Chapter 5 - Lecture7 - Flow in Pipes

WR 211: FLUID
MECHANICS FOR
CIVIL ENGINEERS
P. M. NDOMBA ( Ph.D.)
Senior Lecturer
Water Resources Engineering Department
Assisted by: Ms A. Augustina (Assistant Lecturer)
Contacts: Offices - Block A, Room 201; UDSM ADMIN Block, Room 102;
Email - [email protected];
Time - Anytime official hours 1
5.0 Analysis of Pipe
Flow
2
Course Content:
• 5.0 Flow in pipes:

– Boundary layer theory,
– Laminar and turbulent flow Reynolds
number, Velocity distribution;
– Head losses, Poiseuille equation, Blassius
equation, smooth pipe law, Colebrook and
White, Hazen-Williams, Mannings formula;
– Head loss coefficient for minor losses;
– Pipelines and pipe systems – network
analysis (grid and branched, Hardy cross
and equivalent pipe
3
5.1 Head loss
• Pipes were introduced in the very
earliest days of the practice of
hydraulics.
• Their commonplace use today makes

it of great importance that the laws
governing the flow in them should be
fully understood.
4
5.1 Head loss
• As has been demonstrated in the
previous chapters, some loss of
energy is inevitable in the flow of
any real fluid.
• In the case of flow in a horizontal

uniform pipeline this is evidenced by
the fall of pressure in the direction of
flow (illustrations-equation/sketch).
5
5.1 Head loss
• Clearly, an ability to predict the energy
loss per unit length is essential to
efficient pipeline design.
• This chapter is concerned with an

analysis of the mechanics of flow,
leading directly to an assessment of
losses.
• Flow is in every case deemed to be

steady-that is, the discharge is
constant with respect to time 6
• Recall chapter 4: types of flow – turbulent
and laminar
7
4.2 Types of flow
• (a) Turbulent and
Laminar Flow
• In turbulent flow, the
progression of the fluid particles
is irregular and there is a
seemingly haphazard/random
interchange of position.
• Individual particles are subject

to fluctuating transverse
velocities so that the motion is
eddying/current and
sinuous/twisting rather than
rectilinear.
• If dye is injected at a certain

point, it will rapidly diffuse
throughout the flow stream.
8
4.2 Types of flow
• In the case of turbulent flow
in, say, a pipe, an
instantaneous recording of
the velocity at a section
would reveal a distribution
somewhat as indicated in Fig.
4.1(a).
• The steady velocity, as would
be recorded by normal
measuring instruments, is
indicated in dotted outline,
and it is apparent that
turbulent flow is
characterized by an unsteady
fluctuating velocity
superimposed on a temporal
steady mean. 9
4.2 Types of flow
• In laminar flow all the
fluid particles proceed
along parallel paths and
there is no transverse
component of velocity.
• The orderly progression
is such that each
particle follows exactly
the path of the particle
preceding it without any
deviation.
10
4.2 Types of flow
• Thus a thin filament of

dye will remain as such
without diffusion.
• There is a much greater
transverse velocity
gradient in laminar flow
(Fig 4.1b) than in turbulent
flow (Fig. 4.1a).
• In fact, for a pipe the ratio
of the mean velocity V and
the maximum velocity Vmax
is 0.5 and about 0.85 in
the respective, cases.
11
4.2 Types of flow
• Laminar flow is associated with low velocities

and viscous sluggish fluids.
• In pipeline and open-channel hydraulics, the

velocities are nearly always sufficiently high to
ensure turbulent flow, although a thin laminar
layer persists in proximity to a solid boundary.
12
4.2 Types of flow
• The laws of laminar flow are fully
understood and for simple boundary
conditions the velocity distribution can be
analyzed mathematically.
• Due to its irregular pulsating nature,

turbulent flow has defied vigorous
mathematical treatment, and for the
solution of practical problems it is
necessary to rely largely on empirical or
semi-empirical relationships. 13
5.1 Boundary layer theory
• It is a characteristic of an ideal fluid
that it flows past a solid boundary
with undiminished velocity.
• Real fluid, due to viscous shearing

action that is induced, suffers a
severe retardation in the vicinity of a
boundary and at the boundary itself
the velocity is zero (i.e., no-slip
condition)
14
• End of recap
15
The layer of fluid adjacent to a boundary where the
viscous effects are evident is called boundary layer and
it is of great significance
Flow in the boundary layer can be either laminar or
turbulent
Development of the boundary layer on a thin plate 16

• For a short distance back from the
leading edge the flow is laminar with
the velocity distribution
approximately parabolic and
pressure nearly constant.
17
• At a certain point along the plate the
laminar flow becomes unstable,
some eddying commences and then
after short transitional zone,
turbulence is fully developed.
18
5.2 Laminar Flow-Poiseuille Equation
• Hagen and Poiseuille, working
independently, carried out
experiments on the behaviour of
fluids in small bore pipes
Laminar flow in pipe

19
• It is a horizontal pipe, diameter D, in which the
flow is laminar.
• The mass flow may be visualised as being built up

of a number of concentric cylinders sliding one
upon the other.
20
• We will consider the forces acting on
any one imaginary cylinder radius r
and length L.
• Due to viscous resistance the pressure

(uniform distributed) falls by p in
this distance.
• At the circumference the stress is

 and the velocity is v.
21
• Consider: Frictional resistance =
reduction in pressure force
22
23
Which is the equation of a parabola and the velocity distribution

in the pipe is therefore in the form of a paraboloid.
24
The maximum velocity occurs at the centre line

where r=0, or
25
The results of experiments fully confirm its validity
One important point which is necessary to emphasize is that laminar

flow is unaffected by the nature of the boundary surface. Thus we do not
find in the analysis (Poiseuille’s equation ) any term giving indication of
the wall roughness .
It will be observed that friction gradient hf/L is

directly proportional to the mean pipe velocity, V
26
5.3 Turbulent Flow-Darcy-
Weisbach Formula
• In the period around 1850, Darcy,
Weisbach, and others, as a result of
pipe experiments, deduced a formula
for pipe friction loss which may be
expressed in the form
27
• Derivation from a consideration of the shear
stress at the pipe boundary:
• Frictional resistance = Reduction in

pressure force
28
Weisbach Formula
• The experiments, which were necessarily of somewhat limited
scope and accuracy, tended to indicate that the head loss was
proportional to V raised to a power which was near to 2.
• On the assumption then that
 0  cV 2
29
• Equation 5.8 becomes
p 4cV L 2
  hf
w gD
• Putting And integrating along the pipe we obtain

8c LV 2

hf 
2 gD
Which is in the desired form as presented earlier
30
Weisbach Formula
• It was realised at an early stage that
the friction factor varied not only with
the roughness of the pipe walls but also
with the diameter and velocity.
• This indicated that  was not a simple

coefficient, as once supposed, but an
overall coefficient, which had to
represent the combined effect of
several variables.
31
Weisbach Formula
• Furthermore hf was not dependent on

the square of V but on V raised to some
slightly lesser power, n.
n
h f α cV
32
5.4 The Contribution of Osborne
Reynolds
• 5.4.1 Dye Experiments
• Osborne Reynolds in his classic
experiments at Manchester University
demonstrated most effectively the
characteristics of laminar and turbulent
flow.
33
Osborne Reynolds classic experiments
He showed that,
under suitable
conditions, the
two types of
flow could be
made to occur
in the one pipe.
Fig. 5.2 Behaviour of dye in a pipe
34
Reynolds
5.4.1 Dye Experiments
His apparatus was
extremely simple and
consisted essentially
of a glass tube
through which water
could be passed at
varying velocities.
Provision was made for

the insertion of a thin
jet of aniline dye into
the stream of water
at the upstream end
35
Reynolds
5.4.1Dye Experiments
Commencing with the very
low velocity, it was
found that the dye
remain intact in the
form of a thin slender
thread extending the
whole length of the
tube as in Fig. 5.2(a).
This indicated that the
particles of liquid were
moving in straight
parallel paths and that
the flow was therefore
laminar.
36
Reynolds
5.4.1 Dye Experiments
The velocity of the water
was then gradually
increased and at a
certain point the
thread broke up, as in
Fig.5.2 (b), and the
diffused dye
intermingled with the
water in the tube.
The flow had evidently
passed into the
turbulent state and
was found to remain in
this condition for all
higher velocities. 37
Reynolds
• 5.4.1 Dye Experiments
• By using varying sizes of tube and

differing temperatures of water
Reynolds showed that the critical
velocity at which the change took place
depended on both these factors.
38
5.4.3 Reynolds Number
• Reynolds correctly reasoned that there were
only three factors which could influence the
nature of the flow – the viscosity μ and the
density ρ of the fluid, and the diameter D of
the pipe.
• The roughness might be expected to be a

factor in turbulent flow but not in laminar.
• The possible way in which μ, ρ and D could be

combined to yield the same dimensions
as
the critical velocity was in the form D 39
• By dividing the values of the critical

velocity obtained in its tests by

DReynolds found that for circular pipes
the dimensionless ratio so obtained was
always about 2000.
DV
• He therefore affirmed that if was below

2000 the flow must be laminar, whereas

if it was above the flow must be
turbulent. 40
• In acknowledgement of Reynolds’ most
valuable deduction the dimensionless
DV
 term or DV v is referred to universally as
Reynolds number (R).
41
5.4.4 Similarity of Flow
• The Reynolds number has a much wider
significance than the mere transition
between laminar and turbulent flow,
since it may be regarded as criterion of
similarity of motion in pipes of different
sizes and conveying different fluids.
42
5.4.4 Similarity of Flow
• Reynolds number represents also the
ratio of the inertial to the viscous forces
and as such has an application
extending well beyond that of pipe
flow.
43
A Class Worked Examples
• 1. Determine a suitable diameter for a
pipeline which is to convey 0.057 cumecs of
oil a horizontal distance of 300 m, if the
pressure loss is not to exceed 140 KN/m2. At
the operating temperature the specific
gravity of the oil is 0.9 and the dynamic
viscosity is 1.43 Ns/m2. What power output is
required from the pump?
44
Solution 1
45
Solution1 (contd.)
Power output = ρgh.Q=pressure drop x discharge=Δp.Q

46
A Class Worked Examples
• 2. What is the Reynolds number of a flow of
0.3 m3/s of oil (s.g. = 0.86, μ =0.025 Ns/m2)
through a 500 mm-diameter pipe? (Soln: Re
= 26, 316);
• 3. An oil with S.G. = 0.85 and ν =1.8 x 10-5

m2/s flows in a 15 cm-diameter pipe at 0.8
l/s. Is the flow laminar or turbulent? (soln:
Re=376.66  laminar)
47
5.5 Experimental Investigations on
Friction Losses in Turbulent Flow
• 5.5.1 Blasius
• From an examination of the pipe
friction data, which were beginning to
accumulate at the beginning of the
last century, Blasius came to the
important conclusion that there were
two types of pipe friction in turbulent
flow.
48
• 5.5.1 Blasius
• The first type he associated with
smooth pipes where the viscosity
effects predominate.
49
• 5.5.1 Blasius
• The second type was relevant to rough pipes

where the viscosity and roughness effects
influence the flow and the friction factor is
dependent both on the Reynolds number and
a parameter of relative roughness.
50
• 5.5.1 Blasius
• On the basis of his own researches and of the
experimental data supplied by Saph and
Schoder, Blasius deduced the following
expression for the friction in smooth pipes,
51
• 5.5.1 Blasius
Some years later, Stanton and Pannellm working at the
National Physical Laboratory, investigated in
considerable detail the head loss in smooth brass pipes
conveying both air and water. One dimensionless graph
of
against R the plotted points were found to be compactly situated

along a single curve. The curve in the turbulent zone agreed quite
well with the Blasius formula for values of R up to 105, but after
this point there was increasing divergence, indicating that the
power by which V should be raised was not constant but increased
with R.
52
Nikuradse contribution
• The next major experimental contributions were
made by Nikuradse in the period around 1930.
• He obtained data which enabled the λ-R smooth
pipe curve to be extended to the very high
value of R= 3 x 106.
•
• Nikuradse’s most notable contribution, however,
lay in the field of turbulent flow in rough pipes.
53
• It is extremely difficult to make a quantitative

assessment of roughness.
• A full definition needs to take into account not
only the height of the surface excrescences, k,
but also their spacing and pattern.
• An additional complication is that in practice
the size of the individual excrescences may be
expected to vary
54
• Nikuradse was able to eliminate most of these
possible variables by roughening pipes artificially.
• He did this by gluing uniform sand grains in
closely packed order to the internal surface.
• By so arranging that the excrescence size was
proportional to the pipe diameter, it was possible
to achieve true geometric similarity between
pipes of different size.
55
A judicious combination
of k and D resulted in
six values of k/D,
ranging from
k/D=1/30 to k/D =
1/1014, being
obtained.
The test data produced
the series of smooth
curves (Fig. 5.4)
56
Fig. 5.4 illustrates
several important
points:
1. The laminar portion is

a straight line
(inclined at 45
degrees for equal
vertical and horizontal
scales) and is
unaffected by the
degree of roughness
57
2. Above a certain value
of R, dependent on the
relative roughness, λ is
independent of R and
is constant for a given
value of k/D. This
means that the
viscosity no longer
influences the flow and
that hf is proportional
to V2, which is
sometimes referred to
as the quadratic
relationship
58
3. For values of R above
4000 the fully rough
state is approached by
means of an S-shaped
curve branching off
the smooth pipe
curve.
The rougher the pipe
the lower the value of
R at which the
departure occurs
59
4. The general form of
the curves indicates
that the relationship
between λ, R, and k/D
is complex and can
only be represented
by simple empirical
relationships of the
type, hf = cVn over
very limited ranges
60
Limitation:
The direct practical

value of Nikuradse’s
results was, however,
somewhat limited
since it was difficult
to correlate the
uniform artificial
roughness with the
irregular and wavy
type of roughness
found in
commercial piping.
61
Physical effect
of roughness
Fig. 5.7a: The
excrescences lie well
within the laminar sub-
layer, which smooths
out the flow and
prevents any eddies
from forming.
The pipe behaves as
though it were smooth,
the actual size, shape
and pattern of the
excrescences being
irrelevant
62
Physical effect
of roughness
Fig. 5.7b: The excrescences
are slightly larger than the
thickness of the sub-layer and
therefore protrude out of it
into the turbulent region
.
Some eddying is caused
which absorbs additional
energy and increases the
resistance to flow; the sub-
layer remains, however,
substantially intact.
It is a transitional condition
between the smooth and fully
rough zones.
63
Physical effect
of roughness
Fig. 5.7c: The excrescences
are much larger and have
only one fifth or less of their
height immersed in the sub-
layer. The projections create
considerable disturbance, a
turbulent wake and a train of
vortices being formed. The
sub-layer is almost
completely disrupted and the
form drag due to the
roughness is proportional to
the square of the velocity.
64
Smooth pipes
65
Rough pipes
66
Transitional formula
67
Colebrook and white Transition Formula
68
Transitional formula
Unfortunately, the Colebrook-White formula, as it

stands, is in too complex a form to appeal to
engineers who wish to obtain a rapid solution to their
design problems;
In equation 5.38, D, R, and λ, although interrelated,

are all initially unknown so that a mathematical
solution must proceed on a cumbersome trial and
error basis (numeral methods) 69
70
Moody (1944)
diagram
Simplified the
mathematical
procedure by
Colebrook-white by
reproducing the
transition law
curves on a
standard λ-R type
of diagram (Fig.
5.9).
The relationship
between λ, R, and
k/D can be read off
directly from the
diagram.
(Moody, 1944)
71
5.6. Empirical Formulae
• Prior to the publication of the logarithmic
formulae the only design equations
available were those of the purely
empirical exponential type.
• Simplicity is their chief merit, since they

are particularly amenable to solution by
means of nomograms and charts.
• They have been and still are used

extensively. 72
5.6. Empirical Formulae : Exponential
form of Equation
• For pipes conveying water they

normally take the form
y
V  aD S f x
• Where Sf is the friction gradient, hf/L,

and the coefficient a and the exponents
xNote:
and y are empirical.
The expression is not dimensionless so that care has to be
taken in any conversion of units.
73
5.6.1 Smooth Pipes
• Pipes of aluminum, brass, copper, lead,
alkathene, glass, and asbestos cement
may generally be classified as smooth.
• The Blasius formula (Eq. (5.14)) is

applicable up to R =105 and may be
expressed in the exponential form.
• This follows since
2 0.25 2
0.316 V  v  V
S t  0.25  0.316 
R 2 gD  DV  2 gD
74
5.6.1 Smooth Pipes
• Substituting v = 1.14 mm2/s for water
at 15º C, we obtain
5 4
V  75 D 7
Sf7
75
5.6.2 Pipes in the Transition Zone
• A wide choice of exponential formula is
available. The Hazen-Williams formula
is probably the most common.
V  0.354C H D 0.63
S 0.54
f
• The indices are constant for all classes

of pipe, and variation in roughness is
allowed for by adjustment to the
coefficient CH.
76
5.6.2 Pipes in the Transition Zone
• For instance with spun iron pipes
conveying non-aggressive water the CH
is usually taken as 135.
• When the water is aggressive a value

nearer to 100 may be more appropriate
77
5.6.3 Pipes in the Rough Turbulent Zone
• The Manning formula, which is
extensively employed in connection
with channels, is also applicable to
pipes.
• For pipes the formula becomes

0.397 2 3 12
V D Sf
n
• Where n is a roughness coefficient
78
Worked Example
• A cast iron pipe is to convey 0.152
cumecs of water. If Sf = 1/400
determine the size of pipe which is
required according to the formulae of
• (a) Colebrook-White (k=0,3 mm);

• (b) Hazen Williams (CH = 135); and
• (c) Manning (n = 0.01) assume v = 1.14
mm2/s.
79
Proposed solutions
80
81
V  0.354C H D 0.63
S 0.54
f
82
0.397 2 3 12
V D Sf
n 83
84
5.7 Deterioration of Pipes
• Most water mains which have been in
service for several years suffer some
reduction in carrying capacity due to
incrustation or a coating of slime which
tends to gather on the internal surface.
• The rate of deterioration is dependent

on the chemical constituents of the
water and the pipe material.
85
• When designing a water main, therefore,
it is prudent to take into consideration its
likely condition after a period of years in
service.
• It seems reasonable to conclude,

therefore, that the reduction in carrying
capacity is almost entirely due to the
increase in surface roughness with age;
examinations and tests of pipes confirm
the soundness of this assessment.
86
• By analysing the data from tests on
cast iron pipes Colebrook and White
deduced that the roughness increased
uniformly with age and could be best
expressed by the simple empirical
formula
k T  k o  aT
87
• where ko is the initial effective roughness,
• kT the effective roughness after T years,
• and ‘a’ the annual rate of growth of
roughness
88
Further readings!
89
5.8 Smooth Pipes
• The laminar sub-layer at the wall of a

pipe is illustrated to a greatly enlarged
scale in Fig.
• There is no well-defined line of

demarcation between laminar and
turbulent flow, and the dimension δL,
therefore arbitrarily represents the
distance from the boundary at which the
flow changes from being predominantly
turbulent.
90
5.8 Smooth Pipes
• In the turbulent zone the velocity

distribution is logarithmic.
• The fact that the velocity at the

laminar-turbulent interface is common
to both zones was utilized by Prandtl
and he derived an expression for the
frictional resistance;
91
5.8 Smooth Pipes
1 R 
 2 log
 2.51
• This is the general equation for the

frictional resistance of smooth pipe and is
sometimes referred to as the smooth pipe
law. It provides a much better fit to
experimental data than the Blasius
formula.
• As would be expected the friction factor is

dependent on the Reynolds number. 92
5.8 Rough Pipes
• The derived expression for the friction
factor in the rough turbulent zone is
1 3.7 D
 2. log
 k
• Where k – effective excrescence height

• This is the general equation for the
frictional resistance of rough pipes and
is sometimes referred to as the rough
pipe law or quadratic law
93
5.9 Colebrook and White Transition Formula
• 5.9.1 Correlation with Commercial Pipes
• The work of the Prandtl School
represented a very real advancement in
the knowledge of pipe flow.
• The logical next step was to make it

possible for the results obtained to be
applied to practical design problems.
94
• 5.9.1 Transition Formula

• A much greater understanding of the
behaviour of commercial pipes resulted
from the studies by Colebrook and
White at Imperial College.
• These investigators carried out tests

with air on a pipe 53.5 mm diameter
coated with two sizes of sand grain
arranged in different patterns.
95
• Comparisons was made with the test

records of various types of commercial
pipe and it was found that that closest
agreement was obtained when the
artificial roughness was composed of
isolated grains.
96
• The test results when plotted in the

graphical form of Fig. 5.8 showed that
in spite of the variation in
manufacturing technique a single curve
could be regarded as reasonably
representative.
97
R* Reynolds roughness number v*k/ν, v* - shear velocity
98
• Colebrook and White found that by

judiciously combining the smooth law
and the rough law, a single expression
could be formulated which would
represent the entire turbulent flow
range of the λ-R diagram.
• This expression is: 99

1  k 2.51 
 2 log  
  3.7 D R  
• and is known as the Colebrook-White

Transition law
100

• The transition curve, indicated in Fig. 5.8,
was found to be in reasonable agreement
with the experimental data.
• The curve is asymptotic at approximately

those points which the experimental
results indicate as being appropriate –
that is, at about R* = 0.3 to the smooth
law and at about R* = 60 to the rough
law.
101
• 5.9.2 Importance of the Transition Law

• The Colebrook-White formula was the
successful culmination of many years of
painstaking effort in the lines of thought
first initiated by Prandtl.
• The engineer was now presented with a

design tool of much greater reliability
than the purely empirical formulae which
were all that were hitherto available.
102
• It should be lost sight of, however, that

the Colebrook-White formula is by no
means exact in the same sense as the
Poiseuille formula, since certain
experimental constants are essential to
its derivation.
103
• Nevertheless it fits the experimental

facts so well, which tests of a wider
scope have amply confirmed, that the
evidence points overwhelmingly to its
validity.
104
• 5.9.2 Practical Application
• After the publication of the roughness pipe law, data
concerning the value of the effective roughness k for
various classes of pipe started to accumulate
• Unfortunately, the Colebrook-White formula, as it stands, is

in too complex a form to appeal to engineers who wish to
obtain a rapid solution to their design problems.
• These problems usually involve a determination of the pipe

size required to handle a certain quantity of liquid at a given
friction gradient. In Eq. (5.38), D, R, and λ, although
interrelated, are all initially unknown so that a
mathematical solution must proceed on a cumbersome trial
and error basis.
105
• These problems usually involve a

determination of the pipe size required
to handle a certain quantity of liquid at a
given friction gradient.
• D, R, and λ, although interrelated, are

all initially unknown so that a
mathematical solution must proceed on
a cumbersome trial and error basis. 106
• Moody in 1944 simplified the

mathematical procedure by reproducing
the transition law curves on a standard λ-
R type of diagram (Fig. 5.9).
• The relationship between λ, R and k/D

can be read off directly from the diagram.
107
108

• Ingenious nomographic charts have also
been devised.
• The ‘Universal’ design charts, based on

dimensionless parameters, permit a
direct solution to be effected under a
wide range of conditions.
• In addition, useful tables are available. 109

Table showing the list of Pipe Resistance
Formulae
110
End of further readings
111
Minor Losses in Pipes
112
5.5 Minor Losses in Pipes
• Apart from the loss of head due to friction,
there are other, minor, losses in pipe flow
which arise from changes of section, junctions,
bends, and valves.
• The frictional loss in long pipelines is usually far

in excess of the minor losses and in these
circumstances the latter are often neglected.
• With short pipelines, however, the minor losses

assume a greater relative importance and some
appropriate allowance should be made.
113
• It is usual to regard the minor losses as being
additional to the normal type of frictional
loss.
• As they are associated with a turbulent
dissipation of energy they may be
conveniently expressed in terms of the
velocity head.
2
V
hL  K
2g
114
• Except in the simple cases of a sudden
enlargement and contraction, it is not
possible to make a theoretical assessment,
and so the general procedure is to introduce
a coefficient K, the value of which must be
determined by experiment.
• The expression for energy head loss is then

2
V
hL  K
2g
115
• Appropriate values for K is found listed
in pipe manufacturers' handbooks.
• Obviously, in order to minimise losses,

the transition should be as gradual as
possible.
116
Types of minor losses
• Loss of head due to sudden enlargement;

• Loss of head due to sudden contraction;
• Loss of head at the entry;
• Loss of head due to exit;
• Loss of head due to obstruction;
• Loss of head at bends; and
• Loss of head due to fittings.
117
Enlargement of Sections
An abrupt enlargement is sectional area of flow is shown
in Fig. 5.12;
As a result, the pressure increases from p1 to p2 and the
velocity decreases from V1 to V2;
Separation occurs where the flow stream emerges from
the smaller pipe and normal conditions are not restored
until some distance downstream; The space around the
expanding stream of water is filled with a violent
eddying motion, whose maintenance requires a
continuous supply of energy from the main body of flow.
118
• Now it is reasonable to assume – and
experimental evidence tends to support the
condition – that the pressure on the annular
area (A2-A1) at the enlargement is the same
as the that in the smaller pipe, namely p1 .
119
• The frictional force at the pipe wall is
relatively small in comparison with the force
created by the change of momentum, so we
can utilize the momentum equation and
obtain:
p1A2 – p2A2 = ρQ(V2-V1)
120
121
A common type of sudden
enlargement is the discharge
of a pipe into a large tank or
reservoir (Fig. 5.13).
In Equation 5.50, A2 = ∞, so
that indicating that the
entire velocity energy is
dissipated in eddying and
turbulence. 122
Contraction of Section
At an abrupt contraction (Fig. 5.14) the stream
lines converge upstream of the change of section
and the convergence continues at a diminishing
rate as far as the vena contracta
123
Experiment indicates that the sectional area of the
latter is generally about 0.6A2.
Utilizing this value and bearing in mind that the entire
head loss is accounted for by the subsequent
enlargement from the vena contracta, we can
substitute in Eq. 5.50 to obtain
Thus the value of K, based on the smaller pipe, is

0.44. In practice it is normally taken as 0.5
124
Notes:
1.In the case of long pipes,
the loss of head is very small
as compared to the frictional
loss, and therefore, usually
neglected. K=0.4-0.5
2.The amount of head loss is K=0.78
significantly dependent on
the shape of the entrance
(Figs 5.15 a-c). If the
entrance is well rounded the
entrance loss becomes very
small
K=0.05
125
Homework 3
1. At sudden enlargement of waterline from 0.2 m
to 0.4 m diameter pipe, the hydraulic gradient
rises by 0.8 cm, calculate the rate of flow.
2. A horizontal pipe of diameter d1 has a sudden

expansion to a diameter d2. At what ratio d1/d2
would the differential pressure on either side of
the expansion be a maximum?
126
Table showing the list of Pipe Resistance
Formulae
127

Chapter 5 - Lecture7 - Flow in Pipes

Uploaded by

Copyright:

Available Formats

Chapter 5 - Lecture7 - Flow in Pipes

Uploaded by

Document Information

Original Description:

Original Title

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

Chapter 5 - Lecture7 - Flow in Pipes

Uploaded by

Copyright:

Available Formats

WR 211: FLUID

• 5.0 Flow in pipes:

• Their commonplace use today makes

• In the case of flow in a horizontal

• This chapter is concerned with an

• Flow is in every case deemed to be

• Individual particles are subject

• If dye is injected at a certain

• Thus a thin filament of

• Laminar flow is associated with low velocities

• In pipeline and open-channel hydraulics, the

• Due to its irregular pulsating nature,

• Real fluid, due to viscous shearing

Development of the boundary layer on a thin plate 16

Laminar flow in pipe

• The mass flow may be visualised as being built up

• Due to viscous resistance the pressure

• At the circumference the stress is

Which is the equation of a parabola and the velocity distribution

The maximum velocity occurs at the centre line

One important point which is necessary to emphasize is that laminar

It will be observed that friction gradient hf/L is

• Frictional resistance = Reduction in

• On the assumption then that

• This indicated that  was not a simple

• Furthermore hf was not dependent on

Fig. 5.2 Behaviour of dye in a pipe

Provision was made for

• By using varying sizes of tube and

• The roughness might be expected to be a

• The possible way in which μ, ρ and D could be

• By dividing the values of the critical

2000 the flow must be laminar, whereas

Power output = ρgh.Q=pressure drop x discharge=Δp.Q

• 3. An oil with S.G. = 0.85 and ν =1.8 x 10-5

• The second type was relevant to rough pipes

against R the plotted points were found to be compactly situated

• It is extremely difficult to make a quantitative

1. The laminar portion is

The direct practical

Unfortunately, the Colebrook-White formula, as it

In equation 5.38, D, R, and λ, although interrelated,

• Simplicity is their chief merit, since they

• They have been and still are used

• For pipes conveying water they

• Where Sf is the friction gradient, hf/L,

• The Blasius formula (Eq. (5.14)) is

• The indices are constant for all classes

• When the water is aggressive a value

• For pipes the formula becomes

• (a) Colebrook-White (k=0,3 mm);

• The rate of deterioration is dependent

• It seems reasonable to conclude,

• The laminar sub-layer at the wall of a

• There is no well-defined line of

• In the turbulent zone the velocity

• The fact that the velocity at the

• This is the general equation for the