Professional Documents
Culture Documents
TM
Users Guide
2/24/2003
1b.doc
1:35 PM
Page 1
CAUTION!
AFT Impulse is a sophisticated waterhammer and surge transient
modeling program designed for qualified engineers with experience in
waterhammer analysis and should not be used by untrained individuals.
AFT Impulse is intended solely as an aide for pipe flow analysis
engineers and not as a replacement for other design and analysis
methods, including hand calculations and sound engineering judgment.
All data generated by AFT Impulse should be independently verified
with other engineering methods.
AFT Impulse is designed to be used only by persons who possess a level
of knowledge consistent with that obtained in an undergraduate
engineering course in the analysis of pipe system fluid mechanics and is
familiar with standard industry practice in waterhammer analysis.
AFT Impulse is intended to be used only within the boundaries of its
engineering assumptions. The user should consult the manual for a
discussion of all engineering assumptions made by AFT Impulse.
Information in this document is subject to change without notice. No part of this Users
Guide may be reproduced or transmitted in any form or by any means, electronic or
mechanical, for any purpose, without the express written permission of Applied Flow
Technology.
"AFT Impulse", "AFT Fathom", "AFT Mercury", "Applied Flow Technology", and the
AFT logo are trademarks of Applied Flow Technology Corporation
Chempak is a trademark of Madison Technical Software, Inc.
Microsoft, "Visual Basic", Excel and Windows are trademarks or registered trademarks
of Microsoft Corporation.
2/24/2003
1b.doc
1:35 PM
Page 2
Contents
Summary
Summary..................................................................................................iii
Detailed Contents ..................................................................................... v
Nomenclature....................................................................................... xxix
1. Introduction .................................................................... 1
2. A Walk Through AFT Impulse...................................... 15
3. Fundamental Concepts................................................ 59
4. The Five Primary Windows.......................................... 71
5. Building and Running Models................................... 121
6. Pipe and Junction Specifications Windows ............ 179
7. Customizing AFT Impulse ......................................... 271
8. Steady-State Hydraulic Theory and Loss Models.... 311
9. Theory of Waterhammer & Solution Methodology .. 355
10. Modeling Time and Event Based Transients ......... 413
2/24/2003
1b.doc
1:35 PM
Page iii
2/24/2003
1b.doc
1:35 PM
Page iv
Table of Contents v
Detailed Contents
Summary..................................................................................................iii
Detailed Contents ..................................................................................... v
Nomenclature....................................................................................... xxix
1. Introduction .................................................................... 1
Terminology ............................................................................................. 1
Modeling capabilities ............................................................................... 2
Interface features ...................................................................................... 3
Thermophysical property data .................................................................. 4
Who can use AFT Impulse ....................................................................... 4
Engineering assumptions in AFT Impulse ............................................... 4
Installing AFT Impulse............................................................................. 5
Check hardware and system requirements......................................... 5
Read the README.TXT file............................................................. 5
Run the setup program ....................................................................... 6
Getting started with AFT Impulse ............................................................ 6
Example models ................................................................................. 7
Using online help...................................................................................... 7
Verification models .................................................................................. 7
AFT Impulse overview ............................................................................. 7
Input windows.................................................................................... 8
Output windows ................................................................................. 8
Converting models and databases from AFT Impulse 1.0 and 2.0........... 9
Pipe friction........................................................................................ 9
Infinite pipe junctions ........................................................................ 9
Transient junction data....................................................................... 9
Junction loss models .......................................................................... 9
Reservoir transients............................................................................ 9
Databases ......................................................................................... 10
2/24/2003
1:35 PM
Page v
2/24/2003
1b.doc
1:35 PM
Page vi
3. Fundamental Concepts................................................ 59
Pipes and junctions ................................................................................. 59
Convention for flow entering and exiting............................................... 60
Features for modeling irrecoverable losses ............................................ 60
Incorporating equivalent length data ......................................... 61
Convention for specifying junction base area.................................. 61
Specifying losses.............................................................................. 62
Modeling valve closures with K factors .......................................... 62
2/24/2003
1:35 PM
Page vii
2/24/2003
1b.doc
1:35 PM
Page viii
Table of Contents ix
Customizing features........................................................................ 82
Workspace Preferences ............................................................. 82
Adjusting the Workspace size ................................................... 83
Selective display of pipes and junctions.................................... 83
Scale/flip workspace.................................................................. 84
Renumbering the Workspace objects ........................................ 84
Using the Renumber Wizard ..................................................... 85
Background Graphic ........................................................................ 85
Specifications windows.................................................................... 86
The Model Data window ........................................................................ 86
Display and printing features ........................................................... 88
The Model Data Control window .................................................... 88
Pipe data display........................................................................ 88
Junction data display ................................................................. 88
Show selected pipes and junctions ............................................ 89
Database connections ................................................................ 89
The Output window ................................................................................ 90
Transient output files ....................................................................... 91
The Output Control window ............................................................ 92
Pipe and Junction parameters .................................................... 92
Pipe and Junction Transient parameters.................................... 94
General output ........................................................................... 94
Summaries ................................................................................. 94
Format and action ...................................................................... 95
Show selected pipes/junctions................................................... 96
Database connections ................................................................ 97
Command buttons...................................................................... 97
Output parameter descriptions................................................... 98
Steady flow results ........................................................................... 98
Transient flow results for pipes ....................................................... 99
Transient Max/Min results for pipes ....................................... 100
Summary Max/Min results ...................................................... 101
Transient flow results for junctions ............................................... 103
2/24/2003
1:35 PM
Page ix
2/24/2003
1b.doc
1:35 PM
Page x
Table of Contents xi
Connecting objects ............................................................................... 126
Defining objects.................................................................................... 128
Specifying required property data.................................................. 128
Highlighting required information ................................................. 129
Use status feature ........................................................................... 129
Satisfying connectivity requirements............................................. 129
Inspecting objects ................................................................................. 130
Using the Checklist............................................................................... 132
Analysis Type................................................................................. 133
Specify Steady Solution Control.................................................... 134
Specify Output Control .................................................................. 134
Specify System Properties.............................................................. 134
Define All Pipes and Junctions ...................................................... 134
Section Pipes .................................................................................. 135
Transient Control ........................................................................... 135
Steady Solution Control........................................................................ 135
Output Control...................................................................................... 136
System Properties ................................................................................. 138
System fluid property variation...................................................... 139
Density and dynamic viscosity....................................................... 140
Bulk modulus of elasticity ............................................................. 140
Vapor pressure and cavitation........................................................ 140
The fluid databases ........................................................................ 141
The AFT Standard Database ................................................... 141
The Chempak database............................................................ 141
Accuracy option....................................................................... 142
Viscosity models ............................................................................ 143
Non-Newtonian flow in non-pipe elements............................. 144
Atmospheric pressure..................................................................... 144
Gravitational acceleration .............................................................. 145
Transition Reynolds Numbers ....................................................... 145
Editing the AFT Standard fluid database....................................... 146
Finding object definition status ............................................................ 146
2/24/2003
1:35 PM
Page xi
2/24/2003
1b.doc
1:35 PM
Page xii
2/24/2003
1:35 PM
Page xiii
2/24/2003
1b.doc
1:35 PM
Page xiv
Table of Contents xv
Initial Pressure/Head Guess ........................................................... 204
Displaying junction names and numbers ....................................... 206
Special Conditions ......................................................................... 206
Design Factor ................................................................................. 206
Changing the icon graphic ............................................................. 207
Changing the icon size ................................................................... 207
Specifying Base Area..................................................................... 207
Notes .............................................................................................. 207
Status.............................................................................................. 208
Area Change Specifications window.................................................... 208
Assigned Flow Specifications window................................................. 209
Transient Data ................................................................................ 211
Special Conditions ......................................................................... 211
Special features for steady flow..................................................... 211
Assigned Pressure Specifications window ........................................... 212
Transient Data ................................................................................ 214
Branch Specifications window ............................................................. 214
Transient Data ................................................................................ 216
Check Valve Specifications window.................................................... 217
Specifying losses............................................................................ 218
Forward Velocity to Close Valve................................................... 218
Delta Pressure to Re-Open ............................................................. 218
Transient Data ................................................................................ 218
Control Valve Specifications window.................................................. 219
Control Valve types ....................................................................... 220
PRV/PSV static vs. stagnation pressure .................................. 220
Action if control is lost .................................................................. 221
Special Conditions ......................................................................... 221
Open Percentage Table .................................................................. 222
Transient Data ................................................................................ 222
Failure Transient...................................................................... 222
Control Transient..................................................................... 223
Dead End Specifications window......................................................... 223
2/24/2003
1:35 PM
Page xv
2/24/2003
1b.doc
1:35 PM
Page xvi
2/24/2003
1:35 PM
Page xvii
2/24/2003
1b.doc
1:35 PM
Page xviii
2/24/2003
1:35 PM
Page xix
2/24/2003
1b.doc
1:35 PM
Page xx
2/24/2003
1:35 PM
Page xxi
2/24/2003
1b.doc
1:35 PM
Page xxii
2/24/2003
1:35 PM
Page xxiii
2/24/2003
1b.doc
1:35 PM
Page xxiv
2/24/2003
1:35 PM
Page xxv
2/24/2003
1b.doc
1:35 PM
Page xxvi
2/24/2003
1:35 PM
Page xxvii
2/24/2003
1b.doc
1:35 PM
Page xxviii
Nomenclature
a
wavespeed
BM
BP
CD
discharge coefficient
CM
CP
CV
valve coefficient
CHW
Hazen-Williams factor
diameter of a pipe
modulus of elasticity
friction factor
fT
FB
FW
gravitational constant
head
HGL
Jacobian matrix
loss factor
KB
Ksprinkler
length of a pipe
2/24/2003
1b.doc
1:35 PM
Page xxix
m&
general constant
pressure, static
Po
pressure, stagnation
Pw
wetted perimeter
power
radius
relaxation
R, R
resistance
Rc
coefficient of rigidity
Re
Reynolds Number
pump speed
SB
SC
Sy
yield stress
time
torque on pump
velocity
elevation
, ,
angle
diameter ratio
torque ratio
kinematic viscosity
roughness
2/24/2003
1b.doc
1:35 PM
Page xxx
Nomenclature xxxi
pump speed
density
dynamic viscosity
Subscripts
f
fluid
motor
stagnation
new
old
2/24/2003
1:35 PM
Page xxxi
2/24/2003
1b.doc
1:35 PM
Page xxxii
CHAPTER 1
Introduction
Terminology
The term waterhammer has been used for more than one-hundred
years, and describes a transient phenomenon in piping systems caused by
2/24/2003
1b.doc
1:35 PM
Page 1
Modeling capabilities
AFT Impulse can be used to model a wide variety of pipe system
transients, including:
2/24/2003
1b.doc
1:35 PM
Page 2
Chapter 1 Introduction 3
Interface features
AFT Impulse's graphical interface is based on drag-and-drop operations,
which make it simple to build a model of a generalized pipe system. You
control the arrangement, and you benefit from the direct visual feedback
regarding the layout of your model.
Data is entered for the components in Specifications windows, which are
opened by double-clicking the component of interest. Additional global
editing features simplify making large-scale changes to the model.
AFT Impulse handles both traditional English and SI systems of units.
You assign units to all input parameters by choosing from lists. This
highly flexible approach removes the burden of hand calculated unit
conversions.
2/24/2003
1:35 PM
Page 3
Liquid flow
One-dimensional flow
No chemical reactions
2/24/2003
1b.doc
1:35 PM
Page 4
Chapter 1 Introduction 5
2/24/2003
1:35 PM
Page 5
Required
Processor
Pentium II or higher
30 MB
Floppy drive
Yes
CD-ROM drive
Yes
Monitor resolution
SVGA (800x600)
RAM
64 MB
MS-Windows
2/24/2003
1b.doc
1:35 PM
Page 6
Chapter 1 Introduction 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.
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.
2/24/2003
1:35 PM
Page 7
Output
Graph Results
Figure 1.1
Input windows
The two windows that function exclusively as input windows are the
Workspace window and the Model Data window. These two windows,
one graphical and the other text-based, work together to process model
input data with immense flexibility. The tools provided in these two
windows allow you to model a large variety of pipe networks.
The Visual Report window can function in support of both input and
output data. As an input window, it allows you to see the input data
superimposed on the pipe system schematic created on the Workspace.
Output windows
The two windows that function exclusively as output windows are the
Output window and the Graph Results window. The Output window is
text-based, while the Graph Results window is graphical. These two
windows offer a powerful and diverse range of features for reviewing
analysis results for modeling errors, gaining a deeper understanding of
the pipe system's flow behavior, and preparing the results for
documentation.
As an output window, Visual Report allows you to see the output results
superimposed on the pipe system schematic created on the Workspace.
The five primary windows form a tightly integrated, highly efficient
system for entering, processing, analyzing, and documenting
waterhammer analyses of pipe networks.
2/24/2003
1b.doc
1:35 PM
Page 8
Chapter 1 Introduction 9
Converting models and databases from AFT Impulse 1.0 and 2.0
AFT Impulse 1.0 and 2.0 models can be opened directly by choosing
Open from the File menu. Once saved, they are saved in the version 3.0
format. Importing issues are discussed in the following.
Pipe friction
AFT Impulse 1.0 only worked with constant friction factor models.
When these models are imported into version 3.0, the friction model for
the pipe is set to Explicit Friction Factor or Frictionless, depending on
whether the version 1.0 pipe had a zero or non-zero friction factor.
Reservoir transients
AFT Impulse 1.0 allowed reservoir surface level transients as a delta
from the steady-state. These will all get converted automatically to the
version 3.0 convention of specifying the actual liquid surface level.
2/24/2003
1:35 PM
Page 9
Databases
AFT Impulse 1.0 databases can be converted using the AFT Convert
utility provided with the software.
Combining pipes
AFT Impulse 3.0 will offer to combine pipes for you when you import
AFT Fathom models. This will allow models to run much more
efficiently. When combined, AFT Impulse will create intermediate pipe
elevations at the points where junctions are deleted. All pipe resistance
and junction losses will be maintained in the model.
Databases
AFT Fathom and AFT Mercury databases can be converted using the
AFT Convert utility.
2/24/2003
1b.doc
1:35 PM
Page 10
Chapter 1 Introduction 11
New pump startup method using motor torque and speed data
Pump can control pressure only when set pressure is exceeded high
or low
New relief valve pressure loss model with Cv vs. pressure difference
allows improved modeling of passive relief valves
2/24/2003
1:35 PM
Page 11
Workspace
Output
Graph Results
Pipe Outlet station can be saved into graph sets, and when model is
resectioned the outlet station is automatically updated
User has improved control over graph colors and curve thicknesses
and can now specify defaults
2/24/2003
1b.doc
1:35 PM
Page 12
Chapter 1 Introduction 13
Visual Report
Miscellaneous
2/24/2003
1:35 PM
Page 13
2/24/2003
1b.doc
1:35 PM
Page 14
CHAPTER 2
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".
2/24/2003
1b.doc
1:35 PM
Page 15
Workspace
Toolbox
Minimized
primary
windows
Status Bar
Figure 2.1
2/24/2003
1b.doc
1:35 PM
Page 16
2/24/2003
1:35 PM
Page 17
2/24/2003
1b.doc
1:35 PM
Page 18
Figure 2.2b Walk Through Model with one reservoir and one branch
placed
Figure 2.2c Walk Through Model after placing the Surge Tank
2/24/2003
1:35 PM
Page 19
Figure 2.2d Walk Through Model after placing the second Reservoir
2/24/2003
1b.doc
1:35 PM
Page 20
2/24/2003
1:35 PM
Page 21
2/24/2003
1b.doc
1:35 PM
Page 22
2/24/2003
1:35 PM
Page 23
2/24/2003
1b.doc
1:35 PM
Page 24
Figure 2.2h Walk Through Model with all pipes and junctions
placed
2/24/2003
1:35 PM
Page 25
Figure 2.3
2/24/2003
1b.doc
1:35 PM
Page 26
Figure 2.4
2/24/2003
1:35 PM
Page 27
Figure 2.5
2/24/2003
1b.doc
1:35 PM
Page 28
2/24/2003
1:35 PM
Page 29
Figure 2.6
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.
2/24/2003
1b.doc
1:35 PM
Page 30
A. Define Reservoir J1
To define the first reservoir, open the J1 Reservoir Specifications
window by double-clicking on the J1 icon. Enter in a reservoir surface
elevation of 200 meters. You can assign any unit of length found in the
adjacent drop-down list box of units.
Note: You can also open an object's Specifications window by selecting
the object (clicking on it) and then either pressing the Enter key or by
clicking the Open Pipe/Jct Window icon on the Toolbar.
Enter surface pressure of 1 atmosphere (atm) and a reservoir depth of
50 meters in the table on the Pipe Depth and Loss Coefficient tab.
Note: You can specify preferred units for many parameters (such as
meters for length) in the Parameter & Unit Preferences window.
You can give the component a name, if desired, by entering it in the
Name field at the top of the window. In Figure 2.7, the name of this
reservoir is Supply Tank A. By default the junctions name is the
junction type. The name can be displayed on the Workspace, Visual
Report or in the Output.
2/24/2003
1:35 PM
Page 31
2/24/2003
1b.doc
1:35 PM
Page 32
Figure 2.7
2/24/2003
1:35 PM
Page 33
Inspection
window
Figure 2.8
B. Define Branch J2
Open the J2 Branch Specifications window. In this window, three
connecting pipes will be displayed in the pipe table area. Branches are
connector points for up to twenty-five pipes.
Enter an elevation of 100 meters for the J2 Branch junction (an
elevation must be defined for all junctions). You can also specify a
transient flow source or sink at the junction.
Click the Optional tab and enter an imposed flow rate of -2 m3/sec (the
negative sign means that the flow is out of the junction a flow sink).
Click OK to accept the input and exit the window.
2/24/2003
1b.doc
1:35 PM
Page 34
Figure 2.9
2/24/2003
1:35 PM
Page 35
E. Define Valve J7
Open the J7 Valve Specifications window. Valve junctions connect
with two pipes if they are internal and one pipe if they are exit valves.
This will be an exit valve. Enter an elevation of 25 meters and select the
Exit Valve option at the lower left. Choose Head and enter an exit
pressure of 25 meters, which is equivalent to 1 atm (because exit
atmospheric head is equal to elevation).
Choose the loss model as Cv. This valve will initiate the transient for
this system. Until the valve changes position, the entire pipe system is in
a steady-state condition. Enter a Cv of 2392.6. This is the steady-state
value.
2/24/2003
1b.doc
1:35 PM
Page 36
Figure 2.10 Valve Specifications for J7 initiates the transient for the
system
Select the Transient Data tab. In this transient the valve partially
closes over a period of 10 seconds, then holds steady at the new position
for the remainder of the simulation. Enter the following Cv vs. time data
in the table (see Figure 2.10):
Time
0
Cv
2392.6
10
797.4
100
797.4
2/24/2003
1:35 PM
Page 37
Cv
0
651.2
63
100
The data represents the valve initially closed at time zero. Time zero is
the moment of cracking. The valve then opens and gradually closes
again over a period of 60 seconds.
G. Define Pipe P1
The next step is to specify all the pipes. To open the Pipe Specifications
window, double-click the pipe object on the Workspace.
First open the Pipe Specifications window for Pipe P1 (Figure 2.11).
Choose the Pipe Material as Unspecified, choose the User Specified
Wavespeed option, and select the Friction Model Data Set as
Unspecified with the Friction Model as "Explicit Friction Factor". Enter
a length of 1001.2 meters, a diameter of 1.5 meters, and a friction factor
of 0.012.
Note: This example from Karney assumes all friction factors are
known ahead of time. Normally this will not be the case. In most cases
you will access the roughness for the pipe from the pipe material
database supplied with AFT Impulse, or input your own roughness
2/24/2003
1b.doc
1:35 PM
Page 38
2/24/2003
1:35 PM
Page 39
Length
(meters)
Diameter
(meters)
Friction
Factor
Wavespeed
(meters/sec)
1001.2
2000
1.5
1
0.012
0.013
2000
0.75
0.014
995
502.5
0.5
0.015
1000
502.5
0.5
0.015
1000
1001.2
0.014
996.3
2000.2
0.75
0.013
995.1
100.12
0.014
996.3
2/24/2003
1b.doc
1:35 PM
Page 40
996.3
995.3
2/24/2003
1:35 PM
Page 41
General data
Pipe data
Junction data
Figure 2.12 The Model Data window shows the input data in text
form
Figure 2.13 The Transient Data tab in the Model Data Junction data
area shows all transient data entered
2/24/2003
1b.doc
1:35 PM
Page 42
ni =
Li
ai t
2/24/2003
1:35 PM
Page 43
For Min. and Max. Sections in Controlling Pipe enter 1 section for
each. Limit the Max Error to 1% for the search (usually 5 or 10% is
fine). Check the box for Sort Sectioning by Minimum Error. Then click
the Search button.
A list of possible sectioning is displayed. Click the top line with 1.0000
sections in the controlling pipe. The resulting time step will be displayed
as 0.100492 seconds.
The sectioning and resulting errors in the remaining pipes are displayed
in the table near the bottom. Click OK and the fifth checklist item should
be completed.
2/24/2003
1b.doc
1:35 PM
Page 44
2/24/2003
1:35 PM
Page 45
2/24/2003
1b.doc
1:35 PM
Page 46
Figure 2.16 The Solution Progress window displays the state of the
simulation
2/24/2003
1:35 PM
Page 47
2/24/2003
1b.doc
1:35 PM
Page 48
General results
Pipe results
Junction results
Figure 2.17 The Output window displays output results in text form
2/24/2003
1:35 PM
Page 49
2/24/2003
1b.doc
1:35 PM
Page 50
2/24/2003
1:35 PM
Page 51
Figure 2.19 The Select Graph Data window controls the Graph
Results content
Click the Add >> button to add this station to the list on the right.
From Graph Parameters select Hydraulic Gradeline (HGL) and
choose meters as the Y-Axis units. Then click the Show button to
display the graph (Figure 2.20).
This pressure/head history can be compared to that published in Karney
(1992). This is done in the Verifications documentation provided with
AFT Impulse.
2/24/2003
1b.doc
1:35 PM
Page 52
2/24/2003
1:35 PM
Page 53
2/24/2003
1b.doc
1:35 PM
Page 54
2/24/2003
1:35 PM
Page 55
Figure 2.23 The Visual Report Control window selects content for
the Visual Report window
Click the Visual Report Control button on the Toolbar (or View
menu) and open the Visual Report Control window, shown in Figure
2.23. Default parameters are already selected, but you can modify these
as desired. For now, select Max Pressure Stagnation and Min Pressure
Stagnation in the Pipe Results area. Click the Show button. The Visual
Report window graphic is generated (see Figure 2.24).
It is common for the text in the Visual Report window to overlap when
first generated. You can change this by selecting smaller fonts or by
dragging the text to a new area to increase clarity (this has already been
done in Figure 2.24, as has the selection to show units in a legend). This
window can be printed, copied to the clipboard for import into other
Windows graphics programs, or saved to file.
2/24/2003
1b.doc
1:35 PM
Page 56
Figure 2.24 The Visual Report window integrates results with the
model layout
Conclusion
You have now used AFT Impulse's five primary windows to build a
simple model. Review the rest of this Users Guide for more detailed
information on each of the windows and functions.
2/24/2003
1:35 PM
Page 57
2/24/2003
1b.doc
1:35 PM
Page 58
CHAPTER 3
Fundamental Concepts
2/24/2003
1b.doc
1:35 PM
Page 59
2/24/2003
1b.doc
1:35 PM
Page 60
2/24/2003
1:35 PM
Page 61
Specifying losses
Local losses can be specified in two ways. The most flexible way of
including a loss factor in an AFT Impulse model is to associate it with a
junction specifically defined for that type of loss.
Secondly, you can include loss factors with pipes. When you specify
Additional Losses in the Pipe Specifications window, the loss is
assumed to be distributed evenly along the pipe length, much like
friction losses. For this reason a pipe-associated loss factor is referred to
as a distributed loss. The losses at junctions, on the other hand, are point
losses.
When modeling for waterhammer, it is preferred that losses for all static
components be included in pipes as Additional Losses. This reduces the
number of pipes and model run time.
2/24/2003
1b.doc
1:35 PM
Page 62
2/24/2003
1:35 PM
Page 63
2/24/2003
1b.doc
1:35 PM
Page 64
Instantaneous waterhammer
With few exceptions, it is safe to calculate the maximum possible
waterhammer pressure surge by using the instantaneous waterhammer
equation. The instantaneous waterhammer equation assumes that the
transient event occurs either instantaneously or rapidly enough such that
it is in effect instantaneous.
In such a case, it can be shown (Wylie, et al., 1993, pp. 4) by use of the
mass and momentum equation that the pressure transient is given by the
following equation:
P = a V
(3.1)
where:
P =
pressure surge
density
wavespeed
V =
velocity change
By adding the pressure surge to the existing static pressure, one can
obtain the maximum theoretical pressure in the pipe. However, in some
cases the pressure can exceed the instantaneous prediction. These are
discussed in Chapter 9.
2/24/2003
1:35 PM
Page 65
ft
ft 1 ft 2
lbf s 2
Pmax = 62
* 4000 *10 *
*
+ 50 psig
s
s 144 in 2 32.2 lbm - ft
ft 3
Pmax = 535 + 50
lbm
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.
t = 2
L
a
(3.2)
Conceptual example
Let's consider a simple example of a valve closure. A frictionless pipe is
connected to an upstream reservoir and a downstream valve. Fluid is
2/24/2003
1b.doc
1:35 PM
Page 66
2/24/2003
1:35 PM
Page 67
V = Vsteady
V = Vsteady
V=0
a
V=0
a
Pinstantaneous
Psteady
Pinstantaneous
Psteady
Vsteady
x
V = Vsteady
V = Vsteady
V=0
V=0
P
Psteady
-Vsteady
P
Psteady
Pinstantaneous
x
Pinstantaneous
x
Vsteady
x
-Vsteady
Figure 3.1
2/24/2003
1b.doc
1:35 PM
Page 68
2/24/2003
1:35 PM
Page 69
2/24/2003
1b.doc
1:35 PM
Page 70
CHAPTER 4
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 Output window displays the results of the analysis in text form
and lets you produce attractive, effectively organized printed output.
The Visual Report window merges the output data with the pipe
system layout from the Workspace to present a unique perspective of
2/24/2003
1b.doc
1:35 PM
Page 71
You can change between the five primary windows by selecting them
from the Window menu, the Toolbar, or by clicking the appropriate
icons within the AFT Impulse window.
Model Data
Visual Report
Workspace
Output
Graph Results
Figure 4.1
2/24/2003
1b.doc
1:35 PM
Page 72
2/24/2003
1:35 PM
Page 73
Toolbox
Workspace
Status bar
Figure 4.2
2/24/2003
1b.doc
1:35 PM
Page 74
Spray Discharge
Pump
Volume Balance
General Component
Figure 4.3
If you Double-click the Pipe Drawing Tool it remains active until you
click it again a single time. This allows you to draw a series of pipes
without returning to the Toolbox each time.
Pipe handles and segmenting a pipe
When a single pipe is selected, two black squares (or more if the pipe is
segmented), called handles, appear. When the cursor is placed over a
pipe handle it changes to a crosshair and the pipe end can be moved or
stretched by dragging with the left mouse button.
A pipe can be segmented or bent in up to ten segments (See Figure
4.4). This does not affect the model or the results, only the visual
2/24/2003
1:35 PM
Page 75
Figure 4.4
This feature is useful in graphically bending the pipe around another part
of the model or graphically showing that there are embedded elbows as
Additional Losses.
You can add a segment to a pipe in two ways. The easiest way to add a
single segment is to select Pipe Segments and Add from the Arrange
menu. If there are no segments, then a single segment can also be added
by selecting the Add/Remove Segment from the Toolbar. If the pipe
already has one or more segments, this toolbar selection will remove all
segments.
A more flexible way of adding or removing segments is to use the Pipe
Segment Manager. To open the Pipe Segment Manager, choose Pipe
Segments and Manager from the Arrange menu. A window will display
(Figure 4.5) where you can add, remove and merge segments. A
handle will appear in the middle of the pipe for each new segment. As
the cursor passes over the handle the cursor will change to a crosshair.
Press and hold down the left mouse button and drag the middle of the
2/24/2003
1b.doc
1:35 PM
Page 76
Figure 4.5 The Pipe Segments window allows you to create, merge
and remove pipe segments.
2/24/2003
1:35 PM
Page 77
2/24/2003
1b.doc
1:35 PM
Page 78
Junction icons
Below the four drawing tools are nineteen junction icons. The junctions
allow you to model a large variety of pipe system components important
in waterhammer. To add a junction to a model, drag the desired icon
from the Toolbox and drop it anywhere on the Workspace.
As you move the mouse pointer over the Toolbox, a Tool Tip identifies
the type of junction under the mouse pointer.
The types and order of junction icons on the Toolbox can be modified
through the Toolbox Preferences window. For more details on how to
customize the Toolbox, refer to Chapter 7.
Editing features
Junction and pipe objects on the Workspace can be manipulated as
individual items or as groups. This applies to cutting, copying, pasting,
deleting, duplicating, dragging, locking, and undoing. Pipe objects on the
2/24/2003
1:35 PM
Page 79
2/24/2003
1b.doc
1:35 PM
Page 80
2/24/2003
1:35 PM
Page 81
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.
Customizing features
Workspace Preferences
The Workspace Preferences on the Options menu lets you customize the
appearance and behavior of the Workspace. New defaults can be saved
so AFT Impulse will always appear and behave according to your
preferences.
See Chapter 7 for more information on Workspace Preferences.
2/24/2003
1b.doc
1:35 PM
Page 82
Figure 4.7
2/24/2003
1:35 PM
Page 83
Figure 4.8
2/24/2003
1b.doc
1:35 PM
Page 84
Figure 4.9
Background Graphic
You can load a graphic into the background of the Workspace and build
your model on top of that graphic. The graphic could be, for instance, a
topographic map or a facility drawing. Such graphics are loaded from the
2/24/2003
1:35 PM
Page 85
Specifications windows
Each object on the Workspace has an associated Specifications window.
The Specifications window lists all the input data for the selected object.
There are three ways to open an objects Specifications window from the
Workspace:
1. Double-click the object
2. Select the object and click the Open Pipe/Jct Window icon on the
Toolbar
3. Select the object and press ENTER
The Specifications windows can also be accessed through the Model
Data window.
2/24/2003
1b.doc
1:35 PM
Page 86
Customize view
Pipe data
section
Junction data
section
The Model Data window does not offer any tools to build or add to a
model; all model assembly must occur in the Workspace.
The Model Data window allows you to manipulate existing elements
from the Workspace. Therefore, a complete model can be assembled in
the Workspace window without ever opening the Model Data window.
In fact, for small models this may be the preferred approach.
The Pipes table allows display of all input data for all pipes in the model.
Details of additional losses are displayed in the Pipe Loss Data table,
accessed with an adjacent tab. This table displays all fittings with K
factors grouped into 27 columns.
The other display tab in the pipe area is the Pipe Detail Summary, which
assembles a list of details about each pipe in another format useful to
some engineers.
The junction data area is below the pipe data area. Here the junction data
is separated into tables for each junction type. Click the folder tabs to
see the data for any type of junction.
2/24/2003
1:35 PM
Page 87
2/24/2003
1b.doc
1:35 PM
Page 88
Figure 4.11 The Model Data Control window allows selection of the
parameters to show in the Model Data window
2/24/2003
1:35 PM
Page 89
2/24/2003
1b.doc
1:35 PM
Page 90
2/24/2003
1:35 PM
Page 91
2/24/2003
1b.doc
1:35 PM
Page 92
Figure 4.13 The Output Control window selects parameters for the
Output window. Here the Pipe parameters (for steadyflow results) are shown.
You can change the order of the output by selecting a parameter on the
right-hand list, then clicking the Reorder scroll bar on the far right either
up or down. The list will be reordered accordingly. The output is
presented in the order defined in this list.
The units used for a parameter can be changed by selecting the
parameter then choosing the desired units from the dropdown list below
the list of parameters. By selecting the units of interest, you can obtain
your results in whatever units you find most convenient and meaningful.
This means that you can enter all your input parameters in one set of
units (or a variety of units, for that matter), and have all your output
parameters in a completely different set of units. This feature has
obvious benefits for international applications.
To set selected units for both pipes, junctions and summary reports to
the preferred units (set in Parameter and Unit Preferences), click the Use
Preferred Units button.
2/24/2003
1:35 PM
Page 93
2/24/2003
1b.doc
1:35 PM
Page 94
The second method will force the data to use exactly the number of
digits specified, which may result in the decimal point location changing
from value to value. Using this method with the above example and
specifying four digits, the output will be displayed as:
1235.
0.09876
The third method will align the data on the decimal point only if the
maximum value in the column divided by the minimum value is less than
a limiting order of magnitude that you set. If this ratio is greater than the
limit then each value in the column will be exactly the number of digits
2/24/2003
1:35 PM
Page 95
The fourth method is similar to the third except that if the ratio is greater
than the limit then the entire column will be shown in exponential
format. Using the above example, the column would be shown as:
1.235E+03
9.876E-02
These formatting options are for the entire Output window; however, the
last two are calculated and applied on a column basis. In other words, if
you choose the third option, one column may be aligned but another may
not be.
The output can be forced to always display in exponential notation by
choosing the checkbox.
For transient results, the maximum and minimum values for all selected
parameters are shown in the Output window. The results can show the
max/min values for all pipe stations, or only the max/min values for
anywhere in the pipe. You can choose which type of report you want by
selecting Detailed or Summary in the Max/Min Transient Table
Contents area. For models with a large number of pipe sections, the
Summary report is generally preferable because of the time required to
display the Detailed table.
Show selected pipes/junctions
You can choose which pipes and junctions are to be displayed in the
Output window. Clicking a selected item in the list on the Show Selected
Pipes/Junctions tab will deselect it. Clicking a deselected item will select
it. This feature is especially useful when you are interested only in a
certain area of the model.
You can quickly define the output you want by doing the following:
1. Select the pipes and junctions on the Workspace
2. Open the Output Control window and select the Show Selected
Pipes/Junctions tab
2/24/2003
1b.doc
1:35 PM
Page 96
2/24/2003
1:35 PM
Page 97
2/24/2003
1b.doc
1:35 PM
Page 98
Figure 4.14 The Output window shows steady flow results in the
Pipes table and the various junction tables. Transient
data is shown on the transient tabs.
2/24/2003
1:35 PM
Page 99
Time step
selection tool
slides up and
down
2/24/2003
1b.doc
1:35 PM
Page 100
Figure 4.16 The Transient Max/Min table details the maximum and
minimum values for the entire simulation.
2/24/2003
1:35 PM
Page 101
Figure 4.17 Detailed and Summary versions are available for the
Transient Max/Min table
2/24/2003
1b.doc
1:35 PM
Page 102
2/24/2003
1:35 PM
Page 103
2/24/2003
1b.doc
1:35 PM
Page 104
Figure 4.21 The Select Graph Data window controls the Graph
Results window
2/24/2003
1:35 PM
Page 105
2/24/2003
1b.doc
1:35 PM
Page 106
2/24/2003
1:35 PM
Page 107
2/24/2003
1b.doc
1:35 PM
Page 108
2/24/2003
1:35 PM
Page 109
2/24/2003
1b.doc
1:35 PM
Page 110
2/24/2003
1:35 PM
Page 111
Figure 4.24 Visual Report window integrates input data and output
results with model layout
2/24/2003
1b.doc
1:35 PM
Page 112
Display parameters
The parameters to be shown in the Visual Report window are selected
from the Display Parameters tab. The left list selects the pipe parameters
and the right list selects the junction parameters. In these lists you can
show only the parameters you have selected in the Output Control
window, or all output parameters. You can select more than one
parameter; however, the Visual Report window may get crowded if too
many are selected.
General display
The General Display tab allows you to configure several other items.
2/24/2003
1:35 PM
Page 113
Figure 4.26 The Visual Report with the units and Color Map in the
Legend
The units for all parameters can be shown in a legend instead of next to
the parameter value (see Figure 4.24 or 4.26). This will reduce the
amount of text and crowding shown in the Visual Report window. The
legend can be moved to any location in the Visual Report window. If this
option is chosen while in input-only mode, all values will be converted
to a consistent set of units.
You can change the font by clicking the Fonts button and choosing the
font desired or by using the font size change buttons on the Toolbar.
The number of digits to show in the numerical display can be configured
using the provided drop-down list.
Normally on the Workspace, closed objects are displayed using special
graphics (dashed lines, Xs, etc.). The number of a junction which has
special conditions set, by default, has an X preceding the J (e.g. XJ56).
2/24/2003
1b.doc
1:35 PM
Page 114
2/24/2003
1:35 PM
Page 115
The order of entry is not important. Impulse will sort the entries first by
value and then by the operator, with the > being above the <. The pipes
will be colored using the first match found in the map, starting from the
top of the map. If no match is found, the pipe will be the Workspace pipe
color. Be careful to avoid overlapping operators and values. If, for
example, the map had the following four colors,
2/24/2003
1b.doc
1:35 PM
Page 116
Figure 4.28 The Visual Report Control Color Map allows you to
specify the display colors for pipes according to their
output values.
To avoid overlapping colors, only use less than (<) as the last entry. A
better map would be to use
1) >=60
2) >=20
2/24/2003
1:35 PM
Page 117
2/24/2003
1b.doc
1:35 PM
Page 118
The Toolbars
The Toolbars offer quick access to the features used most frequently in
AFT Impulse. Figure 4.29 shows the Toolbars. There are actually six
toolbars in all. One is the Common Toolbar, and it is always shown. Its
features are common to all five Primary windows. In addition, there is
one Toolbar for each Primary window, offering features appropriate for
work in that window. Each of the functions on the Toolbar is also
available through the menu system.
2/24/2003
1:35 PM
Page 119
2/24/2003
1b.doc
1:35 PM
Page 120
CHAPTER 5
Creating objects
The pipe flow model is assembled on the Workspace by arranging
graphical objects (pipes and junctions) that represent parts of a physical
pipe flow system. In addition to pipe and junction objects, annotation
objects can also be added.
Pipes
To create a pipe, select the Pipe Drawing Tool from the Toolbox and
draw a line on the Workspace to represent the pipe. To draw multiple
pipes more quickly, hold down the CTRL key while drawing each pipe.
This keeps the Pipe Drawing Tool active, so you don't have to select it
for each pipe. Alternatively, you can double-click the Pipe Drawing Tool
to lock it down. You can then continue to draw pipes until you click
the Pipe Drawing Tool a second time.
To draw a vertical or horizontal pipe, hold down the SHIFT key while
drawing the pipe.
2/24/2003
1b.doc
1:35 PM
Page 121
2/24/2003
1b.doc
1:35 PM
Page 122
Figure 5.1
This feature is useful in graphically bending the pipe around another part
of the model or graphically showing that there are embedded elbows as
Additional Losses.
You can add a segment to a pipe in two ways. The easiest way to add a
single segment is to select Pipe Segments and Add from the Arrange
menu. If there are no segments, then a single segment can also be added
by selecting the Add/Remove Segment from the Toolbar. If the pipe
already has one or more segments, this toolbar selection will remove all
segments.
The other way to segment a pipe is to select the pipe on the Workspace
then choose the Pipe Segments and Manager from the Arrange menu. A
window will display where you can add, remove and merge segments. A
handle will appear in the middle of the pipe for each new segment. As
the cursor passes over the handle, the cursor will change to a crosshair.
Press and hold down the left mouse button and drag the middle of the
pipe to the desired location. By holding the SHIFT key the pipe segment
will move only at right angles.
2/24/2003
1:35 PM
Page 123
Junctions
To place a junction, drag the icon from the Toolbox onto the Workspace.
The nineteen available junctions represent various pipe system
components.
When you duplicate a junction using the Duplicate or Copy and Paste
features on the Edit menu, the new junction retains the engineering
information associated with the original junction.
Junctions are shown on the Workspace with default ID numbers, which
you can change in the Junction Specifications window. The letter J
precedes the ID number to signify that it is for a junction. From the
junctions Specifications window you can change the ID number to any
desired value greater than zero but less than 10,000. Identical ID
numbers can be assigned to both a pipe and a junction because the P or
J will distinguish the two.
Unlike pipes, junctions do not have reference positive flow directions.
Junctions for which the flow direction is important (such as pumps and
control valves) derive the flow direction from the flow direction of the
connecting pipes.
Morphing junctions
A junction can be morphed from one type to another. To morph a
junction, hold down the CTRL key and select a junction from the
Toolbox and drop it onto an existing Workspace junction. The junction
type will change to the new junction type, and data that can be kept will
be copied into the new junction.
Splitting pipes
If you hold down the SHIFT key while selecting a junction from the
Toolbox and then drop the junction onto an existing pipe, the pipe will
split into two pipes. The physical length of the original pipe will be
automatically halved, and the new pipe will be assigned the balance.
Thus the sum of the two pipe lengths will equal the length of the original
pipe. Any additional losses in the original pipe will be left in the original
pipe, and the new pipe will have no additional losses.
2/24/2003
1b.doc
1:35 PM
Page 124
Annotations
To create an annotation, select the annotation tool on the Toolbox and
draw the outline of the annotation. When you release the mouse button,
the Workspace Annotation window will display. Here you can enter text
for the annotation, draw lines and arrows, change colors, and draw
outlines. You also can change the colors of the annotation.
Annotations can be cut, copied, pasted, duplicated, and moved like pipe
and junction objects. When you select an existing annotation, resizing
handles appear.
The Annotation Manager assists you in specifying which annotations
will display in which scenarios.
Moving objects
Objects on the Workspace can be moved individually or as groups.
Clicking a Workspace object selects it. Pipe objects change color when
selected. Junction and annotation objects are outlined in red when
selected. This color is configurable in the Workspace Preferences.
To move an object, select it, drag it within the Workspace, and drop it in
the desired location. When an object is dragged beyond the existing
Workspace area, the Workspace is expanded accordingly.
A pipe object can be stretched by grabbing the handles (small black
squares) at the pipe endpoints and moving an endpoint to a new location.
To prevent accidental movement, lock the objects on the Workspace.
The Lock feature is accessed from the Edit menu or the lock button on
the Toolbar.
To group multiple objects for movement or other operations, hold down
the SHIFT key when selecting the objects. Objects can also be grouped
by drawing a box around them using the Selection Drawing Tool on the
Toolbox. You can use the Select Special on the Edit menu to group
objects based on specified criteria.
You can also select all objects in a flow path as follows:
2/24/2003
1:35 PM
Page 125
The endpoints of a flow path are the junctions which connect more than
one flow path, like a branch junction connected to three or more pipes.
You can choose to include or exclude these end point junctions.
Tip: You can undo all pipe and junction movements using Undo on the
Edit menu or Toolbar. You also can press the escape key during the
movement to cancel the move.
Keyboard modifiers
When you drag a junction, the endpoints of any connected pipes are
moved with the junction, thus maintaining the graphical connection.
This functionality may be changed through the Workspace Preferences
window.
Editing objects
The objects you place on the Workspace can be edited with the editing
commands from the Edit menu or the Toolbar. Objects can be cut,
copied, pasted, duplicated, and deleted. These operations can be
performed on individual objects or on groups. The Edit menu provides
one level of undo for each editing operation. Alternately, you can use the
Undo button on the Toolbar.
The Copy Graphics feature on the Edit menu will copy the specified
Workspace objects to the Windows clipboard. This image can then be
pasted into other Windows applications.
Connecting objects
During construction of a new model, objects can be placed anywhere on
the Workspace. In order to assemble a model that is ready to submit to
the Solver, you must connect the objects properly.
Remember that connectivity only exists between junctions and pipes.
There are no junctions that connect to junctions, and no pipes that
connect to pipes.
2/24/2003
1b.doc
1:35 PM
Page 126
Figure 5.2
2/24/2003
1:35 PM
Page 127
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.
2/24/2003
1b.doc
1:35 PM
Page 128
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.
2/24/2003
1:35 PM
Page 129
Inspecting objects
The Inspection feature is a time saving feature for reviewing the data
associated with a Workspace pipe or junction object. Inspection displays
the data for an object in read-only format (see Figure 5.3).
2/24/2003
1b.doc
1:35 PM
Page 130
Figure 5.3
2/24/2003
1:35 PM
Page 131
Figure 5.4
2/24/2003
1b.doc
1:35 PM
Page 132
Status bar
Status light
Figure 5.5
Tip: If you double click the checkbox on the Checklist window or Status
Bar, the appropriate window will be opened. For example, doubleclicking Specify Output Control will open the Output Control window.
Analysis Type
The first item on the Analysis menu is Analysis Type. This drops down
to two menu choices: Transient and Steady Only. By default the
Transient option is selected when a new model is opened. A transient
model always performs a steady-state analysis first, and then proceeds to
the transient solution. The Transient menu selection requires all six
Checklist items to be completed. The Steady Only option requires only
the first four.
2/24/2003
1:35 PM
Page 133
2/24/2003
1b.doc
1:35 PM
Page 134
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.
2/24/2003
1:35 PM
Page 135
Figure 5.6
Output Control
Most engineering analyses involve processing a very large amount of
information, some of which is critical but most of which is of lesser
2/24/2003
1b.doc
1:35 PM
Page 136
2/24/2003
1:35 PM
Page 137
Figure 5.7
System Properties
The System Properties window gives you control over the fluid and
environmental properties that influence your pipe system analysis.
Figure 5.8 shows the System Properties window, which is accessed
through the Analysis menu.
The three fluid properties that are required for all analyses are density,
dynamic viscosity and liquid bulk modulus of elasticity. If only a steadystate analysis is being performed, the liquid bulk modulus of elasticity
will not be used although input is still required. Fluid vapor pressure is
an optional specification.
These properties can be entered by hand when the fluid is selected as
Unspecified, or they can be obtained from the database list of AFT
Standard or Chempak fluids. When selected from the list, density,
2/24/2003
1b.doc
1:35 PM
Page 138
Figure 5.8
2/24/2003
1:35 PM
Page 139
KB =
2/24/2003
1b.doc
1:35 PM
Page 140
2/24/2003
1:35 PM
Page 141
Figure 5.9
The basis of all mixtures is either mass or mole, and is specified on the
System Properties window itself. There are no limits to the number of
components in a particular mixture.
The mixtures you create here are referred to as pre-mixtures because you
are specifying the composition before the model is run.
Accuracy option
There is one accuracy option for Chempak fluids and mixtures. The
State Property Accuracy option influences the number of elements used
in curve fits for calculating property data. Typically you should use the
high accuracy option.
2/24/2003
1b.doc
1:35 PM
Page 142
Figure 5.10 The Create Mixture window (opened from the System
Properties window) allows you to create predefined
mixtures for the model.
Viscosity models
The default viscosity model is Newtonian, which applies to a wide
variety of important liquids including water. Some liquids exhibit a
dependence of viscosity on the fluid dynamics, and these fluids are
called non-Newtonian. When one of these model is selected, additional
input parameters are required (see Figure 5.11).
AFT Impulse offers four non-Newtonian viscosity models. Two of the
models apply to pulp and paper. These are the Duffy model and the
Brecht & Heller model. Also available are models for Power Law fluids
and Bingham Plastics. See Tilton, 1997 for more information on these
viscosity models. More extensive discussion of these viscosity models is
given in Chapter 8.
2/24/2003
1:35 PM
Page 143
Atmospheric pressure
AFT Impulse uses a default atmospheric pressure of 1 standard earth
atmosphere, or 14.696 psia. You are free to change this value. This
pressure is also used to convert absolute pressures to gauge pressures.
Keep in mind that atmospheric pressure varies with altitude. If you are
doing one project at Sea Level and another at 8500 feet you should
2/24/2003
1b.doc
1:35 PM
Page 144
Gravitational acceleration
The default for gravitational acceleration is 1 standard earth
gravitational acceleration, or 32.174 ft/s2. You can change this to a
multiple of standard earth accelerations (that is, number of g's) according
to the design environment for your system. In principle, the body force
on a fluid system does not have to be due to gravity. See Figure 5.12 for
the System Data tab.
2/24/2003
1:35 PM
Page 145
2/24/2003
1b.doc
1:35 PM
Page 146
Section Pipes
In a transient simulation, a common time step must be used for all pipes.
The Method of Characteristics (MOC) requires that each pipe section
satisfies the following:
a=
x
t
(5.1)
where a is the wavespeed in the pipe, x is the length of the section and
t is the time step. The length of a pipe section is just the length of the
pipe divided by the number of sections in that pipe:
x =
L
n
(5.2)
The time step can be obtained by combining Equations 5.1 and 5.2:
t =
L
na
(5.3)
The maximum time is obtained from the pipe with the smallest value of
L/a. This pipe will have one section (n = 1) and thus the time step will
be:
t max =
L
a
(5.4)
min
In any pipe system there will be one pipe that is the controlling pipe.
The controlling pipe is that pipe which has the least number of sections,
frequently only one. Once the controlling pipe is identified, the time step
is determined by solving Equation 5.4 for t. Then the number of
sections in the remaining pipes is obtained from:
ni =
2/24/2003
Li
ai t
1:35 PM
(5.5)
Page 147
ni =
Li
(1 )ai t
2/24/2003
1b.doc
1:35 PM
Page 148
Numbering convention
Once a pipe is broken into sections, computation takes place where the
sections join together. These are called pipe stations. The total number
of stations in a pipe is equal to the number of sections plus 1. The station
at the beginning of the pipe is numbered zero. See Figure 5.14.
Computing
stations
Pipe sections
5
5
J2
P1
P1
1,0
1,1
1,2
J2
1,3
J3
P2
1,4 2,0
P2
2,1
2,2
2,3
Figure 5.15 The pressure drop across the valve at J2 in the AFT
Impulse model at the top is given by the difference
between the solutions at Pipe 1, station 4 and Pipe 2,
station 0.
For junctions with more than one connecting pipe, the principle is the
same (see Figure 5.16).
2/24/2003
1:35 PM
Page 149
J3
P2
P1
J2
P3
J4
P2
1,0
2,0
P1
2,1
J2
1,1
1,3
2,2
2,3
2,4
1,2
3,0
P3
3,1
3,2
3,3
3,4
Figure 5.16 Pipe station convention for AFT Impulse model. Pipes
1, 2 and 3 have 3, 4 and 4 pipe sections, respectively.
Because there is no loss at the J2 branch, the pressure
solution for Pipe 1, computing station 3 (i.e., 1,3) is the
same as both Pipe 2, computing station 4 (i.e., 2,4) and
Pipe 3 computing station 0 (i.e., 3,0). The flow solutions
at these three station will sum to zero at all times (in the
absence of vapor cavitation).
2/24/2003
1b.doc
1:35 PM
Page 150
Transient Control
Figure 5.17 shows the Transient Control window. At the top of the
window is an area called the Transient Solver Control area. This area
offers numerous input fields detailed in the following descriptions.
2/24/2003
1:35 PM
Page 151
2/24/2003
1b.doc
1:35 PM
Page 152
2/24/2003
1:35 PM
Page 153
2/24/2003
1b.doc
1:35 PM
Page 154
Surge tank liquid surface height, inflow and total liquid volume
Pump speed
You can select which junction data you want to save by selecting the
Junctions tab (Figure 5.17) and then selecting the junctions in the list.
2/24/2003
1:35 PM
Page 155
2/24/2003
1b.doc
1:35 PM
Page 156
3 in PVC Pipe, 3.5 in PVC Pipe and 4 in PVC Pipe were created by
first selecting the Base Scenario then clicking on Create Child. North
hydrant open and South hydrant open were, in turn, created by
selecting each of the above children scenarios for the three different pipe
2/24/2003
1:35 PM
Page 157
2/24/2003
1b.doc
1:35 PM
Page 158
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
2/24/2003
1:35 PM
Page 159
Output Control
Transient Control
Database Manager
2/24/2003
1b.doc
1:35 PM
Page 160
Figure 5.20 Attribute values and sources for the 3 in PVC Pipe
scenario in Scenario Manager
2/24/2003
1:35 PM
Page 161
Waterhammer Assistant
The Waterhammer Assistant can be opened from the Help menu or
Toolbar. The Waterhammer Assistant reviews your model and offers
helpful advice (see Figure 5.21).
2/24/2003
1b.doc
1:35 PM
Page 162
Workspace Find
The Find window utility shown in Figure 5.22 helps you find a particular
pipe or junction by its ID number and name. The Find utility becomes
more useful for models that expand over multiple Workspace pages.
You can access the Find utility by selecting Find on the View menu or
by clicking the button on the Toolbar. The Workspace is moved so that
the pipe or junction is located near the center of the screen.
Figure 5.22 The Find feature for the Workspace is helpful for
locating a specific pipe or junction. It is opened from
the View menu or Toolbar.
2/24/2003
1:35 PM
Page 163
Reverse Direction
Each pipe you create has a reference positive flow direction as indicated
by the arrow on the pipe. If you want to reverse the reference positive
flow direction, you can do so by manually moving the pipe endpoints.
However, this procedure is tedious. The Reverse Direction utility toggles
the reference positive flow direction for the selected pipe or pipes.
The Reverse Direction utility is accessed from the Arrange menu or from
the reverse direction button on the Toolbar.
Select Special
Select Special is a tool for selecting or deselecting objects based on
certain criteria (see Figure 5.23) and allows you to quickly select a group
of objects.
2/24/2003
1b.doc
1:35 PM
Page 164
Special Conditions
Some junctions have Special Conditions that alter the normal state of the
junction. For example, a valve can be closed, a pump turned off, or a
relief valve opened.
The special conditions are set in one of three ways:
1. By selecting the junction(s) on the Workspace and choosing Special
Conditions from the Edit Menu
2/24/2003
1:35 PM
Page 165
An X next to the
junction number
indicates a special
condition has been set
2/24/2003
1b.doc
1:35 PM
Page 166
2/24/2003
1:35 PM
Page 167
2/24/2003
1b.doc
1:35 PM
Page 168
2/24/2003
1:35 PM
Page 169
Merging models
Two models can be merged together using the Merge command on the
File menu. Heres how:
If there are duplicate numbers, the objects in the second model will be
changed. The second model will be selected (highlighted) so that you
can move it into position (be careful not to click off the selected portions
or the second model may be deselected).
Tip: To move a selected group of objects, use a junction icon, instead of
a pipe, to drag the group. This will avoid missing the pipe and clicking
on an empty area of the Workspace which will deselect your group.
Merging models with multiple scenarios
If the model you are merging from has multiple scenarios, only the base
scenario can be merged. If you want to merge one of the children of that
model, do the following:
1. Open the model which you want to merge.
2. Open the Scenario Manager.
3. Select the child scenario you want to merge.
4. On the Other Actions button, choose the Save Scenario to File
Without Children.
5. Enter a name and click OK. This process creates a new Impulse
model with the desired scenario as the base scenario.
6. Open the original model and merge in the new model file created in
Step 5.
Print Preview/Special
The Print Preview/Special window shown in Figure 5.25 allows you to
access special print features such as Print Preview and Fit to One Page.
It is accessed from the File menu.
2/24/2003
1b.doc
1:35 PM
Page 170
2/24/2003
1:35 PM
Page 171
Batch mode
AFT Impulse models can be run in batch mode, a feature that is
especially useful for running a group of models during lunch or
overnight.
To run models in batch mode, follow these steps:
1. Create a text file that lists the file names of all models to be run.
This is called the batch list file.
2. Either copy the input model files to the same directory as the batch
list file, or specify the complete path name for the input model files.
Include the disk drive letter in the full path name. If the models are
2/24/2003
1b.doc
1:35 PM
Page 172
2/24/2003
1:35 PM
Page 173
Print Format
With either the Model Data or Output window active, you can open Print
Format from the View menu (see Figure 5.26), allowing you to specify
the content and format of Model Data and Output reports.
You can include Model Reference Information (entered in the Output
Control window) to document your model or make important comments
you want kept with the model.
The various tables in the General section, Pipe section, and Junction
section can be selectively printed. All Pipe and Junction Notes are
shown in the Model Data General section and can be printed at your
option.
Empty columns in the three Model Data tables can be ignored in the
printed report. You can also have each of the Print Content sections start
on a new page. Both of these are formatting options in the Print Format
window.
Figure 5.26 Print Format window specifies the print content of the
Model Data or Output window. It is opened from the
View menu or Toolbar.
2/24/2003
1b.doc
1:35 PM
Page 174
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.
Steady-state solver
The Steady-State Solver obtains a balanced steady-state solution to the
system being modeled. It employs a Newton-Raphson matrix method
that is iterative nature. An initial guess to the steady-state solution is
made, and the Newton-Raphson method progresses the guess towards a
final solution that agrees with the governing equations within some
tolerance. Details are given in Chapter 8.
Control of the Steady-State Solver is given in the Steady Solution
Control window opened from the Analysis menu. Here the user can
specify items such as tolerances, relaxation and matrix methods.
The Steady-State Solver is based on that in AFT Fathom, a
commercial steady-state modeling software developed by Applied Flow
Technology. AFT Impulse contains most of the capability found in AFT
2/24/2003
1:35 PM
Page 175
Transient solver
The Transient Solver models the surge transients in a pipe system using
the Method of Characteristics (MOC). Unlike the iterative NewtonRaphson matrix method used for steady-state solutions, the MOC is a
direct solution method that marches in time. Details are given in Chapter
9.
To function properly, the MOC requires an accurate steady-state solution
to provide the initial conditions. Thus when one runs a transient
simulation in AFT Impulse 3.0, the Steady-State Solver is always run
first. The steady-state solutions are then automatically passed on to the
Transient Solver, which uses them as the initial conditions for the MOC.
This will reduce artificial transient problems to a bare minimum.
Artificial transients
The fundamental cause of artificial transients is mismatched steady-state
conditions. One would hope that artificial transients would never occur
in AFT Impulse 3.0, but that is not the case. If they do occur, there is one
of two causes. First, it may be an indication of a problem with the steady
solution. This could be a problem in the Steady-State Solver, or an input
problem that could not be identified when the model was built.
2/24/2003
1b.doc
1:35 PM
Page 176
2/24/2003
1:35 PM
Page 177
2/24/2003
1b.doc
1:35 PM
Page 178
CHAPTER 6
2/24/2003
1b.doc
1:35 PM
Page 179
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:
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.
2/24/2003
1b.doc
1:35 PM
Page 180
2/24/2003
1:35 PM
Page 181
2/24/2003
1b.doc
1:35 PM
Page 182
2/24/2003
1:35 PM
Page 183
Pipe Model
The Pipe Model tab (Figure 6.1a) allows you to input geometric data for
the pipe.
2/24/2003
1b.doc
1:35 PM
Page 184
2/24/2003
1:35 PM
Page 185
2/24/2003
1b.doc
1:35 PM
Page 186
2/24/2003
1:35 PM
Page 187
2/24/2003
1b.doc
1:35 PM
Page 188
Additional Losses
The additional loss factor is one way to enter loss factors in your model.
The additional loss factor is added to the friction loss calculated by the
Steady-State Solver. This is useful when a pipe section contains elbows
and other fittings.
You can use any of the losses supplied in the Impulse database or you
can enter a specific loss factor. To enter losses, click the Additional
Losses tab (Figure 6.1b) and then click the Additional Losses button to
open the Additional Pipe Losses window (Figure 6.2). This window
allows you to specify losses from a table of standard losses.
Because many of the additional losses are dependent on pipe diameter,
you must specify a diameter before you can open the Additional Pipe
Losses window.
2/24/2003
1:35 PM
Page 189
Figure 6.2
2/24/2003
1b.doc
1:35 PM
Page 190
Design Alerts
A design alert (Figure 6.1c) allows you to specify certain maximum
and/or minimum parameters for the pipe. After running the Steady-State
or Transient Solvers, any exceeded value will be identified in the
Warnings section of the Output window and the specific cell containing
the value in the Pipes section will be highlighted. Design Alerts can be
cross-plotted in the Graph Results window vs. the actual results. One
good example where this is useful is for cross-plots of maximum and
minimum allowed operating pressures.
Optional input
There are several optional input parameters (Figure 6.1d). These are
described in the following sections.
Initial Flow Rate Guess
The Initial Flow Rate Guess is the initial guess at the mass flow solution
through the pipe. Specifying a good initial guess of the flow solution
decreases the number of iterations the Steady-State Solver must execute
in order to arrive at a converged solution. However, specifying initial
guesses takes time and good judgment.
If you do not specify an initial flow guess, AFT Impulse implements a
proprietary method of generating first guesses. This method is usually
sufficient to get the Steady-State Solver going in the right direction so a
converged solution results. However, the method of generating first
guesses is not 100% effective, and there may be times when you will
have to specify the first guess in order to obtain a solution. More
discussion on this subject is given in Chapter 8.
2/24/2003
1:35 PM
Page 191
2/24/2003
1b.doc
1:35 PM
Page 192
2/24/2003
1:35 PM
Page 193
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.
2/24/2003
1b.doc
1:35 PM
Page 194
2/24/2003
1:35 PM
Page 195
Junction Type
Connecting
Pipes
Custom
Database
Window
Format
1/1-25
2-10
Check Valve
Control Valve
Dead End
Area Change
Assigned Flow
Assigned Pressure*
Branch
Gas Accumulator
1-25
Special
Conditions
General Component
Liquid Accumulator
1-25
Pump
1-2
Relief Valve
1-2
Reservoir
1-25
Spray Discharge
1-4
Surge Tank
1-25
Tee/Wye**
1-2
Valve
1-2
Volume Balance
2/24/2003
1b.doc
1:35 PM
Page 196
Format #1:
Junctions with one or two connecting pipes
The first basic window format is for junction types that allow only one
or two connecting pipes (see Figure 6.3). While junctions do not have an
explicitly defined flow direction like pipes do, typically those with two
pipes or less adopt a direction from the connecting pipes. For these
junction types, upstream and downstream pipes are recognized by AFT
Impulse based on the reference positive flow direction of the connecting
pipes. The upstream and downstream pipes are displayed separately near
the top of the window.
In this first basic window type, it is generally important to have the
pipes' reference positive flow directions specified in the physically
correct directions. A good example of this is a Pump junction which
will add pressure to the system in the direction of positive flow through
the connecting pipes. The Pump junction interprets where to add the
pressure based on the directions of the connecting pipes.
To inspect the connected pipe information in this window format,
position the mouse pointer over the connected pipe ID and hold down
the right mouse button.
2/24/2003
1:35 PM
Page 197
Figure 6.3
Format #2:
Junctions with more than two connecting pipes
The second basic window format for junction Specifications windows is
for junctions that allow more than two connecting pipes (see Figure 6.4).
These junctions typically allow up to twenty-five connecting pipes; the
exception is the Tee/Wye junction, which allows only three.
Because the number of pipes connected to a junction may vary, the
second basic format uses a table sized according to the number of
connecting pipes.
An example of this second window format is the Branch Specifications
window, shown in Figure 6.4, which has three connecting pipes. To
determine AFT Impulse's interpretation of the model connectivity for
this second window format, you can review the contents of the table
areas. There are separate loss factors for flow going into the pipe from
the junction and flow going out of the pipe into the junction. The
2/24/2003
1b.doc
1:35 PM
Page 198
Junction Number
Each junction has an ID number (see Figures 6.3 and 6.4) that you are
free to change. However, no two junction numbers can be the same, and
all junction numbers must be greater than zero and less than 10,000.
Duplicate junction numbers are not accepted. The junction numbers you
choose are arbitrary and have no bearing on model connectivity,
direction, or layout. They are merely convenient identifiers.
Junction Name
Each junction can be assigned a name for reference purposes (see
Figures 6.3 and 6.4). The default name is the junction type. The name
does not need to be unique. The name can be displayed on the
Workspace, Visual Report, Model Data and Output windows.
2/24/2003
1:35 PM
Page 199
Figure 6.4
Junction Elevation
Each junction has an entry for the elevation of the junction (see Figures
6.3 and 6.4). Next to each elevation entry is a set of units. For junctions
that use Format #1, the default inlet elevation is entered by the user, and
the outlet elevation is assumed to the same. A different outlet elevation
can be entered by clearing the Same as Inlet checkbox. Different inlet
and outlet elevations will have no effect on the flow solution, but will
offset the local static pressure at the connecting pipe by the hydrostatic
pressure difference.
Junctions that follow Format #2 have a single elevation and offsets from
that elevation (if they exist) for each individual connecting pipe entered
in the connecting pipe table.
Junction elevations are used to account for pressure changes due to
gravity and thus to allow calculation of absolute pressures in pipes. In a
2/24/2003
1b.doc
1:35 PM
Page 200
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.
Pipe connectivity
In each junction Specifications window (see Figures 6.3 and 6.4), you
can see AFT Impulse's interpretation of your system's connectivity. All
junctions must have the proper number of connecting pipes in order to
be completely defined (refer to Table 6.1).
2/24/2003
1:35 PM
Page 201
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
Junction Type
Type of Transient
Assigned Flow
Flowrate
Assigned Pressure
Pressure
Branch
Check Valve
Opening/Closing profiles
Control Valve
Pump
Relief Valve
Opening/Closing profile
Reservoir
Spray Discharge
Surge Tank
Surface Pressure
Valve
The active transient data is entered on the Transient Data tab (Figure
6.5). Table 6.2 lists the type of data that can be entered for each junction
type. If a junction does not have any active transient behavior, there is no
need to enter any data on the Transient Data tab.
In most cases, transient data can be entered as absolute values or as a
percentage of steady-state. The distinction is made when choosing
between Absolute Values and Relative to Steady-State Value (Figure
6.5).
2/24/2003
1b.doc
1:35 PM
Page 202
Figure 6.5
Repeat Transient
If the transient data is periodic, you can enter the data for one cycle of
the period and then tick the check box for Repeat Transient. This will
cause the one cycle of transient data to be repeated once it has reached
the end. The repetition will continue until the end of the simulation.
2/24/2003
1:35 PM
Page 203
2/24/2003
1b.doc
1:35 PM
Page 204
Figure 6.6
2/24/2003
1:35 PM
Page 205
Special Conditions
Some junctions allow you to set Special Conditions (see example in
Figure 6.6). This feature allows you to override the default behavior of
the junction. For example, the special condition for a Valve junction is
to close the valve. The special condition for a Relief Valve is for it to be
open no matter what the pressure. Not all junction types support special
condition settings.
The junction for which the Special Condition is set will have an X
placed before the junction ID number. If the special condition is to close
the junction, by default it will be outlined with a dashed line. These
settings may be changed in the Workspace Preferences window.
Chapter 5 discusses how to apply transient data to junctions with steadystate Special Conditions.
The Vacuum Breaker Valve, Gas Accumulator, Liquid Accumulator,
Surge Tank, and Relief Valve junctions have a unique kind of Special
Condition called Ignore. This makes the steady-state and transient
solvers ignore the presence of the valve. Thus not only is the data
ignored, the junction itself is ignored. This is convenient when locating
one of these devices by trail and error. Rather than having to delete the
junction, it can be ignored.
Design Factor
You can specify a design factor for each junction (see example in Figure
6.6). Design factors allow you to specify multipliers on the junction
pressure loss. These multipliers are applied by the Steady-State and
Transient Solvers during the solution process. In the case of a pump, the
2/24/2003
1b.doc
1:35 PM
Page 206
Notes
Each junction can have associated notes. This is useful for listing
assumptions, reference documents, drawings, measurements, etc. Notes
are entered in the provided area on the Notes tab.
2/24/2003
1:35 PM
Page 207
Status
The Status tab shows which required input data has not been entered.
Figure 6.7
2/24/2003
1b.doc
1:35 PM
Page 208
2/24/2003
1:35 PM
Page 209
Figure 6.8
The Assigned Flow Specifications window follows the first of the two
basic Specifications window formats, displaying the connecting pipe
information in a fixed format. Depending on whether you specify the
junction as an inflow or outflow type, the required connecting pipe
(upstream/inlet or downstream/outlet) will be enabled while the other
pipe is disabled. Note that if the flow is specified as Inflow (to the
system) then the reference positive flow direction of the connected pipe
must be away from the junction (i.e. the junction is upstream of the
pipe).
Tip: If you see that the required pipe in the Specifications window states
None but a pipe is connected to the junction graphically, then check
the Inflow/Outflow setting on the junction or the reference positive flow
direction of the pipe.
2/24/2003
1b.doc
1:35 PM
Page 210
Transient Data
On the transient data tab a flowrate transient can be entered. A special
feature is the ability to model a sinusoidal flowrate transient. Input for
the amplitude and frequency is required. This transient is summed to that
specified in the transient data table. A chopped sine wave takes the
absolute value of the sine function.
More information on transient data, including event transients, is given
earlier in this chapter and in Chapter 10.
Special Conditions
You can set a Special Condition for an Assigned Flow junction, which
will turn the flow off and make it act like a Dead End during steadystate.
2/24/2003
1:35 PM
Page 211
Figure 6.9
2/24/2003
1b.doc
1:35 PM
Page 212
The Assigned Pressure junction type has much in common with the
Reservoir junction. In each case you specify parameters in order to
achieve a known pressure. However, the Assigned Pressure junction
allows you to specify either stagnation or static conditions. This is useful
if, for example, you are modeling a system where the pressure in a pipe
is known and the location is being used as a boundary in the model. If
the measured conditions are for a location with a velocity, then they
represent static conditions and the choice of static properties should be
selected. In this case, the choice of a Reservoir junction would be
inappropriate. Similar to the Reservoir, an Assigned Pressure junction
causes the rest of the system to distribute the flow in a manner consistent
with the defined pressure. You can specify the units for pressure by
selecting from the adjacent drop-down list.
The distance of each connecting pipe (if any difference exists) from the
elevation can be entered in the pipe table.
2/24/2003
1:35 PM
Page 213
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.
2/24/2003
1b.doc
1:35 PM
Page 214
Figure 6.11a
The Branch Specifications window follows the second of the two basic
Specifications window formats. A table on the Loss Coefficients tab
displays the connecting pipe information. This table grows in size to
accommodate up to twenty-five connecting pipes. After you add a fifth
pipe a scroll bar appears, allowing you to review and enter loss factors
for all pipes in the table.
For each connecting pipe, the pipe table lists the reference flow direction
and up to two loss factors. The first loss factor is for physical flow out of
the pipe and into the branch. The second is for physical flow into the
pipe and out of the branch. The loss factors can be specified
independently or left as zero. To edit within the pipe table, simply click
in the appropriate column and row.
The distance of each connecting pipe (if any difference exists) from the
elevation can be entered in the pipe table.
2/24/2003
1:35 PM
Page 215
Figure 6.11b
Transient Data
On the transient data tab a transient source or sink flowrate can be
entered. More information on transient data, including event transients,
is given earlier in this chapter and in Chapter 10.
2/24/2003
1b.doc
1:35 PM
Page 216
A check valve is a device that allows flow in only one direction. AFT
Impulse assumes that the Check Valve is initially open. If the flow
solution indicates that forward flow will not occur, AFT Impulse closes
2/24/2003
1:35 PM
Page 217
Specifying losses
You specify the losses for the Check Valve on the Valve Model tab in
the Specifications window. For convenience, you can specify the
constant loss characteristics of a Check Valve as a valve coefficient (CV)
or as a loss factor (K). Chapter 8 details the relationship between K and
CV.
Transient Data
The closing of a check valve is a complicated transient process because
the mechanical movement is coupled to the hydraulics of the pipe.
Sophisticated models exist in the literature which attempt to model this
complicated process, but these require extensive data about the valve
and typically more simplified models are employed.
AFT Impulse offers a more simplified model, whereby the user can
specify a closing profile transient of Cv or K vs. time. Remember that
when using the K loss model, a value of 1 (negative one) means the
valve is closed. The closing will occur as soon as the closing velocity
criteria is met (discussed above). If no closing transient profile is
entered, the valve will close instantaneously.
You can also specify an opening profile of Cv or K vs. time, which is the
profile the valve will follow if it reopens. If no opening transient profile
is entered, the valve will open instantaneously.
2/24/2003
1b.doc
1:35 PM
Page 218
2/24/2003
1:35 PM
Page 219
2/24/2003
1b.doc
1:35 PM
Page 220
Special Conditions
The Special Condition for a control valve is dependent on the Action If
Control Is Lost setting. If the setting is Always Control (Never Fail) or
Fail Closed, then the Special Condition for the control valve is to close.
If the setting is to Fail Open, then the Special Condition will be to open.
2/24/2003
1:35 PM
Page 221
Transient Data
The Control Valve junction is the only one with two transient tabs. The
first is the valve reaction to a failure, and the second is to the control
point. More information on transient data, including event transients, is
given earlier in this chapter and in Chapter 10.
Failure Transient
Control valves have what might be called inherent event logic. That is,
rather than respond to events at some other location, the failure
transients respond to events of the valve itself. Specifically, the events
are the valve's loss of control and possibly regaining of control.
The failure transient is the valve action when it loses control. If the valve
is set to close upon failure, the closure transient will be used. This is a K
or Cv table vs. time. If the valve were to reopen, the opening transient
would be used. The time scale for the closing or opening of a valve is
relative to when the failure occurs. That is called time zero in a relative
sense.
You can elect to not enter data for the closing or opening transient. If the
valve loses control, the valve action is assumed to be instantaneous. If
the failure action is to fully open, then the closing transient data will
have no meaning and not be used. In a sense the opening transient will
have meaning, but in reality when the valve loses control it is already
2/24/2003
1b.doc
1:35 PM
Page 222
2/24/2003
1:35 PM
Page 223
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.
2/24/2003
1b.doc
1:35 PM
Page 224
2/24/2003
1:35 PM
Page 225
2/24/2003
1b.doc
1:35 PM
Page 226
PA
Orifice
LC
m& A
Connector
Pipe
2/24/2003
1:35 PM
Page 227
2/24/2003
1b.doc
1:35 PM
Page 228
2/24/2003
1:35 PM
Page 229
2/24/2003
1b.doc
1:35 PM
Page 230
2/24/2003
1:35 PM
Page 231
Pump Model
The basic definition of the pump is performed on the Pump Model tab.
You can model a pump with a pump curve or an assigned flow. When
assigning a flow, the total pressure/head developed by the pump is
calculated. This is useful for modeling positive displacement pumps.
The pump operating speed can be entered as a percent of full speed. It is
assumed that the pump curve is entered at full speed. You can use a
speed that is lower or higher than full speed. Impulse will adjust the full
speed pump curve using standard affinity laws relating speed ratios to
pressure rise ratios and flow rate ratios.
2/24/2003
1b.doc
1:35 PM
Page 232
2/24/2003
1:35 PM
Page 233
Pump configurations
Pump data can be entered for multiple configurations. This is available
only for pumps modeled as pump curves. The default is a single
configuration.
A pump configuration is a pump with a specific impeller trim and
operating speed. Multiple impeller trims and operating speeds can be
specified as part of the pump, then a particular combination can be
chosen.
Figure 6.21 shows the Pump Configuration window. Raw data can be
entered, imported from file, or pasted from the clipboard. Then a
polynomial curve can be fitted to the data. Data for head, NPSH, and
efficiency or power can be entered. Also data for the pumps end of
curve and NPSH constant for variable speed can be entered.
If multiple configurations are entered, they are displayed on the Pump
Specifications window in dropdown lists for selection (see Figure 6.22).
2/24/2003
1b.doc
1:35 PM
Page 234
2/24/2003
1:35 PM
Page 235
2/24/2003
1b.doc
1:35 PM
Page 236
Multiple configuration
selections displayed
2/24/2003
1:35 PM
Page 237
Transient Data
As discussed previously, pumps can be modeled as a pump curve or as
an assigned flow. Using assigned flow pumps in steady flow modeling is
useful for sizing a centrifugal pump. In AFT Impulse, however, that is
not appropriate during transient modeling. When modeling a centrifugal
pump transient, a pump with a pump curve is essential. Positive
displacement pumps act as constant flow devices, and the assigned
flowrate feature allows one to adequately model such pumps.
Pumps with curves
Once a pump curve is entered, several options are available for transient
modeling. Modeling a pump with a curve is usually associated with a
centrifugal pump.
No transient from pump, no back flow
If a pump remains at 100% speed during a transient, and the flow is
always positive, then no transient model is required. The appropriate
Transient Model selection is therefore None. In such a case, the pump
performance will be based on the provided head curve which only
applies to positive flow. Of course, if there is a check valve after the
pump which prevents backwards flow, this is the best choice.
No transient from pump, but back flow possible
If a pump remains at 100% speed during a transient, but backwards flow
is possible, the head curve data is insufficient. Therefore four quadrant
2/24/2003
1b.doc
1:35 PM
Page 238
2/24/2003
1:35 PM
Page 239
2/24/2003
1b.doc
1:35 PM
Page 240
2/24/2003
1:35 PM
Page 241
Figure 6.24 Pump with four quadrant data for pump trip modeling
2/24/2003
1b.doc
1:35 PM
Page 242
2/24/2003
1:35 PM
Page 243
2/24/2003
1b.doc
1:35 PM
Page 244
2/24/2003
1:35 PM
Page 245
Special Conditions
Pump junctions in AFT Impulse have two types of Special Conditions.
The first is to turn the pump off and have no flow through it. This is
called "Pump Off No Flow". The second type of Special Condition turns
the pump off but allows flow to go through the pump. This is called
"Pump Off With Flow Through".
2/24/2003
1b.doc
1:35 PM
Page 246
Viscosity corrections
Viscosity corrections to standard pump curves have been published in
the Hydraulic Institute Standards. The method, based on the 14th
edition, 1983, is employed. The method itself consists of a semilogarithmic diagram relating flow rate, head rise and viscosity. The
details of this diagram have been digitized into numerical relationships
inside AFT Impulse 3.0. The corrections to head rise, flow rate and
2/24/2003
1:35 PM
Page 247
2/24/2003
1b.doc
1:35 PM
Page 248
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.
2/24/2003
1:35 PM
Page 249
Transient Data
AFT Impulse allows the user to specify an opening profile transient of
Cv or K vs. time. Remember that when using the K loss model, a value
of 1 (negative one) means the valve is closed. The opening will occur
as soon as the cracking pressure is exceeded. If no opening transient
profile is entered, the valve will open instantaneously.
You can also specify a closing profile of Cv or K vs. time, which is the
profile the valve will follow if it closes. If no closing transient profile is
entered, the valve will close instantaneously.
The zero time for the closing transient is relative to when the valve starts
to close because of backflow. For the opening transient, it is relative to
when the valve starts to open.
More information on transient data is given earlier in this chapter and in
Chapter 10.
2/24/2003
1b.doc
1:35 PM
Page 250
Figure 6.28a
2/24/2003
1:35 PM
Page 251
If a pipe returns to the reservoir above the liquid surface, the depth is
entered as a negative number, or the elevation is specified higher than
that of the liquid surface. Pipes that discharge above the liquid surface
are assumed to have liquid free fall to the liquid surface. AFT Impulse
applies the proper boundary condition for pipes with negative depths. If
the pipe is above the liquid surface, the only appropriate condition is for
the fluid to be flowing from the pipe into the reservoir. Fluid cannot flow
from the reservoir into the pipe as the pipe is above the liquid surface. If
this occurs, AFT Impulse will assume the fluid flowing into the pipe is
the same as the reservoir fluid, solve the system, and then give a warning
in the output.
2/24/2003
1b.doc
1:35 PM
Page 252
Transient Data
On the transient data tab the surface elevation can be changed with time.
More information on transient data, including event transients, is given
earlier in this chapter and in Chapter 10.
2/24/2003
1:35 PM
Page 253
The exit pressure to which the Spray Discharge junction discharges must
be entered.
Two geometries can be modeled. The first is a normal, one hole spray,
where the discharge flow area of the hole is entered. The second is a
sparger which, in principle, works in the same way as a spray. For the
sparger, the flow area of a single hole and the number of holes are
entered. This assumes that the holes are hydraulically close in proximity
and are the same area.
Loss Model
The discharge coefficient for the junction is entered in the appropriate
box. The physical area of the exit flow can also be entered. These two
parameters are used to calculate the discharge flow rate, which depends
on the difference between the internal pressure in the system and the
specified exit pressure.
2/24/2003
1b.doc
1:35 PM
Page 254
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.
2/24/2003
1:35 PM
Page 255
Tank geometry
The Tank Height is the physical top of the tank in relation to the
elevation. If there is a Connector Pipe, then the Tank Height is relative
to the top of the Connector Pipe. This height is used in the case of
overflow. If the surge tank overflows, spillage will occur and the liquid
will exit the pipe system not to be recovered. If you do not want to
model the overflow, specify a very tall tank or do not enter a Tank
Height.
If the Tank Cross-Sectional Area is specified as Constant, the surge tank
is assumed to be a vertical tank with constant cross-sectional area. If the
tank cross-sectional area changes with height, this can be entered by
selecting the variable option and entering the cross-sectional variation in
the table.
The Tank Cross-Sectional Area is used for the purpose of determining
the liquid volume. It will determine how high the liquid will rise or drop,
since a given volume of liquid will flow into the tank during a time step.
2/24/2003
1b.doc
1:35 PM
Page 256
LL
LC
Orifice
m& T
Connector
Pipe
2/24/2003
1:35 PM
Page 257
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.
2/24/2003
1b.doc
1:35 PM
Page 258
Figure 6.32 Surge Tank junctions allow the user to specify an liquid
level for special steady-state behavior.
2/24/2003
1:35 PM
Page 259
2/24/2003
1b.doc
1:35 PM
Page 260
LA
LL
LC
Orifice
m& T
Connector
Pipe
Loss factors
There are two Loss Models available for Tee/Wye junctions: Simple and
Detailed. If you choose the Simple type of Tee/Wye, no loss factors will
2/24/2003
1:35 PM
Page 261
The Detailed type of Tee/Wye will calculate the loss factors as they
depend on the flow split as well as the geometry. Because of the
dependence on the flow split, loss factors usually cannot be calculated
before performing an analysis. AFT Impulse incorporates the best
available models to calculate tee and wye loss factors during the actual
solution. These calculations account for the percentage of the flow split
and the angle of connection. This offers a significant advantage over
hand calculated losses.
If you have specific loss factors that you want to use for a tee/wye
junction, you should use a Branch junction instead.
The Tee/Wye junction functions the same during the steady-state and
transient.
2/24/2003
1b.doc
1:35 PM
Page 262
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
2/24/2003
1:35 PM
Page 263
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.
2/24/2003
1b.doc
1:35 PM
Page 264
2/24/2003
1:35 PM
Page 265
2/24/2003
1b.doc
1:35 PM
Page 266
2/24/2003
1:35 PM
Page 267
Figure 6.38 The Global Pipe Edit window allows you to change data
for many pipes at the same time
2/24/2003
1b.doc
1:35 PM
Page 268
Figure 6.39a
2/24/2003
1:35 PM
Page 269
Figure 6.39b
2/24/2003
1b.doc
1:35 PM
Page 270
CHAPTER 7
2/24/2003
1b.doc
1:35 PM
Page 271
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
2/24/2003
1b.doc
1:35 PM
Page 272
Unit Preferences
You can set the units you prefer to use in AFT Impulse on the Unit
Preferences tab (see Figure 7.2). These units are selected as a first choice
in the unit selection boxes. You can choose any of the units available. To
set a new default, choose the unit type from the list on the right and click
the Set As Preferred Unit button.
2/24/2003
1:35 PM
Page 273
Figure 7.2
You also can customize the units used by AFT Impulse. For each
parameter, AFT Impulse can use many different units. However, if only
a few are ever typically used, then the others can be removed. This will
decrease the number of items in the unit selection boxes, making it easier
to choose the one desired. If a preferred unit is removed from those
available to Impulse, a new one will have to be chosen before leaving
this area. You also must have one set of units available for each
parameter.
To customize the units, first choose the parameter to modify on the list
on the left. Then, among the unit types on the right, check or uncheck the
boxes for the units you want or do not want to see.
2/24/2003
1b.doc
1:35 PM
Page 274
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 well discuss how databases are configured and
administered through your local or wide area network.
2/24/2003
1:35 PM
Page 275
Command buttons
There are eight buttons at the bottom of the Parameter and Unit
Preferences window. Impulse has built-in default parameters, units and
settings which you can choose by clicking the Impulse Default button.
You can also develop your own settings, tailored to your project or
industry, and have these used by default (instead of Impulses defaults).
To make your own default, first select the parameters, units and settings
you would like to use, then click the Set As Default button. Your settings
will be saved and will be used each time any new project is initiated. If
you make changes to the settings, and want to get back to your defaults,
click the User Default button. The default settings are updated only
when you click Set As Default.
You can save the parameter and unit settings to a file by pressing the
Save Preferences button and entering a file name. These setting are
loaded again by pressing the Recall Preferences button and choosing the
file name. The format files you create can be placed on a network for
sharing among a group or company, or incorporated into a companywide database, allowing standardized reporting.
If you have made changes that you dont want to keep, click the Cancel
button. Click OK to use the settings you have defined.
Workspace Preferences
The Workspace Preferences window is accessed from the Options menu.
The features offered in this window allow you to customize the
appearance and behavior of the Workspace.
2/24/2003
1b.doc
1:35 PM
Page 276
Figure 7.3
2/24/2003
1:35 PM
Page 277
2/24/2003
1b.doc
1:35 PM
Page 278
Display Options
From the Display Options tab you can change the symbols used, text
content, and location of the text on the Workspace (see Figure 7.4).
Workspace grid
A grid can be displayed on the background for modeling convenience.
The grid can be fine, medium or coarse. The grid lines can be solid or
dashed. And finally, the pipes and junctions can be snapped to the grid
you specify.
2/24/2003
1:35 PM
Page 279
Figure 7.4
2/24/2003
1b.doc
1:35 PM
Page 280
2/24/2003
1:35 PM
Page 281
2/24/2003
1b.doc
1:35 PM
Page 282
Figure 7.5
Sample Workspace
The tab for Sample Workspace gives you sample of how the Workspace
pipes and junctions will appear if you click the OK button.
Database connections
In the lower left of the Workspace Preferences window there is a
Database checkbox. If this box is checked, your Workspace Preferences
parameters are set up as determined by the database to which you are
connected. This is referred to as an active database. To make it inactive,
uncheck the box or change one of the Workspace Preferences settings
controlled by the database.
2/24/2003
1:35 PM
Page 283
Command buttons
There are eight buttons at the bottom of the Workspace Preferences
window. Impulse has built-in default parameters, units and settings
which you can choose by clicking the Impulse Default button. You also
can develop your own settings, tailored to your project or industry, and
have these used by default (instead of Impulses defaults). To make your
own default, first select the settings you would like to use, then click the
Set As Default button. Your settings will be saved and will be used each
time any new project is initiated. If you make changes to the settings,
and want to get back to your defaults, click the User Default button. The
settings are updated only when you click Set As Default.
You can save the parameter and unit settings to a file by pressing the
Save Preferences button and entering a file name. These setting are
loaded again by pressing the Recall Preferences button and choosing the
file name. The format files you create can be placed on a network for
sharing among a group or company, or incorporated into a companywide database, allowing standardized reporting.
If you have made changes which you dont want to keep, click the
Cancel button. Click OK to use the settings you have defined.
Toolbox Preferences
The Toolbox Preferences window is accessed through the Options menu.
Figure 7.6 shows the layout of the Toolbox Preferences window. The
features in this window allow you to customize the Toolbox in the
manner that is most productive for you.
Once you have changed the appearance and functionality of the Toolbox,
you can set your changes as the user default values by clicking the Set
2/24/2003
1b.doc
1:35 PM
Page 284
Figure 7.6
2/24/2003
1:35 PM
Page 285
Database connections
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.
If the checkbox is disabled, there is no connected database.
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 how AFT Impulse handles transient junction data that does
not cover the entire time range of the simulation (either maintain the
2/24/2003
1b.doc
1:35 PM
Page 286
Figure 7.7
2/24/2003
1:35 PM
Page 287
2/24/2003
1b.doc
1:35 PM
Page 288
Field #
Description
Required
Reference number
Yes
Name of component
Yes
Source
Yes
Diameter
No
Angle
No
Percent open
No
No
K factor
Yes
Page reference
No
2/24/2003
1:35 PM
Page 289
2/24/2003
1b.doc
1:35 PM
Page 290
Component
Bends
Valves
Reference number
Description
100000-100999
Standard
101000-101999
Mitre
102000-102999
Smooth Flanged
200000-200999
Angle
201000-201999
Ball
202000-202999
Butterfly
203000-203999
Cylindrical
204000-204999
Gate
205000-205999
Globe
206000-206999
Plug
207000-207999
Poppet
208000-208999
Three-way
300000-300999
Swing
301000-301999
Lift
302000-302999
Tilting Disk
303000-303999
Stop-check
Orifices
400000-400999
Sharp-edged
401000-401999
Long
500000-500999
Entrances
501000-501999
Exits
600000-600999
601000-601999
Honeycomb
602000-602999
Screen
603000-603999
Tee
Check Valves
Miscellaneous
There are two types of databases that may be customized: local and
external (Note: the AFT Default Internal Database is not included here
since it may not be altered). The local database is located in the
IMP_USER3.DAT file in the Windows directory.
In AFT Impulse you can create multiple external or local custom
databases. The Database Manager window on the Database menu
simplifies this process.
2/24/2003
1:35 PM
Page 291
Figure 7.8
2/24/2003
1b.doc
1:35 PM
Page 292
Figure 7.9
Density, viscosity and bulk modulus are required for all fluids. Vapor
pressure is required if cavitation modeling is desired.
2/24/2003
1:35 PM
Page 293
After you have entered a custom material, each time you open a Pipe
Specifications window the material will be available to you for quick
selection. This minimizes the need to work with handbooks.
2/24/2003
1b.doc
1:35 PM
Page 294
2/24/2003
1:35 PM
Page 295
2/24/2003
1b.doc
1:35 PM
Page 296
Database Manager
Accessed from the Database menu, Database Manager allows you to
2/24/2003
1:35 PM
Page 297
2/24/2003
1b.doc
1:35 PM
Page 298
2/24/2003
1:35 PM
Page 299
Editing databases
The Edit Database tab of Database Manager allows you to create and
delete databases, and edit their contents by copying sections or items
from one database to another. The tab is divided into a Source Database
area on the left and a Changing Database area on the right (see Figure
7.14).
To edit a database, first choose a source database by selecting one of the
four options:
2/24/2003
1b.doc
1:35 PM
Page 300
Figure 7.14 Edit Database tab allows you to select sections from a
specified database and add them to an existing or new
database.
The file and descriptive name of the selected database are displayed as
well as the sections available in that database, i.e. fluids, components or
pipe materials.
The database to be changed and what changes are to be made are
specified in the Changing Database portion of the Edit Database tab.
Delete section deletes from the listed database the section selected
from the Database Sections list
2/24/2003
1:35 PM
Page 301
2/24/2003
1b.doc
1:35 PM
Page 302
2/24/2003
1:35 PM
Page 303
Database
connection
2/24/2003
1b.doc
1:35 PM
Page 304
Save the control formats to separate files. This can be done on the
Output Control window, Visual Report control window, etc.
Copy and paste sections from one file into another file using a text
editor.
2/24/2003
1:35 PM
Page 305
Database Titles/Descriptions
IMPULSE3.DAT
IMP_USER3.DAT
VailMaterials.dat
Development Output Control Settings
Development Output.dat
Final Report Output.dat
My Report Output.dat
My VR Format.dat
This discussion will focus on the first and second methods. Lets assume
that the person responsible for the network database has already created
a new pipe material. Also, assume that the goal is to create a networkwide database to contain this material, as well as two Output Control
settings and a Visual Report format.
The information for the material is copied from the Local User Database
(IMP_USER3.DAT) to a new file using the Database Editor in the
Database Manager (see Figure 7.14). For the source database choose
Local User Database and then select Pipe Materials. Then click Create
New Database and enter a file name, VailMaterials.dat, for this example.
Then enter a description of the database. This description is how AFT
Impulse users will refer to the database and should concisely describe
what is in the database (for this example well use Vail Resort
Materials). Then click Add to copy the pipe materials from the Local
2/24/2003
1b.doc
1:35 PM
Page 306
2/24/2003
1:35 PM
Page 307
2/24/2003
1b.doc
1:35 PM
Page 308
2/24/2003
1:35 PM
Page 309
2/24/2003
1b.doc
1:35 PM
Page 310
CHAPTER 8
2/24/2003
1b.doc
1:35 PM
Page 311
HGL =
P
+z
g
(8.1)
m&
j =1
ij
=0
(8.2)
Pf = f
2/24/2003
1b.doc
L 1
2
V
D2
1:35 PM
Page 312
(8.3)
1
1
V12 + gz1 = P2 + V2 2 + gz2 + Pf
2
2
(8.4)
1
V 2
2
(8.5)
Solving for the frictional pressure drop for a constant area pipes yields:
Po,i Po, j + g ( zi z j ) = Pf
(8.6)
(8.7)
Combining Equations 8.3 and 8.6 and substituting for velocity, V, using
Equation 8.7 gives the mass flow for each pipe:
1/ 2
Po,i Po, j + g ( zi z j )
ij
R
= m& ij
(8.8)
where Rij is the effective flow resistance in the pipe and the subscript ij
refers to the pipe connecting junctions i and j.
f ij Lij
1
+ K ij
Rij =
Dij
2 A 2
ij
(8.9)
Note that R is not the same as R, which is the resistance used in the
Output Control window and is expressed by the following equation:
H = RQ 2
2/24/2003
1:35 PM
Page 313
Po,i Po, j + g ( zi z j )
ij
R
j =1
=0
(8.10)
Po,i Po, j + g ( zi z j )
ij
R
j =1
= m& i, Applied
(8.11)
Po,i Po, j + g ( zi z j )
Fi =
ij
R
j =1
m& i, Applied
(8.12)
The method involves finding all the junction stagnation pressures, Po,i ,
that cause all of the Fi to go to zero, thus satisfying Equation 8.11 at all
junctions.
When applied to a system of equations, the Jacobian matrix contains all
the required derivative information to employ the Newton-Raphson
technique. The Jacobian, JF, is given by:
2/24/2003
1b.doc
1:35 PM
Page 314
F1
F1 F1
P P L P
o, n
o,1 o,2
F2 F2
F
L 2
M
Fn Fn
F
L n
v
v
v
Po, new = Po,old J F1F
(8.13)
2/24/2003
1:35 PM
Page 315
Consider the reservoir in Figure 8.1. When the liquid in a reservoir flows
into connecting pipes, the static pressure immediately drops due to the
increase in velocity (see Figure 8.2).
Po,2
P2
Ploss
Pressure
1
V2 2
2
Location 2
Po,4
1
V42
2
Location 3
P4
Location 4
Distance (x)
Figure 8.2
2/24/2003
1b.doc
1:35 PM
Page 316
2/24/2003
1:35 PM
Page 317
2/24/2003
1b.doc
1:35 PM
Page 318
J5
P3
J1
J2
P1
J3
P2
P4
Figure 8.3
Now consider the systems in Figure 8.4. The first system appears to be
closed, while the second appears to be open. If the same boundary
condition (i.e., surface elevation and pressure) is used for J1, J11 and
J12 in the second system, to AFT Impulse it will appear as an identical
system to the first system. The reason is that AFT Impulse takes the first
system and applies the J1 reservoir pressure as a boundary condition to
pipes P4, P9 and P10. The second system uses three reservoirs to apply
boundary conditions to P4, P9 and P10. But if the reservoirs all have the
same elevation and pressure, the boundary conditions are the same as J1
in the first system. Thus the same boundary condition is used for P4, P9
and P10 in both models, and they appear identical to AFT Impulse.
2/24/2003
1:35 PM
Page 319
P9
J1
J9
P8
P6
J5
P4
P11
J4
J2
P1
P9
J8
J3
P3
P10
J11
J6
P7
J10
P2
J9
P8
J6
P7
J12
P4
J1
P6
J5
P10
P11
J4
J2
P1
Figure 8.4
J3
P3
J10
P2
But how is the flow balanced at J1 in the first system? Looking at the
first system, one sees that to obtain a system balance, whatever flows
into P10 must come back through P4 and P9. Because there is overall
system balance by the Steady-State Solver, it will give the appearance of
a balanced flow at the pressure junction J1. If there is only one boundary
(i.e., junction) where flow can enter or leave the pipe system, then no
2/24/2003
1b.doc
1:35 PM
Page 320
2/24/2003
1:35 PM
Page 321
Figure 8.5
2/24/2003
1b.doc
1:35 PM
Page 322
Roughness-based methods
Roughness-based methods use a pipe roughness value to calculate the
pressure drop. This can be in the form of an absolute roughness (which
has units of length) or a relative roughness that ratios the pipe roughness
to its diameter.
Laminar flow
For laminar flow AFT Impulse uses the standard laminar equation:
f =
64
Re
(8.14)
Note that the laminar friction factor does not depend on the user
roughness.
The System Properties window allows you to change the default laminar
transition Reynolds Number (see Chapter 5).
2/24/2003
1:35 PM
Page 323
9.35
f = 1.14 2 log +
D Re f
Hydraulically smooth
The friction factor in the turbulent range is calculated using ColebrookWhite with roughness = 0 while in the laminar range it continues to be
calculated using Equation 8.14 above.
Hazen-Williams method
AFT Impulse also offers the Hazen-Williams method of specifying
irrecoverable loss information. The Steady-State Solver converts the
Hazen-Williams factor to a Darcy-Weisbach friction factor (Walski
1984, 37). This allows a consistent solution approach to be used for all
pipe system models, while retaining the flexibility of the two approaches
to account for losses.
CHW =
17.25
(VD) 0.081
0.54
2/24/2003
1b.doc
1:35 PM
Page 324
dP 139.212 fQ 2 ( water )
=
L
D5
f =
Re
4 0.86859 ln
1964
.
ln( Re) 38215
.
Re =
10,059Q
d 2 818.56
where:
dP
= length (miles)
= density (lbm/ft3)
water
= diameter (inches)
Re
Q=
d 3(
water )( dP L)
log
4
.
35
+
0.5
2
5.9113( water )
d 2.5 ( dP L)
where:
dP
= length (miles)
2/24/2003
1:35 PM
Page 325
= density (lbm/ft3)
water
= diameter (inches)
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.
Vmax = K C
Vwater = 122
. C 1.4
If V < Vmax, then
dH
= FKV C D
L
where:
dH /L = head loss per length (m water / 100 m pipe)
C
Vmax
2/24/2003
1b.doc
1:35 PM
Page 326
Re =
f =
D 0.205V
C1.157
3.97
Re1.636
dH =
fV 2 LK
D
where:
Re = pseudo Reynolds Number (dimensionless)
D = inner diameter (feet)
V
2/24/2003
1:35 PM
Page 327
= 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.
P
Q =
2 KL
n (1+ 3n ) n
r
1 + 3n
(8.15)
where:
K
= dimensionless constant
Power Law flow is laminar. Substituting Equation 8.15 into 8.14 yields:
A 3+ n V
f =
K
2
m& D
D
n 1
1 + 3n
Sy
H
RV
= 32
+ c 2
L pipe
6gD gD
(8.16)
where:
Rc
= coefficient of rigidity
Sy
= yield stress
2/24/2003
1b.doc
1:35 PM
Page 328
gH = p = f
L1
V 2
D2
(8.17)
f =
64
V
S y Rc
6V + D
f
K non newtonian = K newtonian non newtonian
f newtonian
where the friction factor is for the upstream pipe.
Design factors
In each pipe you can specify a Design Factor for the pipe friction. This is
a multiplier that is applied to the friction factor calculated with the
preceding methods.
Flow relaxation
2/24/2003
1:35 PM
Page 329
Pressure relaxation
Maximum iterations
Matrix method
The default values for the solution control parameters perform well in
the majority of cases. Modify the default parameters only when
necessary or when you are comfortable with numerical convergence
issues.
2/24/2003
1b.doc
1:35 PM
Page 330
Pj , new Pj , old
Pj , new
Convergence = True
Else
Convergence = False
End If
Table 8.1 shows a tolerance example for pressures. Notice how relative
change does not have units, but absolute change has units (of psia).
The default in AFT Impulse is relative tolerance because experience
indicates it is the most robust.
2/24/2003
1:35 PM
Page 331
Iteration #
1
2
3
4
5
6
Pressure
(psia)
100.0
60.0
72.0
66.0
67.0
66.9
Relative
Change
0.6667
0.1667
0.0909
0.0149
0.0015
Abs olute
Change (psia)
40.00
12.00
6.00
1.00
0.10
Figure 8.6
2/24/2003
1b.doc
1:35 PM
Page 332
20.0000
12.0000
14.0000
13.5000
13.6000
13.5900
13.5913
13.5907
13.5909
Relative
Change
0.666667
0.142857
0.037037
0.007353
0.000736
0.000096
0.000044
0.000015
8.000000
2.000000
0.500000
0.100000
0.010000
0.001300
0.000600
0.000200
Meets Abs.
Tolerance ?
Meets Rel.
Tolerance ?
Meets Abs.
OR Rel.
Tolerance ?
Meets Abs.
AND Rel.
Tolerance ?
NO
NO
NO
NO
NO
NO
NO
YES
NO
NO
NO
NO
NO
YES
YES
YES
NO
NO
NO
NO
NO
YES
YES
YES
NO
NO
NO
NO
NO
NO
NO
YES
Relaxation
Another feature of iterative solvers is called relaxation. In short,
relaxation slows the Steady-State Solver's progress toward a solution and
in the process reduces the ability of poorly behaved (non-linear)
components to destabilize the solution. AFT Impulse implements a
proprietary adaptive relaxation scheme that reduces the relaxation
2/24/2003
1:35 PM
Page 333
Relaxation
Value
1
0.5
0.2
0.1
0.05
0
There are two relaxation parameters in AFT Impulse 3.0 (See Figure
8.7): Flow rate relaxation and pressure relaxation.
Flow Rate Relaxation
For poorly-behaved systems, a flow rate relaxation of 0.1 is advised.
Experience shows that using flow rate relaxation factors between 0.5 and
2/24/2003
1b.doc
1:35 PM
Page 334
Figure 8.7
2/24/2003
1:35 PM
Page 335
Maximum Iterations
The Maximum Iterations parameter in the Steady Solution Control
window determines when the Steady-State Solver considers the model
nonconvergent; it does not directly affect the solution. There may be
many reasons why a model does not converge. You want to be sure that
the Steady-State Solver does not continue indefinitely searching for a
solution that cannot be obtained because of an input error or ill-behaved
system.
Most properly defined models will converge in 50,000 iterations. For
larger models, more than 50,000 iterations may occasionally be required,
especially when relaxation is being used. AFT Impulse's default of
50,000 iterations is sufficient for most models.
Matrix Method
AFT Impulses network solution method requires solution of matrices
thousands of times. The default method is Gaussian Elimination, which
is very robust and usually provides the fastest convergence.
For larger systems, Gaussian Elimination with Pivoting and the LU
Decomposition methods can be better and are therefore provided.
2/24/2003
1b.doc
1:35 PM
Page 336
Figure 8.8
2/24/2003
1:35 PM
Page 337
2/24/2003
1b.doc
1:35 PM
Page 338
2/24/2003
1:35 PM
Page 339
Special junction
The Volume Balance junction is a special case. It acts like a pseudotransient junction where the mass flow rate is abruptly changed to
maintain a constant volume flow of two fluids with different densities.
2/24/2003
1b.doc
1:35 PM
Page 340
Junction Type
References
Bend
Area Change
Orifice
Tee/Wye
Valve
Ploss = K
1
V 2
2
(8.18)
2/24/2003
1:35 PM
Page 341
Design factors
Each junction that allows modeling of pressure losses allows input of a
Design Factor. The design factor is multiplied by the K factor
determined by methods discussed in this chapter.
Many junction types allow modeling of K factors as well as other
pressure loss methods. The Design Factor is also applied to these other
methods as a multiplier on the pressure loss.
Area change
Two standard area change loss geometries are available: the conical
transition and the abrupt transition.
The conical expansion correlation (Crane 1988, A-26) is:
Aup
K up = 2.6 sin 1
Adown
2
Aup
K up = 1
Adown
( < 45 degrees)
( > 45 degrees)
K up =
K up =
Adown
Aup
Aup
2
Adown
Aup
( < 45 degrees)
( > 45 degrees)
2/24/2003
1b.doc
1:35 PM
Page 342
Aup
K up = 1
Adown
K up =
A
0.51 down
Aup
Adown
Aup
0.75
Bend
90 degree bends
The three Bend loss correlations are all for turbulent Reynolds Numbers
(Crane 1988, A-29).
The K values for a smooth, flanged bend are provided in Table 8.5.
Table 8.5
r/d
r/d
20 f T
24 f T
1.5
14 f T
10
30 f T
12 f T
12
34 f T
12 f T
14
38 f T
14 f T
17 f T
16
42 f T
50 f T
2/24/2003
1:35 PM
20
Page 343
Nominal
Size
Friction
Factor f T
Nominal
Size
Friction
Factor fT
Nominal
Size
Friction
Factor fT
0.027
0.019
8-10
0.014
0.025
2 , 3
0.018
12-16
0.013
18-24
0.012
0.023
0.017
0.022
0.016
0.021
0.015
K = 30 f T
where fT is the turbulent friction factor given in Table 8.6.
The Mitre bend is given by:
K = 60 f T
Non-90 degree bends
The two non-90 degree bend loss correlations are for turbulent Reynolds
Numbers (Crane 1988, A-29).
A smooth, flanged bend is given by:
K = (n 1) 0.25f T
+ 0 .5 K + K
D
where n is the number of 90 degree bends, K is the loss factor for one 90
degree bend (Table 8.5), and fT is given by Table 8.6.
The Mitre bend is given by Table 8.7.
2/24/2003
1b.doc
1:35 PM
Page 344
2 fT
15
4 fT
30
8 fT
45
15 f T
60
25 f T
75
40 f T
60 f T
90
Valve
Standard valve loss models are used from Idelchik, Miller and Crane.
Orifice
The orifice loss factors are all for turbulent Reynolds Numbers. The
sharp-edged orifice shown in Figure 8.9, is given by the following
(Idelchik 1994, 218):
Aup
K up =
Aorifice
0.7071 Aorifice
Aup
Figure 8.9
0.375
Aorifice
+ 1
Adown
Sharp-edged orifice
2/24/2003
1:35 PM
Page 345
Aup
K up =
Aorifice
K 1 Aorifice
Aup
0.75
Aorifice
K 1 orifice
+21
Adown
Aup
Aorifice
+ 1
Adown
0.375
where:
K = 0.03 + 0.47 10
( 7.7r / Dorifice )
Tee/Wye
The loss factors calculated for tee and wye junctions involve
complicated correlations that depend on the flow split, the ratio of flow
areas, and the angle of the connecting pipes. The models used by AFT
Impulse are taken mostly from chapter 7 of Idelchik's Handbook of
Hydraulic Resistance (1994). A few of the more important equations
that AFT Impulse uses will be presented here, and references will be
given for the others.
2/24/2003
1b.doc
1:35 PM
Page 346
st = straight-through flow
Ac
Ast
As
Ast
s = side branch
Ac
As
2/24/2003
1:35 PM
Page 347
Diverging Case
Side branch (Idelchik 1994, 418):
V 2
Vs
s
K c, s =
= 1 + 2 cos
2
Vc
Vc / 2
Vc
Ps
2/24/2003
1b.doc
1:35 PM
Page 348
As / Ac
Qs / Qc
0.35
0.4
1.1 - 0.7 Qs / Qc
0.35
> 0.4
0.85
> 0.35
0.6
1.0 - 0.6 Qs / Qc
> 0.35
> 0.6
0.6
K c, st =
Q
= st s
Vc2 / 2
Qc
Pst
As / Ac
Qs / Qc
st
0.4
0 - 1.0
0.4
> 0.4
0.5
2 (2 Qs / Qc - 1)
> 0.4
> 0.5
0.3 (2 Qs / Qc - 1)
Converging Case
Side branch (Idelchik 1994, 417):
K c, s =
Ps
Vc2 / 2
Q A 2
A
= A1 + s c 2 c
Ast
Qc As
2
2
Qs
Ac Qs
1
2 cos
As Qc
Qc
Q
A Q
K c, st =
= 1 1 s 2 c s cos
As Qc
Vc2 / 2
Qc
Pst
2/24/2003
1:35 PM
Page 349
Type
Wye
Direction
Reference
Diverging tee
Branch
Diverging tee
Run
Converging tee
Branch
Converging tee
Run
Splitting tee
From branch
Combining tee
To branch
Diverging tee
Branch
Diverging tee
Run
Converging tee
Branch
Converging tee
Run
Splitting tee
From branch
Combining tee
To branch
Diverging tee
Branch
Diverging tee
Run
Converging tee
Branch
Converging tee
Run
Splitting tee
From branch
Combining tee
To branch
Diverging wye
Branch
Converging wye
Run
CV =
Pvalve ( 62.4 / )
2/24/2003
1b.doc
1:35 PM
Page 350
Kvalve
A2
= 1460 2
Cv
Q
P
K sprinkler =
where Q and P can be in the units of your choice. The preference in the
USA is gpm for flowrate and psid for pressure. In Europe it is frequently
liter/min for flow rate and bars for pressure. Note that Ksprinkler is not
dimensionless, but has units associated with it. Also, note that sprinkler
vendor values for Ksprinkler implicitly assume that water is the fluid, so
there is a built-in density of 62.3 lbm/ft3 or 1000 kg/m3. If this sprinkler
is used for something other than water, the Ksprinkler value will need to be
adjusted.
Also note that if the spray is into the atmosphere, as is typical, the P is
the same thing as the gage pressure. That is why the previous equation is
stated by some vendors as just P in the denominator, rather than P.
When stated as just P, the pressure is gage.
Note: The fire sprinkler K is not the same as the dimensionless K value
as given by Equation 8.18 and used extensively throughout AFT
Impulse.
CD A =
2/24/2003
water
1:35 PM
K sprinkler
Page 351
Pumps
Pumps can operate at variable speeds. Affinity laws (also called
homologous laws) allow an estimation of the pump performance at
speeds other than the design speed.
Before discussing the affinity laws, it should be noted that the affinity
laws are an approximation. In many cases pump manufacturers will have
performance data for a number of speeds. These data should be used if
available. When not available, the affinity laws can be used with an
understanding that an approximation is being made.
The affinity laws for pumps are as follows:
H1 s1
=
H 2 s2
Q1 s1
=
Q2 s2
(8.19)
where s is the speed. If the pump data is input as a polynomial, the speed
will affect the curve as follows:
2
4
2
H 2 = s 2 a + s 2b
Q2
Q
Q
Q
+ s 2 c 2 + s 2 d 2 + s 2 e 3
s
s
s
s
3
Q
Q
H 2 = s a + sbQ2 + cQ2 + d 2 + e 23
s
s
2
(8.20)
where H1 is the input curve that typically represents 100% speed and
H2 is the head rise at speed s. In this form, the speed s will be a decimal
and not a percent. That is, 50% speed will result in the appropriate s of
0.5. AFT Impulse determines this automatically in the Steady-State and
Transient Solvers.
If you model the pump as a controlled flow or discharge pressure, AFT
Impulse will backsolve Equation 8.20 for the speed s and include that in
the output Pump Summary.
2/24/2003
1b.doc
1:35 PM
Page 352
Viscosity
The dynamic viscosity in AFT Impulse is assumed to be a function of
temperature only. In general this is a good assumption, although systems
at very high pressure usually start to show some pressure dependence.
These properties are modeled as polynomial curve fits of temperature.
2/24/2003
1:35 PM
Page 353
2/24/2003
1b.doc
1:35 PM
Page 354
CHAPTER 9
Description of waterhammer
Please see Chapter 3 for a summary description of waterhammer in a
simple one-pipe system.
Instantaneous waterhammer
With few exceptions (discussed at the end of the chapter), it is safe to
calculate the maximum possible waterhammer pressure surge by using
the instantaneous waterhammer equation. The instantaneous
waterhammer equation assumes that the transient event occurs either
instantaneously or rapidly enough such that it is in effect instantaneous.
In such a case, it can be shown (Wylie, et al., 1993, pp. 4) by use of the
mass and momentum equation that the pressure transient is given by the
following equation:
P = a V
2/24/2003
1b.doc
1:35 PM
(9.1)
Page 355
pressure surge
density
wavespeed
V =
velocity change
By adding the pressure surge to the existing static pressure, one can
obtain the maximum theoretical pressure in the pipe. However, in some
cases, such as transient cavitation, the pressure can exceed the
instantaneous prediction. The engineer should therefore be cautious in
its application.
Heres an example. Assume flow in a pipe filled with water is
instantaneously stopped by a valve closure. Assume the density is 62
lbm/ft3, the wavespeed is 4000 feet per second, and the initial velocity is
10 feet per second. Also assume that before the transient the static
pressure is 50 psig.
ft
ft 1 ft 2
lbf s 2
Pmax = 62
* 4000 *10 *
*
+ 50 psig
s
s 144 in 2 32.2 lbm - ft
ft 3
Pmax = 535 + 50
lbm
Wavespeed
When a transient event is initiated in a pipe system, the remainder of the
system must adjust to the new conditions. In order to adjust, the
existence of the event must be communicated to the rest of the system.
This communication takes place at the wavespeed of the fluid. The
wavespeed is somewhat analogous to the sonic speed of the liquid.
However, the wavespeed is affected by the pipe structure.
The relationship between these parameters is expressed in the following
equation (Wylie, et al., 1993, pp. 27):
2/24/2003
1b.doc
1:35 PM
Page 356
a2 =
K
1+ c1[(K E )(D e )]
where:
a
= wavespeed
= fluid density
c1 = correction factor
The sonic speed of the liquid is given by the (square root of the)
numerator. The denominator contains the reduction from sonic speed
caused by the pipe properties.
The correction factor, c1, is a function of how the pipe is restrained.
Some examples from Wylie, et al., 1993, page 27-28 are:
1. Thin-walled pipe anchored at upstream end only
c1 = 1 2
2. Thin-walled pipe anchored throughout
c1 = 1 2
3. Thin-walled pipe free to expand throughout
c1 = 1
4. Thick-walled pipe anchored at upstream end only
c1 =
2e
(1 + ) + D (1 2)
D
D+e
c1 =
2/24/2003
2e
(1 + ) + D 1 2
D
D+e
1:35 PM
Page 357
c1 =
2e
(1 + ) + D
D
D+e
7. Circular tunnel
c1 =
2e
(1 + )
D
Given the previous, it should be apparent that in any given pipe the
wavespeed can only be approximated.
t = 2
L
a
(9.2)
Any event that occurs with a time frame shorter than this is equivalent to
an instantaneous event.
Method of Characteristics
The analysis of waterhammer in liquid pipe systems is based on the
transient mass and momentum equations. For a complete development of
the equations please see one of the many excellent references (Wylie, et
al., 1993, Chaudhry, 1987, and Swaffield et al., 1993). Solving the
equations based on a prescribed relationship between time step and
distance step is the essence of the Method of Characteristics which will
be discussed in depth in the following sections.
2/24/2003
1b.doc
1:35 PM
Page 358
Momentum equation
Following from Wylie, et al., 1993, page 22, the momentum equation
can be expressed as
fV V
1 P V
+
+ g sin( ) +
=0
2D
x t
(9.3)
where:
P
= pressure
= velocity
= density
= time
D = diameter
f
= friction factor
a2 V P
+
=0
g x t
(9.4)
where:
a
= wavespeed
Equations 9.3 and 9.4 are two partial differential equations with the two
unknowns P and V and the two independent variables x and t.
Application of the Method of Characteristics will convert these two
partial differential equations to four ordinary differential equations as
follows. To maintain flexibility with g-level, the development will leave
the parameter P in the equations rather than convert to head, H, as in
Wylie, et al., 1993.
2/24/2003
1:35 PM
Page 359
fV V
1 P V
+
+ g sin( ) +
= L1
2D
x t
a 2
V P
+
= L2
x t
(9.5)
(9.6)
L1 + L2 = 0
or,
V
1 P P V
+
+ a 2
+
x
x t t
fV V
+ g sin( ) +
2D
=0
(9.7)
If
dx 1
= = a2
dt
(9.8)
then
1
a
and
dx
= a
dt
From calculus
dP P dx P
=
+
dt x dt t
(9.9)
dV V dx V
=
+
dt x dt t
(9.10)
2/24/2003
1b.doc
1:35 PM
Page 360
V dx
P dx P V
+
+
+
x dt
x dt t t
+ g sin( ) +
fV V
2D
=0
(9.11)
fV V
dP dV
=0
+
+ g sin( ) +
2D
dt dt
or
fV V
1 dP
dV
+
+ g sin( ) +
=0
dt
2D
a dt
(9.12)
fV V
2D
dx = 0
(9.13)
The gravity term contains the term , and the term can be converted as
follows
dz
dx = gdz
dx
g sin( )dx = g
(9.14)
dP + adV + gdz +
fV V
2D
dx = 0
dP +
a
f
dm& + gdz +
dx m& m& = 0
A
2 DA2
2/24/2003
1:35 PM
Page 361
(9.15)
a
dP + A
PA
m& P
zP
dm& + g dz +
m& A
zA
xP
m& m& dx = 0
2 DA2 x
A
(9.16)
f x
2 DA2
(9.17)
m& P m& A = 0
t = 5t
t = 4t
t = 3t
t = 2t
P
t = t
C-
C+
t=0
x=0
A
x = i-1
x=i
x = i+1
x=L
Similarly
2/24/2003
1b.doc
1:35 PM
Page 362
(PP PA ) +
+
f x
2 DA2
a
(m& P m& A ) + g (z P z A )
A
m& P m& A = 0
(9.18)
B=
a
A
Resistance:
R=
f x
2 DA2
=0
(9.19)
(9.20)
where:
BP = B + R m& i 1, old
BM = B + R m& i +1,old
Equations 9.19 and 9.20 are referred to as the compatibility equations
(Wylie, et al., 1993, pp. 42). Since all the parameters in CP, CM, BP, and
2/24/2003
1:35 PM
Page 363
& i, new .
BM are known, the only unknowns in 9.19 and 9.20 are Pi,new and m
Thus there are two equations and two unknowns.
For example, to solve for the pressure at an interior pipe point, the flow
rate can be eliminated and the pressure solved for directly:
C B + C M BP
Pi, new = P M
BP + BM
(9.21)
This value for pressure can be substituted back into equation 9.19 and
9.20 to solve for flow rate, or the flow rate can be solved for directly:
C CM
m& i, new = P
BP + BM
(9.22)
Pnew = Psat
2/24/2003
1b.doc
1:35 PM
Page 364
+ m& up, new m& up, old + m& down, new + m& down, old
) 2t
(9.23)
If no vapor existed previously, the old volume is zero. When the vapor
volume becomes negative, the cavity collapses and the pressure again
rises above vapor pressure. In this case the conventional methods
(Equations 9.19, 9.20) again are applicable.
The DVC method is well understood and documented for interior pipe
points. However, application of the method to general boundary
conditions in multi-pipe systems is limited. The theory for many of the
junctions used in AFT Impulse is developed in Walters, 1991.
Pipe sectioning
The relationship between pipe sectioning and time step is given by
L
n t
a=
(9.24)
where n is the number of sections and L/n is the length of each pipe
section. If the transient event occurs within some particular time step,
the number of sections, n, must be chosen such that it is consistent with
the time step in Equation 9.24.
Specifying the number of sections for multi-pipe systems is more
complicated. Each pipe has its own wavespeed and length. For each pipe
the following must be satisfied:
ni =
Li
ai t
(9.25)
In Equation 9.25 as applied to each pipe in the system, the time step t
must be the same for all pipes. The length, L, is a given for each pipe and
thus cannot be changed. The wavespeed, a, is also a given for each pipe,
although it is known with less certainty than is the length. The free
parameter is then the number of sections, n.
2/24/2003
1:35 PM
Page 365
ni =
Li
(1 )ai t
2/24/2003
1b.doc
1:35 PM
Page 366
CM
P
m& i, new = known
BM
Conversely, if the pressure is known at the downstream end, the positive
compatibility equation is used (Equation 9.19).
m& i, new =
C P Pknown
BP
2/24/2003
1:35 PM
Page 367
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):
CM
P
m& i, new = known
BM
Conversely, if the pressure is known at the downstream end, the positive
compatibility equation is used (Equation 9.19).
m& i, new =
C P Pknown
BP
2/24/2003
1b.doc
1:35 PM
Page 368
+ m& up, new m& up, old + m& down, new + m& down, old
) 2t
& up terms are known because they are specified terms for inflow
The m
& down terms are known. Similar to
junctions. For outflow junctions the m
a pipe interior node, when the vapor volume is negative, the cavity
collapses and the fluid pressure then rises above the vapor pressure.
2/24/2003
1:35 PM
Page 369
Pi, new = C P
Conversely, if the dead end is at the downstream end, the positive
compatibility equation is used (Equation 9.19) with zero flow.
Pi, new = CM
Vvapor , new = Vvapor , old + m& up, new m& up, old
) 2t
& down terms are always zero. For dead ends at the beginning
since the m
& up terms are zero. Similar to a pipe interior node, when
of a pipe, the m
the vapor volume is negative, the cavity collapses and the fluid pressure
then rises above the vapor pressure.
Branch theory
At a branch, there can be multiple pipes (Wylie, et al., 1993, pp. 51-53).
An additional relationship is needed, and that relationship is the
conservation of mass. The total mass flow in and out of the junction
must sum to zero.
In addition, the branch has a single pressure solution, Pj, and this
solution is common to all inflowing pipes.
The compatibility equation is written for each junction. For the pipes
flowing into the branch, the positive equation is used (Equation 9.19)
m& i, new =
C P Pj
BP
2/24/2003
1b.doc
1:35 PM
Page 370
Pj C M
m& i, new =
BM
Now, sum all of the pipe flow rates into the junction
m& i,new = SC S B Pj = 0
(9.26)
where
CP + CM
SC =
BP
SB =
BM
1
1
+
BP
BM
(9.27)
(9.28)
In general, there may be a known flow into the branch (a flow sink) or
into the branch (a flow source). Therefore rewrite Equation 9.26 as
m&
i , new
= SC S B Pj = m& Applied
Pj =
SC + m& Applied
SB
(9.29)
2/24/2003
1:35 PM
Page 371
Vvapor , new = Vvapor , old + + m& out, new + m& out, old
t
2
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):
2/24/2003
1b.doc
1:35 PM
Page 372
m& valve 2
+
1
Rvalve
1
Rvalve
Using the quadratic equation, the solution for positive flow through the
valve is
m& valve = 0
Exit valves
If the valve is an exit valve, then the downstream pressure is known and
2/24/2003
1:35 PM
Page 373
+ m& up, new m& up, old + m& valve, new + m& valve, old
) 2t
t
2
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.
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.
2/24/2003
1b.doc
1:35 PM
Page 374
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).
(9.30)
2/24/2003
1:35 PM
Page 375
(9.31)
BM = B + R m& infinity,old
(9.32)
zi zi +1 = zi 1 zi
Also, in equations 9.31 and 9.32 it is assumed that the pipe property
values B and R remain valid for the infinity point, which is a good
assumption unless the pipe properties change.
m& spray
Pexit
Figure 9.2
2/24/2003
1b.doc
1:35 PM
Page 376
P =
1
V 2
2
2 C D Aspray
m& spray 2
Pj , new =
SC m& spray
SB
sgn (m& )
1
C D 2 Aspray 2 2
m& 2 +
S
1
m& + Pexit C = 0
SB
SB
or
&2
sgn(m& )m +
+
2 C D 2 Aspray 2
SB
m&
2 C D 2 Aspray 2 (S B Pexit SC )
SB
=0
2/24/2003
1:35 PM
Page 377
Vvapor , new = Vvapor , old + + m& out, new + m& out, old
t
2
& spray is the spray outflow term. The m& in terms are obtained
where m
& out terms are
from the positive compatibility equation, while the m
obtained from the negative equation.
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.
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.
H = a + bQ pump + cQ pump 2
where a, b and c are obtained from curve fits of vendor data. This curve
assumes the pump operates at 100% speed. During the transient, the
speed will change.
The affinity law for pumps given by the following equations
2/24/2003
1b.doc
1:35 PM
Page 378
2
Q
= constant1
= constant2
where is the pump speed. Applying these relationships the pump curve
is transformed
= a + b
Q pump
+ c
Q pump 2
H = 2 a + bQ pump + cQ pump 2
Converting to pressure rise
b
P = g 2 a +
m& pump +
m& pump 2
P = 2 a + bQ pump + cQ pump 2
(9.33)
where the primes on the constants were dropped after they were divided
through by g.
By use of the two compatibility equations (Equation 9.19, 9.20),
This is a quadratic equation for the pump flow rate which can be solved
by the quadratic formula
2/24/2003
1:35 PM
Page 379
BP,up + BM , down b
Q pump =
2
c
4b 2 a + C P ,up C M , down
1 1
BP,up + BM , down b 2
)1 2
With the solution for the pump flow rate, the pressure upstream and
downstream of the pump can be obtained from Equations 9.19 and 9.20.
Pumps with higher order curve fits
If the pump is modeled with a third or fourth order curve, the full speed
pump curve is:
= a + b
Q pump
+ c
Q pump 2
+ d
Q pump 3
+ e
Q pump 4
m& pump 3 +
m& pump 4
2/24/2003
1b.doc
1:35 PM
Page 380
BM , down b
m& pump =
2c
4b 2 a + PR CM , down
1 1
(BM ,down b )2
)1 2
2/24/2003
1:35 PM
Page 381
+ m& up, new m& up, old + m& pump, new + m& pump, old
) 2t
+ m& pump, new m& pump, old + m& down, new + m& down, old
) 2t
BM , down b
m& pump =
2
2c
4b 2 a + Psat CM , down
1 1
(BM ,down b )2
)1 2
Similar to a pipe interior node, when the vapor volume is negative the
cavity collapses and the fluid pressure then rises above the vapor
pressure.
2/24/2003
1b.doc
1:35 PM
Page 382
T = I
d
dt
(9.34)
P = T
(9.35)
1
d
dt =
I
T
Assuming an average value of torque over the time step one obtains:
t
1
=
I
T
With the speed known at the previous time step, one can solve for the
new speed as
new = old T
2/24/2003
1:35 PM
t
I
Page 383
new = old
Tnew + Told t
2
I
(9.36)
The old torque is known from the previous step, and thus the new speed
can be solved for by iteration. The mass flow rate and head are obtained
from Equations 9.19 and 9.20 (with the pressure converted to head) and
the pump curve supplied by the user.
Tmotor T fluid = I
d
dt
new = old +
Tmotor , new + Tmotor , old T fluid , new + T fluid , old
2
2
(9.37)
The old motor torque and fluid torque at the previous time step are
known, and the values at the new time step are both functions of the new
speed. The fluid torque is related to the power through Equation 9.35,
and the power can be obtained from the new flow.
The motor torque is entered by the user as a function of speed (see
Figures 9.3 and 9.4), and is therefore known at the new speed. This
process thus requires iteration.
This calculation uses the user provided pump data for head vs. flow and
power vs. flow, and thus only works for positive or zero flows. The mass
flow rate and head are obtained from Equations 9.19 and 9.20 (with the
pressure converted to head) and the pump curve supplied by the user.
The typical shape of a motor torque vs. speed curve is shown in Karassik
(2001, p. 8.25, Figure 26), which is reflected in the Figure 9.4 data.
2/24/2003
1b.doc
1:35 PM
Page 384
Figure 9.3
2/24/2003
1:35 PM
Page 385
H
P
=
H R PR
T
=
TR
Q
m&
v=
=
QR m& R
h=
2/24/2003
1b.doc
(9.38)
N
NR
1:35 PM
Page 386
Figure 9.5
FH ( ) =
FB ( ) =
h
+ v2
2
2 + v2
where
2/24/2003
1:35 PM
Page 387
(9.39a)
= + tan 1
(9.39b)
= tan 1
AFT Impulse allows you to select either definition as the basis for the FH
and FB data in Equation 9.39a.
The parameters FH and FB are properties of the pump, and it is typically
assumed that for pumps of similar specific speed they are invariant (see
Brown & Rogers for a discussion questioning this assumption). Figures
9.5 and 9.6 show data for a pump with a specific speed of 0.46 as given
by Wylie, et al., (1993, pp. 147-148).
Figure 9.6
2/24/2003
1b.doc
1:35 PM
Page 388
Reference
Brown & Rogers, Partial, From Nevada Test, 1980
Kittredge, 1956
Brown & Rogers, from Bureau of Reclamation, 1980
Donsky, 1961
Brown & Rogers, Partial, From Nevada Test, 1980
Brown & Rogers, Partial, From Nevada Test, 1980
Brown & Rogers, Partial, From Nevada Test, 1980
Kittredge, 1956
Thorley, 1996
Thorley, 1996
Thorley, 1996
Thorley, 1996
Brown & Rogers, from Bureau of Reclamation, 1980
Thorley, 1996
Thorley, 1996
Thorley, 1996
Thorley, 1996
Thorley, 1996
Donsky, 1961
Thorley, 1996
Donsky, 1961
With data for FH and FB, the compatibility equations (9.19, 9.20) are
used as follows:
(9.40)
Ppump = PR 2 + v 2 FH
2/24/2003
1:35 PM
Page 389
(9.41)
C P CM vm& R (BP + BM ) + PR 2 + v 2 FH = 0
(9.42)
= NR
0 =
=
60
T0
TR
T
TR
where NR is the rated speed in rpm, and the "0" subscript refers to
conditions at the beginning of the time step.
We can then write,
d = N R
2
2
d = N R
( 0 )
60
60
T=
T + T0
d
= I
2
dt
(9.43)
+ o I
NR
( 0 ) = 0
TR 15t
(9.44)
2/24/2003
1b.doc
1:35 PM
Page 390
( 2 + v 2 )FB + o I NT R 15t ( 0 ) = 0
(9.45)
Equations 9.42 and 9.45 are solved simultaneously using a NewtonRaphson iterative method. Details are given in Wylie, et al., 1993, pp.
149-152.
Tm T f =
Tm + Tm,0
T f + T f ,0
d
=I
dt
2
where the m subscript refers to the motor torque and the f subscript
refers to the fluid torque. Equation 9.44 becomes:
f + f , o m m, o I
NR
( 0 ) = 0
TR 15t
2/24/2003
1:35 PM
Page 391
( ) N
P
I p = 1.5 107
0.9556
(9.46)
P
I m = 118
N
(9.47)
Pj , new =
SC + m& A, new
SB
where SC and SB are given by Equations 9.27 and 9.28. The relationship
between accumulator pressure and volume is assumed to be linear,
K =
P
V V
where K is the effective bulk modulus of the container, and has units of
pressure. The change in volume can be related to flow rate as follows:
t
2
Therefore,
2/24/2003
1b.doc
1:35 PM
Page 392
Pj , new = Pj , old
Pj , new =
2V
m& A, old
K t
2V
SB +
K t
SC + Pj , old
Vvapor , new = Vvapor , old + + m& out, new + m& out, old
t
2
& A is the accumulator flow out of the junction. The m& in terms
where m
are obtained from the positive compatibility equation, while the
m& out terms are obtained from the negative equation.
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.
2/24/2003
1:35 PM
Page 393
Pj , new =
P =
1
V 2
2
or,
P = Pj , new Pexit =
2 Aexit
Substituting
Pexit =
2 Aexit
S
1 2K
2 K
Pexit C = 0
m& exit , new +
2
2
SB
S B Aexit
Aexit
2/24/2003
1b.doc
1:35 PM
Page 394
2/24/2003
1:35 PM
Page 395
C CM P
m& i, new = P
BP + BM
2/24/2003
1b.doc
1:35 PM
Page 396
Pj , new =
SC + m& T , new
SB
LL
Orifice
m& T
LC
Figure 9.7
Connector
Pipe
2 AT
Substitution yields
2/24/2003
1:35 PM
Page 397
SC + m& T , new
m& T ,old t
A
T
+g
m& T , newt
2 AT
C1 C2 m& T , new = 0
& T , new
Solving for m
m& T , new =
m&
t
SC
Psurface g LL,old T , old + C1
SB
2 AT
gt
1
C2 +
+
2 AT S B
This flow rate can be back substituted to obtain the junction pressure and
surge tank liquid height.
Pj , new =
2/24/2003
1b.doc
SC + m& A, new
SB
1:35 PM
Page 398
PA
Orifice
Connector
Pipe
m& A
LC
Figure 9.8
The gas volume and pressure in the accumulator can be related by the
thermodynamic law
PAV An = C A
where n is the polytropic constant (1.4 for isentropic air, 1.0 for
isothermal) and CA is a constant.
Therefore,
2/24/2003
1:35 PM
Page 399
t
2
SC + m& A, new
= PA, new C1 C2 m& A, new
SB
or
PA, new =
SC + m& A, new
+ C1 + C2 m& A, new
SB
S
C3 = C + C1
SB
C4 =
1
+ C2
SB
t
2
t
2
xi +1 = xi
F ( xi )
F ( xi )
where xi is the current value of x, xi+1 is the new value, F is the function
of x to drive to zero, and F is the derivative.
In this case
)(
2/24/2003
1b.doc
1:35 PM
Page 400
dF
C7
= F =
+ C4 C5 + C6 m& A, new n
dm& A, new
C5 + C6 m& A, new
where
C7 = nC AC6
& A, new is calculated. This value can then be backBy iteration, m
substituted to obtain the new junction pressure and new gas volume.
Sonic inflow
m& =
2/24/2003
C d Ain Po
RTo
1:35 PM
1
1 +
( +1) (2(1 ))
Page 401
(9.48a)
Subsonic inflow
( +1)
2
Pi
2 Pi
1 Po
P
C A P
m& = d in o
RTo
(9.48b)
Sonic outflow
C A P
1
m& = d out i 1 +
2
RTL
( +1) (2(1 ))
(9.48c)
Subsonic outflow
m& =
C d Aout Pi
RTL
2
( +1)
Pi
2 Pi
1 Po
Po
(9.48d)
where:
Pi =
Po =
atmospheric pressure
To =
atmospheric temperature
TL =
liquid temperature
CdAin
CdAout =
gas constant
Equation of state
Besides the compatibility equations (9.19 and 9.20), we need another
equation to form a complete set. The ideal gas equation of state is used:
PV = mRT
Substituting terms one obtains
2/24/2003
1b.doc
1:35 PM
Page 402
Pi
t
Pi 2Vold
1
m& down, new + m& down,old m& up, new m& up, old
+
RTL t
L
2mold
) =
Pi 2Vold
C P
1 Pi C M
RTL t
BP
L BM
2mold
+ m& gas , old + m& gas, new
t
(9.49)
1
Pi 2
+
L BM
C
+ Pi M
L BM
2m
RTL old
t
2/24/2003
L BP
m& d
m& u
2V
CP
+ old old + old
t
L BP
L
L
1:35 PM
Page 403
& sonic is the sonic inflow which can be calculated from Equation
where m
9.49a based completely on input data.
Subsonic Inflow
Equation 9.49 is modified using Equation 9.48b and solved iteratively
using Newton-Raphson. Defining Pr = Pi/Po obtains:
1
1
Pr 2 Po 2
+
L BM L BP
m& d
m& u
2V
C
C
+ Pr Po old M P + old old
L BM L BP
L
L
t
2m
2R 2
Pr
+ Cd Ain PoTL
Pr ( +1) = 0
(
)
1 To
Defining terms for Newton-Raphson solution yields
2/24/2003
1b.doc
1:35 PM
Page 404
)0.5 + C2 Pr 2 + C3Pr + C4 = F
0.5
dF
0.5C1 (Pr a Pr b ) (aPr a 1 bPr b 1 )+ 2C2 Pr + C3 =
dP
C1 Pr a Pr b
a=
+1
b=
C1 = Cd Ain PoTL
2R
( 1)To
1
1
+
C2 = Po 2
B
B
L M
L P
m& d
m& u
2V
C
C
C3 = Po old M P + old old
L BM L BP
L
L
t
2m
2/24/2003
1:35 PM
Page 405
1
1
+
Pr 2 Po 2
L BM L BP
m& d
m& u
2V
C
CP
+ Pr Po old M
+ old old
L BM L BP
L
L
t
2m
+ Pr (C D Aout Po )
2RTL 2
Pr
Pr ( +1) = 0
1
C1 Pr a Pr b
)0.5 + C2 Pr + C3 + C4 P1 = F
r
0.5C1Pr Pr a Pr b
+ C 2 C4
a=
1
Pr
dF
dPr
+1
b=
C1 = Cd Aout Po
2RTL
1
1
1
+
C2 = Po 2
L BM L BP
m& d
m& u
2V
C
CP
+ old old
C3 = Po old M
L BM L BP
L
L
t
2m
2/24/2003
1b.doc
1:35 PM
Page 406
1
1
+
Pi 2
L BM L BP
m& d
m& u
2V
C
CP
+ Pi old M
+ old old
L BM L BP
L
L
t
( +1) (2(1 ))
+ Pi C D Aout RTL 1 +
2m
1
z
2
orifice
Figure 9.9
2/24/2003
1:35 PM
Page 407
F1 F3 F friction Forifice + Fg = A2 L2
dV2
dt
or,
F1 F3 F friction Forifice + Fg = L2
dm& 2
dt
Now,
F1 = P1 A2
F3 = P3 A2
F friction =
Forifice =
fL2 m& 2 2
A2
D2 2 A2 2
m& 2 2
2 C D 2 Ao 2
A2
Fg = A2 gz
Substituting terms and dividing by area A2
P1 P3
1
fL2 1
L dm& 2
& 2 2 gz = 2
m& 2 2
m
D2 2 A2 2
A2 dt
2 C D 2 Ao 2
Taking average values over time, and denoting current time by a prime,
P1 + P1 P3 + P3 fL2 1
m& 2 m& 2
2
2
D2 2 A2 2
1
L (m& m& 2 )
& 2 m& 2 gz = 2 2
m
A2
t
2 C D 2 Ao 2
Collecting terms and simplifying
P1 P3 = C1 + C2 m& 2
where
2/24/2003
1b.doc
1:35 PM
Page 408
C1 = P3 P1 + 2 gz
C2 =
2 L2
m& 2
A2 t
1
2 L2
fL2 1
m& 2 +
m& 2 +
2
2
2
D2 A2
A2 t
C D Ao
(9.50)
(9.51)
C1 = P3 P1 + 2 gz
C2 =
2 L2
m& 2
A2 t
2 L2
fL2 1
m& 2 +
D2 A2 2
A2 t
C1 = P3 P1
C2 =
C D 2 Ao 2
m& 2
C1 = 0
C2 = 0
2/24/2003
1:35 PM
Page 409
2/24/2003
1b.doc
1:35 PM
Page 410
P = aV
(9.52)
where:
P =
pressure surge
density
wavespeed
V =
velocity change
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.
2/24/2003
1:35 PM
Page 411
2/24/2003
1b.doc
1:35 PM
Page 412
CHAPTER 10
Time-based transients
Time-based transients occur in accordance with the time of the transient
simulation. The start and stop times of the simulation are specified in the
Transient Control window. The initiation of time-based transients are
pre-specified before a model is run.
For example, consider a valve closure transient. Assume the valve starts
to close at two seconds into the simulation, and the time it takes for the
valve to close is one second. After that, the valve stays closed. If the
initial valve Cv is 250, the transient would appear as in Figure 10.1.
In the "Initiation of Transient" area there are four options. For timebased transients, the option is specified as Time. In the Transient Data
area the data is entered. Here the Cv profile of the valve will follow the
profile entered and start to close at 2 seconds.
2/24/2003
1b.doc
1:35 PM
Page 413
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.
2/24/2003
1b.doc
1:35 PM
Page 414
The data in the Transient Data table has a slightly different meaning than
does a time-based transient. Here time zero is relative to the time at
which the event criteria is first met. For example, if Pipe 2 reaches 150
psig at 3.65 seconds, the valve will start its transient at 3.65 seconds.
The valve will close (i.e., Cv = 0) at 1 second after the event initiation,
or 4.65 seconds.
2/24/2003
1:35 PM
Page 415
2/24/2003
1b.doc
1:35 PM
Page 416
Sequential events
Sequential dual events progress from the first event to the second and no
further. For example, these could be two high pressure settings. When
the first setting is reached the valve opens partially, and when the second
setting is reached, which would likely be at a higher pressure than the
first event, the valve opens the rest of the way. Similar to a single event
transient, if the pressure drops and then rises again, the events would not
be initiated again.
Sequential dual events are specified by selecting Dual Event Sequential
in the Initiation of Transient area of the junction window (Figure 10.3).
Similar to Cyclic Dual Events, two event tabs appear where the first and
second event can be specified. The transients that are initiated when the
events occur are specified in the table of the First Transient and Second
Transient tab.
2/24/2003
1:35 PM
Page 417
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.
2/24/2003
1b.doc
1:35 PM
Page 418
2/24/2003
1:35 PM
Page 419
Repeat transient
For transients that are periodic, you can specify that the transient repeats
itself. This is specified by selecting the "Repeat Transient" checkbox
(see Figure 10.1, 10.2 or 10.3). When Repeat Transient is selected, the
first and last data points must match. After the transient reaches the final
data point, it returns to the first data point and starts over.
Event messages
Events that are initiated during the simulation are displayed in the two
Event Messages area of the Output window (see Figure 10.4). The first
is a list of event messages sorted by junction. The second is a list of
event messages sorted by time. If no events occur, the Event Messages
tab will be hidden.
2/24/2003
1b.doc
1:35 PM
Page 420
Figure 10.4 The Event Messages area shows a listing of all events
that occur during the simulation sorted by junction and
time.
2/24/2003
1:35 PM
Page 421
2/24/2003
1b.doc
1:35 PM
Page 422
Figure 10.6 All junction transient data is shown in the Model Data
window on the Transient Data tab.
2/24/2003
1:35 PM
Page 423
2/24/2003
1b.doc
1:35 PM
Page 424
CHAPTER 11
General considerations
Surge pressures that are excessively high can burst pipes and damage
equipment. Surge pressures that reduce the local pressure can result in
pipes being crushed (due to atmospheric pressure exceeding the internal
liquid pressure), cavitation and liquid column separation (which can then
result in large pressure spikes when the cavity collapses), and
subatmospheric pressures that are unacceptable for drinking water
pipelines.
The surge pressures result from changing conditions in the system,
which change the fluid velocity. Options include slowing down the
system changes and thus reducing velocity or providing locations which
can deliver extra fluid when the pressure decreases or accept extra fluid
when the pressure rises.
If using a surge suppression device, it is generally desirable to locate the
device as close as possible to the cause of the transient. The practicality
of the different options depend highly on the particular case of interest.
Methods for surge reduction are discussed in detail in Swaffield et al.
(1993, Chapter 6), Wylie et al. (1993, Chapter 10), and Chaudhry (1987,
Chapter 10).
2/24/2003
1b.doc
1:35 PM
Page 425
Flexible hose
Flexible hose can sometimes be used in small systems, which reduces
wavespeed and hence pressure surge.
Parallel valves
Rather than use one large valve to stop the flow, perhaps two or more
valves can be used in parallel. The timing of the valve closures is
staggered, thus allowing a portion of the flow to be stopped with each
valve.
Relief valves
Relief valves can be used to avoid high pressure situations. In addition,
inward relief valves can be used for low pressure situations.
2/24/2003
1b.doc
1:35 PM
Page 426
Surge tank
Surge tanks can be an effective way to reduce surge pressures if the
steady-state system pressures are relatively low. High steady-state
pressures result in high hydraulic gradelines. Since the liquid level in a
surge tank rises to the local hydraulic gradeline, they would have to be
very tall if located in a high pressure region.
Gas accumulator
Gas accumulators can be effective in spreading out the wave front of a
transient and thus reducing peak pressures. Gas accumulators change the
frequency response of the system and can in fact amplify pressure surge
if not sized and located properly.
2/24/2003
1:35 PM
Page 427
2/24/2003
1b.doc
1:35 PM
Page 428
CHAPTER 12
2/24/2003
1b.doc
1:35 PM
Page 429
2/24/2003
1b.doc
1:35 PM
Page 430
2/24/2003
1:35 PM
Page 431
Figure 12.2 Model Data window display for the model in Figure
12.1. From this perspective it is much easier to see that
one of the pipes (#4) has length data entered in miles
rather than feet. This is not automatically an input error,
as AFT Impulse accepts pipe input with mixed
engineering units. However, more frequently than not
the user intends to use consistent units, so this
probably is an error.
2/24/2003
1b.doc
1:35 PM
Page 432
Pressure (psid)
10
8
6
4
2
0
0
50
100
150
200
150
200
Pressure (psid)
10
8
6
4
2
0
0
50
100
Figure 12.3 Top graph shows a third or fourth order pump curve fit
with an inversion at 150 ft3/sec. AFT Impulse interprets
flow rates to the right of this inversion as acceptable
which can cause convergence problems or erroneous
results. All pump curve fits should include sufficient
data such that the curve crosses the zero pressure rise
axis as shown in the bottom curve.
2/24/2003
1:35 PM
Page 433
2/24/2003
1b.doc
1:35 PM
Page 434
2/24/2003
1:35 PM
Page 435
2/24/2003
1b.doc
1:35 PM
Page 436
2/24/2003
1:35 PM
Page 437
Application example
A real life example is shown in Figure 12.4. The system represents part
of a tank farm for refined petroleum products. Gasoline flows from a
refinery some 60 miles away into the tank farm. The supply pipeline is
represented by pipe P1. Pipes P2-P9 are relatively short local pipes at the
tank farm with changing diameters. This model is for gasoline, and the
gasoline comes into junction J10 and flows initially to the tank at J106.
The transient of interest is when the J201 valve is opened and the J104
valve is closed. This will divert the incoming gasoline from J106 to
J212. The total scale of the tank farm piping is less than a mile.
The transient lasts for about 30 seconds. During the opening and closing
of the valves, the transient waves propagate out to J106 and J212, as
well as to the supply junction J1. Using the wavespeed of gasoline it can
be determined that any waves that propagate to pipe P1 will not reflect
back to the tank farm within 30 seconds, and thus will not impact the
transient. However, during the transient procedure, gasoline still flows
from junction J1 into the system.
One approach is to model pipe P1 as the full 60 mile length. Thus all the
pipe stations along 60 miles must be calculated for the entire 30 second
simulation. A second approach is to specify pipe P1 as very short (13
feet long in this case, chosen for modeling convenience based on other
considerations) and to specify J1 as an infinite pipe junction.
Both approaches will give the same transient results. However, the first
approach would take over 100 times as long to run (a 1.9 hour run vs. a 2
minute run when using an infinite pipe). Again, the difference in runtime
is that with an infinite pipe, no computation is required along the 60 mile
pipe except for two stations at the entrance.
2/24/2003
1b.doc
1:35 PM
Page 438
2/24/2003
1:35 PM
Page 439
Supply
Tank (J1)
Transfer
Pump (J3)
Shoreline
Elevation
Lake Michigan
An engineer is tasked with sizing the pump for this system. The engineer
has an idea of the discharge head needed for the pump, and so builds an
AFT Impulse model as shown in Figure 12.6.
2/24/2003
1b.doc
1:35 PM
Page 440
P3>
J1
J3
P1>
J4
P2>
Discharge
Tank (J2)
Transfer
Pipeline
Supply
Tank (J1)
Transfer
Pump (J3)
Shoreline
Elevation
2/24/2003
1:35 PM
Page 441
P2>
J1
J3
P1>
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.
2/24/2003
1b.doc
1:35 PM
Page 442
Examples
The simplest example is the system shown in Figure 12.9. In AFT
Impulse terms, the system has two assigned flow junctions.
Inlet
Q = 1 ft3/s
El = 0 ft
L = 100 ft
D = 2 in.
f = 0.02
Outlet
Q = 1 ft3/s
El = 0 ft
Figure 12.9 Model with two assigned flows. This model does not
have a unique solution.
Obviously, the flow in the pipe is known, but, what is the pressure at the
inlet? At the outlet? It cannot be determined because there is no
reference pressure. The reference pressure is that pressure from which
2/24/2003
1:35 PM
Page 443
2/24/2003
1b.doc
1:35 PM
Page 444
L = 100 ft
D = 2 in.
f = 0.02
Inlet
Q = 1 ft3/s
El = 0 ft
L = 100 ft
D = 2 in.
f = 0.02
Inlet
P = 10 psig
El = 5 ft
L = 100 ft
D = 2 in.
f = 0.02
Inlet
P = 10 psig
El = 5 ft
Outlet
P = 0 psig
El = 5 ft
Outlet
Q = 1 lbm/s
El = 0 ft
Outlet
P = 0 psig
El = 5 ft
Figure 12.0 a-c Top two models with one pressure and one assigned
flow, bottom model with two pressures. All of these
models have a unique solution.
Second, if the user was allowed to specify two flows as in Figure 12.9,
he/she could specify them with different flow rates. Clearly, they must
have the same flow or an inconsistency occurs. The basic reason the
inconsistency would be possible is because Figure 12.9 does not have a
unique solution.
Figure 12.10a-c shows three other model configuration possibilities. It is
not possible to specify inconsistent conditions for any of the Figure
12.10 models, and they always have a unique solution no matter what
input is specified by the user.
The four models in Figures 12.9 and 12.10a-c contain all logical
possibilities. In all four cases there are four things we want to know: the
pressure and flow at the inlet, and the pressure and flow in the outlet. In
each case we know two of these four. The problem is that in the Figure
12.9 model the lack of a reference pressure makes it impossible to
determine the inlet and outlet pressures, even though we know the flow.
The three models in Figure 12.10a-c all have at least one pressure, thus
all four of the desired parameters can be determined. Table 12.1
summarizes this. Also note in Table 12.1 that in the Figure 12.10c case
the flow can be determined, but it is by iteration.
2/24/2003
1:35 PM
Page 445
Inlet
Pressure
Fig. 12.9
Fig. 12.10a
Fig. 12.10b
Fig. 12.10c
Inlet
Flow
X
Outlet
Pressure
Outlet Flow
OK ?
No
Yes
X
X
Yes
Yes
Comments
No reference pressure, cannot
calculate pressures
1 known flow, can calculate
everything directly
1 known flow, can calculate
everything directly
2 pressures, can iterate for flow
Inlet
Q = 1 ft3/s
El = 0 ft
Outlet
Q = 0.5 ft3/s
El = 0 ft
Figure 12.11
2/24/2003
1b.doc
1:35 PM
Page 446
Inlet
P = 10 psig
El = 10 ft
J1
J2
J3
Outlet
Q = 1 ft3/s
El = 0 ft
FCV
Q = 1 ft3/s
Inlet
P = 10 psig
El = 5 ft
J1
J2
J3
Outlet
P = 0 psig
El = 5 ft
FCV
Q = 1 ft3/s
Figure 12.12a-b The model at the top does not have a reference
pressure after the Flow Control Valve so the
pressure drop across the FCV cannot be determined.
The top model does not have a unique solution. The
bottom model has a pressure upstream and
downstream, and a unique solution exists.
2/24/2003
1:35 PM
Page 447
Figure 12.13
2/24/2003
1b.doc
1:35 PM
Page 448
Inlet
P = 10 psig
El = 5 ft
Pump with Known (assigned) Flow
Q = 1 ft3/s
2/24/2003
1:35 PM
Page 449
Inlet
P = 10 psig
El = 5 ft
Inlet
P = 10 psig
El = 5 ft
Outlet
P = 0 psig
El = 5 ft
FCV
Q = 1 ft3/s
Pump with
Assigned Flow
Q = 1 ft3/s
FCV
Q = 1 ft3/s
FCV
Q = 1 ft3/s
Outlet
P = 20 psig
El = 5 ft
Figure 12.15 a-b Neither model above has a unique solution. The
top model has two flow control valves in series.
The bottom model has a pump modeled as an
assigned flow in series with an FCV. In both cases
either a third pressure junction is needed between
the two middle junctions, or one of the junctions
must be changed from a flow controlling device.
2/24/2003
1b.doc
1:35 PM
Page 450
Inlet
P = 10 psig
El = 5 ft
Outlet
P = 20 psig
El = 5 ft
Pump with
Known Flow
Q = 1 ft3/s
Inlet
P = 10 psig
El = 5 ft
Pump with
Known Flow
Q = 1 ft3/s
FCV
Q = 1 ft3/s
PDCV
dP = 5 psid
Inlet
P = 10 psig
El = 5 ft
Outlet
P = 20 psig
El = 5 ft
Outlet
P = 20 psig
El = 5 ft
Pump with Curve FCV
Q = 1 ft3/s
dP = f(Q)
Figure 12.16a-c. The model at the top is the same as Figure 12.15b
and does not have a unique solution. To size the
pump, change the model to the second one shown
above, which uses a pressure drop control valve
(PDCV) rather than an FCV. Once the second model
is run, the pump is sized, a pump curve exists, and
the model at the bottom can be run using an FCV.
2/24/2003
1:35 PM
Page 451
P5^
^P8
J5
J8
FCV
Q = 100 gpm
FCV
Q = 100 gpm
P4^
J1
J3
J2
P1>
P2>
P7^
J7
J4
P3>
P6>
Figure 12.17
This brings up the question: which FCV should you turn into a PDCV?
When FCVs are put in parallel, frequently the pipe design has one of
the FCVs further away from the pump than the others. Because of the
additional piping leading to this FCV, it will be the weakest link in the
chain of parallel FCVs, by virtue of having the lowest pressure drop
across it. The most remote FCV should be chosen as the PDCV. If the
most remote FCV is chosen, when the minimum pressure drop required
is applied to the PDCV all other FCVs in the parallel system will have
greater than the minimum pressure drop thus satisfying the pressure
drop requirement for those FCVs as well.
What if you do not know which FCV is the most remote? In this case
make your best guess, change it to a PDCV at the minimum required
pressure drop, run the model, then verify whether all other FCVs have a
pressure drop that meets or exceeds the requirement. If not, then the
FCV with the smallest pressure drop as determined by the first run is in
reality the weakest FCV. Choose this FCV, change it to a PDCV, then
2/24/2003
1b.doc
1:35 PM
Page 452
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.
2/24/2003
1:35 PM
Page 453
Inlet
Q = 1 ft3/s
El = 10 ft
Outlet
P = 0 psig
El = 10 ft
PRV
Pdown = 10 psig
Figure 12.19
Figure 12.20 shows the possible cases with PRVs and PSVs, while
Table 12.2 comments on the six cases. In summary, at least one known
pressure is always needed on the side of a pressure control valve
opposite of the controlled side.
2/24/2003
1b.doc
1:35 PM
Page 454
Inlet
Q = 1 ft3/s
El = 10 ft
Outlet
P = 0 psig
El = 10 ft
PRV
Pdown = 10 psig
Inlet
P = 30 psig
El = 10 ft
Outlet
Q = 1 ft3/s
El = 10 ft
PRV
Pdown = 10 psig
Inlet
P = 30 psig
El = 10 ft
Inlet
Q = 1 ft3/s
El = 10 ft
Inlet
P = 30 psig
El = 10 ft
Inlet
P = 30 psig
El = 10 ft
PRV
Pdown = 10 psig
PSV
Pup = 20 psig
PSV
Pup = 20 psig
PSV
Pup = 20 psig
Outlet
P = 0 psig
El = 10 ft
Outlet
P = 0 psig
El = 10 ft
Outlet
Q = 1 ft3/s
El = 10 ft
Outlet
P = 0 psig
El = 10 ft
2/24/2003
1:35 PM
Page 455
Inlet
Pressure
Inlet
Flow
X
Outlet
Pressure
X
Outlet Flow
X
X
OK ?
Comments
No
Yes
Yes
Yes
X
No
Yes
2/24/2003
1b.doc
1:35 PM
Page 456
J5
a
P1v
J2
P4^
J3
J4
P2>
P3>
Open
J1
J7
P6>
J5
P1v
J2
P4^
XJ4
J3
P2>
J6
P5>
P3>
J7
P6>
J6
P5>
Closed
Figure 12.21 a-b Top model has reference pressures at J1 and J2.
Bottom model has closed valve at J4 which
isolates the J5 and J6 assigned flows. There is no
reference pressure for J5 and J6 and no unique
solution exists for the bottom model.
2/24/2003
1:35 PM
Page 457
2/24/2003
1b.doc
1:35 PM
Page 458
Figure 12.22
Artificial transients
An artificial transient occurs when the steady-state results do not match
the transient calculation initial step. During the first time step AFT
Impulse tries to bring the mismatched conditions to what it considers
steady conditions which generates a transient pressure. This transient is
not a true physical transient, but a false artificial one based on
mismatched data. If undetected, the artificial transient will interact with
the physical transient being modeled and corrupt the results.
Before AFT Impulse initiates the transient simulation, it performs a
zero time step computation and compares the predicted results with
the results from the steady-state solution. If the difference exceeds the
artificial transient criteria (0.5% by default), then AFT Impulse will
display a warning in the output. If Stop Run if Artificial Transient
Detected is selected, the simulation will be immediately terminated. This
2/24/2003
1:35 PM
Page 459
2/24/2003
1b.doc
1:35 PM
Page 460
APPENDIX A
2/24/2003
1b.doc
1:35 PM
Page 461
2/24/2003
1b.doc
1:35 PM
Page 462
Appendix A
2/24/2003
1:35 PM
Page 463
2/24/2003
1b.doc
1:35 PM
Page 464
Appendix A
The F5 function key also behaves this way for junction specifications
windows.
2/24/2003
1:35 PM
Page 465
2/24/2003
1b.doc
1:35 PM
Page 466
APPENDIX B
Limitations
The size of model you can create with AFT Impulse is limited by the
amount of RAM on your PC. When solving a system, AFT Impulse
creates a square matrix that is double-precision. The size of the matrix is
determined by the total number of tee/wye and branch junctions in your
model.
For example, if you have 100 tee/wye and branch junctions, a 100x100
array is created. With double precision, the amount of RAM needed is
16x100x100, or 160KB. A 1000x1000 array would require 16MB.
In addition to this memory requirements, a relatively small amount of
memory is required for other Steady-State Solver and modeling
activities. AFT Impulse itself requires approximately 15MB of RAM to
run.
Window 95, 98 and ME are limited to displaying a maximum of
approximately 350-400 junction icons as a result of the way in which
these operating systems handle graphics resources. Windows NT, 2000
and XP, on the other hand, has no such limit. One can determine the
amount of free graphics resources by selecting Help from within AFT
Impulse, then clicking on About AFT Impulse and then System Info.
Among the information displayed will be GDI Memory Available
(Graphical Device Interface). As this value approaches zero, Windows
95, 98 and Me will become unstable and, ultimately, crash.
It is important to note that this Windows 95, 98 and ME limitation does
not limit the number of junctions or, as a result, the model size, since
junctions may also be displayed as a solid box, an outline box or not at
all (the junctions are still there even if not displayed, and may be edited
2/24/2003
1b.doc
1:35 PM
Page 467
2/24/2003
1b.doc
1:35 PM
Page 468
APPENDIX C
Installation Issues
Customization files
As you customize AFT Impulse, two new files are created in your
WINDOWS directory:
IMPULSE3.INI
IMP_USER3.DAT
The IMPULSE3.INI file contains the information to customize the
features in AFT Impulse. The IMP_USER3.DAT contains all custom
database information. If you invest significant time in customizing AFT
Impulse, especially in the area of custom databases, you are strongly
advised to keep backup copies of these files.
2/24/2003
1b.doc
1:35 PM
Page 469
2/24/2003
1b.doc
1:35 PM
Page 470
Network installations
AFT Impulse 3.0 requires a number of auxiliary files to run properly.
Because of how 32-bit Windows operating systems work (Windows 9x
and NT), certain files must be registered in the Windows registry. For
software installed directly on the client PC, this situation works well.
However, it creates a dilemma and an inconvenience for software
installed on a local or wide area network.
2/24/2003
1:35 PM
Page 471
2/24/2003
1b.doc
1:35 PM
Page 472
GRAPHS32.OCX
GSW32.EXE
GSWDLL32.DLL
GSWAG32.DLL
2/24/2003
1:35 PM
Page 473
2/24/2003
1b.doc
1:35 PM
Page 474
2/24/2003
1:35 PM
Page 475
2/24/2003
1b.doc
1:35 PM
Page 476
APPENDIX D
2/24/2003
1b.doc
1:35 PM
Page 477
2/24/2003
1b.doc
1:35 PM
Page 478
APPENDIX E
2/24/2003
1b.doc
1:35 PM
Page 479
Contacting AFT
Telephone support
Applied Flow Technology can be reached at:
(719) 686-1000 (Voice)
(719) 686-1001 (FAX)
Web site
You can download the latest maintenance releases of Impulse, find out
what new things are happening at AFT, and get the latest information by
visiting our website at:
http://www.aft.com
E-Mail support
AFT can also be reached by E-mail at:
support@aft.com or www.aft.com
Mail support
You can send mail to AFT at:
Applied Flow Technology
P.O. Box 6358
Woodland Park, CO 80866-6358 USA
You can send courier mail to AFT at:
Applied Flow Technology
400 W. Hwy. 24, Suite 201
Woodland Park, CO 80863 USA
2/24/2003
1b.doc
1:35 PM
Page 480
References
2/24/2003
1b.doc
1:35 PM
Page 481
2/24/2003
1b.doc
1:35 PM
Page 482
Glossary
2/24/2003
1b.doc
1:35 PM
Page 483
2/24/2003
1b.doc
1:35 PM
Page 484
Glossary 485
fluid transient Same as waterhammer.
Graph Results window The graphical window that allows preparation
of plots.
head In AFT Impulse, synonymous with piezometric head and
Hydraulic Grade Line.
head loss The irrecoverable loss in head that occurs in a pipe system.
Head loss is equal to the frictional pressure loss converted into units of
length.
HGL Abbreviation for Hydraulic Grade Line.
history The information on how the solution scheme progresses.
Hydraulic Grade Line The sum of the pressure head and gravity head.
This quantity is relative to the elevational coordinate system specified by
the engineer.
hydraulic transient Same as waterhammer.
ID number The unique number associated with a pipe or a junction.
The ID number is always greater than zero but less than 10,000.
incompressible A fluid whose density changes are small enough that a
constant density can be safely assumed.
infinite pipe A junction which does not reflect and pressure or flow
waves. Available in Assigned Flow and Assigned Pressure junctions.
inspection A feature in AFT Impulse that allows you to obtain
read-only information about objects.
instantaneous waterhammer A waterhammer event that occurs so
rapidly that it is effectively instantaneous, thus yielding the highest
possible pressure surge.
iteration A single step of the Steady-State Solver, during which a new
solution is obtained based on the previous solution state.
Jump A feature that allows lateral changes to be made between
Specifications windows.
junction In AFT Impulse, a pressure/head solution point that connects
and balances flow from pipes.
lock To fix an object's position on the Workspace in order to prevent
accidental movement.
2/24/2003
1:35 PM
Page 485
2/24/2003
1b.doc
1:35 PM
Page 486
Glossary 487
PRV See Pressure Reducing Valve.
Pump configuration A particular pump which contains data for
rotational speed and impeller size.
reference positive flow direction The flow direction you assign to a
pipe, which by convention is positive. Flow in the opposite direction is
considered negative.
relaxation A parameter that affects how the iterative solution scheme
approaches a pipe flow solution.
relief valve A type of valve that is typically closed but opens to release
fluid when a set pressure occurs. Used as a protective device.
resistance A convenient term relating the head loss to flow rate (R =
H/Q2).
Scenario Manager Window that allows one to create and manage
multiple pipe flow model scenarios. These include different equipment,
sizes, and operating conditions.
Solver Either Steady-State or Transient. The Steady-State Solver is the
part of AFT Impulse that contains the pipe steady flow network solution
method. The Transient Solver contains the transient waterhammer
solution algorithms.
Specifications window
are entered.
2/24/2003
1:35 PM
Page 487
2/24/2003
1b.doc
1:35 PM
Page 488
Example models 7
Index
Getting started 6
installation 5
A
Absolute roughness 62, 186, 187, 295,
323
summary of capabilities 2
customizing 288
design factors 192
in Solver 340
AFT Mercury
Converting models from 10
specifying 62
splitting pipes 463
window 189
ADDLOSS.DAT 289
Aligning objects 81
Analysis type
steady-state vs. transient 175
AFT Fathom
Annotation Manager 78
AFT Impulse
converting custom databases from 1.0
9
Annotations 125
engineering assumptions 4
Area Change
2/24/2003
1:35 PM
Page 489
Autosave 286
Specifying 207
BATCH.LOG 173
example 443
Bends
loss factors 343
loss factors 62
source/sink indicator on Workspace
280
loss factors 62
Specifications window 34, 214
loss factors 62
2/24/2003
1:35 PM
Page 490
Index 491
compatibility equations 363, 367, 368,
369, 370, 373, 375, 379
C
Calculator 175
transient 153
graphing 52
Check Valve
in output file 92
theory 395
Check Valve junction 340
Delta Pressure to Re-Open 218
Forward Velocity to Close Valve
218
Continuity equation
steady-state 312
mixtures 29
transient 359
1:35 PM
Colebrook-White 324
2/24/2003
Page 491
types 220
D
Darcy-Weisbach 187, 312, 324
Database Manager 160, 291, 297
databases 290
fluids 292
IMP_DBUSER.LIB 305
Making databases available 298
2/24/2003
1:35 PM
Page 492
Index 493
sharing databases using
DATABASE.LIB 307
DATABASE.LIB 298, 304, 307, 309
Databases 32, 290
Drag-and-drop 3
E
Editing
on Workspace 79
EGL 106
Elbow
See Bend 343
Elevation 106, 200
default 273
Event messages 103, 420
Event transients 414
Example models 7
object status 80
2/24/2003
1:35 PM
Page 493
non-Newtonian 326
Relative roughness 323
fluid transients
orifice 227
See waterhammer
fluidhammer
See waterhammer
Fouling 187
Hazen-Williams 324
2/24/2003
1:35 PM
Page 494
Index 495
Global Pipe Edit window 267
Glossary 483
impedance 363
Impulse Examples.hlp 7
Graphics Server
problems loading 473
Gravitational acceleration 5, 145
Grid
On Workspace 24, 115, 279
Grouping objects 80, 125
generation 205
hydraulic transients
See waterhammer
Hydraulically smooth 62, 188
2/24/2003
1:35 PM
Page 495
Notes 207
Number 199
Pump 232
Junction icons
Status 208
Branch 214
Valve 263
connectivity 126
Elevation 200
creating 124
format 197
defining 128
displaying on Workspace 83
Name 199
2/24/2003
1:35 PM
Page 496
Index 497
displaying transient results in Output
window 103
editing 79, 126
finding 163
database connections 89
Elasticity 231
2/24/2003
1:35 PM
Page 497
Additional Losses 87
customizing 88
Pipe table 87
NPSHR 237
NPSP 140
NPSPR 237
NPSP See NPSH
Momentum equation
One-dimensional flow 59
steady-state 312
transient 359
Orifice
N
Network databases 201. See Database
Manager
Newton-Raphson method 3, 312, 314,
321
transient solution for gas
accumulators 400
1:35 PM
Page 498
database connections 97
Format and Action 95
General Output Control 94
Pipe and Junction parameters 92, 94
2/24/2003
scenarios 160
Index 499
Show Selected Pipes/Junctions 96
using saved format files in Database
Manager 302
customizing 287
Event messages 420
exporting output 90
number formatting 95
setting defaults 97
updates 104
Overview of AFT Impulse 7
P
Parallel Pipes
elevation 273
2/24/2003
1:35 PM
Page 499
Notes 194
displaying on Workspace 83
editing 125
finding 163
ID numbers 122
loss factor 62
Status 194
wavespeed 189
segmenting 75
Pipe table
stretching 125
Point loss 62
connectivity 126
creating 121
Popup menu 80
defining 128
2/24/2003
1:35 PM
Page 500
Index 501
Pressure
NPSHR 237
Primary windows 7, 71
2/24/2003
1:35 PM
Page 501
R
README.TXT 5
References 481
loss factors 62
Relative location 18
Exit 394
Reynolds number
Internal 394
2/24/2003
1:35 PM
Page 502
Index 503
same as parent junction 201
same as parent pipe 184
Section Pipes window 43, 135, 147,
152, 176, 366, 457, 458
junctions 206
No reflections - infinite pipe 169
Pipe/Junction Symbol on Workspace
281
Transient 169
Workspace appearance 278
Specifications windows 86, 179
accessing 179
inspecting in 131
Jump feature 180
Splitting pipes 124, 463
Spray Discharge junction 338
Solver
discussion of Steady-State and
Transient 46
2/24/2003
1:35 PM
Page 503
e-mail 480
Spreadsheet
surge flow
exporting data to 90
See waterhammer
Stagnation pressure
explanation 315
Static pressure
explanation 315
Status Bar 132
Steady Solution Control window 134,
135
absolute tolerance 331
convergence 330
false convergence 335
flow rate relaxation 334
Matrix Method 336
maximum iterations 336
parameters 329
pressure relaxation 335
relative tolerance 331
relaxation 333
relaxation example 334
tolerance 330
tolerance example 332
Submerged pumps 381
Support 479
2/24/2003
1:35 PM
Page 504
Index 505
balance verification 321
surge transients
See waterhammer
System pressure 155, 442
System Properties window 28, 134,
138
AFT Standard fluid database 141
atmospheric pressure 144
Chempak database 141
creating mixtures 141
Terminology 1
time step 365, 458, 459
Time-based transients 413
Tip of the Day 286
scenarios 160
Start and stop time 151
Tolerance
1:35 PM
2/24/2003
Page 505
Transient Data
branch 371
Unspecified fluid 28
pump 381
2/24/2003
1:35 PM
Page 506
Index 507
variable pipe resistance 152
Verification
Verification against published results
7
Verifying steady-state network
solutions 321
View Results button See Solution
Progress window
Viscosity corrections 247
Visual Report Control window 112,
288
scenarios 160
2/24/2003
1:35 PM
Page 507
2/24/2003
1:35 PM
Page 508