Nikuradse’s experiment
No scientific way of specifying pipe
roughness
Experiments have shown that friction not
only depends on size and shape of the
roughness projections but also on their
spacing
An expression of frictional factor based on
fluid properties is required
For Laminar flow, it is given by????
(equating Darcy-Wesbach and Hagen-Poiseuille eqs)
f=64/RN
Calculation of head loss due to friction in
pipes
To analyse turbulent velocity profile
To compute shear stress on pipe wall
J.Nikuradse (1933) conducted an experiment to
develop the relationship for frictional factor
Coated several different sizes (2.5, 5 and 10 cm
dia) of pipes with sand grains by sieving them into
different grain sizes of reasonable uniform
diameters (0.1 to 1.6 mm).
Denoted the sand grain diameter by e (the arbitrary
absolute roughness)
Also defined arbitrary relative roughness over pipe
diameter by e/D and produced graphs of ‘f’ against
Re for relative roughness ranging from 1/1014 =
0.000985 to 1/30 = 0.0333.
The identified regions will comprise of
Laminar flow (f=64/Re) , f (R e )
Smooth turbulent region where, f (R e )
f 0.316
Re 0.25
Transitional turbulent (region where f varies both
with Re and relative roughness , most pipes lie in
this region)
f (R e, e / D )
Rough turbulent (f remains constant for a given
relative roughness, it is independent of Re)
f (e / D )
If the thickness of the viscous sublayer
L e (i.e. eu * / v 5 ), where e is the height of
the roughness projection, the viscous sublayer
completely submerges the effect of e.
Prandtl gave the following equation for the
calculation of frictional factor for such a case
1 Re f
2 log
f 2 . 51
The above equation is implicit in f and is difficult
to solve, and can only be solved by iteration or a
graph of f versus Re.
C.F. Colebrook (1939) suggested the
following explicit relation for calculating f
1 Re
1.8 log
f 6.9
For fully rough pipe flow, where Re is quite
high and correspondingly L is quite small,
or in other words,
eu* / v 70
1 3.7
2 log
f e / D (von Karman eq)
For transitional case
5 eu* / v 70
neither smooth not fully rough pipe equation
applies. Colebrook (1939) gave the following
relation for such cases
1 e/ D 2.51
2 log
f 3.7 Re f
for e=0, it reduces to smooth pipe flow equation
and for large Re it reduces to rough pipe flow
equation. But it is implicit in f.
Haaland S. E. gave the following explicit
relation for f
1 e/D
1 . 11
6 .9
1 .8
f 3 . 7 R e
So for commercially available pipes, frictional
factor f can be calculated using
Using empirical equations (already studied)
Using Moody’s chart or Stanton diagram
◦ Simple approach
◦ Hit and Trial
Solve all the related problems in Chapter 8 along with the
defined as under. 10th Edition
EXERCISE
8.2.1 8.5.3 8.9.2 8.11.1 8.13.2 8.16.1 8.18.3
8.2.2 8.5.4 8.10.1 8.11.2 8.13.3 8.16.2
8.2.3 8.6.1 8.10.2 8.11.3 8.13.4 8.17.1
8.3.1 8.6.2 8.10.3 8.12.1 8.15.1 8.17.2
8.3.2 8.8.1 8.10.4 8.12.2 8.15.2 8.17.3
8.5.1 8.8.2 8.10.5 8.12.3 8.15.3 8.18.1
8.5.2 8.9.1 8.10.6 8.13.1 8.15.4 8.18.2
PROBLEMS
All the problems from 8.1
to 8.47, 8.49 till 8.53
Assumptions
All pipes are sufficiently long
Minor head losses and velocity head can be
neglected
hL h f
Continuity and energy equations require that flow
entering the junction equals the flow leaving it
Pressure head at J is common to all pipes (P is
common to all)
Q 0 Q Q
in out
Q1 Q2 Q3
Elevation of P must lie between the surfaces of
reservoirs A and C.
h2 and Q2 will be zero if P is at the same elevation
as reservoir B.
If P is above level of B then water must flow into B
and
Q1 Q2 Q3
If P is below the level of B then flow must be out of
B and
Q1 Q2 Q3
Given:
all pipe lengths
all pipe diameters
surface elevation of two
reservoirs (A and B)
Given 1 2 3
flow to or from one of
these (Q1) Length L1 L2 L3
Required: Diameter D1 D2 D3
flow from two reservoirs Elevation E1 E2 -
(Q2 and Q3) Discharge Q1 - -
surface elevation of third
Required
reservoir (C)
Discharge Q2 Q3
Elevation E3
Solution:
determine h1 using pipe friction equation and Moody’s diagram
This fixes the elevation of P so h2 can be determined.
Knowing h2 one can determine flow Q2 using Colebrook/Haaland
and Darcy Weisbach equation
Q3 can be determined from continuity
Q1 Q2 Q3
This enables the determination of h3 and surface elevation of C.
Solve Sample Problem 8.14
Page No 329
Fluid Mechanics with Engineering
Applications (10th Edition)
E. John Finnemore and Joseph B. Franzini
Given:
all pipe lengths
all pipe diameters
surface elevation of
two reservoirs (e.g. A and C)
flow to or from third
reservoir (say Q2). Given 1 2 3
Required: Length L1 L2 L3
surface elevation of
Diameter D1 D2 D3
third reservoir (B)
flow from two Elevation E1 - E3
reservoirs (A and C) Discharge - Q2 -
Required
Discharge Q1 Q3
Elevation E2
Solution
We know that h1 h3 h13
Assume an elevation of P, which yields values of h1
and h3. Q1 and Q3 can be determined as in Case 1.
If the relation at J is not satisfied, adjust the value of
P. This can be done by plotting the result of each
assumption on a graph (ΣQ vs P)
Inflows to J will be taken as positive and outflows as
negative. Select the value of P for ΣQ=0.
Determine h2 from Q2 with Haaland and Darcy
Weisbach equation.
Sample Problem 8.15
Page No 330
Fluid Mechanics with Engineering
Applications (10th Edition)
E. John Finnemore and Joseph B. Franzini
Three reservoir
problem
Given
all pipes lengths
all pipe diameters
elevations of all
three reservoirs Given 1 2 3
Required Length L1 L2 L3
Diameter D1 D2 D3
Flows in all three
Elevation E1 E2 E3
pipes
Discharge - - -
Required
Discharge Q1 Q2 Q3
Elevation
Solution
It is not clear whether the flow is into or out of reservoir
B. At first it is assumed that there is no flow in pipe 2
and the peizometeric level in P is same as B
Q1 and Q3 can be determined using head losses h1 and h3
in the Haaland and Darcy Weisbach equations
If Q1 Q3 elevation at P must be raised to satisfy
continuity at J and Q1 Q2 Q3
If Q1 Q3 then P must be lowered to satisfy continuity
at J causing water to flow out from B.
Then Q1 Q2 Q3
From here on, the solution proceeds along the same
lines as case 2.
Sample Problem 8.16
Page No 332
Fluid Mechanics with Engineering
Applications (10th Edition)
E. John Finnemore and Joseph B. Franzini
For municipal distribution systems, pipes are
frequently interconnected so that the flow to any
given outlet may come from several different paths
It is impossible to tell the direction of flow in such
pipes
Flow in a network of pipes, however complicated it
must satisfy continuity and energy relations as
1. Flow through any junction must equal the flow out
of it (Qin = Qout)
2. Flow in each pipe must satisfy pipe friction laws
3. Algebraic sum of head losses around any closed
loop must be zero (ΣhL=0)
hL is positive clockwise and negative anticlockwise
Pipe networks are too complicated to solve analytically
Method of successive approximations introduced by
Hardy Cross (1936) is used.
The method consists of following steps
i. By careful inspection, assume the most reasonable
distribution of flows that satisfies condition (1)
ii. Write condition (2) for each pipe
hL KQ n
K and n are constant for each pipe. If minor losses
are considered, then
1
n=2 (SI units), n=1.852 (BG units) K 2
C
Where C = A√{(2.g.hL)/(f.L/D) + ∑k}
Note: minor losses are included only within pipe or
loop but neglected at junctions
iii. Compute sum of head losses around each
elementary loop to satisfy (3).
hL 0
hL KQ n
Clockwise-positive, anticlockwise-negative
The condition might not be satisfied in first trial
iv. Adjust flow in each loop by a correction ΔQ to
balance the head in that loop
So KQ 0
n
Q Q0 Q then for each case
KQ n K (Q0 Q) n K (Qon nQon 1 Q) ......)
If ΔQ is small compared with Q0, the terms after the
second one can be ignored so
n 1
hL KQ QKnQ
n
0 0
For a loop,
L 0 and because ΔQ is same for
h KQ n
all pipes in loop
KQ n
0 Q KnQ n1
0 0
Q
0
KQ n
h L
0
KnQ n 1
n
h L
Q o
-ve sign shows that whenever there is an excess of
head loss around a loop in clockwise direction, ΔQ
must be subtracted from clockwise Q0 and added
to counter clockwise ones and vice versa.
v. The losses are still not balanced even after the
application of first correction to each loop due to
the interaction of one loop with another. So the
procedure is repeated until the corrections
become negligible.
CEP has been attached in Teams
Submit the CEP after 3 weeks
Sample problem 8.19, page 342.
If the flow into and out of a two-loop pipe system are as
shown in Fig. S8.19, determine the flow in each pipe
using only a basic scientific calculator. The K values for
each pipe were calculated from the pipe and minor loss
characteristics and from an assumed value of f, and n = 2
Figure shows a pipe made
up of sections of different
diameters
This pipe must satisfy the
equations of continuity
and energy given by
Q Q1 Q2 Q3
h h1 h2 h3
For if the Q is given the problem is straight
forward. Head loss may be directly found by
adding contributions from various sections
i.e. by using the Equation of the head loss
For if Pipe Material or ‘e’ is given use Darcy’s
Weisbach Approach to calculate individual
head loss contributions after finding e/D, V, R
and f for each pipe.
Two approaches are used for the solution
equivalent velocity head method (only this will be
discussed here)
equivalent length method (Le)
According to the first approach
2 2 2
f LV f2 L V f LV
h
1 1 1
2 2 3 3 3
2 gD1 2 gD2 2 gD3
Q1 Q2 A1V1 A2V2
Q1 Q3 A1V1 A3V3
Used to increase the
discharge capacity of a
system
Governing equations
Q Q1 Q2 Q3
h h1 h2 h3
A B
Two approaches are used for solution
Approximate approach
Exactly or relatively exact approach (will be
studied)
For the cases where head loss hL is given
Total discharge can be calculated by
calculating the individual discharge from all
the pipes using
hL KQ n
If pipe material or in other words if ‘e’ is
given then the solution becomes more
accurate
Neglecting minor losses and inserting values
of all individual discharges
1/ n 1/ n 1/ n
h f1 h f2 h f3
Q
K1 K2 K3
Since all head losses are equal
1/ n 1
1/ n
1
1/ n
1
1/ n
1
1/ n
Q h f h f
1/ n
K1 K2 K 3 K
Darcy Weisbach equation (considering minor
losses) will be written as
2
L V
hL f k
D 2g
Solving for V and Q
2 ghL
Q1 A1V1 A1 C1 hL
f1
L1
k1
D1
Q C1 hL C2 hL C3 hL C1 C2 C3 hL