Pipes

How to Use: Incompressible Pipes (Transient) Page 1 of 17
Incompressible Pipes
How to Use: Transient
To select a particular item, click on the required link below:
General Points
Available Pipe Models
Determining the Pipe Model
Rigid Pipe Model
Compliant Pipe Model
Elastic Pipe Model
Variable Wave Speed Pipe Model
General Points
1. The pipe profile curve allows you to represent the elevational profile of the pipe from node to
node. The elevation data is NOT used in the calculation.
2. The profile curves can be overlayed onto transient result plots, thus enabling a pictorial
representation of pipe pressure along the pipe length, relative to pipe elevation. A profile curve
can only be applied to cylindrical pipes.
3. In a transient simulation, the four pipe types, circular, rectangular, hexagonal and prismatic
are all treated by the simulation in a similar manner.
Available Pipe Models
There are four available pipe models; namely:
Rigid
Elastic
Compliant, and
Bubbly.
Each of these four models is discussed as follows:
mk:@MSITStore:C:\PROGRA~2\FLOWMA~1\FLOWMA~1\Help\Pipes.chm::/FTSTPipe... 6/5/2018
Rigid Pipe Model
1. This models incompressible flow through a pipe where fluid and pipe elasticity are ignored. The
rigid pipe model can reduce computational time for mass oscillation types of simulation, and
allows you to account for frictional and fluid inertia effects in pipes where elastic effects are
negligible (e.g. short lengths in long systems).
2. The rigid model should be used when:
a. The pipe length is short compared to the distance travelled by a pressure wave during one
time step, and
b. The change in flow conditions takes place over a time considerably longer than the
pipeline period 2L/a (i.e. the time required for a pressure wave to travel from one end of
the pipe line to the other and return).
Compliant Pipe Model
This models flows through a pipe when one or more of the following is important:
1. There is slow transient behaviour and the rigid model is not suitable.
2. The energy stored in the pipe because of the compliance of the fluid and/or the pipe walls is
important.
3. There is a large pressure drop across a relatively short pipe.
4. Long pipes are being pressurized slowly.
Dynamic and Continuity Equations
Integration of the compliant pipe dynamic and continuity equations (see the simulation Description
section for details) assumes that the mass flow rate varies slowly with time. However, at a
discontinuity in the gradient of the Pressure v Flow curves, small instabilities in mass flow rate
have been observed. Tests carried out by Flowmaster indicate that this instability reduces after a
few time steps.
Compliant pipes should only be used for slow transients. A slow transient is defined as one which
occurs in a time much greater than the pipeline period 2L/a, where L is the pipe length and a is the
wave speed. Compliant pipes should not be used in networks where zero or near zero flow
conditions will exist.
Bulk Modulus
1. The Bulk Modulus of the pipe (βp) is optional.
2. The effective Bulk Modulus, βe, is a combination of the fluid Bulk Modulus (determined from
the fluid properties) and, if set the Bulk Modulus of the pipe, βp which is set on the Component
Data Form. If the Liquid Properties Data Sheet fluid Bulk Modulus (βf) is not set, then the fluid
Bulk Modulus is calculated from the wave speed (α) which is given in the Component Data
Form:
Equation 1
3. The effective Bulk Modulus is then calculated as follows:
Equation 2
If βp is not set, then βe is simply equal to βf:
Equation 3
Where:
βe = Effective Bulk Modulus

βf = Fluid Bulk Modulus
βp = Bulk Modulus of the pipe*
α = Wave speed
*This is not the Bulk Modulus of the pipe material. It is the volumetric stiffness of the pipe
envelope:
Where:
t = Pipe wall thickness

E = Young's (Elastic) Modulus of the pipe material
D = Internal diameter of the pipe
C = Pipe constraint parameter typically 1, (1 - µ/2) or (1 - µ2) where µ = Poisson's Ratio. For
most engineering applications, it is reasonable to appropriate C = 1, in view of other
uncertainties which arise (such as Fluid Stiffness due to gas content).
General Rules
1. If βf is specified in the Liquid Properties Data Form, then it is used.
Elastic Pipe Model
1. This models flow through an elastic pipe, in which pressure and flow variations require a finite
time to travel the full length of the pipe. The elasticity of the fluid and the pipe are accounted
for by a combined speed of sound for both fluid and pipe.
2. Pressure waves travel through the pipe at the speed of sound. The passage of these waves is
calculated using the Method of Characteristics, assuming a constant wave speed.
3. Pressure changes in the system are directly related to changes in velocity (flow rate). In
transient events, flow rate changes cause pressure fluctuations (and vice versa). For rapid
transient events (where events occur faster than the periodic time 2L/a), the pressure rise is
given as follows:
Equation 4
or
Equation 5
Where:
∆P = Pressure change in fluid

∆H = Head change in fluid
∆v = Velocity change
α = Wave speed
g = Gravity
ρ = Liquid density
Modelling
Flowmaster uses the Method of Characteristics to calculate the propagation of pressure waves
along pipes. A fixed Distance v Time grid is used. Flowmaster will perform the following checks on
this model:
1. To ensure that the Number of Pipe Reaches (S) is an integer, then:
a. The input wave speed is adjusted as follows:
Equation 6
Where:
ψ = The amount of wave speed adjustment. The maximum amount of adjustment is

limited to ± 15%
b. A warning message is issued to warn users of the wave speed adjustment.
2. If after the wave speed adjustment (Max ± 15%), S is still not an integer, then the Rigid pipe
model is used, and a warning message is issued.
Time variations in pressure and flow rates are calculated through the pipe for vertical, horizontal or
inclined pipes, taking frictional effects into account. Gas release and re-absorption resulting from
transient activity is not taken into account.
Unsteady Friction Modelling
The Problem
Fast transients in fluid flow in pipes and ducts often occur in water systems during valve closure or
pump shutdown. Accurate predictions of pressure responses in such systems are important for safe
design.
Traditionally, in fast transient modelling, the steady or 'quasi-steady' friction models are
incorporated into the modelling algorithm. This is also true for Flowmaster. Flowmaster supports
both steady and quasi-steady friction models for transient flow. In the pipe model, the 'Friction
Data' sub-form is designed for this purpose. The following table shows the friction options and the
model that is used:
Table 1: Friction Option v Model

No Friction Option Model Used
1 Colebrook-White Equation Approximation Quasi-steady Model.
2 Hazen-Williams Equation
3 Fixed Friction factor Steady Friction Model.
The steady or quasi-steady friction is satisfactory for most engineering flows, namely; slow
transients. However for fast transients, (e.g. water hammer), the steady or quasi-steady friction
models have shown significant discrepancies in the attenuation and phase shift of pressure waves
when the computational results are compared with known measurements.
Unsteady Friction Model
To improve the accuracy of fast transient modelling, the following unsteady friction model is
introduced:
Equation 7
Where:
α = Wave speed
D = Pipe diameter
f = Friction factor
fq = Steady or quasi-steady friction factor
k = Friction coefficient
V = Flow velocity
t = Time
x = Distance
The above equation gives the correct sign of convective term for all possible flow and water
hammer movement directions, for either acceleration or deceleration phases.
The friction coefficient, k, is evaluated as:
Equation 8
Where: C* is Vardy's shear decay coefficient which is dependent on the flow regime:
For laminar flow it is defined as:
For turbulent flow, it is defined as:
Equation 9
Application of the Unsteady Friction Model
The unsteady friction model is incorporated into the elastic pipe model. It is recommended that the
model is only used to model fast transient events, such as water hammer.
As can be seen from Equation 7, the unsteady friction factor is dependent on the instantaneous
convective acceleration ∂V/∂x. For reasons of modelling accuracy, it is reasonable to assume that
the time step is properly selected so that the right number of internal nodes are available.
It is difficult to quantify the number of internal nodes required, as this dependent on the problem
being analysed. However, a good approach is to check the solution dependency on the time step, if
possible.
Method of Characteristics
The Method of Characteristics uses a Distance-Time grid to track the passage of the pressure
waves. The particular implementation in Flowmaster is such that a fixed distance time grid is used.
You MUST specify the time step to be used for the simulation. This is done on the simulation Form
(see 'Simulation Description: Transient' for details). There is a relationship between time step,
length of pipe and speed of sound such that:
Equation 10
Where:
S = No. of internal calculation reaches in the pipe

L = Specified length of the pipe
α = Specified wave speed
∆t = Specified time step
It is relatively simple to select a simulation time step, ∆t that satisfies Equation 10 for single pipe
systems. However, when we are dealing with complex pipe systems of two or more pipes, then the
selection of ∆t requires a certain amount of care. The following approaches can be used to select
the time step.
Approach 1 - Using the Time Step Calculator
The Time Step Calculator is a wizard that calculates the time step, based on the following
information:
Default wave speed
Allowed % diff in wave speed
Minimum number of reach lengths
Minimum pipe length
Convergence factor.
Using this information (which you can adjust), the wizard calculates the 'best' time step for all the
Elastic Pipes in the network, providing they are longer than the minimum pipe length. For details
about how to use the wizard, click-on the following link:
Using the Time Step Calculator - Overview
Approach 2 - Adjusting the Wave Speed
As recommended in Reference 1, for complex pipe systems, it is often necessary to adjust the
input wave speed in order to satisfy Equation 10. In as much as the wave speed is probably not
known with great accuracy, it may be permissible to adjust the input wave speed within a certain
limit (15%). In the first instance, you should try to find a time step that satisfies, or at least
approximates Equation 10.
If the input time step satisfies Equation 10, then it is used and no further actions are required.
However, if the input time step does not satisfy Equation 10 for a particular pipe, then the wave
speed for that pipe can be adjusted in the following way:
1. Use the input time step to calculate the number of internal reaches:
Equation 11
The calculated S' is not an integer, if Equation 10 is not satisfied for the particular pipe.
2. Calculate the new adjusted wave speed as follows:
Equation 12
3. Next, check to see if the wave speed adjustment is within the 15% tolerance, as recommended
in Reference 1. If this tolerance is not acceptable, then you should use approach 1.
Equation 13
Although there is essentially no limit on the magnitude of the wave speed adjustment, when
using the Time Step Calculator, it is strongly recommended that you keep within the ± 15%
range.
4. The maximum number of Internal Sections is not restricted. However, you should be aware
that the flows and pressures (and mass ratios of gases for variable wave speed models) are
calculated and recorded in the database for each internal node. Obviously, this will increase
the amount of file space used and slow down access to graphs and other results.
5. It is possible to vary the wave speed a to meet the criterion. This may lead to spurious
pressure surges being created as the pressure wave travels between pipes of different wave
speed. (This of course may be relevant where the pipe material changes.)
6. To avoid using a very small time step, you may adjust the wave speed in the pipe in
accordance with Equation 10, providing that the adjustment is not significantly larger than the
uncertainty of the wave speed.
7. Flowmaster does not require the inclusion of pipes between every component. You should
therefore consider whether any 'short' pipes need to be included in the network, or whether
they could be modelled as Rigid Pipes (see Rigid Pipe Model for details), or as Discrete Losses.
You should remember that the determination of the wave speed for the system is generally
very inaccurate. Therefore, unnecessary detail in the network model will give no benefit to the
accuracy of the simulation.
8. In situations where peak pressures are very important, it is recommended that consideration is
given to modelling the system with different values of wave speed. Generally, high values of
wave speed will give the 'worst' maximum pressure predictions. However, in certain
circumstances this may not be the case.
Figure 1 shows the wave velocity in pipes for various pipe materials:
Figure 1: Pressure Wave Velocity in Pipes filled with Water (Kwater = 2070 N/m2)
The wave speed will be influenced by the following factors:
Pipe material
Pipe wall thickness/diameter ratio
Pipe support
Fluid bulk modulus
Fluid density
Gas content
Temperature.
If the bulk modulus is NOTSET, then the wave speed is calculated as follows:
Equation 14
If the bulk modulus is SET, then the wave speed can be calculated from the following equation:
Equation 15
Where:
ρ = Fluid density
βf = Fluid Bulk Modulus
βp = Bulk Modulus of the pipe which is calculated as:
Where:
D = Internal diameter of pipe

t = Pipe wall thickness
E = Young's (Elastic) Modulus of pipe material
βp = Pipe constraint parameter typically 1, (1 - µ/2) or (1 - µ2) where µ = Poisson's Ratio. For
most engineering applications, it is reasonable to appropriate C = 1, in view of other
uncertainties which arise (such as Fluid Stiffness due to gas content).
The Method of Characteristics uses a time distance grid as shown in Figure 2. A description of the
method is beyond the scope of this guide. This can be found in most text books on transients.
The value of calculated wave speed is then applied to Equation 10 to determine the value for S. If
this value does not meet the same S-criterion required for the user-specified wave speed, then the
following message is generated:
Pipe no. 0001 cannot use variable wave speed, because the ambient wave speed
(1234.56 m/s), length and t/step are incompatible.
Figure 2: Method of Characteristics - Time Distance Grid
Conditions are known at t = 0 along the length of the pipe. Perturbations are then introduced into
the system (valve movement, pump trip) and the information is propagated along the pipe using
Characteristic Lines which have a slope of the wave speed (α). In this way, it is possible to predict
conditions at the 0 + ∆t, the next time step.
Once these conditions are known, the whole process can be repeated to progress to the next time
step. A detailed description can be found in the following reference:
'Fluid Transients in Systems', Wylie & Streeter, Published by Prentice Hall 1993, ISBN 0-13-
322173 -3.
Vaporization
As pressure in the fluid falls, it will reach a physical limit below which it cannot drop, i.e., the
vapour pressure of the fluid. This may occur anywhere in a piping network. In Flowmaster
vaporization is allowed to occur at nodes and in gas admission valves (see the 'Valves' help file for
details).
Vaporization is not modelled within a pipe but a warning is output during the simulation and is
written to the warnings and log files should pressure fall to this limit. However, the pressure in the
pipe will be allowed to drop below this limit. Where pipe vaporization warnings are issued the
results from this simulation should be carefully examined.
The vapour pressure can be set as a data item for each pipe in the network.
If it is necessary to model vaporization, for example at a high point on a pipeline, a node should be
included in the model. This will then allow modelling of vapour cavity formation, growth and
collapse to occur, providing the correct nodal data is set.
The node data should be input as shown in Table 2 to model vapour cavity formation and growth at
a node. When set to Auto Vaporization, the node will appear red in the network diagram.
Table 2: Node Data Settings for Modelling Auto Vaporization

Node Level Level Datum Temp. Wet/Dry Gap. Press Auto Vap.
n Value 0.0 NOTSET Wet Value Yes
If Auto Vaporization is set to 'Yes' and no value for vapour pressure is given, then the vapour
pressure value for the liquid in the simulation will be used.
Line Pack
In long pipelines the frictional pressure gradient may cause significant pressure increases over the
initial 'surge pressure'. Line Pack will be automatically modelled by Flowmaster provided a frictional
value has been entered.
Variable Wave Speed Model
General
Within the high-pressure parts of diesel injection systems, fuel pressure changes rapidly between
vapour pressure and ~1500 bar during the injection cycle.
Under very low pressure, vapour cavities may form in the fuel, or free gas may be released from
solution. The bulk modulus of the gas is much smaller than liquid bulk modulus, i.e., the gas has a
much higher compressibility and this significantly reduces the wave speed.
Over higher pressure ranges (~5 to 1500 bar), vapour cavities collapse and free gas is absorbed or
compressed to a small fractional volume, so that the contribution from the gas towards the overall
compressibility of the mixture is small. Variations in liquid bulk modulus with temperature and
pressure then dominate in determining the wave speed.
Modelling
1. This models an elastic pipe filled with a bubbly gas-liquid mixture with a modest gas content
(in the order of 1% by volume or less). This can occur when low pressure waves cause
dissolved gas to come out of solution. Elasticity effects are modelled by the Method of
Characteristics modified to take into account variations in wave speed, which is a function of
the void fraction and the pressure inside the pipe.
2. The model calculates the quantity of gas released from solution when the mass of dissolved
gas is greater than the equilibrium value. The quantity of gas re-absorbed when the gas
pressure rises sufficiently, assuming that the free gas is evenly distributed is also modelled.
3. Time variations in pressure and flow rates are calculated through the pipe for vertical,
horizontal or inclined pipes, taking frictional effects into account.
4. Values for gas release and absorption factors are difficult to determine. They will be dependent
on the gas and fluid properties, ambient temperature and pressure conditions, and the rate of
change of pressure caused by the pressure waves.
5. Generally, gas release will damp pressure surge activity and reduce peak pressures. However,
this effect is generally found after the passage of the first pressure wave. Therefore, initial
peak pressures will generally be similar to those obtained using the elastic pipe model.
6. It is not recommended that variable wave speed model be incorporated with vapour cavity
modelling. The two phenomena are generally not encountered at the same time in real
systems.
7. You should remember that gas will find it far easier to come out of solution than to go back
into solution. Therefore, the two factors for gas release and gas absorption should not be the
same.
Equilibrium Constant (E)
This is defined by the following equation:
Equation 16
Where:
W = Molecular Weight ratio of the gas to the liquid
h = Henry's Law constant
Henry's Law constant is given in atmospheres/mole fraction and is temperature dependent.

Therefore the value of h used MUST correspond to the temperature entered on the 'Ambient
Conditions' sub-form.
Henry's Law constant is also dependent on partial pressure, when the partial pressure of the solute
is > 1 atm. The value of h is only valid over a small pressure range, you should always select an
appropriate value and consider the validity of the results.
Gas-Liquid Equilibrium Data
Figures 3 shows values of h for different Aqueous Solutions, for different values of temperature.
Figure 4 shows values of Solubility for Hydrocarbon Solvents, for different values of temperature.
Figure 3: Solubility of Various Gasses in Water
Figure 4: Solubility Bands for different Hydrocarbon Solutions
Gas Release Factor (Kr) and Gas Absorption Factor (Ka)
The variable wave speed model is invoked when the Gas Release and the Gas Absorption Factors
are both SET. Gas release may occur when Kr ≥ 10-10 if dissolved gas exists, and gas absorption
may occur when Ka ≥ 10-10 if free gas exists. Once gas release or absorption is occurring, the gas
is assumed to be instantaneously distributed along the entire length of the pipe, irrespective of
where the pressure wave may have travelled to. The value of Kr for air/water solutions is difficult to
deduce from available literature, and appears to depend very strongly on the degree of agitation of
the liquid. (This is also true CO2 in water as in the behaviour of 'fizzy' drinks).
Extremely agitated air/water solutions give values as high as Kr = 0.2 s-1, whereas steady flows of
less than 1 m/s in small bore (1 inch, 25 mm) pipes gives values down to Kr = 0.001 s-1. A value of
Kr = 0.01 s-1 is thought to be satisfactory for large scale industrial piping systems, but a range of
probable values should always be tested, where this parameter is considered to significantly affect
the results of the simulation. Typically, Ka << Kr, and in most cases Kr can be set to zero.
Vapour Cavitation
You should NOT attempt to model gas absorption and gas release at the same location as
modelling vapour cavity formation, growth and collapse at nodes.
Free Gas
Free gas has a dramatic effect on the wave speed of the fluid. For example, for water, 0.1% air will
reduce the wave speed from 1200 m/sec to 600 m/sec.
Enhanced Liquid Properties
To enable fluid viscosity, density and bulk modulus to vary with the local temperature and
pressure, the following three surfaces have been developed. They are:
Bulk Modulus v T & P
Dynamic Viscosity v T & P
Density v T & P
The surfaces will use the ambient temperature specified on the 'Ambient Conditions' sub-form,
providing that the Heat Transfer option has not been set. This operation is different from previous
versions of Flowmaster where the fluid properties were obtained from a constant value, or from a
curve at ambient temperature.
If you implement surfaces for fluid properties that do not match the current constant value, or
curves at the ambient temperature which is set on the 'Ambient Conditions' sub-form, then the
'Ambient' fluid property values used by components will be different.
Data Priority
When set, these surfaces take precedence over any curves, Eagle data or other data items which
define the same properties. Tables 3 and 4 summarize the logic used for calculating Bulk Modulus,
Dynamic Viscosity and Density:
Table 3: Data Settings for Settings for Dynamic Viscosity, µ
µvTvP µvT Constant

Action
Surface Curve µ
Use µ vs T vs P Surface,
SET
Warning: Fluid name - Surface used for viscosity
SET
NOTSET Use µ v T v P Surface,
SET
SET Use µ v T v P Surface,
NOTSET
NOTSET Use µ v T v P Surface,
No warning issued
Use µ v T Curve,
Warning: Fluid name - Curve used for Viscosity
SET
SET Note: This warning message is a change to existing
functionality
NOTSET NOTSET Use µ v T Curve,
No warning issued
SET Use constant µ

NOTSET No warning issued
NOTSET Error halt
The above settings are also applicable for Density, ρ, but substituting ρ for µ.
Table 4: Data Settings for Settings for Bulk Modulus, βf
βf vs T vs P Surface Constant βf Action
SET Use βf vs T vs P Surface

Warning: Fluid Name - Surface used for bulk modulus
SET
NOTSET Use βf vs T vs P Surface
No warning issued
SET Use constant βf

NOTSET No warning issued
NOTSET Error Halt

Pipes

Uploaded by

Copyright:

Available Formats

Pipes

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pipes

Uploaded by

Copyright:

Available Formats

How to Use: Incompressible Pipes (Transient) Page 1 of 17

How to Use: Transient

To select a particular item, click on the required link below:

Available Pipe Models

There are four available pipe models; namely:

Each of these four models is discussed as follows:

Rigid Pipe Model

2. The rigid model should be used when:

Compliant Pipe Model

3. There is a large pressure drop across a relatively short pipe.

4. Long pipes are being pressurized slowly.

Dynamic and Continuity Equations

1. The Bulk Modulus of the pipe (βp) is optional.

3. The effective Bulk Modulus is then calculated as follows:

If βp is not set, then βe is simply equal to βf:

βe = Effective Bulk Modulus

t = Pipe wall thickness

1. If βf is specified in the Liquid Properties Data Form, then it is used.

Elastic Pipe Model

∆P = Pressure change in fluid

1. To ensure that the Number of Pipe Reaches (S) is an integer, then:

a. The input wave speed is adjusted as follows:

ψ = The amount of wave speed adjustment. The maximum amount of adjustment is

b. A warning message is issued to warn users of the wave speed adjustment.

Unsteady Friction Modelling

Table 1: Friction Option v Model

Unsteady Friction Model

The friction coefficient, k, is evaluated as:

For laminar flow it is defined as:

For turbulent flow, it is defined as:

Application of the Unsteady Friction Model

S = No. of internal calculation reaches in the pipe

Approach 1 - Using the Time Step Calculator

Default wave speed

Allowed % diff in wave speed

Minimum number of reach lengths

Minimum pipe length

Using the Time Step Calculator - Overview

Approach 2 - Adjusting the Wave Speed

2. Calculate the new adjusted wave speed as follows:

The wave speed will be influenced by the following factors:

Pipe wall thickness/diameter ratio

Fluid bulk modulus

D = Internal diameter of pipe

Figure 2: Method of Characteristics - Time Distance Grid

Table 2: Node Data Settings for Modelling Auto Vaporization

n Value 0.0 NOTSET Wet Value Yes

Variable Wave Speed Model

Equilibrium Constant (E)

This is defined by the following equation:

W = Molecular Weight ratio of the gas to the liquid

h = Henry's Law constant

Henry's Law constant is given in atmospheres/mole fraction and is temperature dependent.

Gas-Liquid Equilibrium Data

Figure 3: Solubility of Various Gasses in Water

Figure 4: Solubility Bands for different Hydrocarbon Solutions

Gas Release Factor (Kr) and Gas Absorption Factor (Ka)

Enhanced Liquid Properties

Bulk Modulus v T & P