Professional Documents
Culture Documents
10 Analysis of Pipe Flows: The Moody Chart
10 Analysis of Pipe Flows: The Moody Chart
| |
= =
|
\
l
1/
1
n
c
u r
V R
| |
=
|
\
( ) ( )
* *
2 1
0
L
p p g h =
.
ENG243 Fluid Mechanics 2010 page: 2
The Colebrook formula
Using Moody chart to read off the value of f was done in
the pre-calculator days. There are many equations that
curve-fitting the Moody chart and allow more accurate
value of f to be calculated. Amongst these, the Colebrook
formula is believed to give the best result.
10
1 / 2.51
2.01
3.7
log
D
D
f Re f
(
= +
(
(
note that f appears on both sides of the equation hence an
iteration process is required to obtain its value.
To begin with the iteration, the first value of f is derived
from the Miller formula:
2
0
10
/ 5.74
0.25
3.7
log
0.9
D
D
f
Re
(
= +
`
(
)
or from Eqn 8.35b
1.11
10
1 / 6.9
1.8
3.7
log
D
D
Re f
(
| |
= +
(
|
\
(
Note that the
logarithm is base 10
Actually f from one
of these two
equations is
sufficient in most
practical purposes.
The iteration process begins with a given value of Re
D
and
/D.
The first estimated value of f is derived from Eqn 8.35b or
Miller formula.
This value of f is then substituted into the right hand side of
the Colebrook formula so that the f in the left hand side is
the updated value.
We continue to use the Colebrook formula in a loop until a
desirable accuracy is achieved.
Three iterations will
produce sufficient
accuracy in most
situations.
Minor Losses -- The Loss Coefficient -- K
There are losses of total pressure (related to energy) for
flow over various pipe connections such as pipe bends,
nozzles, pipe junctions, diffusers, ... etc.
For a drop of total pressure (p) over these rather short
connections, we define a loss coefficient, K, such that:
2
1
2
p K V =
and
2
.minor
2
L
V
h K
g
=
In the case where the velocities are different between the
upstream and downstream station, the value of K will
depend on the choice of the velocity in one of the two
sides. For most situation, the higher velocity is chosen to
define K
Section 8.4.2
Here V is the
average velocity;
that is Q/A.
ENG243 Fluid Mechanics 2010 page: 3
Consider a pipe system as shown above. Let the subscripts
of the minor lost coefficient K
n
be related to the numeric as
indicated in the diagram and let the length
ab
l be the length
of the pipe between (a) and (b).
The energy equation is given by:
( )
* * 2 2
1
2 1 2 . .minor 1 2 L major L
p p gh gh V V o + =
where
( ) [ ]
.
2
1
2 1 1 3 3 4 4 5 5 6 6 7 7 8
/
L major
gh
V f D
=
+ + + + + l l l l l l
and
( ) [ ]
2
1
2 .minor 1 1 3 4 5 6 78 L L
gh V K K K K K K K = + + + + + +
the total head loss in this system is:
( ) ( ) ( )
* * 2 2
1
2 1 2 . .minor 1 2
/ /
L major L
H p p g h h V V g = = +
the power consumed by the system is: ( ) Q g H Watt
Example 8.8
Note that K
L
is based
on the pipe velocity
V
1.
K
1
is referring to the
entrance loss (if
any).
Velocities appear in
the energy equation
because the velocity
at inlet and exit are
not the same.
Piezometric
pressures are use in
the equation.
The problem will be
simpler if the friction
factor, f, is given.
On the other hand,
the pipe roughness is
required and
together with Q (to
get Re) to allow f be
evaluated from the
Colebrook formula.
Section 8.4.2 contains values of K for various pipefittings
and configurations.
The minor loss coefficient, K, can be converted to an
equivalent length using the relationship:
/
e
K f D = l that is: /
e
KD f = l
Hence the minor losses in the problem above may be
written in terms of equivalent lengths:
( ) [ ]
2
1
2 .minor 1 1 3 4 5 6 78
/
L L
gh V f D = + + + + + + l l l l l l l
The equivalent
length is meaningful
if the system has a
constant pipe
diameter.
ENG243 Fluid Mechanics 2010 page: 4
Section 8.5 illustrates a wide range of pipe flow
applications.
Section 8.5; 8.5.1
Multiple pipe systems
(1) Pipes in series
For pipe of different diameters connected in series, there
will be minor losses at each pipe junction. These minor
loss coefficients can be found from commercial catalogues
for pipefittings and these coefficients may be base on the
upstream velocity of the junction or downstream.
Lets assume that the minor losses of pipe junction are
based on the upstream velocity. Let n be the subscript that
denotes the properties associate with the nth pipe. For
three in series as shown below the total head loss (major +
minor) is expressed as:
( ) [ ]
( ) [ ] ( ) [ ]
2
1
2 1 1 1 1 1 2
2 2
1 1
2 2 2 2 2 2 2 3 3 3 3 3
/ /
/ / / /
L
H V f D K g
V f D K g V f D g
= +
+ + +
l
l l
and again there is a mass conservation equation:
( ) ( ) ( )
2 2 2
4 4 4 1 1 2 2 3 3
D V D V D V
= +
(2) Pipe in parallel as shown above
In this case, the head loss for all the three pipes are the
same and all equal to
A B
z z (i.e., pressure at A & B are
atmospheric). Hence:
[ ]
2
1
/
2
L n n n n n entrance n exit
h V f D K K
g
= + + l
for n = 1, 2 & 3
Section 8.5; 8.5.1
The energy equation
between A and B is
given by:
2 2
2 1
*
1
2
L
p
g
V V H
g
=
( +
neglecting the
entrance and exit
losses.
Note that velocities
at A and B are zero.
ENG243 Fluid Mechanics 2010 page: 5
(3) Y- connections Figure 8.36
Here the flow in pipe 1 is spilt into pipes 2 & 3. Strictly
there will be minor loss between the flow from pipe 1 to
pipe 2 and from pipe 1 to pipe 3. However if these pipes
are very long the minor losses are small in comparison with
the major losses and hence neglected in this study.
The first equation is the mass conservation:
1 2 3
Q Q Q = +
Nest is to consider the head loss equations. There are two
loops: from A to B via pipes 1 & 2 and from A to B via
pipes 1 & 3. These two head loss equations are:
2 2
1 1 1 2 2 2
1 2
0
2 2
A B
A B
p p V f V f
z z
g g g D g D
( ( (
+ + + =
( ( (
l l
and
2 2
3 2 3 1 1 1
1 3
0
2 2
A B
A B
V f p p V f
z z
g g g D g D
( ( (
+ + + =
( ( (
l l
Rewrite the conservation equation as:
1 1 2 2 3 3
AV AV AV = +
These three equations enable the solution for the three
unknown velocities.
Section 8.5.2
Note that here we
write out pressure
and height instead of
the piezometric
pressure.
Usually the pressure
at A and B are
atmospheric and are
set to zero.
If the friction
coefficients are not
given but have to be
derived from
Colebrook formula
through the use of
pipe roughness, the
problem becomes
more difficult to
solve.
ENG243 Fluid Mechanics 2010 page: 6
Pipe flow rate measurement
One of the traditional methods of measure flow rate in pipe
is to use an obstruction in the pipe and calibrate the flow
rate against the pressure drop over the obstruction.
Flow through an obstruction, as shown above, is
asymmetric: On the contraction side (side 1) the flow
follows the boundary. However, at the expansion side
(side 2) the flow separates from the wall and become a jet
within the fluid in the pipe. Hence that is a large loss of
pressure head over the relatively short section.
The most common method is the use of an orifice plate.
The orifice plate is a simple device and can be manufacture
in a mechanical workshop. The plate has a concentric
circular hole. The plate is sharpened as shown above. In
according to flow measurement standard, pressure tappings
are located at D upstream and D/2 downstream of the plate.
The volume flow rate is related to the difference of
piezometric pressure by the formula:
( ) 2 * Q CA g p =
and the calibration constant, C, is given by:
2.5
2.1 8
0.75
91.71
0.5959 0.0312 0.184
Re
d
C
= + +
where = (d/D)
2
Section 8.6
The exit flow will
follow the wall if the
wall expands at an
angle not more than
7
0
as in the case of
venturi meter.
This type of flow
measurement is
economical to make
but it may incur
unacceptable loss of
pressure head.
Detail for flow
measurement in pipe
flow can be found in
Australian standard
AS2360 and British
standard BS1042
D
D/2
d
Flow
D
p*
D
ENG243 Fluid Mechanics 2010 page: 7
The Venturi meter
The venturi meter is well contoured such that flow are
contracted and expanded with very little head loss.
However there is a pressure drop between the inlet section
and at the throat of the venturi. This is due to the increase
of kinetic energy. For an ideal case with no head loss, the
piezometric pressure drop between the inlet and the throat
is given by:
( )
2
2 2 2
2
1 1
* 1
2 2
d
d D D
D
V
p V V V
V
| |
= =
|
\
As
2
d
D
V D
V d
| |
=
|
\
hence
4
2
4
1
* 1
2
D
D
p V
d
| |
=
|
\
the flow rate ( )
2
4 d
Q D V
=
In practice, there may be a need for a discharge coefficient,
C, to cover all the losses etc. Hence:
actual
Q CQ = .
The venturi meter
incurs very little
head loss to the flow.
However, it takes up
more space which
may not be practical
in many cases. In
addition it cost much
more than an orifice
plate.
The Rotameter
The rotameter is an ingenious device. It directs flow
vertically upward along a slightly diverging cylinder. A
float will be suspended in the cylinder at a position where
the fluid momentum on it is balanced by its weight. See
Figure 8.48. The height of the float indicates the flow rate.
Other flow meters and sensors
Recently many elector-mechanical flow meters have been
developed. In general the flow rate are calibrated with
voltage output of the device and can be interfaced with
computer data logging systems.
2
4
4
2 *
4
1
p
Q D
D
d
=
| |
|
\
ENG243 Fluid Mechanics 2010 page: 8