stagnation pressure at each junction,

and column matrix vF contains the F values (Equation 8.12) at each

junction. The updated solutions for oPv

are obtained from the following

Newton-Raphson equation:

FJPP Foldonewovvv 1

,,−−= (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.



1

42 3

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).

Distance (x)

Location 2 Location 4

Po,2

P2

222

1Vρ

P4

Po,4

242

1Vρ

Location 3

ΔPloss

Pres

sure

Figure 8.2 Steady-state pressure profile along horizontal pipe shown in Figure 8.1.


8/10/2007 4:10:28 PM Page 331


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)ρV2

2.

The appropriateness of either stagnation or static pressure as a boundary condition depends on whether there is a velocity associated with the pressure that is specified. Is the pressure in a vessel or large (lower velocity) header, or is it in a pipe somewhere in the middle of a system? If the first, stagnation is more appropriate; if the second, static.

Pressure boundary conditions with no velocity are more common across all piping industries, and are usually appropriate. However, this is not always the case.

When to use static pressure Now let's consider the application of a true static pressure boundary condition. A static boundary condition has a unique velocity associated with it. From Equation 8.4 it is clear that application of the static pressure is not sufficient to specify location 1. The velocity and elevation is also required.

In AFT Impulse, elevation data is entered for junctions. The pipes adopt the elevation of the junction to which they are connected.

But where does the velocity information for Equation 8.4 come from? One might respond that it can be obtained from the flow rate. But this raises another question: what if the flow rate isn’t known? Put another way, what if the flow rate is what we are solving for?

When the user supplies a stagnation pressure boundary condition, AFT Impulse can use Equation 8.6. Here no velocity information is required. A stagnation pressure boundary can thus be supplied to multiple pipes that might have different velocities. That is why the same reservoir junction can connect with multiple pipes.

However, if the user specifies a static pressure as the boundary condition, to apply Equation 8.4 a unique velocity must be supplied.



Thus it is only possible to connect a static pressure to a single pipe with a unique velocity.

Important: A static pressure boundary has a unique velocity associated with it, and can thus connect to only one pipe.

Where would one find a static pressure boundary? The best example is when the boundary condition is inside a pipe. A pipe system model can start and end anywhere it is convenient for the user. It may be convenient to not start the model at the physical boundary (such as a tank) but at a particular location in the pipe system. This could be, for example, at the location of a pressure measurement. Or it could be at the boundary of the pipe system for which your company is responsible, with another company responsible for what is on the other side of that boundary.

If one needs to model such a situation, the Assigned Pressure junction allows one to model either a static or stagnation pressure. The default stagnation pressure allows connection of up to 25 pipes. If static pressure is chosen, only one connecting pipe is allowed.

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

Open vs. closed systems

In order to model a closed system, only one pressure junction is used in the model. Typically this would be either a Reservoir or Assigned Pressure. As elaborated in Chapter 12, pressure type junctions are an infinite source of fluid and do not balance flow. How then can one be used to balance flow in a closed system?

To answer this question, it is worth considering how AFT Impulse views a closed system model. AFT Impulse does not directly model closed systems, and in fact does not even realize a closed system is being modeled.


8/10/2007 4:10:28 PM Page 333


Consider the system shown in Figure 8.3. This is an open system. Fluid is taken from J1 and delivered to J4 and J5. Because AFT Impulse’s steady-state solution engine solves for a mass balance in the system, all flow out of J1 must be delivered to J4 and J5. Because the flow is steady-state, no fluid can be stored in the system; what goes in must go out.

P1 P2

P3

P4

J1 J2 J3

J4 J5

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.



P1 P2

P3P4

P10

P6

P7P8

P9

P11

J1

J2

J3

J4

J5

J6J9J8

J10

P1 P2

P3

P4

P10

P6

P7P8P9

P11

J1

J2

J3

J4

J5

J6J9J8

J10

J11

J12

Figure 8.4 The first system is closed, and the second open. In both systems the flow into P10 is the sum of P4 and P9. If the second system has the same conditions at J1, J11 and J12, the two system will appear identical to AFT Impulse.

But how is the flow balanced at J1 in the first system? Looking at the first system, one sees that to obtain a system balance, whatever flows into P10 must come back through P4 and P9. Because there is overall system balance by the Steady-State Solver, it will give the appearance of a balanced flow at the pressure junction J1. If there is only one boundary (i.e., junction) where flow can enter or leave the pipe system, then no


8/10/2007 4:10:28 PM Page 335


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

Verifying network solutions

As discussed previously, AFT Impulse uses a Newton-Raphson matrix method to obtain a system level mass balance. When the Steady-State Solver is finished, all branching sections are balanced for flow. Once solved, it is a simple matter to go back and sum up all mass flows at each branch to verify that these equations are satisfied. In fact, AFT Impulse does just that.

After the Steady-State Solver converges, AFT Impulse loops over each junction and adds up the inflow and outflow of mass and compares the sum to the solution tolerances. If it appears that any junction is out of tolerance, a warning is given in the output.

In the Solution Balance Summary table AFT Impulse provides additional information on the balance at each junction. This table can be displayed by selecting the option in the Output Control window General Output tab. Figure 8.5 shows an example of the Solution Balance Summary table.

Pressure drop calculation methods

AFT Impulse can model both Newtonian and non-Newtonian fluid behavior. A non-Newtonian fluid is one where the viscosity depends on the fluid dynamics. For example, as the velocity changes a non-Newtonian fluid’s viscosity can change.

On the other hand, a Newtonian fluid’s viscosity does not depend on the fluid dynamics, and is a function only of the thermodynamic state (i.e., temperature and pressure). Most industrial fluids follow the Newtonian fluid behavior model, although there are important industrial applications where the fluids are non-Newtonian.



Figure 8.5 Solution Balance Summary table shown in the Output window offers a balance report of all junctions in the model.

Pressure drop calculation methods for Newtonian fluids

AFT Impulse offers the Darcy-Weisbach loss model approach as the default method for describing pipe frictional losses. Standard textbooks give additional discussion on this method. AFT Impulse uses the Moody approach to calculating friction factors; note that the Moody approach differs from the Fanning friction factor by a factor of 4.


8/10/2007 4:10:28 PM Page 337


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:

Ref

64= 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).

Turbulent flow The Colebrook-White iterative friction factor equation is used to obtain friction factors in the turbulent flow regime.

235.9

log214.1

−

⎥⎥⎦

⎤

⎢⎢⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛+−=

fReDf

ε 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 Colebrook-White with roughness = 0 while in the laminar range it continues to be calculated using Equation 8.14 above.

Hazen-Williams method

AFT Impulse also offers the Hazen-Williams method of specifying irrecoverable loss information. The Steady-State Solver converts the Hazen-Williams factor to a Darcy-Weisbach friction factor (Walski 1984, 37). This allows a consistent solution approach to be used for all pipe system models, while retaining the flexibility of the two approaches to account for losses.

( )Cf VDHW =

17 250 54 0 081

.. .

where V is in ft/s and D is in ft.

Resistance

The pipe resistance relates head loss to volumetric flow rate. In equation form, the head loss is:

2RQH =Δ

Using standard relationships, the friction pressure drop is related to volumetric and mass flow rate as follows:

ρρ

22 mRg

gQRPf&

==Δ

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):

( )5

2212.139

D

fQ

L

dP waterρρ=


8/10/2007 4:10:28 PM Page 339


( )

f =

−⎛⎝⎜

⎞⎠⎟

⎡

⎣⎢

⎤

⎦⎥

1

4 0868591964 38215

2

. lnRe

. ln Re .

( )56.818

059,102 −

=ν

νd

QRe

where:

dP = pressure drop (psid)

L = length (miles)

f = friction factor, MIT

Q = volumetric flow rate (barrels/hour)

ρ = density (lbm/ft3)

ρwater = density of water (62.3 lbm/ft3)

D = diameter (inches)

Re = Reynolds number, MIT

ν = kinematic viscosity (Seconds Saybolt Universal)

Miller Turbulent method

The Miller Turbulent method is appropriate for light hydrocarbons and is given by the following equation (Pipe Line Rules of Thumb Handbook):

( )( )

( )( )Q

d dP L d dP L

water

water=⎛

⎝⎜⎜

⎞

⎠⎟⎟ +

⎡

⎣⎢⎢

⎤

⎦⎥⎥

2.5

0 5

3

25 91134 35

.log .

.ρ ρ

ρ ρμ

where:

dP = pressure drop (psid)

L = length (miles)

Q = volumetric flow rate (barrels/hour)

ρ = density (lbm/ft3)

ρwater = density of water (62.3 lbm/ft3)



d = diameter (inches)

μ = dynamic viscosity (centipoise)

Frictionless pipes

Frictionless pipes are convenient for connecting junctions that may not have a physical pipe between them. There are, however, some limitations where a frictionless pipe may be located in a model. An example is a frictionless pipe that connects two Assigned Pressure junctions. Such a pipe would have an infinite flow rate. In such cases AFT Impulse identifies inappropriate placement of frictionless pipes and informs the user when a model is run.

Pressure drop calculation methods for non-Newtonian fluids

Duffy method for pulp and paper stock

The Duffy method includes 20 different stock selections to obtain the constants necessary to apply the method (TAPPI 1988):

V K Cmax = ′ σ

V Cwater = 122 1 4. .

If V < Vmax, then

dH

LFKV C D= α β γ

where:

dH /L = head loss per length (m water / 100 m pipe)

C = pulp consistency (% dryness)

Vmax = max velocity (for Duffy Equation) in above equation (m/s)

Vwater = water velocity in above equation (limit at which pulp effects are negligible), (m/s)

D = inner pipe diameter (mm)


8/10/2007 4:10:28 PM Page 341


F = correction factor for temperature or pipe roughness (dimensionless)

K’, K, σ, α, β, γ = constants depending on pulp type (dimensionless)

If V > Vmax but less than Vwater, then use above head loss equation with V = Vmax. If V > Vwater then just use normal friction calculation for water.

The constants used for the specific type of paper stock are displayed in the General Results section of the output. In addition, you can optionally display in the output the Vmax and Vwater terms, the dryness coefficient entered, and the Safety Factor either entered or calculated.

Brecht & Heller method for pulp and paper stock

The Brecht & Heller method includes eight different stock selections for the constants (Ingersoll-Dresser Pumps, 1995). The Brecht & Heller method is recognized to be more conservative than Duffy. That is, the pressure drop predictions are higher.

157.1

205.0

C

VDeR

ρ=′

636.197.3

eRf

′=

dHfV LK

D=

2

where:

Re’ = pseudo Reynolds Number (dimensionless)

D = inner diameter (feet)

V = average velocity (ft/s)

ρ = stock density (lbm/ft3)

C = pulp consistency (% dryness)

f = pseudo friction factor (dimensionless)

K = constant depending on pulp type (dimensionless)

L = length (feet)



The constants used for the specific type of paper stock are displayed in the General Results section of the output. In addition, you can optionally display in the output the special Reynolds Number and Friction Factor terms.

Power Law non-Newtonian

This correlation is taken from Darby (2001), pp. 166-167. Note that Darby uses the Fanning friction factor, which is different from the Moody friction factor used in AFT Impulse by a factor of 4. The equations below have been modified from Darby accordingly.

Defining:

K = constant with units of dynamic viscosity

n = dimensionless constant

Then:

( ) ( ) 81881

−− ++−=

TrT

Lff

ffαα

where:

plLf Re

64=

( )npl

Tn

f39.287.11

21

Re

2728.0+

−=

( ) ( )nplTr nf 757.0414.04 Re24.5exp1016.7 +− −×=

plcpl ReRe −=Δ

( )

( )( )nnn

plnnK

VD

132

8Re

2

+=

− ρ

Δ−+=

41

1α


8/10/2007 4:10:28 PM Page 343


( )nplc −+= 18752100Re

This friction factor can then be used in Equation 8.3.

Bingham Plastic non-Newtonian

This correlation is taken from Darby (2001), pp. 168-169. Note that Darby uses the Fanning friction factor, which is different from the Moody friction factor used in AFT Impulse by a factor of 4. The equations below have been modified from Darby accordingly.

Defining:

Rc = coefficient of rigidity

Sy = yield stress

Then:

( ) mmT

mL fff

1+=

Re

400007.1 +=m

where:

⎟⎟⎠

⎞⎜⎜⎝

⎛−+=

73

4

Re3

He

Re6

He1

Re

64

ff L

cR

DVρ=Re

He is the Hedstrom number:

2

2

Hec

y

R

SD ρ=

193.0Re

10a

Tf =

( )( )He109.2exp146.0188.5 5−×−+−=a



Non-Newtonian flow through non-pipes At your option in the System Properties window, you may apply a correction to hydraulic losses at elements other than pipes. This includes junctions and pipe fittings & losses.

The correction is as follows:

newtonian

newtoniannonnewtoniannewtoniannon f

fKK −

− =

where the friction factor is for the upstream pipe.

Design factors

In each pipe you can specify a Design Factor for the pipe friction. This is a multiplier that is applied to the friction factor calculated with the preceding methods.

Steady Solution Control parameters

AFT Impulse provides you control over a number of parameters that influence numerical convergence:

• Pressure solution tolerance

• Mass flow rate solution tolerance

• Flow relaxation

• Pressure relaxation

• Maximum iterations

• Matrix method

The default values for the solution control parameters perform well in the majority of cases. Modify the default parameters only when necessary or when you are comfortable with numerical convergence issues.


8/10/2007 4:10:28 PM Page 345


Solution tolerance specification

The nature of tolerance specification for pressure and mass flow rate is the same. A discussion of tolerance specification will therefore cover both areas.

AFT Impulse uses an iterative approach to obtain pipe flow solutions; it can obtain a converged solution to as many significant digits as you want (within the computer's numerical capability). However, for practical engineering problems, each digit after about the fourth significant digit generally has low confidence associated with it because of the uncertainties in the input data. The more digits, the lower the confidence. Further, obtaining those additional digits in your solution requires additional computer resources, which typically means longer run times. At a certain point, obtaining additional significant digits becomes counterproductive.

Therefore, the Steady-State Solver needs to know when to stop iterating and assume that a sufficient solution has been obtained. Because you know the nature of your pipe system much better than the Steady-State Solver, you are in a better position to specify the appropriate level of convergence. You do this by setting the tolerances in the Steady Solution Control window.

AFT Impulse bases the solution tolerance specifications on absolute change or relative change of the unknown parameter. As the Steady-State Solver iterates, AFT Impulse continuously updates the solution. Each time a solution update is obtained, AFT Impulse compares the new solution to the old solution for each individual pipe and junction. Presumably, when the correct solution is being approached, the corrections made during each iteration become smaller and smaller. When all corrections for an iteration are smaller than the tolerance you specify, AFT Impulse considers the solution converged.

The absolute change method is generally more reliable for iterative solvers like AFT Impulse because it is less sensitive to the magnitude of the solution, whether close to zero or very large. However, in principle, specifying absolute tolerance requires some knowledge of the final solution. If you specify mass flow convergence as sufficient when the flow no longer changes by 1 lbm/sec each iteration, then your solution will be compromised if the magnitude is also near 1 lbm/sec. In this case it would be better to set the tolerance 3 or 4 orders of magnitude below the solution.



In programming terms, the logic for absolute tolerance looks as follows:

If End

False eConvergenc

Else

True eConvergenc

Then )Junctions All(For If ,,

=

=

<− absoldjnewj TOLPP

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 End

False eConvergenc

Else

True eConvergenc

Then )Junctions All(For If,

,,

=

=

<−

relnewj

oldjnewjTOL

P

PP

Table 8.1 shows a tolerance example for pressures. Notice how relative change does not have units, but absolute change has units (of psia).

The default in AFT Impulse is relative tolerance because experience indicates it is the most robust.

AFT Impulse 4.0 offers options for combining absolute and relative tolerance. You can tell AFT Impulse to assess convergence based on whether either an absolute or relative criteria is satisfied, or whether both absolute and relative criteria are satisfied.


8/10/2007 4:10:28 PM Page 347


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 #Pressure

(psia)Relative Change

Absolute Change (psia)

1 100.0 — —2 60.0 0.6667 40.003 72.0 0.1667 12.004 66.0 0.0909 6.005 67.0 0.0149 1.006 66.9 0.0015 0.10

Figure 8.6 Tolerance input on Steady Solution Control window.

Figure 8.6 shows the tolerance input area in the Steady Solution Control window. Table 8.2 shows an example of how the four different criteria would be applied to flow rate iterations.



Table 8.2 Example of four tolerance methods for a mass flow rate solution. Absolute tolerance = 0.0005 lbm/s, Relative tolerance = 0.0001.

Iteration #

Mass Flow Rate (lbm/s)

Relative Change

Absolute Change

(lbm/sec)

Meets Abs. Tolerance ?

Meets Rel. Tolerance ?

Meets Abs. OR Rel.

Tolerance ?

Meets Abs. AND Rel.

Tolerance ?

1 20.0000 — — — — — —2 12.0000 0.666667 8.000000 NO NO NO NO3 14.0000 0.142857 2.000000 NO NO NO NO4 13.5000 0.037037 0.500000 NO NO NO NO5 13.6000 0.007353 0.100000 NO NO NO NO6 13.5900 0.000736 0.010000 NO NO NO NO7 13.5913 0.000096 0.001300 NO YES YES NO8 13.5907 0.000044 0.000600 NO YES YES NO9 13.5909 0.000015 0.000200 YES YES YES YES

How tolerances relate to solution accuracy The tolerances you set do not have very much to do with how close you are to the true solution. Do not make the mistake of concluding that the tolerance you specify means AFT Impulse’s solution is that close to the true solution. As in Table 8.2, the tolerance is how close the current iteration values are to the previous iteration values. That is all it means.

If there is concern that the AFT Impulse results may not be close enough to the true answer, you have the option of lowering the tolerances, rerunning the models, then comparing the results to the previous run. After performing such a comparison, if you are satisfied that the results are not changing in a significant way then you can feel confident in the model predictions.

Relaxation

Another feature of iterative solvers is called relaxation. In short, relaxation slows the Steady-State Solver's progress toward a solution and in the process reduces the ability of poorly behaved (non-linear) components to destabilize the solution. AFT Impulse implements a proprietary adaptive relaxation scheme that reduces the relaxation dynamically in response to highly non-linear features of the problem. However, you are free to override AFT Impulse's adaptive approach.

When applied to the flow solution, the relaxation is used as follows:


8/10/2007 4:10:28 PM Page 349


( ) oldoldnewnew mmmrm &&&& +−=

where r is the relaxation. The solution after each iteration is modified by weighting it with the previous solution and applying the relaxation factor. The relaxation must always be greater than zero and less than or equal to 1. The march toward a solution is thus relaxed, because the impact of the new solution is minimized by the magnitude of the relaxation parameter. In essence, by specifying small relaxation values, you reduce the speed of the march while causing the march to go more smoothly. This is sometimes the only practical approach for highly non-linear problems.

A relaxation value of 1 is the same as specifying no relaxation at all. Each new solution is used directly for the next iteration step. A relaxation of zero is not accepted because no updates would occur to the unknown parameters and no solution would be obtained. See Table 8.3 for an example.

Table 8.3 Example of different relaxation values. Old Flow Rate is that at the previous iteration. Ideal New Flow Rate is that determined by the matrix solution. Actual New Flow Rate is what is actually accepted for the new iteration.

Relaxation Value

Old Flow Rate (lbm/s)

Ideal New Flow Rate (lbm/s)

Actual New Flow Rate (lbm/s)

1 10 20 200.5 10 20 150.2 10 20 120.1 10 20 11

0.05 10 20 10.50 10 20 10

There are two relaxation parameters in AFT Impulse 4.0 (See Figure 8.7): Flow rate relaxation and pressure relaxation.

Flow Rate Relaxation For poorly-behaved systems, a flow rate relaxation of 0.1 is advised. Experience shows that using flow rate relaxation factors between 0.5 and 1 has limited usefulness. If a flow rate relaxation of 1 does not work, you should generally try flow rate relaxation parameters between 0.1 and 0.5. If these still do not work, there are probably other areas where your



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.


8/10/2007 4:10:28 PM Page 351


Maximum Iterations

The Maximum Iterations parameter in the Steady Solution Control window determines when the Steady-State Solver considers the model nonconvergent; it does not directly affect the solution. There may be many reasons why a model does not converge. You want to be sure that the Steady-State Solver does not continue indefinitely searching for a solution that cannot be obtained because of an input error or ill-behaved system.

Most properly defined models will converge in 50,000 iterations. For larger models, more than 50,000 iterations may occasionally be required, especially when relaxation is being used. AFT Impulse's default of 50,000 iterations is sufficient for most models.

Matrix Method

AFT Impulse’s network solution method requires solution of matrices thousands of times. The default method is Gaussian Elimination, which is very robust and usually provides the fastest convergence.

For larger systems, Gaussian Elimination with Pivoting and the LU Decomposition methods can be better and are therefore provided.

Solution Progress window and iteration history

The Solution Progress window (Figure 8.8) shows you how the current state of the solution compares to the tolerances you specify. It also gives information on your Steady Solution Control settings and gives you helpful feedback on what stage the Steady-State Solver is in.

When the Solution Progress window is displayed during execution, you have the option of pausing or canceling the Steady-State Solver in the middle of a run. If you pause, you can change solution control parameters during a run and resume execution with the new parameters. You also can check the iteration history.



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.


8/10/2007 4:10:28 PM Page 353


There are different warning levels depending on the importance of the warning.

1. WARNING – A warning that actually says WARNING offers important information about the converged solution or the model itself. The information should always be reviewed.

2. CAUTION – A cautionary warning is of less importance than a full warning. The included information may or may not be of interest to the user.

3. Informational – The informational warning is for information only and is not serious.

A complete list of the warnings you may receive, with a brief explanation given of each, is given in the Help System. Whenever any warnings occur that are WARNING level, the General Results section of the Output Window will be displayed with the text in red.

Modeling irrecoverable losses

Chapter 6 discusses in detail the features offered by AFT Impulse to model irrecoverable losses. Chapter 3 discusses the basic conventions used by AFT Impulse for defining losses.

As discussed in Chapter 3, AFT Impulse allows you to specify the base area for modeling losses at junctions. By default the upstream pipe area is used.

Junction pressure losses are modeled as point losses (and point additions for pumps). They are not smoothed out over a length. Pipe losses, both frictional losses and additional minor losses, are smoothed out over the pipe length and are more accurately referred to as distributed losses.

It is important to note that AFT Impulse does not include the frictional part of the loss when assigning a loss factor. Frictional lengths in fittings are commonly lumped into the adjacent pipes. An example is a large radius elbow. The loss due to the bend is the loss factor K; the friction in the elbow is an independent effect. The magnitude of this distinction is frequently negligible.



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 ChangeCrane 1988, A-26 and Idelchik 1994, 208, 216

Orifice Idelchik 1994, 218, 220

Tee/WyeIdelchik 1994, chapter 7 and Miller

ValveIngersoll-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:

2

2

1VKPloss ρ=Δ (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.


8/10/2007 4:10:28 PM Page 355


Design factors

Each junction that allows modeling of pressure losses allows input of a Design Factor. The design factor is multiplied by the K factor determined by methods discussed in this chapter.

Many junction types allow modeling of K factors as well as other pressure loss methods. The Design Factor is also applied to these other methods as a multiplier on the pressure loss.

Area change

Two standard area change loss geometries are available: the conical transition and the abrupt transition.

The conical expansion correlation (Crane 1988, A-26) is:

2

12

sin6.2 ⎟⎟⎠

⎞⎜⎜⎝

⎛−⎟

⎠⎞

⎜⎝⎛=

down

upup A

AK

μ (μ < 45 degrees)

2

1 ⎟⎟⎠

⎞⎜⎜⎝

⎛−=

down

upup A

AK (μ > 45 degrees)

The conical contraction correlation (Crane 1988, A-26) is:

2

12

sin8.0

⎟⎟⎠

⎞⎜⎜⎝

⎛

⎟⎟⎠

⎞⎜⎜⎝

⎛−⎟

⎠⎞

⎜⎝⎛

=

up

down

up

down

up

A

A

A

A

K

μ

(μ < 45 degrees)

2

2sin15.0

⎟⎟⎠

⎞⎜⎜⎝

⎛

⎟⎟⎠

⎞⎜⎜⎝

⎛−

=

up

down

up

down

up

A

A

A

A

K

μ

(μ > 45 degrees)

The abrupt expansion correlation (Idelchik 1994, 208) is:



2

1 ⎟⎟⎠

⎞⎜⎜⎝

⎛−=

down

upup A

AK

The abrupt contraction correlation (Idelchik 1994, 216) is:

2

75.0

15.0

⎟⎟⎠

⎞⎜⎜⎝

⎛

⎟⎟⎠

⎞⎜⎜⎝

⎛−

=

up

down

up

down

up

A

A

A

A

K

Bend

90 degree bends The three Bend loss correlations are all for turbulent Reynolds Numbers (Crane 1988, A-29).

The K values for a smooth, flanged bend are provided in Table 8.5.

Table 8.5 K values for 90 degree smooth, flanged bends

r/d K r/d K

1 20 f T 8 24 f T

1.5 14 f T 10 30 f T

2 12 f T 12 34 f T

3 12 f T 14 38 f T

4 14 f T 16 42 f T

6 17 f T 20 50 f T


8/10/2007 4:10:28 PM Page 357


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

½” 0.027 2” 0.019 8”-10” 0.014

¾” 0.025 2 ½”, 3” 0.018 12”-16” 0.013

1” 0.023 4” 0.017 18”-24” 0.012

1 ¼” 0.022 5” 0.016

1 ½” 0.021 6” 0.015

The standard threaded elbow is given by:

TfK 30=

where fT is the turbulent friction factor given in Table 8.6.

The Mitre bend is given by:

TfK 60=

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:

( ) KKD

rfnK T +⎟

⎠⎞

⎜⎝⎛ +−= 5.025.01 π

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.



Table 8.7 K values for Mitre bends

μ (deg.) K

0 2 f T

15 4 f T

30 8 f T

45 15 f T

60 25 f T

75 40 f T

90 60 f T

μ

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):

2375.02

11707.0

⎥⎥⎥

⎦

⎤

⎢⎢⎢

⎣

⎡

⎟⎟⎠

⎞⎜⎜⎝

⎛−+⎟

⎟⎠

⎞⎜⎜⎝

⎛−⎟

⎟⎠

⎞⎜⎜⎝

⎛=

down

orifice

up

orifice

orifice

upup A

A

A

A

A

AK

Figure 8.9 Sharp-edged orifice

The round-edged orifice, shown in Figure 8.10, is given by the following (Idelchik 1994, 220):


8/10/2007 4:10:28 PM Page 359


⎢⎢⎢

⎣

⎡

⎟⎟⎠

⎞⎜⎜⎝

⎛−+⎟

⎟⎠

⎞⎜⎜⎝

⎛−′

⎟⎟⎠

⎞⎜⎜⎝

⎛=

275.02

11down

orifice

up

orifice

orifice

upup A

A

A

AK

A

AK

⎥⎥⎥

⎦

⎤

⎟⎟⎠

⎞⎜⎜⎝

⎛−′⎟⎟

⎠

⎞⎜⎜⎝

⎛−+

375.0

112up

orifice

down

orifice

A

AK

A

A

where:

( )orificeDrK

/7.71047.003.0

−•+=′

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.



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.

Ac Ast AcAst

As As

c = combined flow st = straight-through flow s = side branch

Figure 8.11 Idelchik’s nomenclature for diverging and converging flow through Tee/Wye junction

In the AFT Impulse Tee/Wye window, a simpler and more general nomenclature is used. AFT Impulse refers to the a, b and c pipes. This allows AFT Impulse to cover all the possibilities in Table 8.10 without specifying exactly which pipe is the combined flow pipe. In some cases, you may not know all the flow directions in a system, and might not be able to specify which is the combined flow pipe.

AFT Impulse’s input method allows you to specify which pipe is the branch pipe. Idelchik calls this the “s” pipe for “side branch”. AFT Impulse calls it the c pipe. If the tee angle is 90 degrees, it is not important which pipe you call the a pipe and which the b pipe. The distinction between a and b takes on more importance for non-90 degree pipes. In this case, AFT Impulse allows you to specify the angle of the branch pipe.

Figure 8.12 shows the Tee/Wye Specification window. The α angle itself is specified on the Tee/Wye Model tab, where there are other options for how to model the tee.


8/10/2007 4:10:28 PM Page 361


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):

⎥⎥

⎦

⎤

⎢⎢

⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛−⎟⎟

⎠

⎞⎜⎜⎝

⎛+=

Δ= αλ

ρcos21

2/

2

2,c

s

c

s

c

ssc V

V

V

V

V

PK

where λ is found from Table 8.8.



Table 8.8 Diverging case λ values

As / Ac Qs / Qc λ

≤ 0.35 ≤ 0.4 1.1 - 0.7 Qs / Qc

≤ 0.35 > 0.4 0.85

> 0.35 ≤ 0.6 1.0 - 0.6 Qs / Qc

> 0.35 > 0.6 0.6

Straight run, constant area (Idelchik 1994, 419):

2

2,2/

⎟⎟⎠

⎞⎜⎜⎝

⎛=

Δ=

c

sst

c

ststc Q

Q

V

PK τ

ρ

where τst is given in Table 8.9:

Table 8.9 Diverging case τst values

As / Ac Qs / Qc τ st

≤ 0.4 0 - 1.0 0.4

> 0.4 ≤ 0.5 2 (2 Qs / Qc - 1)

> 0.4 > 0.5 0.3 (2 Qs / Qc - 1)

Converging Case Side branch (Idelchik 1994, 417):

2/2,c

ssc

V

PK

ρΔ

=

⎟⎟⎟

⎠

⎞

⎜⎜⎜

⎝

⎛⎟⎟⎠

⎞⎜⎜⎝

⎛−⎟⎟

⎠

⎞⎜⎜⎝

⎛−−⎟⎟

⎠

⎞⎜⎜⎝

⎛+′= αcos2121

222

c

s

s

c

c

s

st

c

sc

cs

Q

Q

A

A

Q

Q

A

A

AQ

AQA

Straight run, constant area (Idelchik 1994, 417):

αρ

cos2112/

22

2, ⎟⎟⎠

⎞⎜⎜⎝

⎛−⎟⎟

⎠

⎞⎜⎜⎝

⎛−−=Δ=

c

s

s

c

c

s

c

ststc Q

Q

A

A

Q

Q

V

PK


8/10/2007 4:10:28 PM Page 363


References for all tee and wye junction correlations are given in Table 8.10.

Table 8.10 References for Tee/Wye junction correlations

Geometry Type Direction Reference

Constant area, sharp Diverging tee Branch Idelchik 1994, 451

Diverging tee Run Idelchik 1994, 453

Converging tee Branch Idelchik 1994, 429-432

Converging tee Run Idelchik 1994, 429-432

Splitting tee From branch Miller 1990, 323*

Combining tee To branch Miller 1990, 316*

Constant area, r/D = 0.1 Diverging tee Branch Idelchik 1994, 459-461

Diverging tee Run Idelchik 1994, 453





Area sum, sharp Diverging tee Branch Idelchik 1994, 452

Diverging tee Run Idelchik 1994, 454-455





Wye Diverging wye Branch Idelchik 1994, 451

Converging wye Run Idelchik 1994, 473

*Estimates have been made on available data

K and CV (valve coefficient)

The definition for CV is as follows (Lyons 1982, 133):

( )C

Q

PV

valve

=Δ 62 4. / ρ

where Q is in gpm, ΔP is in psid, and ρ is in lbm/ft3.



It can be shown that K and Cv are related by the following approximate expression (Lyons 1982):

KA

Cvalvev

= 14602

2

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:

P

QKsprinkler Δ

=

where Q and ΔP can be in the units of your choice. The preference in the USA is gpm for flowrate and psid for pressure. In Europe it is frequently liter/min for flow rate and bars for pressure. Note that Ksprinkler is not dimensionless, but has units associated with it. Also, note that sprinkler vendor values for Ksprinkler implicitly assume that water is the fluid, so there is a built-in density of 62.3 lbm/ft3 or 1000 kg/m3. If this sprinkler is used for something other than water, the Ksprinkler value will need to be adjusted.

Also note that if the spray is into the atmosphere, as is typical, the ΔP is the same thing as the gage pressure. That is why the previous equation is stated by some vendors as just P in the denominator, rather than ΔP. When stated as just P, the pressure is gage.

Note: The fire sprinkler K is not the same as the dimensionless K value as given by Equation 8.18 and used extensively throughout AFT Impulse.

The Ksprinkler is related to the discharge coefficient. It can be shown that the relationship is given by

sprinklerwater

D KAC2

ρ=


8/10/2007 4:10:28 PM Page 365


Pumps

Pumps can operate at variable speeds. Affinity laws (also called homologous laws) allow an estimation of the pump performance at speeds other than the design speed.

Before discussing the affinity laws, it should be noted that the affinity laws are an approximation. In many cases pump manufacturers will have performance data for a number of speeds. These data should be used if available. When not available, the affinity laws can be used with an understanding that an approximation is being made.

The affinity laws for pumps are as follows:

2

1

2

1

2

2

1

2

1

s

s

Q

Q

s

s

H

H

=

⎟⎟⎠

⎞⎜⎜⎝

⎛=

ΔΔ

(8.19)

where s is the speed. If the pump data is input as a polynomial, the speed will affect the curve as follows:

4

23

322

222

2

4

323

222

222222

41

231

221

21

221

22

41

31

2111

s

Qe

s

QdcQsbQasH

s

Qes

s

Qds

s

Qcs

s

QbsasH

eQsdQscQsbQsasHsH

eQdQcQbQaH

++++=Δ

⎟⎠⎞

⎜⎝⎛+⎟

⎠⎞

⎜⎝⎛+⎟

⎠⎞

⎜⎝⎛++=Δ

++++=Δ=Δ

++++=Δ

(8.20)

where ΔH1 is the input curve that typically represents 100% speed and ΔH2 is the head rise at speed s. In this form, the speed s will be a decimal and not a percent. That is, 50% speed will result in the appropriate s of 0.5. AFT Impulse determines this automatically in the Steady-State and Transient Solvers.

If you model the pump as a controlled flow or discharge pressure, AFT Impulse will backsolve Equation 8.20 for the speed s and include that in the output Pump Summary.



Chempak thermophysical property database

The optional Chempak Database offers approximately 700 fluids to the user. In addition, it offers non-reacting mixture calculations. Information on the theory and calculation methods used in Chempak is given in Appendix D.

Accuracy options When performing interpolations for properties, a 2 point or 4 point interpolation scheme can be used. The 2 point scheme implemented in AFT Impulse is the Standard Accuracy option. The 4 point scheme is the high accuracy option.

ASME steam tables database

The ASME Steam Tables Database offers users access to water properties obtained from the IAPWS Industrial Formulation 1997 for the Thermodynamic Properties of Water and Steam (ISPWS-IF97), (see ASME Press in References). These properties are both pressure and temperature dependent.

AFT Standard fluids

The AFT Standard Database offers about 10 common fluids to the user. Information on how to use these fluids within AFT Impulse are given in Chapter 5.

Viscosity

The dynamic viscosity in AFT Impulse is assumed to be a function of temperature only. In general this is a good assumption, although systems at very high pressure usually start to show some pressure dependence. These properties are modeled as polynomial curve fits of temperature.


C H A P T E R 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:

VaP Δ−=Δ ρ (9.1)



where:


ρ = density

a = wavespeed


By adding the pressure surge to the existing static pressure, one can obtain the maximum theoretical pressure in the pipe. However, in some cases, such as transient cavitation, the pressure can exceed the instantaneous prediction. The engineer should therefore be cautious in its application.

Here’s an example. Assume flow in a pipe filled with water is instantaneously stopped by a valve closure. Assume the density is 62 lbm/ft3, the wavespeed is 4000 feet per second, and the initial velocity is 10 feet per second. Also assume that before the transient the static pressure is 50 psig.

psig 585

50535

psig 50ft-lbm 2.23

slbf*

in 144

ft 1*

s

ft10*

s

ft4000*

ft

lbm62

max

max

2

2

2

3max

=+=

+−=

P

P

P

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):

Chapter 9 Theory of Waterhammer and Solution Methodology 369

8/10/2007 4:10:28 PM Page 369


( )( )[ ]eDEKc

Ka

1

2

1+= ρ

where:

a = wavespeed

K = fluid bulk modulus of elasticity

ρ = fluid density

E = pipe modulus of elasticity

D = pipe inner diameter

e = pipe wall thickness

c1 = correction factor

The sonic speed of the liquid is given by the (square root of the) numerator. The denominator contains the reduction from sonic speed caused by the pipe properties.

The correction factor, c1, is a function of how the pipe is restrained. Some examples from Wylie, et al., 1993, page 27-28 are:

1. Thin-walled pipe anchored at upstream end only

211 μ−=c

2. Thin-walled pipe anchored throughout

21 1 μ−=c

3. Thin-walled pipe free to expand throughout

11 =c

4. Thick-walled pipe anchored at upstream end only

( ) ( )2112

1 μμ −+

++=eD

D

D

ec

5. Thick-walled pipe anchored throughout

( ) ( )21 11

2 μμ −+

++=eD

D

D

ec



6. Thick-walled pipe free to expand throughout

( )eD

D

D

ec

+++= μ1

21

7. Circular tunnel

( )μ+= 12

1 D

ec

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:

a

Lt 2=Δ (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.


8/10/2007 4:10:28 PM Page 371


Momentum equation

Following from Wylie, et al., 1993, page 22, the momentum equation can be expressed as

02

)sin(

1 =+++D

VfVg

t

V

x

P α∂∂

∂∂

ρ (9.3)

where:

P = pressure

V = velocity

ρ = density

x = distance along pipe

t = time

g = acceleration due to gravity

D = diameter

f = friction factor

α = angle of pipe slope

Mass continuity equation

Following from Wylie, et al., 1993, page 27, the continuity equation can be expressed as

0

2=+

t

P

x

V

g

a

∂∂

∂∂

(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.



The equations are identified as L1 and L2 .

12)sin(

1L

D

VfVg

t

V

x

P =+++ α∂∂

∂∂

ρ (9.5)

22

L

t

P

x

Va =+

∂∂

∂∂ρ (9.6)

Combine these two equation linearly using a parameter, λ, (Wylie, et al., 1993, pp. 27)

0 21 =+ LL λ

or,

⎟⎠⎞

⎜⎝⎛ ++⎟

⎠⎞

⎜⎝⎛ +

x

Va

t

V

t

P

x

P

1 2

∂∂λρ

∂∂ρ

∂∂

∂∂

λλ

02

)sin( =++

D

VfVg

ραρ (9.7)

If

2 1

adt

dx λλ

== (9.8)

then

a

1±=λ

and

adt

dx ±=

From calculus

t

P

dt

dx

x

P

dt

dP

∂∂

∂∂ += (9.9)

t

V

dt

dx

x

V

dt

dV

∂∂

∂∂ += (9.10)

Substituting Equation 9.8 into 9.7


8/10/2007 4:10:28 PM Page 373


⎟⎠⎞

⎜⎝⎛ ++⎟

⎠⎞

⎜⎝⎛ +

dt

dx

x

V

t

V

t

P

dt

dx

x

P

∂∂ρ

∂∂ρ

∂∂

∂∂λ

02

)sin( =++

D

VfVg

ραρ (9.11)

and substituting Equations 9.9 and 9.10 into 9.11

02

)sin( =++⎟

⎠⎞

⎜⎝⎛+⎟

⎠⎞

⎜⎝⎛

D

VfVg

dt

dV

dt

dP ραρρλ

or

02

)sin(

1 =+++±D

VfVg

dt

dV

dt

dP

a

ραρρ (9.12)

Now multiply Equation 9.12 by (a dt), which also equals dx

02

)sin( =+++± dx

D

VfVdxgadVdP

ραρρ (9.13)

The gravity term contains the term α, and the term can be converted as follows

gdzdxdx

dzgdxg )sin( ρραρ =⎟

⎠⎞

⎜⎝⎛= (9.14)

where dz is the vertical change in pipe elevation. Therefore, substituting 9.14 into 9.13

02

=+++± dx

D

VfVgdzadVdP

ρρρ

Further substitution of the mass flow rate for velocity yields

02

2

=+++± mmdxDA

fgdzmd

A

adP &&&

ρρ (9.15)

Integrating Equation 9.15 with the positive sign along the C+ characteristic from point A to P (Figure 9.1) yields



02

2=+++ ∫∫∫∫

P

A

P

A

P

A

P

A

x

x

z

z

m

m

P

P

dxmmDA

fdzgmd

A

adP &&&

&

& ρρ (9.16)

where the subscript P denotes the current solution point at x = i at time t = Δt, and subscript A denote the values at point x = i -1 at the previous time step, t = 0. A similar equation can be written for the C- characteristic from point B to P.

The term on the right before the equal sign in Equation 9.16 can be integrated assuming a second order accurate representation of m& to obtain

( ) ( ) ( )

02

2=Δ

+−+−+−

AP

APAPAP

mmDA

xf

zzgmmA

aPP

&&

&&

ρ

ρ (9.17)

t = 0

t = Δt

t = 2Δt

t = 3Δt

t = 4Δt

t = 5Δt

x = 0 x = L

A B

P

x = i x = i+1x = i-1

C+ C-

Figure 9.1 Method of Characteristics grid


8/10/2007 4:10:28 PM Page 375


Similarly

( ) ( ) ( )

02

2=Δ+

−+−+−−

AP

APAPAP

mmDA

xf

zzgmmA

aPP

&&

&&

ρ

ρ (9.18)

Introducing two convenient parameters:

Impedance:

A

aB

=

Resistance:

22

DA

xfR

ρΔ=

Equations 9.17 and 9.18 then become

( ) ( ) ( ) 0 =+−+−+− APAPAPAP mmRzzgmmBPP &&&& ρ

( ) ( ) ( ) 0 =+−+−+−− APAPAPAP mmRzzgmmBPP &&&& ρ

Substituting more general variables for A, B and P, and simplifying the equations yields

newiPPnewi mBCP ,, &−= (9.19)

newiMMnewi mBCP ,, &+= (9.20)

where:



( )1,1,1 −−− −−+= iioldioldiP zzgmBPC ρ&

( )1,1,1 +++ −+−= iioldioldiM zzgmBPC ρ&

oldiP mRBB ,1−+= &

oldiM mRBB ,1++= &

Equations 9.19 and 9.20 are referred to as the compatibility equations (Wylie, et al., 1993, pp. 42). Since all the parameters in CP, CM, BP, and BM are known, the only unknowns in 9.19 and 9.20 are Pi,new and newim ,& .

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:

MP

PMMPnewi BB

BCBCP

++=, (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:

MP

MPnewi BB

CCm

+−=,& (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


8/10/2007 4:10:28 PM Page 377


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:

satnew PP =

oldvapornewvapor VV ,, =

( )ρ2,,,,t

mmmm olddownnewdownoldupnewupΔ++−−+ &&&&

(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

tn

La

Δ= (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:



ta

Ln

i

ii Δ

=

(9.25)

In Equation 9.25 as applied to each pipe in the system, the time step Δt must be the same for all pipes. The length, L, is a given for each pipe and thus cannot be changed. The wavespeed, a, is also a given for each pipe, although it is known with less certainty than is the length. The free parameter is then the number of sections, n.

In any pipe system there will be one pipe that is the controlling pipe. The controlling pipe is that pipe which has the least number of sections, sometimes only one. Once the controlling pipe is selected, the time step is determined by solving Equation 9.24 for Δt. Then the number of sections in the remaining pipes is obtained from Equation 9.25.

However, this presents a dilemma. Because the number of sections in remaining pipes is derived from Equation 9.25, typically the number of sections, n, will not be a whole number. Since partial sections in a pipe cannot be modeled, an alternative must be found.

Because the wavespeed for each pipe is the least certain parameter, an uncertainty in wavespeed up to 15% is possible. Therefore, the wavespeed in each pipe is allowed to depart slightly from its original calculated value in order to cause n to be a whole number.

Therefore, Equation 9.25 becomes

( ) ta

Ln

i

ii Δ±

= 1 ψ

where ψ is the accepted uncertainty in wavespeed up to ±0.15 (Wylie, et al., 1993, pp. 54).

Two other things should be noted about pipe sectioning. Because each pipe section is explicitly solved along the characteristic lines, breaking pipes into continuously smaller sections offers no improvement in accuracy. That is, using twice the number of sections in a pipe will not yield a more accurate prediction (see Chapter 12). This is in contrast to other kinds of finite difference methods where more increments offers improved accuracy.

The other item of note is the effect of breaking the model into additional increments. According to Equation 9.24, when the number of sections is


8/10/2007 4:10:28 PM Page 379


doubled, the time step must cut in half. This results in increasing runtime by a factor of 4 for each doubling of the number of sections.

The Section Pipes window, discussed in Chapter 5, automates the process of obtaining acceptable roundoff error when assigning sections.

Effects of flow resistance

The resistance to flow inside the piping system is a result of friction and irrecoverable losses through valves and fittings. Without resistance the pressure waves would never damp out.

It is beneficial to include all of the resistance in the system. However, properly locating the resistance is not nearly as important in waterhammer analysis as it is in conventional steady flow analyses. What this means is that it is usually more efficient to “lump” the resistances together into a single pipe, rather than use a separate junction to model the resistance for each individual component.

For example, it is usually best to lump the valve resistance into the adjacent pipes unless there are valve position changes during the simulation. This is true of all types of local resistances (i.e., valves and fittings).

The effect on the results of lumping a resistance is usually minor, while the benefit of faster computation time is major. The faster computation results from the fact that the user can use less pipe sections to model the same system.

Assigned Pressure theory

With an assigned pressure, the pressure is known at the pipe endpoint. The unknown is then the flow rate, which can be obtained by solving the appropriate compatibility equation.

If the pressure is known at the upstream end of the pipe, the negative compatibility equation is used (Equation 9.20).

M

Mknownnewi B

CPm

−=,&



Conversely, if the pressure is known at the downstream end, the positive compatibility equation is used (Equation 9.19).

P

knownPnewi B

PCm

−=,&

If the known pressure changes with time, then a current pressure is obtained for the current time step and used in the equations.

Assigned Pressure vapor cavitation theory

Assigned Pressure junctions cannot reach vapor pressure because the pressure is specified and is therefore always above vapor pressure.

Reservoir theory

A reservoir is similar to an assigned pressure. With the known reservoir height, the pressure can be obtained at the junction and the flow can be solved directly (Wylie, et al., 1993, pp. 43):

surfaceRnewiknown PgHPP +== ρ,

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 compatibility equation is used (Equation 9.20).

M

Mknownnewi B

CPm

−=,&


P

knownPnewi B

PCm

−=,&

If the reservoir height changes with time, then a current height is obtained for the current time step and used in the equations.


8/10/2007 4:10:28 PM Page 381


Reservoir vapor cavitation theory

Reservoir junctions cannot reach vapor pressure because the pressure is specified and is therefore always above vapor pressure.

Assigned Flow theory

With an assigned flow, the flow rate is known at the pipe inlet or outlet (Wylie, et al., 1993, pp. 43). The unknown is then the pressure, which can be obtained by solving the appropriate compatibility equation.

If the flow rate is known at the upstream end of the pipe, the negative compatibility equation is used (Equation 9.20).

knownMMnewi mBCP &+=,


knownPPnewi mBCP &−=,

If the known flow rate changes with time, then a current flow rate is obtained for the current time step and used in the equations.

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:

oldvapornewvapor VV ,, =

( )ρ2,,,,t

mmmm olddownnewdownoldupnewupΔ++−−+ &&&&

The upm& terms are known because they are specified terms for inflow

junctions. For outflow junctions the downm& terms are known. Similar to

a pipe interior node, when the vapor volume is negative, the cavity collapses and the fluid pressure then rises above the vapor pressure.



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.

Pnewi CP =,

Conversely, if the dead end is at the downstream end, the positive compatibility equation is used (Equation 9.19) with zero flow.

Mnewi CP =,

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:

( )ρ2,,,,t

mmVV oldupnewupoldvapornewvaporΔ−−+= &&

since the downm& terms are always zero. For dead ends at the beginning

of a pipe, the upm& terms are zero. Similar to a pipe interior node, when

the vapor volume is negative, the cavity collapses and the fluid pressure then rises above the vapor pressure.

Branch theory

At a branch, there can be multiple pipes (Wylie, et al., 1993, pp. 51-53). An additional relationship is needed, and that relationship is the conservation of mass. The total mass flow in and out of the junction must sum to zero.

In addition, the branch has a single pressure solution, Pj, and this solution is common to all inflowing pipes.


8/10/2007 4:10:28 PM Page 383


The compatibility equation is written for each junction. For the pipes flowing into the branch, the positive equation is used (Equation 9.19)

P

jPnewi B

PCm

−=,&

where Pj is the branch junction pressure, as yet unknown.

For the pipes flowing out of the branch, the negative equation is used (Equation 9.20)

M

Mjnewi B

CPm

−=,&

Now, sum all of the pipe flow rates into the junction

0, =−=∑ jBCnewi PSSm& (9.26)

where

P MC

P M

C CS

B B= +∑ ∑ (9.27)

∑∑ +=MP

B BBS

11 (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

AppliedjBCnewi mPSSm && −=−=∑ ,

Now solve for the junction pressure

B

AppliedCj S

mSP

&+= (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:

( )( )

( ) ρ2,,

,,

,,

,,t

mm

mm

mm

VV

oldBnewB

oldoutnewout

oldinnewin

oldvapornewvaporΔ

⎟⎟⎟⎟

⎠

⎞

⎜⎜⎜⎜

⎝

⎛

+−

++

−−

+= ∑∑

&&

&&

&&

where Bm& is the branch source/sink term defined as positive flow into

the branch. The inm& terms are obtained from the positive compatibility

equation, while the outm& terms are obtained from the negative equation.

Similar to a pipe interior node, when the vapor volume is negative, the cavity collapses and the fluid pressure then rises above the vapor pressure.

Tee/Wye theory

The Tee/Wye junction is a static element, and its solution method is essentially the same as the Branch junction. The only difference is that the loss factors are updated dynamically similar to steady flow.

Valve theory

A valve junction has a pressure loss across the valve. When the valve closes this loss becomes infinite (Wylie, et al., 1993, pp. 57). With two connecting pipes, the upstream and downstream pipe pressures are obtained from the compatibility equations (Equations 9.19, 9.20):

upnewiupPupPupnewi mBCP ,,,,,, &−=

downnewidownMdownMdownnewi mBCP ,,,,,, &+=

From a mass balance:


8/10/2007 4:10:28 PM Page 385


valvedownnewiupnewi mmm &&& == ,,,,

The pressure drop is related to the flow rate according to the following:

2,,,, valvevalvedownnewiupnewi mRPP &=−

where

Rvalve = Valve Resistance

Substituting the compatibility equations (Equations 9.19, 9.20) in and eliminating the pressures

( )valveupPupP mBC &,, −

( ) 2,, valvevalvevalvedownMdownM mRmBC && =+−

Rearranging terms,

( )valveupPupPvalve

valve mBCR

m && ,,2 1 −−

( ) 01

,, =++ valvedownMdownMvalve

mBCR

&

Using the quadratic equation, the solution for positive flow through the valve is

( )downMupPvalvevalve BBRm ,,1 +−=&

( ) ( )downMupPvalvedownMupPvalve CCRBBR ,,2

,,2 41 −+++

For negative flow, the solution is

( )downMupPvalvevalve BBRm ,,1 +=&

( ) ( )downMupPvalvedownMupPvalve CCRBBR ,,2

,,2 41 −−+−

After solving for flow rate, valvem& , the upstream and downstream

pressures can be obtained from the compatibility equations.

If the valve closes then 1/Rvalve = 0 and



0=valvem&

Exit valves

If the valve is an exit valve, then the downstream pressure is known and

upPvalvevalve BRm ,1−=&

( )exitupPvalveupPvalve PCRBR −++ ,2

,2 41

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.

oldupvapornewupvapor VV ,,,, =

( )ρ2,,,,t

mmmm oldvalvenewvalveoldupnewupΔ++−−+ &&&&

olddownvapornewdownvapor VV ,,,, =

( )ρ2,,,,t

mmmm olddownnewdownoldvalvenewvalveΔ++−−+ &&&&

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.

upPvalvevalve BRm ,1−=&

( )satupPvalveupPvalve PCRBR −++ ,2

,2 41

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.


8/10/2007 4:10:28 PM Page 387



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.

tititi PPPP Δ−+Δ−+Δ− +−= 3,22,1,newinfinity, 33

tititi mmmm Δ−+Δ−+Δ− +−= 3,22,1,newinfinity, 33 &&&&

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),

newiPPnewi mBCP ,, &−=

and the negative characteristic compatibility equation (Equation 9.20)

newiMMnewi mBCP ,, &+= (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:

( )1oldinfinity,oldinfinity, +−−−= iiM zzgmBPC ρ& (9.31)

oldinfinity,mRBBM &+= (9.32)

In Equation 9.31, use the following approximation which is perfectly accurate for straight pipes

iiii zzzz −=− −+ 11

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.


8/10/2007 4:10:28 PM Page 389


Pexit

spraym&

Figure 9.2 Spray discharge schematic

The relationship for pressure drop across the nozzle is as follows

2

2

1VP ρ=Δ

( )2

2,2

1)sgn( spray

sprayDsprayexitnewj m

ACmPPP &&

ρ=−=Δ

where, for clarity, spraym& is defined positive flowing out of the system.

In addition, the spray junction is similar to a branch junction, so the following equation applies with the flowrate defined positive out.

B

sprayCnewj S

mSP

&−=,

where SC and SB are given by Equations 9.27 and 9.28.

Combining the preceding equations and eliminating flow rate,

( ) 01

2

1sgn 2

22=−++

B

Cexit

BsprayD S

SPm

Sm

ACm &&&

ρ

or

( )0

2

2)sgn(

22

222

=−

+

+

B

CexitBsprayD

B

sprayD

S

SPSAC

mS

ACmm

ρ

ρ&&&



This is a quadratic equation, and the mass flowrate can be obtained by use of the quadratic formula.

Note: If the spray junction pressure falls below the exit pressure, reverse flow will occur. If the exit pressure is the atmosphere, the reverse flow will be air. However, Impulse assumes the inflow is at the pipe system density. This can lead to erroneous results if ignored. Reverse flow can be determined by graphing the flow at the spray junction or looking at the summary of maximum and minimum flows in the Output window.

Spray Discharge vapor cavitation theory

When the calculated pressure at a Spray Discharge junction drops below vapor pressure, a vapor cavity forms at the junction. This is similar to a Branch junction.

The vapor volume calculation is as follows

( )( )

( ) ρ2,,

,,

,,

,,t

mm

mm

mm

VV

oldspraynewspray

oldoutnewout

oldinnewin

oldvapornewvaporΔ

⎟⎟⎟⎟

⎠

⎞

⎜⎜⎜⎜

⎝

⎛

++

++

−−

+= ∑∑

&&

&&

&&

where spraym& is the spray outflow term. The inm& terms are obtained

from the positive compatibility equation, while the outm& terms are

obtained from the negative equation.


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.


8/10/2007 4:10:28 PM Page 391


Modeling pump as a known speed

Assuming that the pump curve can be modeled by a quadratic, then

2pumppump cQbQaH ++=Δ

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

constant12

=Δα

H

constant2=αQ

where α is the pump speed. Applying these relationships the pump curve is transformed

2

2

2 αααpumppump Q

cQ

baH ′+′+′=Δ

22pumppump QcQbaH ′+′+′=Δ αα

Converting to pressure rise

⎟⎟⎠

⎞⎜⎜⎝

⎛ ′+

′+′=Δ 2

22

pumppump mc

mb

agP &&ρρ

ααρ

22pumppump cQbQaP ++=Δ αα (9.33)

where the primes on the constants were dropped after they were divided through by ρg.

By use of the two compatibility equations (Equation 9.19, 9.20),

pumpPPnewup mBCP &−=,

pumpMMnewdown mBCP &+=,



and substituting these equations into Equation 9.33

( ) ( )22

pumppump

pumpPPpumpMM

mcmba

mBCmBC

&&

&&

++=

−−+

αα

This is a quadratic equation for the pump flow rate which can be solved by the quadratic formula

⎟⎟⎠

⎞⎜⎜⎝

⎛ −+=

c

bBBQ

downMupPpump 2

,, α

( )

( ) ⎥⎥⎥

⎦

⎤

⎢⎢⎢

⎣

⎡

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

−+

−+−−×

21

2,,

,,24

11α

α

bBB

CCab

downMupP

downMupP

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:

432pumppumppumppump eQdQcQbQaH ++++=Δ

and with the speed included,

4

4

3

3

2

2

2 αααααpumppumppumppump Q

eQ

dQ

cQ

baH ′+′+′+′+′=Δ

Combine this with the compatibility equation as in the previous section to obtain:

( ) ( ) pumppumpPPpumpMM mbamBCmBC &&& αα +=−−+ 2

42

32pumppumppump m

em

dmc &&&

αα+++

This equation is then solved by Newton-Raphson iteration to obtain the flow rate, pumpm& .


8/10/2007 4:10:28 PM Page 393


The above models are accessed using the Without Inertia option.

Backflow through the pump – one quadrant If the pump cannot supply enough head to overcome the system requirements at some time during the transient, backflow will occur through the pump. When this occurs, the pump curve loses meaning.

To address this situation, AFT Impulse makes an approximation when using the Without Inertia model. In the case of backflow it is assumed that the head rise for all negative flow rates is equivalent to the head rise at zero flow for the given curve at the given speed. With the head rise known, the negative flow rate through the pump is obtained by using both compatibility equations with the known head rise. After the flow rate is known, the pump suction and discharge pressures are calculated.

To avoid this approximation, a four quadrant model can be used.

Backflow through the pump – four quadrants The model “Startup – Four Quadrant Known Speed” uses the four quadrant pump model, which allows backflow through the pump for pump startup modeling with known speed. If the pump operates at full speed and backflow is possible, the model “No Transient – Four Quadrant” can be used.

Flow rates greater the pump maximum – one quadrant It is possible for the system transients to cause flow rates through the pump greater than the maximum flow from the curve. In such cases, the Without Inertia model assumes the head rise is zero and calculates the flow rate from the compatibility equations.

To avoid this approximation, a four quadrant model can be used.

Flow rates greater the pump maximum – four quadrants The model “Startup – Four Quadrant Known Speed” uses the four quadrant pump model, which allows flows greater the pump maximum to be modeled for pump startups with known speed. If the pump operates at full speed and flows greater the pump maximum are possible, the model “No Transient – Four Quadrant” can be used.



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:

⎟⎟⎠

⎞⎜⎜⎝

⎛ −=

c

bBm downM

pump 2, α

&

( )( ) ⎥

⎥⎥

⎦

⎤

⎢⎢⎢

⎣

⎡

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

−

−+−−×

21

2,

,24

11α

α

bB

CPab

downM

downMR

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.

oldupvapornewupvapor VV ,,,, =

( )ρ2,,,,t

mmmm oldpumpnewpumpoldupnewupΔ++−−+ &&&&

olddownvapornewdownvapor VV ,,,, =

( )ρ2,,,,t

mmmm olddownnewdownoldpumpnewpumpΔ++−−+ &&&&

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.


8/10/2007 4:10:28 PM Page 395


⎟⎟⎠

⎞⎜⎜⎝

⎛ −=

2,

2 ρ

υα

c

bBm downM

pump&

( )( ) ⎥

⎥⎥

⎦

⎤

⎢⎢⎢

⎣

⎡

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛

−

−+−−×

21

2,

,24

11ρα

α

bB

CPab

downM

downMsat

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.

Modeling pump with unknown speed

There are four pump models in AFT Impulse 4.0 that account for pump inertial effects and predict the speed over time. These are:

• Trip With Inertia and No Back Flow or Reverse Speed

• Startup With Inertia and No Back Flow or Reverse Speed

• Trip With Inertia – Four Quadrant

• Startup With Inertia – Four Quadrant, Known Motor Torque/Speed

Predicting the response of a pump during a trip or startup requires additional data for the pump. Use of the affinity laws for pumps is essential. If backflow is possible, "four quadrant" pump curve data is required. The first two models mentioned above assume the presence of a check valve which prevents backflow, or just that backflow never occurs. In such cases, the pump will usually operate in the first quadrant. Thus four quadrant data is not required.

The third and fourth models allow backflow, and thus require four quadrant data. All four models will be discussed in the following sections.

Trip With Inertia and No Back Flow or Reverse Speed

When no backflow or reverse speed occurs through the pump, the normal pump curve data for flow vs. head can be used to model the pump response. Pump power data is also required as a function of flow.



The method is presented by Watters (1979, p. 169). The unbalanced torque is given by:

dt

dIT

ω−= (9.34)

where T is the torque on the pump, I is the moment of inertia of the pump and entrained liquid, and dω/dt is the change in angular speed with time.

The pump power is related to the pump torque by:

ωTP = (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

∫∫ −=T

ddt

I

ω1

Assuming an average value of torque over the time step one obtains:

ωΔ−=ΔTI

t 1

With the speed known at the previous time step, one can solve for the new speed as

I

tToldnew

Δ−= ωω

Adding in the definition of torque from Equation 9.35

I

tTT oldnewoldnew

Δ+−=2

ωω (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.


8/10/2007 4:10:28 PM Page 397


Startup With Inertia and No Back Flow or Reverse Speed

This model uses a similar approach to that just discussed, but adds motor torque into the calculation. This is described, for example, in Brown & Rogers (1980).

dt

dITT fluidmotor

ω=−

Integrating over a time step:

I

tTTTT oldfluidnewfluidoldmotornewmotor

oldnew

Δ⎟⎟⎠

⎞⎜⎜⎝

⎛ +−

+

+=

22,,,,

ωω

(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.



Figure 9.3 Motor torque vs. speed data is used for startup modeling

This calculation uses the user provided pump data for head vs. flow and power vs. flow, and thus only works for positive or zero flows. The mass flow rate and head are obtained from Equations 9.19 and 9.20 (with the pressure converted to head) and the pump curve supplied by the user.

The typical shape of a motor torque vs. speed curve is shown in Karassik (2001, p. 8.25, Figure 26), which is reflected in the Figure 9.4 data.


8/10/2007 4:10:28 PM Page 399


Figure 9.4 Example graph of motor torque vs. speed

Trip With Inertia – Four Quadrant

The method used for modeling a pump trip is described in most waterhammer textbooks (Chaudhry, 1987, Chapter 4, Swaffield, et al., 1993, pp. 154-166, Wylie et al., 1993, Chapter 7). From the conventional method, four dimensionless parameters are defined:

R

RR

R

RR

N

N

m

m

Q

Qv

T

T

P

P

H

Hh

=

==

=

ΔΔ=

ΔΔ=

α

β

&

& (9.38)



where the R subscript refers to rated conditions, usually those at the best efficiency point. In summary, h is the ratio of the pump head to the rated head, β is the ratio of the pump torque to the rated torque, v is the ratio of the flowrate to the rated flowrate, and α is the ratio of the synchronous speed to the rated synchronous speed.

Figure 9.5 Pump window with pump data entered for specific speed of 0.46.

Two further dimensionless parameters are created to relate these parameters:

22

22

)(

)(

vF

v

hF

B

H

+=

+=

αβθ

αθ

(9.39a)

where


8/10/2007 4:10:28 PM Page 401


απθ v1tan−+= (9.39b)

Some references choose to define θ slightly differently:

v

αθ 1tan−=

Figure 9.6 Variation of FH and FB for a pump with specific speed of 0.46 (see Wylie, et al., 1993).

AFT Impulse allows you to select either definition as the basis for the FH and FB data in Equation 9.39a.

The parameters FH and FB are properties of the pump, and it is typically assumed that for pumps of similar specific speed they are invariant (see Brown & Rogers for a discussion questioning this assumption). Figures 9.5 and 9.6 show data for a pump with a specific speed of 0.46 as given by Wylie, et al., (1993, pp. 147-148).

AFT Impulse provides data of FH and FB vs. θ for pumps of twenty-one specific speeds. See Table 9.1 for a list of all specific speeds available.



If you should need to construct data for other sets, please consult the three textbooks referenced above for additional information. Note that the textbooks do not give a method for constructing these data sets, but they contain references that you can pursue.

Table 9.1 Four quadrant data available in AFT Impulse

-- Specific Speed -- Reference(-) (gpm units) (metric units)

0.3 810 15.7 Brown & Rogers, Partial, From Nevada Test, 19800.39 1060 20.5 Kittredge, 19560.42 1140 22.1 Brown & Rogers, from Bureau of Reclamation, 19800.46 1270 24 Donsky, 19610.48 1320 25.5 Brown & Rogers, Partial, From Nevada Test, 19800.55 1490 28.8 Brown & Rogers, Partial, From Nevada Test, 19800.57 1565 30.3 Brown & Rogers, Partial, From Nevada Test, 19800.71 1935 37.4 Kittredge, 19560.79 2160 41.8 Thorley, 19961.21 3300 64 Thorley, 19961.36 3725 71.9 Thorley, 19961.44 3940 76.1 Thorley, 19961.61 4400 85.1 Brown & Rogers, from Bureau of Reclamation, 19801.83 5000 97 Thorley, 19961.9 5200 101 Thorley, 1996

2.32 6340 123 Thorley, 19962.48 6790 131 Thorley, 19962.54 6940 134 Thorley, 19962.78 7600 147 Donsky, 19613.2 8760 170 Thorley, 1996

4.94 13500 261 Donsky, 1961

With data for FH and FB, the compatibility equations (9.19, 9.20) are used as follows:

( ) pumpPMPMpump mBBCCP &++−=Δ (9.40)

But from Equations 9.38 and 9.39,

( ) HRpump FvPP 22 +=Δ α (9.41)

Rpump mvm && =

Substituting Equations 9.41 into 9.40 obtains:


8/10/2007 4:10:28 PM Page 403


( ) ( ) 022 =+++−− HRMPRMP FvPBBmvCC α& (9.42)

Since the value of FH depends on θ, which depends on α and v (see Equation 9.39b), iteration is required to find the appropriate value from the tabulated data supplied (e.g., Figure 9.5).

We need one more equation to solve for all parameters in Equation 9.38. We can get this from the pump torque. The change in rotational speed with torque was given in Equation 9.34. Defining several more terms,

R

R

R

T

T

T

T

N

=

=

=

β

β

απω

00

60

2

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,

( )060

2

60

2 ααπαπω −== RR NdNd

Equation 9.34 can be written

dt

dI

TTT

ω−=+=2

0 (9.43)

Substituting these terms into Equation 9.43 obtains

( ) 015 0 =−

Δ−+ ααπββ

tT

NI

R

Ro (9.44)

Using Equations 9.38 and 9.39 in Equation 9.44 obtains

( ) ( ) 015 0

22 =−Δ

−++ ααπβαtT

NIFv

R

RoB (9.45)



Equations 9.42 and 9.45 are solved simultaneously using a Newton-Raphson iterative method. Details are given in Wylie, et al., 1993, pp. 149-152.

Startup With Inertia – Four Quadrant, Known Motor Torque/Speed

For pump startup where backwards flow occurs, four quadrant data can be used along with the known motor torque vs. speed to calculate the time to reach full speed. This model is especially useful for cases where a pump is being started with backflow already established.

The methodology is similar to that just discussed Trip With Inertia – Four Quadrant section. However, Equations 9.43, 9.44 and 9.45 must be modified to include the motor torque. Including the motor torque changes these equations as follows:

dt

dI

TTTTTT

ffmmfm

ω=+

−+

=−22

0,0,

where the m subscript refers to the motor torque and the f subscript refers to the fluid torque. Equation 9.44 becomes:

( ) 015 0,, =−

Δ−−−+ ααπββββ

tT

NI

R

Rommoff

and Equation 9.45 becomes:

( ) ( ) 015 0,,

22 =−Δ

−−−++ ααπβββαtT

NIFv

R

RommofB

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:

( ) 9556.0

37105.1 ⎟⎟

⎠

⎞⎜⎜⎝

⎛=N

PI p (9.46)


8/10/2007 4:10:28 PM Page 405


where P is the power in kW, N is the rotational speed in rpm and I is in kg-m2.

An estimate of the motor rotational inertia is

48.1

118 ⎟⎠⎞

⎜⎝⎛=

N

PIm (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 liquid- filled container in a system (Wylie, et al., 1993, pp. 124-125). The interaction with the pipe system occurs as the container stretches and contracts in response to transient pressures.

The liquid accumulator is similar to a branch, and thus the same equations are involved.

B

newACnewj S

mSP ,

,&+

=

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,

VV

PK

ΔΔ=′

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:

( )ρ2,,t

mmV oldAnewAΔ+−=Δ &&

Therefore,

( )ρV

mmtKPP oldAnewA

oldjnewj 2,,

,,&& +Δ′

−=



Substituting and eliminating flow rate

tK

VS

mtK

VPS

P

B

oldAoldjC

newj

Δ′+

−Δ′

+= ρ

ρ

2

2,,

,

&

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:

( )( )

( ) ρ2,,

,,

,,

,,t

mm

mm

mm

VV

oldAnewA

oldoutnewout

oldinnewin

oldvapornewvaporΔ

⎟⎟⎟⎟

⎠

⎞

⎜⎜⎜⎜

⎝

⎛

+−

++

−−

+= ∑∑

&&

&&

&&

where Am& is the accumulator flow out of the junction. The inm& terms

are obtained from the positive compatibility equation, while the

outm& terms are obtained from the negative equation.


Relief Valve theory

Relief valve junctions stay closed until the cracking pressure is reached, at which point it will open and relieve the pressure. The valve can relieve either to an outside ambient pressure or into another pipe in the system. There are therefore two models required to address each case.

Internal relief valve

An internal relief valve is similar to a regular valve except that it remains closed until the cracking pressure is exceeded. One significant difference is if the Variable Cv model is used. This is distinct to the relief valve junction.


8/10/2007 4:10:28 PM Page 407


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.

B

newexitCnewj S

mSP ,

,&−

=


The pressure drop across the exit is

2

2

1VP ρ=Δ

or,

2,2,

2newexit

exitexitnewj m

A

KPPP &

ρ=−=Δ

Substituting

2,2

,

2newexit

exitexit

B

newexitC mA

KP

S

mS&

&

ρ=−

−

This is a quadratic equation in which newexitm ,& can be obtained from the

quadratic formula.

0221

2,22

, =⎟⎟⎠

⎞⎜⎜⎝

⎛−++

B

Cexit

exitnewexit

exitBnewexit S

SP

A

Km

A

K

Sm

ρ&&

If using the variable Cv model, the K factor above itself becomes a function of ΔP, which requires iteration to solve for the mass flow rate.

Inline Relief Valve

When this valve opens, it is similar to a Spray Discharge junction with two pipes connected. When it is closed it acts like a branch junction with two pipes. If the Variable Cv model is used, the solution requires iteration.



Variable Cv model

For passive relief valves the valve Cv typically depends on the pressure difference across the valve, as the pressure difference is what causes the valve to open. When the pressure difference is below the cracking pressure, the valve remains closed. When it rises above the cracking pressure, it opens. As the pressure difference increases, it opens further until it is fully open. Any further pressure increases do not increase the Cv, and so the fully open Cv is used for all higher pressures.

When working with the variable Cv model, the cracking pressure is entered in pressure/head difference (see Figure 9.7). In addition, the first Cv data must equal zero (see Figure 9.8), and the corresponding pressure/head to zero Cv must be the same as the cracking pressure/head difference (shown as 700 kPa in both Figures 9.7 and 9.8).

In Figure 9.8, the first data point represents the valve closed at the cracking pressure, with any greater pressure difference resulting in a positive Cv, which means that flow will occur. The maximum pressure difference in the table (770 kPa in Figure 9.8) represents the pressure difference to fully open the valve. The Cv of 150 thus represents the Cv when the valve is fully open. Any pressure differences greater than 770 kPa will result in a Cv of 150, the maximum value.

Figure 9.7 Variable Cv model must use pressure/head difference


8/10/2007 4:10:28 PM Page 409


Relief Valve vapor cavitation theory

Internal relief valves Internal Relief Valve junction vapor cavitation is identical to that of a regular valve. Initially the valve is closed and acts like a regular valve. If it cracks open, then it acts like an open valve.

Exit relief valves Exit Relief Valve junction vapor cavitation acts like a dead end junction.

Inline relief valves Inline Relief Valve junction vapor cavitation acts like a branch junction with two pipes.



Figure 9.8 First data point must have Cv of zero with the pressure/head loss equal to the cracking pressure/head difference (as shown in Figure 9.7)

Check Valve theory

A check valve is identical to a regular valve, except that it closes when reverse flow occurs.

Check Valve vapor cavitation theory

The check valve vapor cavitation model is identical to a regular valve, except that it closes when reverse flow occurs.


8/10/2007 4:10:28 PM Page 411


Control Valve theory

We can use the methods for known flow or pressure to calculate the conditions at a control valve. AFT Impulse flow and pressure control valves allow the control setting to be changed with time. Whether constant or changing, the flow or pressure is known and can be used to solve the equations.

Flow control valves

With the known flowrate, one can substitute into Equations 9.19 and 9.20 and solve for the upstream and downstream pressures, respectively, at the valve. These can then be compared to the failure settings to see if loss of control occurs. For instance, if the upstream pressure was less than the downstream, that would indicate loss of control.

Once the valve loses control, it acts like a regular valve as discussed earlier in the chapter.

Pressure control valves

Pressure control can be specified on the upstream or downstream side of the valve. A discussion of one these control locations will apply to the other. Let's consider the downstream pressure control case – a Pressure Reducing Valve (PRV).

With the known pressure downstream, one can substitute the known pressure into Equation 9.20 and solve for the flowrate through the valve. This flowrate can then be substituted into Equation 9.19 to solve for the upstream pressure. The upstream pressure can then be compared to the control pressure and failure settings to see if loss of control occurs. For instance, if the upstream pressure was less than the control pressure, that would indicate loss of control.

Once the valve loses control, it acts like a regular valve as discussed earlier in the chapter.

Pressure drop control valves

The flow through a constant pressure drop is obtained by subtracting Equation 9.20 from 9.19 and solving for flowrate.



MP

MPnewi BB

PCCm

+Δ−−=,&

Surge Tank theory

A surge tank is an open tank to the atmosphere (Figure 9.9) and is used as a surge suppression device (Wylie, et al., 1993, pp. 128-129).

The flow balance on the surge tank junction is similar to a branch junction, with the deficit equaling the flow out of the surge tank, Tm& .

Similar to a branch, it can be shown that for all connecting pipes to the junction

B

newTCnewj S

mSP ,

,&+

=


LL

LC

ConnectorPipe

Orifice

Tm&

Figure 9.9 Surge tank schematic

The optional connector pipe and orifice effect is incorporated by the lumped inertia relationship as follows:

newTnewLsurfacenewj mCCgLPP ,21,, &−−+= ρ


8/10/2007 4:10:28 PM Page 413


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:

tA

mmLL

T

oldTnewToldLnewL Δ

+−=

ρ2,,

,,&&

Substitution yields

⎟⎟⎠

⎞⎜⎜⎝

⎛ Δ+−+−

+

T

oldToldLsurface

B

newTC

A

tmLgP

S

mS

ρρ

2,

,, &&

02 ,21, =−−

Δ+ newT

T

newT mCCA

tmg &&

Solving for newTm ,&

⎟⎟⎠

⎞⎜⎜⎝

⎛+Δ+−

+⎟⎟⎠

⎞⎜⎜⎝

⎛ Δ−−−

=

BT

T

oldToldLsurface

B

C

newT

SA

tgC

CA

tmLgP

S

S

m1

2

2

2

1,

,

,ρ

ρ&

&

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 9.10) is an enclosed tank filled with gas and has a free surface. It is used as a surge suppressor (Wylie, et al., 1993, pp. 125-128).

The flow balance on the accumulator junction is similar to a branch junction, with the deficit equaling the flow out of the accumulator, Am& .

Similar to a branch, it can be shown that for all connecting pipes to the junction,

B

newACnewj S

mSP ,

,&+

=


The potential connector pipe and orifice effect is incorporated by the lumped inertia relationship as follows

newAnewAnewj mCCPP ,21,, &−−=

where C1 and C2 are defined in conjunction with lumped inertia. Refer to Equations 9-50 and 9-51 later in this chapter.

LC

ConnectorPipe

Orifice

PA

Am&

Figure 9.10 Gas accumulator schematic


8/10/2007 4:10:28 PM Page 415


The gas volume and pressure in the accumulator can be related by the thermodynamic law

An

AA CVP =

where n is the polytropic constant (1.4 for isentropic air, 1.0 for isothermal) and CA is a constant.

Therefore,

An

newAnewA CVP =,,

and

( )ρ2,,,,t

mmVV oldAnewAoldAnewAΔ+−= &&

By substitution

newAnewAB

newAC mCCPS

mS,21,

, &&

−−=+

or

newAB

newACnewA mCC

S

mSP ,21

,, &

&++

+=

Combining equations and simplifying,

( )( ) An

newAnewA CmCCmCC =++ ,65,43 &&

where

ρ

ρ

2

2

1

6

,,5

24

13

tC

tmVC

CS

C

CS

SC

oldAoldA

B

B

C

Δ=

Δ+=

+=

+=

&



This is a non-linear equation that can be solved using Newton-Raphson iteration. Newton-Raphson has the following form:

( )( )i

iii xF

xFxx

′−=+1

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

( )( ) An

newAnewA CmCCmCCF −++= ,65,43 &&

When the value of newAm ,& has been found that causes F to equal zero,

the equation will be satisfied. To solve by Newton-Raphson the derivative of F is given by,

( )nnewAnewAnewA

mCCCmCC

CF

md

dF,654

,65

7

,&

&&++

+=′=

where

67 CnCC A−=

By iteration, newAm ,& is calculated. This value can then be back-

substituted to obtain the new junction pressure and new gas volume.

Gas Accumulator vapor cavitation theory

Any short-lived vapor cavities that develop at a Gas Accumulator junction are assumed to have a negligible effect compared to the accumulator itself. Conceptually, vapor bubbles would rise by buoyancy to the accumulator surface and evaporate into the accumulator gas. Their effect on the system would thus be negligible compared to that of the accumulator gas. Therefore, vapor cavitation is ignored at gas accumulators.

Vacuum Breaker Valve theory

A vacuum breaker valve (sometimes called an air inlet valve) allows air to flow into the pipe during low pressure events, and expels the air when


8/10/2007 4:10:28 PM Page 417


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

( )( )γγγγ−+

⎟⎠⎞

⎜⎝⎛ −+=

12)1(

2

11

o

oind

RT

PACm& (9.48a)

Subsonic inflow

( )

⎥⎥

⎦

⎤

⎢⎢

⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛−⎟⎟

⎠

⎞⎜⎜⎝

⎛−

=+ γγγ

γγ 12

1

2

o

i

o

i

o

oind

P

P

P

P

RT

PACm& (9.48b)

Sonic outflow

( )( )γγγγ−+

⎟⎠⎞

⎜⎝⎛ −+=

12)1(

2

11

L

ioutd

RT

PACm& (9.48c)

Subsonic outflow

( )

⎥⎥

⎦

⎤

⎢⎢

⎣

⎡⎟⎟⎠

⎞⎜⎜⎝

⎛−⎟⎟

⎠

⎞⎜⎜⎝

⎛−

=+−− γγγ

γγ 12

1

2

o

i

o

i

L

ioutd

P

P

P

P

RT

PACm& (9.48d)

where:

Pi = pressure in the pipe at the junction node

Po = atmospheric pressure

To = atmospheric temperature

TL = liquid temperature



CdAin = flow area for inflow

CdAout = flow area for outflow

R = gas constant

γ = specific heat ratio cp/cv

Equation of state

Besides the compatibility equations (9.19 and 9.20), we need another equation to form a complete set. The ideal gas equation of state is used:

mRTVP =

Substituting terms one obtains

( )

( )⎟⎠⎞

⎜⎝⎛ +Δ+

=⎟⎠⎞

⎜⎝⎛ −−+Δ+

oldgasnewgasold

oldupnewupolddownnewdownoldL

i

mmt

m

QQQQt

VRT

P

,,

,,,,

2

2

&&

Eliminating the volumetric flowrate using the mass flowrate obtains

( )

⎟⎠⎞

⎜⎝⎛ ++

Δ

=⎟⎟⎠

⎞⎜⎜⎝

⎛−−++

Δ

oldgasnewgasold

oldupnewupolddownnewdownL

old

L

i

mmt

m

mmmmt

V

RT

P

,,

,,,,

2

12

&&

&&&&ρ

Using Equations 9.19 and 9.20 to eliminate mass flowrate yields:

⎟⎠⎞

⎜⎝⎛ ++

Δ

=⎟⎟⎠

⎞⎜⎜⎝

⎛⎟⎟⎠

⎞⎜⎜⎝

⎛−−−+−+

Δ

newgasoldgasold

oldupP

iPolddown

M

Mi

L

old

L

i

mmt

m

mB

PCm

B

CP

t

V

RT

P

,,

,,

2

12

&&

&&ρ

(9.49)

Now we can use Equations 9.48 to eliminate newgasm ,& as follows:


8/10/2007 4:10:28 PM Page 419


Sonic Inflow The pressure at the junction can be solved directly from Equation 9.49 as follows:

02

2

112

=⎟⎠⎞

⎜⎝⎛ ++

Δ−

⎟⎟⎠

⎞⎜⎜⎝

⎛

Δ+−+−−+

⎟⎟⎠

⎞⎜⎜⎝

⎛+

sonicgasold

L

old

L

u

L

d

PL

P

ML

Mi

PLMLi

mmt

mRT

t

Vmm

B

C

B

CP

BBP

old

oldold

&&

&&

ρρρρ

ρρ

where sonicm& is the sonic inflow which can be calculated from Equation

9.49a based completely on input data.

Subsonic Inflow Equation 9.49 is modified using Equation 9.48b and solved iteratively using Newton-Raphson. Defining Pr = Pi/Po obtains:

( )( )( ) 0

1

2

2

2

11

12

22

=−⎟⎟⎠

⎞⎜⎜⎝

⎛−

+

⎟⎠⎞

⎜⎝⎛ +

Δ−

⎟⎟⎠

⎞⎜⎜⎝

⎛−+−−

Δ+

⎟⎟⎠

⎞⎜⎜⎝

⎛+

+ γγγγ

γ

ρρρρ

ρρ

rro

Loind

gasold

L

L

u

L

d

PL

P

ML

Moldor

PLMLor

PPT

RTPAC

mt

mRT

mm

B

C

B

C

t

VPP

BBPP

old

oldold

&

&&

Defining terms for Newton-Raphson solution yields



( ) FCPCPCPPC rrb

ra

r =+++− 432

25.0

1

( ) ( )r

rb

ra

rb

ra

r dP

dFCPCbPaPPPC =++−− −−−

32115.0

1 25.0

( )

⎟⎠⎞

⎜⎝⎛ +

Δ−=

⎟⎟⎠

⎞⎜⎜⎝

⎛−+−−

Δ=

⎟⎟⎠

⎞⎜⎜⎝

⎛+=

−=

+=

=

old

oldold

gasold

L

L

u

L

d

PL

P

ML

Moldo

PLMLo

oLoind

mt

mRTC

mm

B

C

B

C

t

VPC

BBPC

T

RTPACC

b

a

&

&&

2

2

11

1

2

1

2

4

3

22

1

ρρρρ

ρρ

γγ

γγγ

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:


8/10/2007 4:10:28 PM Page 421


( ) ( )[ ] 01

2

2

2

11

12

22

=−−

+

⎟⎠⎞

⎜⎝⎛ +

Δ−

⎟⎟⎠

⎞⎜⎜⎝

⎛−+−−

Δ+

⎟⎟⎠

⎞⎜⎜⎝

⎛+

+−− γγγγγ

ρρρρ

ρρ

rrL

ooutDr

gasold

L

L

u

L

d

PL

P

ML

Moldor

PLMLor

PPRT

PACP

mt

mRT

mm

B

C

B

C

t

VPP

BBPP

old

oldold

&

&&

Defining terms for Newton-Raphson solution yields

( ) FP

CCPCPPCr

rb

ra

r =+++− −− 1432

5.01

( ) ( )rr

br

ar

br

arr

dP

dF

PCC

bPaPPPPC

=−+

+−− −−−−−−−

242

115.01

1

5.0

⎟⎠⎞

⎜⎝⎛ +

Δ−=

⎟⎟⎠

⎞⎜⎜⎝

⎛−+−−

Δ=

⎟⎟⎠

⎞⎜⎜⎝

⎛+=

−=

+=

=

old

oldold

gasold

L

L

u

L

d

PL

P

ML

Moldo

PLMLo

Looutd

mt

mRTC

mm

B

C

B

C

t

VPC

BBPC

RTPACC

b

a

&

&&

2

2

11

1

2

1

2

4

3

22

1

ρρρρ

ρρ

γγ

γγγ

where the function F is solved for Pr.



Sonic Outflow Using Equation 9.48c, the pressure at the junction can be solved directly from Equation 9.49 as follows:

( ) ( )( )

02

2

11

2

11

121

2

=⎟⎠⎞

⎜⎝⎛ +

Δ−

⎟⎟

⎠

⎞

⎜⎜

⎝

⎛⎟⎠⎞

⎜⎝⎛ −++

⎟⎟⎠

⎞⎜⎜⎝

⎛−+−−

Δ+

⎟⎟⎠

⎞⎜⎜⎝

⎛+

−+

old

oldold

gasold

L

LoutDi

L

u

L

d

PL

P

ML

Moldi

PLMLi

mt

mRT

RTACP

mm

B

C

B

C

t

VP

BBP

&

&&

γγγγ

ρρρρ

ρρ

Lumped inertia pipe with orifice theory

AFT Impulse uses a pipe with lumped inertia (Wylie, et al., 1993, pp. 122-124) as a connector for some elements, although it is not a complete element by itself. Such an element may also include an orifice (or a check valve, which will be treated as an orifice) as shown in Figure 9.11.

1

3

2

orifice

Δz

Figure 9.11 Lumped inertia with friction schematic


8/10/2007 4:10:28 PM Page 423


Summing forces acting on the lumped section, 2,

dt

dVLAFFFFF gorificefriction

22231 ρ=+−−−

or,

dt

mdLFFFFF gorificefriction

2231

&=+−−−

Now,

zgAF

AAC

mF

AA

m

D

fLF

APF

APF

g

oDorifice

friction

Δ=

=

=

==

2

222

22

222

22

2

2

233

211

2

2

ρρ

ρ

&

&

Substituting terms and dividing by area A2

dt

md

A

Lgzm

ACm

AD

fLPP

oD

2

2

2222

222

22

231 22

1

2

1 &&& =−−−− ρ

ρρ

Taking average values over time, and denoting current time by a prime,

22222

23311

2

1

22mm

AD

fLPPPP&& ′−+′

−+′

ρ

( )

t

mm

A

Lgzmm

AC oD Δ−′

=−′− 22

2

222222

1 &&&& ρ

ρ

Collecting terms and simplifying

22131 mCCPP ′+=′−′ &

where



22

2131

22 m

tA

LgzPPC &

Δ−+−= ρ (9.50)

tA

Lm

ACm

AD

fLC

oD Δ++=

2

222222

22

22

211&&

ρρ (9.51)

If a connector pipe exists but no orifice then

tA

Lm

AD

fLC

mtA

LgzPPC

Δ+=

Δ−+−=

2

222

22

22

22

2131

21

22

&

&

ρ

ρ

If an orifice exists but no connector pipe

2222

131

1m

ACC

PPC

oD

&ρ

=

−=

If neither exist then

0

0

2

1

==

C

C

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


8/10/2007 4:10:28 PM Page 425


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.



Wylie, et al., (1993, pp. 4) gives the instantaneous waterhammer equation as:

VaP Δ−=Δ ρ (9.52)

where:


ρ = density

a = wavespeed


To apply this equation, one would typically determine the steady-state velocity and assume the flow stops suddenly. The -ΔV is thus equal to the steady flow velocity.

Following is a description of cases where the pressure surge can exceed that predicted by the instantaneous waterhammer equation.

Cavitation

When a positive pressure surge occurs, it is always followed by negative pressure surge that drops the pressure. If the drop in pressure causes cavitation, then the pressure spike that occurs when the cavity collapses can cause the local pressure to exceed that predicted by Equation 9.52.

The potential to cavitate can be assessed by comparing the pressure rise in Equation 9.52 to the steady-state pressure. If the pressure surge is greater than the steady-state pressure, then cavitation is possible. If it is less than the steady-state pressure and the fluid has a low vapor pressure (like water), then it will probably not cavitate. If the fluid has high vapor pressure, this will need to be subtracted from the steady-state pressure and then compared to the pressure surge.

Pipe size changes

When using Equation 9.52, the case considered is usually a valve closure. Equation 9.52 is frequently applied with the ΔV obtained based on the pipe diameter at the valve. However, if smaller diameter pipes exist upstream that carry the same flowrate, their ΔV will be larger because they have less area. This will result in a larger pressure surge in the smaller pipe.


8/10/2007 4:10:28 PM Page 427


To be safe, Equation 9.52 needs to be applied to the pipe with the largest ΔV potential in the entire system.

Multiple pressure surge sources

If there is potential for multiple elements to generate pressure surge at the same time, the surge pressures can interact and add together, thus exceeding the individual surge pressures.

Line pack due to friction

In systems with high friction, pressure can increase above the instantaneous values because of something called line pack. Even systems with modest friction can experience this. In brief, the friction in the system prevents the flow from coming to a complete stop in cases such as a closing valve. With the closed valve, the fluid continues to move towards the closed valve, albeit at a lower velocity, increasing pressures above the instantaneous prediction. See Wylie (1993) for more information.

Elevation changes

If the pressure surge propagates to an area of the system at a lower elevation, the peak pressure can be higher because of the higher static pressure.


C H A P T E R 1 0

Modeling Time and Event Based Transients

In general, transients can be initiated in one of two ways. They can be initiated according to a prescribed time schedule or according to certain events that happen in the system. These two categories are referred to as time-based and event-based transients.

Event-based transients break down further into three categories: Single event, cyclic dual events and sequential dual events. The behavior of time-based and the three event-based transients are described in this chapter.

Time-based transients

Time-based transients occur in accordance with the time of the transient simulation. The start and stop times of the simulation are specified in the Transient Control window. The initiation of time-based transients are pre-specified before a model is run.

For example, consider a valve closure transient. Assume the valve starts to close at two seconds into the simulation, and the time it takes for the valve to close is one second. After that, the valve stays closed. If the initial valve Cv is 250, the transient would appear as in Figure 10.1.

In the "Initiation of Transient" area there are four options. For time-based transients, the option is specified as Time. In the Transient Data area the data is entered. Here the Cv profile of the valve will follow the profile entered and start to close at 2 seconds.



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

Chapter 10 Modeling Time and Event Based Transients 431

8/10/2007 4:10:28 PM Page 431


criteria is such that the transient will be initiated when the outlet static pressure in Pipe 2 exceeds 1000 kPa. If the pressure at this location never reaches 1000 kPa, the transient specified in the Transient Data area will not occur.

Figure 10.2 Single event transient specifies how a junction will respond if and when some condition is satisfied.

The data in the Transient Data table has a slightly different meaning than does a time-based transient. Here time zero is relative to the time at which the event criteria is first met. For example, if Pipe 2 reaches 1000 kPa at 3.65 seconds, the valve will start its transient at 3.65 seconds. The valve will close (i.e., Cv = 0) at 1 second after the event initiation, or 4.65 seconds.

Dual event transients: cyclic and sequential

Dual event transients contain two different transients for the junction.



Cyclic events Cyclic dual events can repeatedly switch from one to the other based on the behavior of the model. For example, these could be high and low pressures at some location. When the low pressure setting is reached a valve transient is initiated which closes the valve. The pressure may then build up to a point where the high pressure setting is reached. In this case a second valve transient is initiated which opens the valve. Should the pressure drop to the low pressure setting again, the valve will again close. The two settings will continue to cycle the valve on and off until the end of the simulation.

Cyclic dual events are specified by selecting Dual Event Cyclic in the Initiation of Transient area of the junction window (Figure 10.3). When selected, two event tabs appear. Here the first and second event can be specified. In this case the first event is a pressure at Pipe 2 inlet less than 500 kPa. The transient that is initiated when this event occurs is in the table of the First Transient tab in Figure 10.3. Here the valve closes one second after the event.

A second event, not shown, is specified on the Second Event tab, with the accompanying transient specified on the Second Transient tab, also not shown. Typically, the first data of the second transient should match the final data of the first transient. In other words, if the valve is closed as in Figure 10.3 by setting the Cv value to 0, the first data point of the second transient should be zero and then proceed to some higher value which represents the opening of the valve.

More often than not the two events will specify the same parameter at the same location. For example, if the first event specified a pressure at a certain pipe, the second event would also specify pressure at the same pipe location. However, this is not required. You can specify the first event as pressure and the second as flowrate, for example, or you could specify the two events to be at different locations in the pipe system.


8/10/2007 4:10:28 PM Page 433


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.



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.


8/10/2007 4:10:28 PM Page 435


Thought experiment to further clarify event transients

To further clarify the capabilities of event transient modeling, here we will discuss how one could use a regular AFT Impulse Valve junction to simulate a check valve.

As discussed previously, check valves have inherent event logic. That logic assumes that the valve closing transient is initiated when backflow begins. Once a certain pressure differential exists across the closed check valve, the valve reopens.

One could use a regular valve junction to simulate a check valve as follows. First, the valve would need to be specified as Dual Event Cyclic in the Initiation of Transient area. The first event would be a flowrate that is less than or equal to zero at the inlet of the valve's discharge pipe. A Cv transient could then be entered to simulate how the check valve closes.

The second event would be a differential pressure across the check valve that is greater than or equal to the check valve's reopening pressure differential. A reopening Cv profile could be entered in the second transient data table.

With the above considerations, the valve junction and check valve junction would respond identically.

Other transient features

Absolute vs. relative transient data

In the Transient Data area (see Figure 10.1, 10.2 or 10.3) it can be seen that there are two options: "Absolute Values" and "Relative To Steady-State Values". In the case of a valve junction, the first option allows you to put in actual Cv values (for example) vs. time. The second option allows you to put in a percentage of the steady-state Cv value. While not shown here, the steady-state value is entered on the Loss Model tab.

For junctions other than valves, different parameters are specified as the transient parameter. For example, the transient could be the steady-state pump speed, pump flowrate, spray discharge area, or boundary pressure.



Repeat transient

For transients that are periodic, you can specify that the transient repeats itself. This is specified by selecting the "Repeat Transient" checkbox (see Figure 10.1, 10.2 or 10.3). When Repeat Transient is selected, the first and last data points must match. After the transient reaches the final data point, it returns to the first data point and starts over.

Add time offset

There may be times you want to shift the time data points in the transient data table. For instance, assume you have a valve specified as a time-based transient where the valve starts to close after five seconds. Further assume that you wish to change the transient so that it starts to close after seven seconds. In such a case, you want to shift all time values by two seconds. If you select the Edit Table button (see Figure 10.1, 10.2 or 10.3), there is menu choice for Add Time Offset. If you select this, you are prompted to enter an offset time, and all times in the table will shift by the specified time.

Graphing the transient data

It is helpful to see the transient data in graph form, and this is easily done by selecting the Show Graph button (see Figure 10.1, 10.2 or 10.3).

Event messages

Events that are initiated during the simulation are displayed in the two Event Messages area of the Output window (see Figure 10.4). The first is a list of event messages sorted by junction. The second is a list of event messages sorted by time. If no events occur, the Event Messages tab will be hidden.


8/10/2007 4:10:28 PM Page 437


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.



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.


8/10/2007 4:10:28 PM Page 439


Figure 10.6 All junction transient data is shown in the Model Data window on the Transient Data tab.


C H A P T E R 1 1

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).



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.

Chapter 11 Reducing Surge Pressures 443

8/10/2007 4:10:28 PM Page 443


Vacuum breaker valves for low pressures

Vacuum breaker valves, also called air inlet valves, open to permit air into the system during low pressure conditions. These are usually located at high elevation points along the pipeline.

Surge tank

Surge tanks can be an effective way to reduce surge pressures if the steady-state system pressures are relatively low. High steady-state pressures result in high hydraulic gradelines. Since the liquid level in a surge tank rises to the local hydraulic gradeline, they would have to be very tall if located in a high pressure region.

Gas accumulator

Gas accumulators can be effective in spreading out the wave front of a transient and thus reducing peak pressures. Gas accumulators change the frequency response of the system and can in fact amplify pressure surge if not sized and located properly.


C H A P T E R 1 2

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



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

Chapter 12 Special Topics: Including Troubleshooting Models 447

8/10/2007 4:10:28 PM Page 447


this case, one of the pipe lengths was accidentally entered in miles rather than feet.

Figure 12.1 Workspace layout of a pipe system with an input error in one of the pipes (error not shown).

Poor pump curve fits

The Pump Specification window allows curve fits of pump data up to fourth order. Depending on the number of data points entered, pump curve fits that look like the one in Figure 12.3a are possible.

AFT Impulse interprets flow rates past the inversion as acceptable operating points, which can lead to strange results or non-convergence. You should always take care that your pump curves cross the flow rate axis as shown in Figure 12.3b. If you attempt to generate a curve fit like that shown in Figure 12.3b, AFT Impulse will warn you. In addition, it will warn you when you try to run the model.



Figure 12.2 Model Data window display for the model in Figure 12.1. From this perspective it is much easier to see that one of the pipes (#4) has length data entered in miles rather than feet. This is not automatically an input error, as AFT Impulse accepts pipe input with mixed engineering units. However, more frequently than not the user intends to use consistent units, so this probably is an error.

Use the Output window Sort feature

The Sort Output feature in the Output window is discussed in Chapter 5. If a model runs but the results look incorrect, use the Sort feature in the Output Window to look for extremes in velocity, pressure drop or other parameters. An input error (like incorrect diameter) may be easier to see by looking at its effect on the results.


8/10/2007 4:10:28 PM Page 449


0

2

4

6

8

10

12

0 50 100 150 200

Flow (ft3/sec)

Pre

ssu

re (

psid

)

0

2

4

6

8

10

12

0 50 100 150 200

Flow (ft3/sec)

Pre

ssu

re (

psi

d)

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.

a

b



Tee/Wye junction complexity

As discussed in Chapter 8, tee/wye junctions contain powerful and sophisticated models for determining the pressure losses at flow splits and converging flow streams. This power comes at the cost of additional model complexity. The complexity can lead to convergence problems in models with numerous tees. Here are a couple of recommendations.

Tee/Wye junctions have two loss options: Simple and Detailed. The Simple loss model ignores losses at the tee. The Detailed model uses the correlation summarized in Chapter 8. The potential convergence problems occur when using the Detailed loss model.

The potential problem is not with the Detailed tee model itself. Rather, it is due to multiple tees coupling together in such a way that AFT Impulse cannot converge on all of the flow splits.

You can quickly and easily check if a convergence problem is being caused by tees. Using the Global Junction Edit window, you can change all tees to the Simple loss model and rerun the model. If it now converges, then the problem was due to the use of Detailed tee modeling.

Experience has shown that this problem can frequently be resolved by using absolute tolerance criteria in the Steady Solution Control window rather than relative tolerance. Tolerance criteria is discussed at length in Chapter 8.

You also have the option of just using the Simple (lossless) tee model for all tees. This is how tees have been traditionally handled, so lossless tees do have some utility in engineering calculations.

Lower the flow rate relaxation

The Steady Solution Control window allows you to change the numerical relaxation parameter for both flow rate and pressure. A full explanation of relaxation is given in Chapter 8. It is recommended that you use only 0.5 or 1.0 for pressure relaxation. But lower values of flow rate relaxation can be very helpful.

A good first step is to lower the flow rate relaxation to 0.05 or 0.1. By lowering the flow rate relaxation you improve AFT Impulse's ability to obtain convergence. The cost is slower run times. In some cases, accepting the slower run times that come with flow rate relaxation is the only means of obtaining convergence. You can lower the relaxation


8/10/2007 4:10:28 PM Page 451


values as low as you want, but it is recommended it not be lowered below 0.01.

Try absolute tolerance

Most models converge more reliably with relative tolerances, which is the default in AFT Impulse’s Steady Solution Control window. But some models converge much better with absolute tolerance. What this usually means is that there is at least one element of the model that is close to the real answer in absolute terms but does not lock in on a relative (percentage) basis.

If a model will not converge with relative tolerance, you can try changing to absolute tolerance. Even better, try the option to converge on either absolute or relative, which will offer the most flexibility.

An explanation of absolute and relative tolerance is given in Chapter 8.

Make initial flow rate guesses for pipes

If you do not provide any initial guesses, AFT Impulse will make its own guesses. On occasion, the AFT Impulse guesses are far away from reality. For some models, providing a crude initial guess for all the pipe flow rates in the system can help the model converge.

The crude guesses can be as simple as selecting all the pipes and initializing all of them to 1 kg/sec, even though some operate at 0.1 kg/sec and others at 10 kg/sec.

Review the Iteration History

The Iteration History window is discussed in Chapter 8. You can use this to see which pipes or junctions are farthest out of tolerance, which may direct your troubleshooting efforts toward these elements.

Change boundary conditions

Changing the boundary pressures or flow rates can sometimes give you insight into what is giving the model trouble.



Turn off parts of the model

Using the Special Conditions feature, you can turn off elements in your model and, in effect, shut down entire sections. Then you can focus on getting selected sections to run to convergence.

Break a large model into submodels

AFT Impulse provides convenient features for deleting parts of your model and merging parts back in. A section of your model can be quickly deleted by selecting it on the Workspace and choosing Delete from the Edit menu. If the section is instead Cut, it can be Pasted into a new model of its own for testing or holding. At a later time, the removed sections can be added back into the current model by using the Merge feature on the File menu.

The advantage of breaking your model into sub-models is that you can investigate the convergence characteristics of each part of your model independently. If you can get each sub-model to converge, you should be able to get the original model to converge also. If one of your sub-models will not converge, it is doubtful your original model will converge either.

Control valve failure issues in steady-state

AFT Impulse offers modeling of PRV’s, PSV’s and FCV’s. Based on the interaction with the pipe system, control valves can end up in a situation where they cannot control to the desired control points. When this occurs, it is an indication that the desired control point cannot be obtained unless the control valve acts like a pump and adds pressure to the system. To diagnose which control valves (if any) need to add pressure to reach their control point, set their failure state to Always Control (Never Fail).

It is very important to grasp that when a valve fails to control and the failure action is set to Always Control (Never Fail), the resulting solution in the Output is only useful for diagnostics and does not represent a physically meaningful answer. Why is that? Because in order to maintain the control pressure or flow the valve had to add pressure. If a control valve is operating in this mode, you will be clearly notified in the Output window, both in the Warnings and in the Valve Summary.


8/10/2007 4:10:28 PM Page 453


The following failure options are available for control valves:

1. Maintain control by adding pressure (useful for diagnostics only)

2. Fail open or fail closed if upstream pressure is insufficient

3. Fail open or fail closed if downstream pressure is excessive (applicable to pressure control valves only)

In such cases where failure to control exists, one or more valves may fail simultaneously. Once failed, AFT Impulse will not return a valve to its control capability.

Infinite pipe junctions

An infinite pipe junction is a special junction that accepts pressure waves but never reflects any waves (Wylie, et al., 1993, pp. 121-122). This junction is useful in modeling the part of a pipe system which is so long that its communication time results in no interaction with the rest of the system.

This feature is offered as part of the Assigned Flow and Assigned Pressure junctions. During steady-state a fixed flow or pressure exists at the junction, but during transient flow both the pressure and flow change in a manner consistent with the fact that no waves are reflected. The theory for an infinite pipe is given in Chapter 9.

Judicious use of infinite pipe junctions can significantly reduce model run time. The reason is that if there is a pipe which is so long it will not reflect any waves during the simulation, the extra computation for that pipe can take a lot of time. The infinite pipe junction is a way of avoiding any computations for the pipe but still maintaining its effect on the system.

Application example

A real life example is shown in Figure 12.4. The system represents part of a tank farm for refined petroleum products. Gasoline flows from a refinery some 100 km away into the tank farm. The supply pipeline is represented by pipe P1. Pipes P2-P9 are relatively short local pipes at the tank farm with changing diameters. This model is for gasoline, and the gasoline comes into junction J10 and flows initially to the tank at J106. The transient of interest is when the J201 valve is opened and the J104



valve is closed. This will divert the incoming gasoline from J106 to J212. The total scale of the tank farm piping is less than a mile.

Figure 12.4 Example of where an infinite pipe junction can be used. Here J1 is specified as an infinite pipe. Use of an infinite pipe reduces the model run time by a factor of 100.

The transient lasts for about 30 seconds. During the opening and closing of the valves, the transient waves propagate out to J106 and J212, as well as to the supply junction J1. Using the wavespeed of gasoline it can be determined that any waves that propagate to pipe P1 will not reflect back to the tank farm within 30 seconds, and thus will not impact the transient. However, during the transient procedure, gasoline still flows from junction J1 into the system.


8/10/2007 4:10:28 PM Page 455


One approach is to model pipe P1 as the full 100 km length. Thus all the pipe stations along 100 km must be calculated for the entire 30 second simulation. A second approach is to specify pipe P1 as very short (13 feet long in this case, chosen for modeling convenience based on other considerations) and to specify J1 as an infinite pipe junction.

Both approaches will give the same transient results. However, the first approach would take over 100 times as long to run (a 1.9 hour run vs. a 2 minute run when using an infinite pipe). Again, the difference in runtime is that with an infinite pipe, no computation is required along the 100 km pipe except for two stations at the entrance.

How pressure junctions work in steady-state

Pressure junctions (e.g., Reservoirs or Assigned Pressures) in AFT Impulse are an infinite source or sink of fluid. This means that they can draw or discharge as much fluid as is necessary to maintain the specified pressure. Some engineers have difficulty grasping this concept, and misuse pressure junctions in AFT Impulse. This section offers a physical example of how pressure junctions work to clarify for those having difficulty with the concept.

Consider the physical pumping system shown in Figure 12.5. This system pumps water from a tank near the shore of Lake Michigan to a tank on a nearby hill.

An engineer is tasked with sizing the pump for this system. The engineer has an idea of the discharge head needed for the pump, and so builds an AFT Impulse model as shown in Figure 12.6.



Lake Michigan

ShorelineElevation

TransferPipeline

TransferPump (J3)

DischargeTank (J2)

SupplyTank (J1)

Figure 12.5 Physical pumping system to transfer water from shoreline supply tank (J1) to discharge tank (J2) at top of hill

J1

J2

J3 J4P1 P2

P3

J1

J2

J3 J4P1 P2

P3

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.


8/10/2007 4:10:28 PM Page 457


Lake Michigan (J4)

ShorelineElevation

TransferPipeline

TransferPump (J3)

DischargeTank (J2)

SupplyTank (J1)

Figure 12.7 Physical system represented by model in Figure 12.6.

Why? Because the reservoir junction (J4), which is an infinite source of fluid, has been located between the pump (J3) and the discharge tank (J2). The J4 reservoir isolates the supply tank and pump (J1 and J3) from the discharge tank (J2). The J4 reservoir is very much like placing Lake Michigan between the supply and discharge piping.

If the engineer changes the surface elevation in J4, it is like changing the surface elevation of Lake Michigan. It will change the flow rate in the pipes, but no matter how high the water level of Lake Michigan, it does not change the fact that the pumping system is isolated from the discharge tank. The flow rate in the piping from J4 to J2 is entirely dependent on the difference in surface elevations between J2 and J4, and is not influenced by the J3 pump in any way.

How then should the physical system in Figure 12.5 be modeled? It should be modeled as shown in Figure 12.8.



J1

J2

J3P1

P2

J1

J2

J3P1

P2

Figure 12.8 Proper model of physical system shown in Figure 12.5.

If the goal is to size the pump at J3, the model is Figure 12.8 can be used with the pump modeled as an assigned flow pump.

The role of pressure junctions in steady-state

In modeling a generalized pipe network, it is possible to construct models that do not have a unique solution. A common occurrence of this is when a model contains one or more sections completely bounded by known flow rates. A series of examples with accompanying explanation will clarify situations where this occurs. The length of this topic is proportional to the frequency with which engineers get confused over the issue. It became apparent that an exhaustive discussion would be very helpful for some engineers. Do not dismiss this topic too quickly because of its length.

Before beginning, it will be helpful to consider an aspect of the philosophy of computer modeling. At the risk of stating the obvious, it must be recognized by the user that a computer model cannot calculate anything that cannot, in principle, be calculated by hand. All the computer does is accelerate the calculation process. Some users expect the model to generate independent information and become frustrated when the software requests additional information from them. But if the user was doing the calculation by hand that same information would still be required. The difference is that in the case of the hand calculation, the engineer would be forced to think through why the information is needed.


8/10/2007 4:10:28 PM Page 459


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 = 0.05 m3/s El = 0 m

Outlet Q = 0.05 m3/s El = 0 m

L = 30 m D = 5 cm f = 0.02

Figure 12.9 Model with two assigned flows. This model does not have a unique solution.

Obviously, the flow in the pipe is known, but, what is the pressure at the inlet? At the outlet? It cannot be determined because there is no reference pressure. The reference pressure is that pressure from which other pressures in the system are derived. There can be one or more reference pressures, but there always has to be at least one.

The model in Figure 12.9 can be built with AFT Impulse. If you try to run this model, AFT Impulse will inform you that it cannot run it because of the lack of a reference pressure. In AFT Impulse there are numerous junctions that can act as a reference pressure: the reservoir, assigned pressure, spray discharge, pump with submerged option, and valve and relief valve after cracking (with the exit valve option) behave as reference pressures because that is their nature. Three other junction types, surge tank, gas accumulator, and assigned flow (when the steady-state flow is zero) have a special feature that allows them to act as reference pressures. This feature is found on each junction's Optional tab, and is discussed at length in Chapter 6.

There are a couple other things worth noting about Figure 12.9. First is that the model in Figure 12.9 has redundant boundary conditions. When



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.

Inlet P = 0.1 MPa El = 1.5 m


L = 30 m D = 5 cm f = 0.02

Inlet P = 0.1 MPa El = 1.5 m

Outlet P = 0 MPa(g) El = 1.5 m

L = 30 m D = 5 cm f = 0.02

L = 30 m D = 5 cm f = 0.02


Inlet Q = 0.05 m3/s El = 0 m

a

b

c

Figure 12.10a-c Top two models with one pressure and one assigned flow, bottom model with two pressures. All of these models have a unique solution.

Second, if the user was allowed to specify two flows as in Figure 12.9, he/she could specify them with different flow rates. Clearly, they must have the same flow or an inconsistency occurs. The basic reason the inconsistency would be possible is because Figure 12.9 does not have a unique solution.

Figure 12.10a-c shows three other model configuration possibilities. It is not possible to specify inconsistent conditions for any of the Figure


8/10/2007 4:10:28 PM Page 461


12.10 models, and they always have a unique solution no matter what input is specified by the user.

The four models in Figures 12.9 and 12.10a-c contain all logical possibilities. In all four cases there are four things we want to know: the pressure and flow at the inlet, and the pressure and flow in the outlet. In each case we know two of these four. The problem is that in the Figure 12.9 model the lack of a reference pressure makes it impossible to determine the inlet and outlet pressures, even though we know the flow. The three models in Figure 12.10a-c all have at least one pressure, thus all four of the desired parameters can be determined. Table 12.1 summarizes this. Also note in Table 12.1 that in the Figure 12.10c case the flow can be determined, but it is by iteration.

Table 12.1 Four logical possibilities of boundary conditions for single pipe systems

CASE Inlet Pressure

Inlet Flow

Outlet Pressure

Outlet Flow OK ? Comments

Fig. 12.9 X X No No reference pressure, cannot calculate pressures

Fig. 12.10a X X Yes 1 known flow, can calculate everything directly

Fig. 12.10b X X Yes 1 known flow, can calculate everything directly

Fig. 12.10c X X Yes 2 pressures, can iterate for flow

As we consider multi-pipe systems, there are a host of other possibilities that present themselves. All other configuration possibilities which lack a reference pressure ultimately boil down to the same problem that exists in Figure 12.9.

The pipe lengths, diameters and friction factors are not included in the remaining models because they do not influence the main point of this topic. It is assumed that these parameters can be obtained for each pipe and the resulting pressure drop calculated with standard relationships.

Consider the model in Figure 12.11. With three boundaries having a known flow, this clearly has the same problem as the model in Figure 12.9. No unique pressures at any location can be calculated.



Inlet Q = 0.05 m3/s El = 0 m



Figure 12.11 Three known flows at boundaries that lack a reference pressure, similar to the model in Figure 12.9. This model does not have a unique solution.

A pressure at any one of the three boundaries in Figure 12.11 would be sufficient to allow a unique solution of the system. It is also possible that there could be two pressures and one flow, or three pressures. As long as there is at least one pressure, a unique solution exists.

Now consider the models in Figure 12.12. The Figure 12.12a model does have one reference pressure, but in this case there is still a problem. The Flow Control Valve (FCV) is controlling the flow to 0.05 m3/s, and the downstream flow is demanding 0.05 m3/s. The section of the system preceding the FCV can be solved (because there is a pressure upstream), but the section after the FCV does not have a unique solution. Why? In order to obtain a unique solution, the pressure drop across the FCV must be known. But any pressure drop could exist and satisfy the conditions of this model. It is thus not possible to determine the pressure at the outlet flow demand because it depends on the FCV pressure drop which is not known. The model in 12.12b does have a pressure downstream of the FCV, and thus there is a unique pressure drop across the FCV and a unique solution exists for the model in Figure 12.12b.

If you input the remaining data for the Figure 12.12a model and run it in AFT Impulse, you will get the message shown in Figure 12.13. AFT Impulse identifies the part(s) of the model where a known pressure is needed.


8/10/2007 4:10:28 PM Page 463


Inlet P = 0.1 MPa(g) El = 3 m

Inlet P = 0.1 MPa(g) El = 1.5 m



FCV Q = 0.05 m3/s

FCV Q = 0.05 m3/s

J1 J3 J2

J1 J3 J2

a

b

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.



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.


8/10/2007 4:10:28 PM Page 465




Pump with Known (fixed) Flow Q = 0.05 m3/s

Figure 12.14 A pump modeled as an assigned flow for sizing behaves identically to the FCV model in Figure 12.12a, for which no unique solution exists.

Increasing in model complexity, Figure 12.15 shows some additional examples of models without a unique solution. In particular, Figure 12.15b has a pump with an assigned flow that can add as much pressure as it wants, and the downstream FCV can then take out as much as it wants. There are an infinite number of possible solutions.

While it would be highly unusual for a system to have two FCV’s in series, it is worth considering the model in Figure 12.15b further. It is common to have an FCV in series with a pump. How does one size the pump in such a case? There are several ways to do this, one of which is to change the pump from an assigned flow to an actual pump curve. Trying multiple pump curves will guide you to the best pump.

But there is a better way. Consider for a moment the pressure drop across the FCV in Figure 12.15b. In the installed system, is any pressure drop acceptable? Usually not. It is typical to have pressure drop limits based on the system design and the FCV itself. A common requirement is a minimum pressure drop across the FCV. If no minimum pressure drop exists, let’s say we choose a pump that results in the pressure drop across the FCV being 70 Pa. If the system is built and the pump slightly underperforms its published curve, the FCV will not be able to control to its flow control point. Even if the pump does operate exactly on the published curve, eventually fouling in the pipes will result in increased pipe resistance (and pressure drop) and again the FCV will not be able to control to its flow control point.





FCV Q = 0.05 m3/s

FCV Q = 0.05 m3/s


Outlet P = 0.15 MPa(g) El = 1.5 m

FCV Q = 0.05 m3/s

Pump with Fixed Flow Q = 0.05 m3/s

a

b

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.


8/10/2007 4:10:28 PM Page 467




FCV Q = 0.05 m3/s



PDCV dP = 0.05 Mpa





FCV Q = 0.05 m3/s

Pump with Curve dP = f(Q)

a

b

c

Figure 12.16a-c. The model at the top is the same as Figure 12.15b and does not have a unique solution. To size the pump, change the model to the second one shown above, which uses a pressure drop control valve (PDCV) rather than an FCV. Once the second model is run, the pump is sized, a pump curve exists, and the model at the bottom can be run using an FCV.

Let’s make a slight addition in complexity to the model in Figure 12.16. In this case we have two flow control valves in parallel (see Figure 12.17). To size this pump, we apply the process described in Figure 12.16. We choose one of the FCV’s, make it a PDCV, size the pump, choose a pump with an actual pump curve and enter it into the model, then return the PDCV to an FCV.



FCV Q = 100 gpm

FCV Q = 100 gpm

Pump – known flow Q = 200 gpm

P1> P2>

P4^

P6>

P7^

P5^ ^P8

P3>

J1 J2 J4

J8

J6

J5

J7J3

Figure 12.17 A pump with an assigned flow in series with two parallel FCV’s. The model in its current form does not have a unique solution. To size the pump, first change the FCV at J8 to a PDCV.

This brings up the question: which FCV should you turn into a PDCV? When FCV’s are put in parallel, frequently the pipe design has one of the FCV’s further away from the pump than the others. Because of the additional piping leading to this FCV, it will be the weakest link in the chain of parallel FCV’s, by virtue of having the lowest pressure drop across it. The most remote FCV should be chosen as the PDCV. If the most remote FCV is chosen, when the minimum pressure drop required is applied to the PDCV all other FCV’s in the parallel system will have greater than the minimum pressure drop — thus satisfying the pressure drop requirement for those FCV’s as well.


8/10/2007 4:10:28 PM Page 469


What if you do not know which FCV is the most remote? In this case make your best guess, change it to a PDCV at the minimum required pressure drop, run the model, then verify whether all other FCV’s have a pressure drop that meets or exceeds the requirement. If not, then the FCV with the smallest pressure drop as determined by the first run is in reality the weakest FCV. Choose this FCV, change it to a PDCV, then change the original PDCV back to an FCV since it is not the weak link. If the pipe system leading to each FCV is truly identical, and all FCV’s have identical pressure drops, any of the FCV’s will serve as the PDCV.

The preceding process can be extended to cases where there are three or more FCV’s in parallel, and also cases with more than one pump in parallel supplying the FCV’s.

In Figure 12.17 the most remote valve is J8, the one on the right. If the minimum pressure drop is 0.05 MPa, change J8 to a PDCV with a 0.05 MPa drop. This results in the output shown in Figure 12.18a. Notice how the pressure drop across the other FCV at J5 (6.837 psid) is greater than 0.05 MPa because it is closer to the pump. Also notice how the flow through J8 is still 25 m3/hr even though it is not controlling flow. Once again, the reason for this is that the pump is controlling the flow to 100 m3/hr, and with the J5 FCV controlling to 50 m3/hr, the excess must go through J8. The pump is thus sized and its pump head rise is shown in Figure 12.18b (39.06 meters).

a

b

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 PRV is used to control pressure downstream of the valve, while a PSV controls pressure at the valve inlet. One thing that PRV’s and PSV’s have in common with FCV’s is that the pressure drop cannot be known ahead of time. The pressure drop depends on the balance of the pipe system.

Consider the simple system shown in Figure 12.19. With the pressure controlled downstream of the PRV, the pressure upstream is not known. Without knowledge of the PRV’s pressure drop, the pressure upstream of the PRV cannot be determined and thus no unique solution exists.

Outlet P = 0 MPa(g) El = 3 m

Inlet Q = 0.05 m3/s El = 3 m

PRV Pdown = 0.1 MPa(g)

Figure 12.19 If a PRV is downstream of a known flow the PRV pressure drop cannot be calculated. This model does not have a unique solution.

Figure 12.20 shows the possible cases with PRV’s and PSV’s, while Table 12.2 comments on the six cases. In summary, at least one known pressure is always needed on the side of a pressure control valve opposite of the controlled side.


8/10/2007 4:10:28 PM Page 471



Inlet Q = 0.05 m3/s El = 3 m PRV

Pdown = 0.1 MPa(g)






PSV Pup = 0.15 MPa(g)







Inlet Q = 0.05 m3/s El = 3 m



a

b

c

d

e

f

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.



Table 12.2 Summary of six possibilities for PRV and PSV configurations.

CASE Inlet Pressure

Inlet Flow

Outlet Pressure

Outlet Flow OK ? Comments

Fig. 12.20a - PRV X X No No reference pressure upstream of PRV, cannot calculate

Fig. 12.20b - PRV X X Yes 1 known flow, can calculate everything directly

Fig. 12.20c - PRV X X Yes 2 pressures, can iterate for flowFig. 12.20d - PSV X X Yes 1 known flow, can calculate

everything directlyFig. 12.20e - PSV X X No No reference pressure downstream

of PSV, cannot calculate

Fig. 12.20f - PSV X X Yes 2 pressures, can iterate for flow

If a pump is being sized in series with a PRV or PSV, modeling the pump as an assigned flow will not permit a unique solution if the pump is upstream of a PRV or downstream of a PSV. The reasons for this are the same as those in the previous section on FCV’s.

In combination with the previous examples, it should be apparent how the non-unique solution problem can occur in more complicated systems. In all such cases, the reasoning will boil down to the same problems already discussed.

Closing parts of a system

A model may be built which has sufficient information to obtain a unique solution. Then a user closes a junction and suddenly the problem has no unique solution. This is shown in Figure 12.21a-b. The top model has a unique solution and will run fine. In the bottom model the valve is turned off and the section of the system downstream of the valve has only known flows. No reference pressure exists in this section because it is isolated from the known pressures in the section on the left.


8/10/2007 4:10:28 PM Page 473


Open

P1v

P2> P3>

P4^

P5>P6>

J7 J6J3 J4

J5J1

J2

a

Closed

P1v

P2> P3>

P4^

P5>P6>

J7 J6J3 XJ4

J5J1

J2

b

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.



To satisfy the Method of Characteristics (MOC) requirement, the time step and number of pipe sections are related by Equation 5.1. If more sections are desired, then shorter time steps are required. For example, to use ten sections in a pipe rather than five requires twice as many computations because there are twice as many sections. However, the time step must be reduced by half (Equation 5.3); thus to model the pipe for the same time duration (say 1 second) requires twice as many time steps. Therefore, using double the number of sections results in four times the number of computations.

A related issue is that of the controlling pipe (see Chapter 5 for more information on controlling pipes). If the controlling pipe is very short, this requires more sections in all other pipes in the model. Hence it may be prudent to either neglect the shortest pipe or identify a way to combine it into a longer pipe, which will lead to a different controlling pipe. The Section Pipes window will automatically combine pipes of common diameter for you (see Figure 12.22).

For example, consider a case where the controlling pipe is 1 foot long and the next shortest pipe is 10 feet long. If the controlling pipe can be neglected or combined, then the 10 foot pipe becomes the controlling pipe. This will result in a factor of 10 fewer pipe sections and factor of 10 fewer time steps, or a run time reduction by a factor of 100!

Artificial transients

An artificial transient occurs when the steady-state results do not match the transient calculation initial step. During the first time step AFT Impulse tries to bring the mismatched conditions to what it considers steady conditions which generates a transient pressure. This transient is not a true physical transient, but a false “artificial” one based on mismatched data. If undetected, the artificial transient will interact with the physical transient being modeled and corrupt the results.


8/10/2007 4:10:28 PM Page 475


Figure 12.22 The Section Pipes window alerts you to pipes that could be combined and possibly reduce model runtime.

Before AFT Impulse initiates the transient simulation, it performs a “zero time step computation” and compares the predicted results with the results from the steady-state solution. If the difference exceeds the artificial transient criteria (0.5% by default), then AFT Impulse will display a warning in the output. If Stop Run if Artificial Transient Detected is selected, the simulation will be immediately terminated. This feature is especially helpful for simulations that take a long time to run. For example, it is wasteful to run a large model for an hour only to find at the end that an artificial transient input error was present at the very beginning, thus nullifying the results.

The artificial transient detection feature is an automatic way for the user to identify problems of this nature. A manual process can also be used, and it is instructive to consider the manual process to better understand artificial transients.

In some cases where pipes are initially not flowing, the first transient calculation is some very small but non-zero flowrate. This can appear to



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.


A P P E N D I X A

Keyboard Modifiers and Shortcuts

The keyboard modifiers and shortcuts for the Workspace are summarized here for convenient reference. These shortcuts are displayed when you click the Shortcuts button on the Toolbox, and also display at the bottom of the Toolbox in abbreviated form.

Selection Drawing Tool

Double-clicking tool If you double-click the Selection Drawing Tool, the Selection Drawing Tool remains active after you have completed the selection. It remains active until you click it again a single time

Control key If you hold down the CTRL key when completing the selection drawing (just before releasing the mouse button), the Selection Drawing Tool remains active. The CTRL key thus performs a similar function to double-clicking the tool.

Shift key 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.



Alt key 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.

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.

Appendix A Keyboard Modifiers and Shortcuts 479

8/10/2007 4:10:28 PM Page 479


Zoom Select Tool

Shift key If you hold down the SHIFT key down while selecting the zoom area, the zoom state will always be 100%.

Panning the Workspace

If the Workspace has been enlarged beyond the default one screen size (Workspace scrollbars will be visible), then the Workspace can be panned by holding down the CTRL key, pressing the mouse down on the Workspace, and then dragging in the direction you want to move the Workspace.

Dragging junctions from Toolbox

Control key A junction can be “morphed” from one type to another. To morph a junction, hold down the CTRL key while selecting a junction from the Toolbox. Drop the junction onto an existing Workspace junction. The junction type will change to the new junction type, and data that can be kept will be copied into the new junction.

Shift key If you hold down the SHIFT key while selecting a junction from the Toolbox and then drop the junction onto an existing pipe, the pipe will split into two pipes. The physical length of the original pipe will be automatically halved, and the new pipe will be assigned the balance. Thus the sum of the two pipe lengths will equal the length of the original pipe. Any fittings/losses in the original pipe will be left in the original pipe, and the new pipe will have no fittings/losses.



Inspecting pipes and junctions

Shift key If you hold down the SHIFT key when you perform an inspection, only the items not yet defined are displayed.

Control key If you hold down the CTRL key when you perform an inspection, the inspection window will show the output data for the pipe or junction.

Moving junctions with connected pipes

Control key By default, when you move an existing junction icon, any connected pipes will move along with the junction and retain the connection. If you hold down the CTRL key as you release the junction, only the junction will move. All pipe connections will be broken. This behavior can be reversed such that connections will not be retained unless you hold down the CTRL key. This reversal can be specified on the Workspace Preferences window.

Pipe and junction specifications window

F2 function key When in a pipe or junction specifications window you can toggle the highlight feature by pressing the F2 function key. You can also toggle this from the Options menu.

F5 function key When in a pipe specifications window, if you press the F5 function key, you will jump to the next highest numbered pipe. If you press the CTRL key while pressing the F5 function key, you will jump to the next lowest numbered pipe.

Appendix A Keyboard Modifiers and Shortcuts 481

8/10/2007 4:10:28 PM Page 481


The F5 function key also behaves this way for junction specifications windows.


A P P E N D I X 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



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.


A P P E N D I X C

Installation Issues

Customization files

As you customize AFT Impulse, two new files are created in your application data folder (e.g., C:\Documents and Settings\Your Account\Application Data\Applied Flow Technology\AFT Impulse\4.0):

IMPULSE4.INI

IMP_USER4.DAT

The IMPULSE4.INI file contains the information to customize the features in AFT Impulse. The IMP_USER4.DAT contains all custom database information. If you invest significant time in customizing AFT Impulse, especially in the area of custom databases, you are strongly advised to keep backup copies of these files.

Special file control features

What is the Control.aft File?

Control.aft is read as AFT Impulse begins to load and instructs AFT Impulse where certain key files are located. Typically there is little need to change the settings, however this file can be used to override the internal defaults and customize your working environment.



Control.aft is typically located in the application folder. However, on networks, it may be necessary to locate it in the Windows\System folder of each client. If AFT Impulse finds the file in the Windows\System folder it will use it and ignore any file in the application folder.

Control.aft can be edited using any text-based editor. Notepad, supplied as part of the Windows installation, is a good choice. Be careful to avoid changing any of the headings or key words. If these are changed AFT Impulse will not find the important data.

Specifying where Chempak is located

Before AFT Impulse can use the Chempak Fluid Database it needs to know where the data is located. This paths is specified in the control.aft file.

When Chempak is installed the setup procedure automatically updates any control.aft file specified with the path to the data.

You can manually edit the file by opening it in Notepad (or any text editor) and adding the following lines to the [SEARCH PATHS] section (replace C:\AFT Products\Chempak Data with the path information based on your system):

ChempakDataPath = C:\AFT Products\Chempak Data

This will instruct AFT Impulse to look in the C:\AFT Products\Chempak Data directory for the data.

Network installations

AFT Impulse 4.0 requires a number of auxiliary files to run properly. Because of how 32-bit Windows operating systems work (Windows 9x, NT/2000/XP), certain files must be registered in the Windows registry. For software installed on a client PC this is accomplished directly during setup. For software installed on a local or wide area network, however, the installation becomes two-step process.

The first step is to install AFT Impulse on the network using the “Network (auxiliary files on client)” in the setup program. This performs a typical network installation except the auxiliary files are not installed or registered.

Appendix C Installation Issues 487

8/10/2007 4:10:28 PM Page 487


The second step is for each client to run the Client Setup program installed into a subfolder below AFT Impulse named “Network Client Setup”. This process uses standard installation protocol whereby the auxiliary files are copied to the client's SYSTEM folder only if they do not exist, or are newer versions than those that already exist. The Client Setup will also install the sample and verification files on the client and will create shortcut links on the Start Menu referencing the applications and files on the network server. Finally, it will update the IMPULSE4.INI file with details of the setup. Although this option is not a perfect solution, we believe it is the best because the client PC registry points only at local files (as far as AFT Impulse is concerned).

Problems loading the Graphics Server

AFT Impulse uses a third party tool for graphing called the Graphics Server. There are four files that are used to load and run the Graphics Server:

• GRAPHS32.OCX

• GSW32.EXE

• GSWDLL32.DLL

• GSWAG32.DLL

Because of how the Graphics Server is written, multiple copies of GSW32.EXE or GSWDLL32.DLL in your Path will cause problems when your try to load AFT Impulse 4.0.

In practice, the most common problem is a network installation where all AFT Impulse files have been installed into the network folder along with AFT Impulse. Thus, GSW32.EXE and GSWDLL32.DLL will exist in the AFT Impulse network folder. A client PC which happens to have copies of either or both of these files on the local hard drive (usually in the WINDOWS\SYSTEM folder on Windows 95/98 and WINNT\SYSTEM32 on Windows NT) will fail to load properly. The most likely reason copies exist on the local hard drive is another application installed locally uses them.

A complete list of your options is given below, along with an explanation of the impact of each option. In the following explanations, the term conflicting files means the GSW32.EXE and GSWDLL32.DLL



files. Your WINDOWS\SYSTEM folder is just called your SYSTEM folder.

1. Delete the conflicting files on your hard drive – This will solve the problem for AFT Impulse 4.0, but may cause other software you use to stop working. This is not a permanent solution in that if you install other software in the future that installs copies of the conflicting files into your SYSTEM folder, the current problem with AFT Impulse will occur again.

2. Change the names of the conflicting files on your hard drive – This will solve the problem for AFT Impulse 4.0, but may cause other software you use to stop working. However, when you try to run the other software and it fails to load because of these missing files, you can change the names back to the original names. This will then cause AFT Impulse 4.0 to stop working until you rename them once again, but will allow you to run your other software that needs these files. This will also require you to remember which files you renamed. This is not a permanent solution in that if you install other software in the future that installs copies of the conflicting files into your SYSTEM folder, the current problem with AFT Impulse will occur again.

3. Move the conflicting files on your hard drive into the application folder that uses them – This will solve the problem for AFT Impulse 4.0 and still allow the other software you use to run properly. However, you will need to know which software uses the conflicting files so you can copy the files into their folder. If you remember what you do, you can move the files out of the SYSTEM folder (or wherever conflicting files exist) and into a temporary folder. When a software application that used to work fails because these files are missing, you can copy the files into its folder at that time. This is not a permanent solution in that if you install other software in the future that installs copies of the conflicting files into your SYSTEM folder, the current problem with AFT Impulse will occur again.

4. Allow AFT Impulse 4.0 to automatically rename the files for you – This is not a perfect solution, but it does have advantages. This option enables a feature whereby AFT Impulse searches for the conflicting files in both your SYSTEM folder and WINDOWS folder before the full AFT Impulse application is launched. If it finds them, it automatically renames them and thus only the copies in the

Appendix C Installation Issues 489

8/10/2007 4:10:28 PM Page 489


AFT Impulse network folder are found. When you exit AFT Impulse, the files will be renamed back to their original names. The advantages are, 1) you can continue to run your other software that uses these files and AFT Impulse with no conflicts (as long as they are not run at the same time) and, 2) even if you install other software in the future, this option will continue to work for you. The disadvantages are, 1) you are permitting AFT Impulse to change files on your local hard drive every time it is run, which understandably makes some users uncomfortable and, 2) you cannot run AFT Impulse 4.0 at the same time as the other software which needs these files.

5. Reinstall AFT Impulse with the Network option to run auxiliary files on the client (recommended) – This option is the best long-term option. To do this, first uninstall AFT Impulse from the network and reinstall it with the Network option. See Information on Software Installation for more help.


A P P E N D I X D

Chempak Technical Information

© Copyright 1998 Madison Technical Software Inc. Used with permission.

The theoretical basis for Chempak is a lengthy topic and has been included in the AFT Impulse 4.0 Help system. To learn more about Chempak, open the Help system and search for Chempak.


A P P E N D I X E

Obtaining Technical Support

In order to receive technical support, the Licensing Agreement return card must be signed by an authorized representative of your company and returned to Applied Flow Technology within 30 days of purchase. AFT Impulse is a sophisticated software application intended for trained engineers, and in order to provide effective support, Applied Flow Technology needs to know who is using the software.

AFT's three levels of support

1. Free Software Support for One Year

AFT provides free support for the first year to those who sign and return the Licensing Agreement. This support is for resolving installation problems and clarifying usage issues.

2. Annual Support, Upgrade and Maintenance (SUM)

After one year, AFT provides support to those sites that purchase an annual SUM. This support will also be for resolving questions on how to use specific features in the program.

3. Waterhammer Modeling Consulting Support

AFT offers specialized consulting support to users who have general waterhammer analysis questions. For questions that do not pertain strictly to AFT Impulse, but instead pertain to the general nature of how



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:

[email protected] 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


References

ASME Press, ASME International Steam Tables for Industrial Use (CRTD-Vol. 58), published by the American Society of Mechanical Engineers, Three Park Avenue, New York, NY 10016

Brown R.J. , and D.C. Rogers, “Development of Pump Characteristics from Field Tests”, Journal of Mechanical Design, ASME, vol. 102, pp. 807-817, Oct., 1980.

Carthew, G.A., Goehring, C.A., van Teylingen, J.E., "Development of dynamic head loss criteria for raw sludge pumping", Journal Water Pollution Control Federation (WPCF), Volume 55, Number 5, (May 1983).

Chaudhry, M. H., Applied Hydraulic Transients, 2nd edition, Van Nostrand Reinhold Company, New York, NY, 1987.

Crane Co., Flow of Fluids Through Valves, Fittings, and Pipe, Technical Paper No. 410, Crane Co., Joliet, IL, 1988.

Donsky, B., "Complete Pump Characteristics and the Effects of Specific Speeds on Hydraulic Transients", ASME Journal of Basic Engineering, pp. 685-699, Dec. 1961.

Fox, R. W., and A.T. McDonald, Introduction to Fluid Mechanics, 3rd edition, Wiley & Sons, New York, NY, 1985.

Idelchik, I. E., Handbook of Hydraulic Resistance, 3rd edition, CRC Press, Boca Raton, FL, 1994.

Ingersoll-Dresser Pumps, Cameron Hydraulic Data, Liberty Corner, NJ, 1995.

Jeppson, R. W., Analysis of Flow in Pipe Networks, Ann Arbor Science Publishers, Ann Arbor, MI, 1976.



Karassik, I.J., J.P. Messina, P. Cooper and C.C. Heald, Pump Handbook, 3rd Edition, McGraw Hill, New York, NY, 2001.

Karney, B.W., and D. McInnis, "Efficient Calculation of Transient Flow in Simple Pipe Networks", ASCE Journal of Hydraulic Engineering, Vol. 118, No. 7, July, 1992.

Kittredge, C.P., “Hydraulic Transients in Centrifugal Pump Systems”, ASME Journal of Basic Engineering, Vol. 78, No. 6, pp. 1307-1322, Aug. 1956.

Lyons, J. L., Lyons' Valve Designers' Handbook, Van Nostrand Reinhold, New York, NY, 1982.

Miller, D. S., Internal Flow Systems, 2nd edition, Gulf Publishing Company, Houston, TX, 1990.

Pipe Line Rules of Thumb Handbook, Gulf Publishing, Houston, TX.

Swaffield, J. A., and Boldy, A. P., Pressure Surge in Pipe and Duct Systems, Avebury Technical, Hampshire, England, 1993.

TAPPI, Generalized method for determining the pipe friction loss of flowing pulp suspensions, TAPPI, TIS 0410-14, Atlanta, GA, 1988.

Tilton, J. N., Perry's Chemical Engineer's Handbook, Seventh Edition, Chapter 6, McGraw-Hill, New York, NY, 1997.

Thorley, A.R.D., and A. Chaudry, “Pump Characteristics for Transient Flow”, Pressure Surges and Fluid Transients, BHR Group, 1996.

Walski, T. M., Analysis of Water Distribution Systems, Van Nostrand Reinhold Company, New York, NY, 1984.

Walters, T.W., "Analytical Approaches to Modeling Transient Vaporous Cavitation in Multi-Pipe Fluid Systems", ASME FED-Vol. 116, New York, NY, 1991.

Watters, G.Z, Modern Analysis and Control of Unsteady Flow in Pipelines, Ann Arbor Science, Ann Arbor, MI, 1979.

White, F. M., Fluid Mechanics, McGraw Hill, New York, NY, 1979.

Wylie, E.B., V.L. Streeter & L. Suo, Fluid Transients in Systems, Prentice Hall, Englewood Hills, New Jersey, 1993.


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.



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.

Glossary 499

8/10/2007 4:10:28 PM Page 499


fluid transient Same as waterhammer.

Graph Results window The graphical window that allows preparation of plots.

head In AFT Impulse, synonymous with piezometric head and Hydraulic Grade Line.

head loss The irrecoverable loss in head that occurs in a pipe system. Head loss is equal to the frictional pressure loss converted into units of length.

HGL Abbreviation for Hydraulic Grade Line.

history The information on how the solution scheme progresses.

Hydraulic Grade Line The sum of the pressure head and gravity head. This quantity is relative to the elevational coordinate system specified by the engineer.

hydraulic transient Same as waterhammer.

ID number The unique number associated with a pipe or a junction. The ID number is always greater than zero and up to 30,000.

incompressible A fluid whose density changes are small enough that a constant density can be safely assumed.

infinite pipe A junction which does not reflect and pressure or flow waves. Available in Assigned Flow and Assigned Pressure junctions.

inspection A feature in AFT Impulse that allows you to obtain read-only information about objects.

instantaneous waterhammer A waterhammer event that occurs so rapidly that it is effectively instantaneous, thus yielding the highest possible pressure surge.

iteration A single step of the Steady-State Solver, during which a new solution is obtained based on the previous solution state.

Jump A feature that allows lateral changes to be made between Specifications windows.

junction In AFT Impulse, a pressure/head solution point that connects and balances flow from pipes.

lock To fix an object's position on the Workspace in order to prevent accidental movement.



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.

Glossary 501

8/10/2007 4:10:28 PM Page 501


PRV See Pressure Reducing Valve.

Pump configuration A particular pump which contains data for rotational speed and impeller size.

reference positive flow direction The flow direction you assign to a pipe, which by convention is positive. Flow in the opposite direction is considered negative.

relaxation A parameter that affects how the iterative solution scheme approaches a pipe flow solution.

relief valve A type of valve that is typically closed but opens to release fluid when a set pressure occurs. Used as a protective device.

resistance A convenient term relating the head loss to flow rate (R = H/Q2).

Scenario Manager Window that allows one to create and manage multiple pipe flow model scenarios. These include different equipment, sizes, and operating conditions.

Solver Either Steady-State or Transient. The Steady-State Solver is the part of AFT Impulse that contains the pipe steady flow network solution method. The Transient Solver contains the transient waterhammer solution algorithms.

Specifications window The window in which an object's properties are entered.

stagnation pressure Sum of static and dynamic pressures.

station A location in a pipe where a transient computation occurs. Form computational reasons pipes are broken into sections. Because computation is required at the pipe inlet and outlet, there will always be one more station than there are sections.

steady-state The condition of a system in which the rate of change of all parameters is negligibly small.

surge flow Same as waterhammer.

tolerance The value at which the Steady-State Solver should consider that convergence has been obtained.

Toolbox The area on the Workspace window that offers tools for building pipe flow models.



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.


Index

A Absolute roughness 62, 200, 201, 310,

337

Absolute tolerance See Steady Solution Control window

Affinity laws 246

for pumps 365, 391, 395

AFT DEFAULT INTERNAL DATABASE 313

AFT Fathom

Converting models from 10

AFT Impulse

engineering assumptions 4

Example models 7

Getting started 6

installation 5

New features in version 3.0 11

Overview 7

qualifications for use 4

summary of capabilities 2

using online help 7

AFT IMPULSE LOCAL USER DATABASE 313

AFT Impulse window 14, 71

AFT Mercury

Converting models from 10

AFT Standard fluid 27, 139

Change Fluid Data window 308

Database 4, 145, 366

Database editing 150

Aligning objects 81

Analysis type

steady-state vs. transient 189

Animate Using Output File See Animation

Animate Using Solver See Animation

Animation 53, 109, 160

Annotation Manager 79, 129

Annotation tool 15, 78

Annotations 129

Area Change

as Fittings & Loss 204

Area Change junction

abrupt transition 221, 355

conical transition 221, 355

loss factors 222, 355

Specifications window 221

Artificial transients 44, 159, 474

manual detection 476

Stop run when detected 159

ASME ’97 Water See ASME Steam Tables

ASME Steam Tables 28, 139, 145, 147, 366

Assigned Flow junction



contrast with pressure-type junction 59

example 459

loss factors 62

Special features for steady flow 224


Transient Data 224

waterhammer theory 381

Assigned Pressure junction

How pressure junctions work 455

loss factors 62

periodic data 227


the role of pressure junctions 458

Transient Data 227

use in open and closed systems 332

using static and stagnation pressure 332


Atmospheric pressure 149

Autosave 302

Auxiliary Graph Formatting window 304

B Base area 61, 221, 230, 242, 274, 360

Specifying 220

Batch runs 187

Bends

loss factors 356

Bernoulli equation 327

Bingham Plastic See non-Newtonian

Black and white junction icons 186

Branch junction 16, 273

flow source or sink 228

loss factors 62

Specifications window 32, 227

Transient Data 229


Brecht & Heller method See non-Newtonian

C CAESAR II Force Files 113, 165

Calculator 189

Cavitation 144, 308

transient 159

Check Valve

theory 410

Check Valve junction

Delta Pressure to Re-Open 231

Forward Velocity to Close Valve 231

inherent event logic 434

loss factors 230

Special Condition check valve 230


Transient Data 231

Index 505

8/10/2007 4:10:28 PM Page 505


Checklist 24, 136

Chempak 27, 139, 146, 366

accuracy options 366

mixtures 28

specifying Chempak file locations 486

chopped sine wave 224, 227

Clipboard 115

Colebrook-White 337, 338

Color Map See Visual Report window

column separation See vapor cavitation

communication time 42, 65, 66, 370, 387, 453

compatibility equations 376, 380, 381, 382, 383, 385, 387, 391

Component Database 311

computing stations 154, 424

graphing 52

in output file 95

max/min results in output 99

numbering convention 154

results in output window 101

Connectivity 130, 132, 134

graphical connection 131

in Junction Specification window 214

in Pipe Specifications window 198

model connection 132

Constant Pressure Drop Valve See Pressure Drop Control Valve

Continuity equation

steady-state 326

transient 371

Control Valve junction

action if setpoint not achievable 233

Control Transient 236

failure to control issues 452

Failure Transient 235

Open Percentage Table 235

sizing pumps with flow control valves 464

Special Conditions 234


static and stagnation pressure 332

Transient Data 235

types 232


Control.aft file 485

controlling pipe 42, 44, 153, 378, 474

Conventions See also Base area

flow entering and exiting 60

Convergence 351

Copy 80, 126, 128, 130

Create Mixture window 146

Crude oil 338

CTRL key See Keyboard modifiers, See Keyboard modifiers

Custom databases See Databases

Customize Graph window 112



Customizing AFT Impulse

databases 305

files created 485

fluids 306

graph styles 304

junction (component) databases 311

Model Data window 90

Output window 95, 303

Parameter Preferences 287

pipe fittings & losses 305

pipe materials 309

Toolbox Preferences 300

Unit Preferences 289

Visual Report window 304

Workspace window 83, 292

Customizing AFT Impulse See Visual Report Control window

Customizing AFT Impulse See Output Control window

Customizing AFT Impulse See Workspace Preferences window

Customizing AFT Impulse See Toolbox Preferences window

Customizing AFT Impulse See Parameters and Unit Preferences window

Customizing AFT Impulse See Model Data Control window

Customizing AFT Impulse See Databases

Cut 80, 130

cyclic dual events 432, See Event transients

D Darcy-Weisbach 201, 326, 338

Database Manager 171, 306, 312

Benefits of shared databases 316

Connecting and disconnecting databases 313

creating network databases 319

DATABASE.LIB 313, 318, 321

Editing databases 314

Enterprise network databases 318

IMP_DBUSER.LIB 319

Making databases available 312

sharing databases using DATABASE.LIB 321

DATABASE.LIB 313, 318, 321, 323

Databases 29, 305

Benefits of shared databases 316

Component database 214, 244, 311

Connecting to external shared databases 323

Creating an enterprise-wide network database system 318

custom junctions 311

fluid See AFT Standard fluid or Chempak

fluid database 145, 150, 306

fluid properties 150, 305

Index 507

8/10/2007 4:10:28 PM Page 507


Friction Data Set 309, 310

IMP_USER4.DAT 287, 305, 315, 320, 485

pipe materials 199, 309

special control files 317

Specifying common locations 302

Dead End junction



Defining objects

Checklist - Define All Pipes and Junctions 28, 139

List Undefined Objects window 29, 133

object status 80

Delete 80, 130

Density 144

Design Alerts See Pipe Specifications window

cross-plotting against results 108

Design factors

defaults 288

junctions 219, 355

pipes 206, 344

Discrete Vapor Cavity Model 376, 424

Distributed loss 62, 353

Drag-and-drop 3

Duffy method See non-Newtonian

Duplicate 80, 126, 128, 130

Dynamic Pressure 331

Dynamic viscosity 144

E Editing

on Workspace 80

EGL 108

Elbow

See Bend 356

Elevation 108, 212

default 289

Event messages 104, 436

Event transients 430

Example models 7

Exit valve 386, See Valve junction

Explicit friction factor 62, 202

Extended Model Check 189

F Fanning friction factor 336

FCV See Flow Control Valve

Find utility 177

Fittings & Losses 63, See Losses, See Losses

window 203

Flip workspace See Workspace window

Flow Control Valve 232, 452, 462, 463, 466



Flow rate relaxation See Steady Solution Control window

Fluid database See AFT Standard fluid or Chempak

Fluid Database window 306, 308

fluid transients

See waterhammer

fluidhammer

See waterhammer

Forces

Defining force sets 161

displaying transient force data in Graph Results 112

exporting transient force data from Graph Results 165

including friction in force balance 112

including momentum in force balance 112

Fouling 201

Friction Data Set 200, 309, 310

Friction factor 150, 201

changes during transient flow 159

Friction models 62, 335

Absolute roughness 337

Frictionless pipe 340

Hazen-Williams 338

Hydraulically smooth 338

Miller Turbulent method 339

MIT Equation 338

non-Newtonian 340

Relative roughness 337

Resistance 338

Frictional Losses See Losses

Frictionless pipe 62, 202, 340

G Gas Accumulator junction

connector pipe 240

flow restrictor 239

Graphing transient data 241

in output file 161

Initial Gas Volume 237

lumped inertia 422

Maximum and minimum volumes 239

orifice 239

Polytropic Constant 237



surge reduction 443


Gaussian Elimination 351

With pivoting 351

General Component junction

loss factors 242


General Preferences 302

Global Pipe Edit window 278, 279

Index 509

8/10/2007 4:10:28 PM Page 509


Glossary 497

Graph Results window 8, 52, 72, 106, 304

Animation 53, 109

Import Graph Data 106

Save Graph Data 106

Select Data 52

Select Graph Data 106

Graphics Server

problems loading 487

Gravitational acceleration 5, 149

Grid

On Workspace 23, 118, 296

Group Manager 82

Grouping objects 80, 129

Groups 81, 82

used in graphing 108

H Hazen-Williams 202, 338

H-Equation method 325

HGL 108, 326

Hide Empty Columns


Highlight feature 30, 133, 194

Homologous pumps See Affinity Laws

Hydraulic Grade Line See HGL

hydraulic transients

See waterhammer

Hydraulically smooth 62, 202

Hydrocarbons - roughness 339

I Icon size 294

ID numbers 20, 28, 126, 128, 151, 197, 297

ID Reduction - Scaling 200

IMP_DBUSER.LIB 319, 322, 323

IMP_USER4.DAT 287, 303, 305, 315, 320, 485

impedance 375

Impulse Examples.hlp 7

IMPULSE4.INI 287, 303, 485

Incrementing Numbers 86

Infinite Pipe

example 453


theory 387

Initial guesses

flow guess 205

generation 218

pressure guess 217

using for troubleshooting models 451

Inspection feature 31, 134, 210, 212

Showing only undefined items 480

Installation 485

Network installations 486



Intermediate Pipe Elevations 207

Irrecoverable losses See Losses

Iteration history 351, 451

J Jacobian matrix 328

Jump 132, 193, 214

Junction icons

printing black and white 186

Junction label position 297

Junction Results table 93

Junction Specifications window 208

Area Change 221

Assigned Flow 222

Assigned Pressure 225

Branch 227

Check Valve 229

connected pipes 214

Control Valve 231

Copy Data From Junction 214

Database list 214

Dead End 236

design factor 219

displaying name 219

Elevation 212

format 210

Gas Accumulator 237

General Component 241

Initial pressure guess 217

Jumping to next higher or lower number junction 133

Liquid Accumulator 243

Name 212

Notes 220

Number 212

Pump 244

Relief Valve 259

Reservoir 261


Spray Discharge 263

Status 220

Surge Tank 265

Tee/Wye junction 271

Vacuum Breaker Valve 274

Valve 273

Volume Balance 276

Junctions 59

Base area 220

comparison of types 208

connectivity 130

creating 128

customizing icons 300

defining 132

displaying data in Output window 99

displaying in Model Data 90

displaying in Visual Report 118

Index 511

8/10/2007 4:10:28 PM Page 511


displaying on Workspace 83

displaying transient data in Graph Results 108

editing 80, 130

finding 177

icons on Toolbox 79

ID numbers 128

keyboard shortcuts 479

loss factor 62

negative flow rates 210

placing on Workspace 128

pressure losses 353

specifying data 132

K Keyboard modifiers 477

CTRL 73, 75, 125, 130

SHIFT 73, 75, 78, 80, 125, 129

L Laminar flow 150, 337

Legend 117, 121

Liquid Accumulator junction

Elasticity 244

Initial Volume 244



Lock 80, 129

Loss factor 60, 62, 203, 210, 211, 221, 227, 230, 234, 242, 261, 272, 354

loss models 34, 35, 354

Losses 60, 62, 327, 336, 353, See Fittings & Losses

LU Decomposition 351

Lumped inertia pipe with orifice

theory 422

M Matrix Method 351

Merging models 184

with multiple scenarios 184

Method of Characteristics 42, 370, 371, 474

reliability with cavitation 425

Miller Turbulent Method 202, 339

minor losses See Losses

MIT Equation 202, 338

Mixtures 4, 28, 146, 366

Model Data Control window 90

database connections 92

Scenario display 91

showing junction data 90

showing pipe data 90

showing selected pipes and junctions 91

using saved format files in Database Manager 317

Model Data window 8, 40, 71, 87, 88



customizing 90

opening pipe and junction specifications windows from 132, 193

Pipe Detail Summary 89

Pipe Fittings & Losses 88

Pipe table 88

print content 188

scenario display 91

transient junction data 438


Viewing scenario differences 169

Model Status Light 136

modulus of elasticity 201

Momentum equation

steady-state 326

transient 371

Moody friction factor 336

morph 128, 479

N Network databases 214, See Database

Manager

Newton-Raphson method 3, 326, 328, 335

transient solution for gas accumulators 416

non-Newtonian 148, 335, 340

Bingham Plastic 148, 343

Brecht & Heller method 341

Duffy method 340

junction losses 344

Pipe Fittings & Losses 344

Power Law 148, 342

pulp and paper 340, 341

NPSH 144, 248, 249

constant for variable speed 248

NPSHR 249

NPSP 144

NPSPR 249

NPSP See NPSH

O Object status 80, 134, 151, 208, 220

One-dimensional flow 59

Open Pipe/Jct Window 87, 132, 193

Orifice

loss factors 358

round edged 358

sharp edged 358

Output Control window 25, 51, 95, 138, 140, 303, 335


Format and Action 98

General Output Control 97

Pipe and Junction parameters 95

Pipe transient parameters 97

scenarios 171

Index 513

8/10/2007 4:10:28 PM Page 513


Show Selected Pipes/Junctions 99


Output file 47, 160

transient data 94

Output window 8, 48, 71, 93

customizing 303

Event messages 436

number formatting 98

print content 188

setting defaults 100

updates 104


Overview of AFT Impulse 7

P Parallel Pipes

in Pipe Specifications window 206

Parameter and Unit Preferences window 287


default roughness model 202

Design Factors 288

elevation 289


setting preferred units 290

unit preferences 289

Use Most Recent Pipe Size Data 288


Paste 80, 126, 128, 130

PDCV See Pressure Drop Control Valve

Piezometric head 325

Pipe Detail Summary


Pipe Drawing tool 14, 20, 74, 125, 478

Pipe Fittings & Losses

customizing 305

design factors 206

in Solver 353


Non-Newtonian flow 344

specifying 62

Pipe flow direction 222

Pipe Loss Data table 88

Pipe Material Database window 199, 309

Pipe Results table 93

pipe sectioning 41, 377, 378, See Section Pipes

effect on run time 473

Pipe Specifications window 37, 39, 194

connected junctions 198

Copy Data From PIpe 198

Copy Previous 198

Design Alerts 205



displaying name 205

Fittings & Losses 203

Fluid Properties 208

friction model 201

ID Reduction - Scaling 200

initial flow 205

jumping to next higher or lower number pipe 133

material 199

Notes 208

optional input 205

Parallel Pipes 206

pipe diameter 200

Pipe Length 199

Pipe Line Size and Color 206

pipe modulus of elasticity 201

Pipe Name 197

Pipe Number 197

pipe Poisson ratio 201

pipe support 203

pipe wall thickness 200

roughness model 201

Status 208

wavespeed 203

pipe stations See computing stations

Pipe Support 203

Pipe table


Pipes 20, 59

connectivity 130

creating 125

defining 132

Design factors 206

displaying data in Output window 99

displaying in Model Data 90

displaying in Visual Report 118

displaying on Workspace 83

displaying transient data in Output window 101

editing 129

finding 177

flow direction 60, 126, 128, 178, 198, 210

Global Editing 278

handles 76, 129

ID numbers 126

length on Workspace 126, 199

loss factor 62

losses 60, 61, 62, 336, 338, 353

number of connections 126

Reverse direction 178

segmenting 76

specifying data 132

Splitting 128, 479

stretching 129

Point loss 62

Index 515

8/10/2007 4:10:28 PM Page 515


Poisson Ratio 201

Popup menu 81

positive flow direction 21

Power Law See non-Newtonian

Pre-mixtures 146

Pressure

defining system pressure 166, 458

Pressure Drop Control Valve 233, 466

Pressure Reducing Valve 232, 452, 470

static and stagnation pressure discussion 332

Pressure relaxation See Solution Control window

Pressure Sustaining Valve 233, 452, 470, 471

static and stagnation pressure discussion 332

Primary windows 7, 71

Print Content 90, 93, 188

Print Preview 185

Print Special 185

PRV See Pressure Reducing Valve

PSV See Pressure Sustaining Valve

pulp and paper stock See non-Newtonian

Pump junction 464

as assigned flow 255

check valve 246

controlled discharge pressure 248, 365

controlled flow 248, 365

Controller transient 255

efficiency and power usage 249

Estimating the pump inertia 254

Graphing pump data 258

in output file 161

Known speed transients 250

NPSHR 249

problems from poor curve fits 447

pump configurations 246

Pump Startup With Inertia 254

Pump trip with inertia – four quadrant 253

Pump trip with inertia and no backflow 252

sizing with flow control valves 464

Special Conditions 257, 258


Submerged pump 246

Transient Data 249

variable speed 245, 248, 365

viscosity corrections 258


pump speed 391

Pumps 144, 365

efficiency 248, 249

four quadrant 253, 254, 399, 404



power usage 248, 249

Pump configurations 246

sizing method 455

sizing with flow control valves 464

submerged 394

viscosity corrections 258

R README.TXT 5

Reference Information 97, 188

Reference positive flow direction

See positive flow direction

Reference Pressure 459

References 495

Relative location 16

Relative roughness 62, 200, 202, 337

Relative tolerance See Steady Solution Control window

Relaxation See Steady Solution Control window


Relief Valve junction

entering transient data 260

Exit 407


Internal 406

Internal vs. exit relief valves 259

loss factors 62, 230


surge reduction 442

Transient valve cracking 260


Renumber Increment 86

Renumber Wizard 86

Renumbering See Workspace window, See Workspace window

Repeat transient 436

Reservoir junction 15, 60

Difference between a Reservoir and Surge Tank 270

How pressure junctions work 455

loss factors 62

Specifications window 29, 34, 261

the role of pressure junctions 458

Transient Data 263

use in open and closed systems 332


resistance 375, 379

variable during transient 63

Resistance

Hydraulic resistance friction model in pipe 62, 202, 338

Resistance Curve 242

Reverse Direction 178

Reynolds number

effects during transient flow 159

transition 150, 337

Index 517

8/10/2007 4:10:28 PM Page 517


S Scale/flip workspace See Workspace

window

Scaling See ID Reduction - Scaling

Scenario Manager 166

re-establishing broken links 176

Scenario logic 172


Scenarios

merging models with multiple scenarios 184

Model Data window ancestor display 91

same as parent junction 214

same as parent pipe 198

Section Pipes window 41, 139, 152, 157, 190, 379, 473, 474

Select Flow Path 80, 81, 129

Select Graph Data window 107

Select Special 80, 178

Selection Drawing tool 14, 73, 80, 129, 477

selecting objects completely or partially in box 294

sequential dual events 433, See Event Transients

Setup program 6

SHIFT key See Keyboard modifiers, See Keyboard modifiers, See Keyboard modifiers

Show list box

in Output window 93

Show Object Status 151

single event transients 430

Solution Balance Summary table 335

Solution Progress window 46, 93, 351

iteration history 351

View Output button 93

Solver

discussion of Steady-State and Transient 46

output file for transient data 47, 94

Recommendations on Steady Solver 190

Steady-State 138, 189, 325, 344

Transient 190, 370

Transient Solver control 157

verification of steady-state 335

working with the steady and transient solvers 189

Sort output feature 93


Sparger 264

Special Conditions 180, 452

Ignore 184

junctions 219

No reflections - infinite pipe 183

Transient 183

Workspace appearance 293



Specifications windows 87, 193

accessing 193

inspecting in 135

Jump feature 194

Splitting pipes 128, 479

Spray Discharge junction

Graphing Spray Discharge data 265

in output file 161


Transient Data 265


Stagnation pressure

at pressure control valves 332

explanation 329

Static pressure

at pressure control valves 332

explanation 329

Status Bar 136

Steady Solution Control window 138, 139

absolute tolerance 346

convergence 345

false convergence 350

flow rate relaxation 349

Matrix Method 351

maximum iterations 351

parameters 344

pressure relaxation 350

relative tolerance 346

relaxation 348

relaxation example 349

tolerance 345

tolerance example 347

Steam Tables See ASME Steam Tables

Submerged pumps 394

Support 493

Annual Support 493

e-mail 494

surge flow

See waterhammer

surge suppression 441

Surge Tank junction 17

as a branch during steady-state 269

as a finite size reservoir during steady-state 270

as a pressurizer during steady-state 270

connector pipe 267

Difference between a Reservoir and Surge Tank 270

flow restrictor 267

Graphing Surge Tank data 271

in output file 161

lumped inertia 422

Model as dipping tube vessel 270

Only Allow Outflow 268

orifice 267

Index 519

8/10/2007 4:10:28 PM Page 519



Specifications window 33, 265

surge reduction 443

Tank area 266

Tank geometry 266

Tank height 266

Transient Data 268


surge transients

See waterhammer

System pressure 166, 458

System Properties window 26, 138, 142

AFT Standard fluid database 145

ASME Steam Tables 145

atmospheric pressure 149

Chempak database 146

creating mixtures 146

fluid list 307

gravitational acceleration 149

scenarios 171

Transition Reynolds number 150

T Technical Support See Support

Tee/Wye junction 211

convergence problems and recommendations 450

loss factors 62, 272, 359

Simple, no loss 272



Terminology 1

time step 378, 474, 475

Time-based transients 429

Tip of the Day 302

Title 97, 117

Tolerance

balance verification 335


Toolbars 122

Toolbox 72, 79, 125, 128

customizing See Toolbox Preferences window

shortcuts See Workspace

Toolbox Preferences window 300

Current Toolbox/Unused Tools 300




ToolTips 79

Transfer Results to Initial Guess 93, 186

Transient Control window 44, 139, 157, 190, 241, 255, 258, 265, 271, 276, 429, 473

artificial transients 474



Defining force sets 161

Estimated file size and run time 162

Include in transient output file 160

Model transient cavitation 159

Save Output to File Every Time Step 158

Saving junction output 161

Saving pipe output 160

scenarios 171

specifying content for output file 95

Start and stop time 157

Stop run if artificial transient detected 159

variable resistance 64, 158

Transient Data

Absolute vs. relative 435

Repeat transient 436

Transient output file See Output file

Transient special conditions See Special Conditions

Transient vapor cavitation See vapor cavitation

Transition flow 150, 337

Transition Reynolds number 150

Transport properties 366

Troubleshooting steady-state models 446

Turbulent flow 150, 337

U Undefined Objects window 29, 133,

151

Undo 80, 130

Unit Preferences See Parameter and Unit Preference window

Unspecified fluid 27

V Vacuum Breaker Valve junction

Graphing Vacuum Breaker Valve data 276

in output file 161



surge reduction 443

Valve coefficient 363

Valve junction 19, 60

Exit valve 273

loss factors 62, 356, 358


Specifications window 34, 35, 273

Transient Data 274


vapor cavitation 44, 156

transient 376, 424

vapor cavitation waterhammer theory

assigned flow 381

assigned pressure 380

Index 521

8/10/2007 4:10:28 PM Page 521


branch 384

check valve 410

dead end 382

gas accumulator 416

infinite pipe 388

liquid accumulator 406

pump 394

spray discharge 390

surge tank 413

valve 386

Vapor pressure 144

Variable fluid properties 143, 208

variable pipe resistance 63, 158

Verification

Verification against published results 7

Verifying steady-state network solutions 335

View Output button See Solution Progress window

Viscosity corrections 258

Visual Report Control window 116, 304

scenarios 171


Visual Report window 8, 55, 71, 114, 304

Color Map 119

Display Parameters 117

General Display 117

input-only mode 119

legend 117, 121

Save Options 121

Show Selected Pipes and Junctions 118

Visual Report Control 116

Volume Balance junction


W Walk Through Example 13

wall thickness 200

Water Data from ASME Steam Tables 28

Water properties from the ASME Steam Tables See ASME Steam Tables

waterhammer 1, 64

description 64

instantaneous 65, 367

Waterhammer Assistant 176

wavespeed 38, 42, 64, 66, 152, 153, 203, 368, 370, 377, 378, 424

assumption that it remains constant 5

Calculated 199

User Specified 200

Web site 494

Workspace See Workspace window

grid 296



Keyboard modifiers 477

Reverse Direction 21

shortcuts 301, 477

symbols 296

Workspace Preferences window 129, 151, 180, 183, 219, 292

Action When Dragging Junctions 294

Allowable Workspace Label Movements 297

Background Picture Scaling 298

colors and fonts 298


display options 295

displaying name and/or ID numbers 297

icon size 294

Pipe endpoint adjustments 294

pipe line options 292

Popup menu 297

Sample Workspace 298


special conditions 293

Special Conditions Graphics 298

symbols 296

Transient indicators on the Workspace 437


Workspace window 8, 14, 71, 72

Background Picture Scaling 298

changing junction icons 220

colors and fonts 298

customizing 83, 292

Displaying name and/or ID numbers 297

editing 80

finding pipes and junctions 177

flipping 84

pipe lengths on 199

Popup menu 297

relative locations 16

Renumber Wizard 86

renumbering 85

scaling 84

setting the junction label position See Workspace Preferences window

size 83

Special Conditions Graphics 298

zoom feature 83

Z Zoom feature 83, 115

Zoom Select tool 15, 77, 479

Zoom to Fit 115