ANALYSIS OF VISCOUS FLOW IN PIPES

The viscous flow in pipe systems may be analyzed by classifiying them in the following
manner:

i) Series system of pipes,

ii) Parallel system of pipes,
iii) Pipe networks,
iv) Interconnected reservoir systems.

In designing or analysis of pipe flow systems, there are six primary parameters
involved, which are
1) the total head loss of the system, hf,
2) the volumetric flow arte of the fluid, Q
3) the diameter of the pipe, d
4) the length of the pipe, L
5) the roughness of the pipe, ef
6) the fluid density r and viscosity, m

Usually, one of the first three parameters is to be determined, while the reamaining
ones are either known or can be specified by the designer. The method of performing
the design or the analysis is different depending on what is unknown.

Series Sytem of Pipes

Parallel Sytem of Pipes

Pipe networks
Interconnected reservoir systems

7 Head Loss 2 1
i) Series Pipe System

If the pipe system is arranged such that the fluid flows through a single path without
branching, it is referred as a series pipe system.

In a series pipe system, the total head loss is the sum of the head losses in each
serially connected pipe, and the volumetric flow rate in pipes should be the same.

hl  hla  hlb  hlc Q  Qa  Qb  Qc

Series pipe system analysis and design problems can be classified into three classes as
follows:

Class I system: The total head loss, hl, of the system is to be determined.
Class II system: The volumetric flow rate , Q, of the system is to be determined.
Class III system: The diameter of the pipes, d, of the system is to be determined.

7 Head Loss 2 2
CLASS I SYSTEMS

In class I systems, the volumetric flow rate Q, the diameter of the pipes d, the length of
the pipes L, the roughness of the pipe es, and fluid properties density r and viscosity m
are known and the total head loss hl is to be determined.

Total hand loss between two points of the system is sum of the major and all local
(minor) losses, i.e.
L V2 V2
hl  hl ,major  hl ,min or hl  f K
D 2g 2g
To complete these calculations, friction factor for each pipe and loss coefficient for all
fittings need to be determined. The following procedure may be followed:

The procedure for solving Class I problems

1) Determine the relative roughness, es/d, for each pipe.

2) Calculate the average velocity in each pipe, V=4Q/pd2
3) Calculate Reynolds number, Re=rVd/m
4) Determine the friction factor f
5) Calculate the major head loss, hl=f(L/d)(V2/2g)
6) Determine minor head loss coefficients K and then calculate the minor head
losses for each variable are portions, hl=K(V2/2g).
7) Evaluate the total head loss of the system by adding the major and minor losses
calculated in stepes 5 and 6.

7 Head Loss 2 3
Example: In the system shown, water at 20 oC is flowing at the volumetric flow rate of
0.015 m3/s. The suction line is made from commercial steel pipe of diameter 4 inch and
length is 15 m. The discharge line is also commercial steel pipe and its diameter is 2
inch and length is 200 m. The entrance to the suction line from the reservoir is through
a square edged inlet. The elbow is 90o standad elbow and the globe valve is fully open.
Calculate the power supplied to the pump if its efficiency is 76%.

Solution:

To be completed in class

7 Head Loss 2 4
7 Head Loss 2 5
7 Head Loss 2 6
CLASS II SYSTEMS
In class II systems, the total head loss of the system hl, the diameter of the pipes d, the
length of the pipes L, the roughness of the pipe es, and fluid properties density r and
viscosity m are known and the volumetric flow rate of flow Q is to be determined.

Since the volumetric flow rate is unknown, the fluid velocity and hence Reynolds
number can not be calculated directly. Thus, the friction coefficient can not be
determine directly. Hence an iterative solution procedure is applied to determine the
volumetric flow rate of the fluid. The main steps of this iterative solution procedure is
given below:

The procedure for solving Class II problems

1) Determine the relative roughness, es/d, for each pipe.
2) Assume a friction factor f for each pipe from the Moody diagram using the relative
roughness es/d in fully rough region. The reason for using the fully rough region in
Moody diagram is that in this region friction factor is independent from the
Reynolds number. If this assumption is true no iteration is necessary.
3) Calculate the major head loss, hl=f(L/d)(V2/2g) in terms of the unknown velocity.
4) Determine minor head loss coefficients K and then calculate the minor head losses
for each variable are portions, hl=K(V2/2g) in terms of the unknown velocity.
5) Evaluate the total head loss of the system by adding the major and minor losses
calculated in stepes 4 and 5.
6) Realate the unknown velocities in each pipe by using the continuity equation.
7) Express the total head loss hl of the system in terms of one of the unknown
velocities, and then solve the unknown velocities.
8) Calculate Reynolds number, Re=rVd/m for each pipe.
9) Determine the improved friction factor f for each pipe from the Moody diagram
using Reynolds number and es/d.
10) Compare the asumed and the improved values of the friction factor and repeat
steps 3 through 9 as many times as needed in order to obtain desired accuracy in
the friction factor.
11) Evaluate volumetric flow rate of the fluid, Q=pd2V/4

7 Head Loss 2 7
Example
The piping system, which is shown in the figure is used to transfer water at 25oC from
one storage tank to the other. The larger pipe is a 6 inch commercial steel pipe having a
total length of 30 m. The smaller pipe is a 2 inch commercial steel pipe having a total
length of 15 m. The entrance to the larger pipe from the storage tank is through an
inward projecting pipe. 90o standard elbows are used, and the gate valve is half open.
Determine the volumetric flow rate through the system.

Solution:

To be completed in class

7 Head Loss 2 8
7 Head Loss 2 9
7 Head Loss 2 10
7 Head Loss 2 11
CLASS III SYSTEMS
Class III system are true design problems. In class III systems, the total head loss of the
system hl, the volumetric flow rate Q, the length of the pipes L, the roughness of the
pipes es, and fluid properties density r and viscosity m are known and the pipe diameter
d is to be determined.

Since the diameter is unknown, the fluid velocity and hence the Reynolds number can
not be calculated directly. Thus, the friction coefficient can not be determine directly.
Hence, an iterative solution procedure is applied to determine the pipe diameter. The
main steps of this iterative solution procedure is given below:

The procedure for solving Class III problems

1) Assume the diameter d of each pipe
2) Determine the relative roughness, es/d, for each pipe.
3) Calculate the average velocity V=4Q/pd2 for each pipe.
4) Calculate Reynolds number, Re=rVd/m for each pipe.
5) Determine the friction factor f for each pipe from the Moody diagram using
Reynolds number and es/d.
6) Calculate the major head loss for each pipe using hl=f(L/d)(V2/2g).
7) Determine minor head loss coefficients K and then calculate the minor head loss for
each variable are portions using hl=K(V2/2g).
8) Evaluate the total head loss of the system by adding the major and minor losses
calculated in stepes 6 and 7.
9) Compare the calculated total head loss with the given head loss. If the calculated
head loss is less than the given head loss, then decrease pipe diameter. If the
calculated head loss is more than the given head loss, then increase the pipe
diameters. Repeat steps (2) through (8) until the calculated and the given head
losses are close enough.

7 Head Loss 2 12
Example:
In a chemical processing system, benzene at 20 oC is taken from the bottom of a larger
tank and transferred by gravity to another part of the system, as shown in the figure.
The length of the pipe between two tanks is 7 m and it is commercial steel pipe. A filter
is installed in the line, and is known to have a loss coefficient of 8.5. The gate valve is
fully open. Determine the diameter of the pipe which would allow a volumetric flow
rate of 0.0025 m3/s through this system.

To be completed in class

7 Head Loss 2 13
7 Head Loss 2 14
7 Head Loss 2 15
7 Head Loss 2 16
Parallel System of Pipes

If the pipe system causes the flow to branch into two or more lines, then it is referred as
a parallel system of pipes. A typical parallel system of pipes is shown in Figure. The flow
in the main line splits into three branches at section 1, and then rejoins at section 2.

The head loss in each branch between sections 1 and 2 must be equal, that is
ℎ𝑓1−2 = ℎ𝑓𝑎 = ℎ𝑓𝑏 = ℎ𝑓𝑐

Also, one should observe that the volumetric flow rate in the main line is equal to the
sum of the volumetric flow rates through each branch. Hence
𝑄 = 𝑄𝑎 + 𝑄𝑏 + 𝑄𝑐
Usually, two types of problems occur in parallel system of pipes. These are:

i) Class I Systems for which the head loss in each branch is known, and the
volumetric flow rate in each branch and in the main line are to be determined.
ii) Class II Systems for which the total volumetric flow rate is known, and the
volumetric flow rate and the head loss in each branch are to be determined.

Parallel system of pipes

7 Head Loss 2 17
Class I Systems

In Class I problems, the pressure drop or the head loss across the parallel branches is
known, and it is desirable to determine the volumetric flow rates in each branch and in
the main line.

The procedure for solving Class I problems may be given as follows:

1) Determine the relative roughness, es/d, for each branch by using the pipe
diameter, d, and the pipe material.
2) Assume a friction factor, f, for each branch from the Moody diagram by using the
relative roughness, ɛs/d, in the fully rough region. The reason for the assumption
of, the flow in the fully rough region is that no iteration is necessary, if this
assumption is true.
3) Calculate the major head losses for each branch by using the Darcy-Weisbach
equation, hf = f (L/d)(V2/2g) in terms of the unknown branch velocities.
4) Determine the minor head loss coefficients K and then calculate the minor head
losses for the variable area portions in each branch by using hf= kV2/(2g) in terms
of the unknown branch velocities.
5) Evaluate the total head loss ,hf, of each branch by adding up the major and the
minor head losses from steps (3) and (4).
6) Equate this head loss to the given head loss across the parallel branches and solve
for the unknown branch velocities.
7) Calculate the Reynolds number, Re = ρVd/μ for each branch,
8) Determine the improved value of the friction factor, f, for each branch from the
Moody diagram by using the Reynolds number, Re, and the relative roughness,
ɛs/d.
9) Compare the assumed and the improved values of the friction factor for each
branch and repeat steps (3) through (8) as many times as needed in order to
obtain the desired accuracy in the friction factor.
10) Evaluate the volumetric flow rate of the fluid, Q = πd2 V/4, in each branch and in
the main line.

7 Head Loss 2 18
Example 10.4
The arrangement, which is shown in Figure, is used to supply lubricating oil to the
bearing of a large machine. The bearings act as restrictions to the flow. The head loss
coefficients for the journal bearings in branches x and y are 11 and 4, respectively. The
length of 1" commercial steel branch x is 10 m, while the length of 1 ½ inch commercial
steel branch y is 5 m. Each of the four bends in the tubing has a radius of 100 mm. The
pressures before and after the branching are 275 kPa and 195 kPa respectively. The
density and the absolute viscosity of the lubricating oil are 881 kg/m3 and 2.2x10-3
Pa.s. Determine:
a) the volumetric flow rate through each bearing, and
b) the total volumetric flow rate.

Solution

To be completed in class

7 Head Loss 2 19
7 Head Loss 2 20
7 Head Loss 2 21
7 Head Loss 2 22
Class II Systems

In Class II problems, the volumetric flow rate in the main line is known, and it is
desirable to determine the volumetric flow rates in each branch and the head loss
across the parallel branches.

The procedure for solving Class II problems may be given as follows

1) Determine the relative roughness, ɛs/d, for each branch by using the pipe
diameter, d, and the pipe material.
2) Assume a friction factor, f, for each branch from the Moody diagram by using the
relative roughness, ɛs/d in the fully rough region. The reason for the assumption
of the flow in the fully rough region is that no iteration is necessary, if this
assumption is true.
3) Calculate the major head losses for each branch by using the Darcy-Weisbach
equation, hf= f(L/d)(V2/2g) in terms of the unknown branch velocities.
4) Determine minor loss coefficients k and then calculate the minor head losses for
the variable area portions in each branch by using hf = kV2/2g in terms of the
unknown branch velocities.
5) Evaluate the total head loss, hf, of each branch by adding up the major and the
minor head losses from steps (3) and (4).
6) Equate the the total head loss of each branch and express all unknown branch
velocities in terms of one unknown branch velocity.
7) As long as the volumetric flow rate in the main line is the sum of the volumetric
flow rates of each branch, then express it in terms of the unknown branch
velocities.
8) Solve for the unknown branch velocities.
9) Calculate the Reynolds number, Re = ρVd/μ, for each branch.
10) Determine the improved value of the friction factor, f, for each branch from the
Moody diagram by using the Reynolds number, Re, and the relative roughness,
ɛs/d.
11) Compare the assumed and the improved values of the friction factor for each
branch and repeat steps (2) through (10) as many times as needed in order to
obtain the desired accuracy in the friction factor.
12) Evaluate the volumetric flow rate of the fluid, Q = πd2V/4 in each branch.

7 Head Loss 2 23
Example
0.0063 m3/s of water at 15°C is flowing in a 2" commercial steel pipe at section 1, as
shown in Figure. The heat exchanger in branch x has a head loss coefficient of 12.5.
Branch y is a bypass line, which is composed of 1 1/4" commercial steel pipe with a
length of 6 m. All three valves are wide open, and the elbows are standard.
Determine;
a) the volumetric flow rate of water in each branch and
b) the pressure drop between points 1 and 2.

Solution

To be completed in class

7 Head Loss 2 24
7 Head Loss 2 25
7 Head Loss 2 26
7 Head Loss 2 27
Pipe Networks
In municipal water distribution systems, the pipelines are frequently branching and
looping in a complex manner which is often referred as a pipe network. A simple pipe
network is shown in Figure. The most important factor which makes the analysis of pipe
networks difficult, is the uncertainty about the direction of the flow in different pipes
one should note that, it is almost impossible to tell the direction of the flow, such as in
pipe c of Figure.

The fluid flow in a pipe network must satisfy the basic principles of the conservation of
mass and the conservation of energy. These principles can be summarized as follows
i) At each junction of the pipe network, the sum of the volumetric flow rates into the
junction must be equal to the sum of the volumetric flow rates out of the junction.
ii) The algebraic sum of the head losses around any closed loop must be zero.

The pipe network problems are generally solved by a trial-and-error procedure. The
most practical and widely used method of flow analysis in pipe networks is that of
successive approximations, which is developed by Hardy Cross. In the solution
procedure, which is outlined below, the pipe network should be divided into several
closed loops.

1) Determine the relative roughness, ɛs/d, for each pipe from Figure 10.4 by using the
pipe diameter, d, and the pipe material.
2) Assume a friction factor, f, for each pipe from the Moody diagram by using the
relative roughness, ɛs/d in the fully rough flow region.
3) Express the total head loss, hf, for each pipe in the form of
ℎ𝑓 = 𝐾𝑄2
where K is the equivalent resistance to the flow in the pipe, and Q is the unknown
volumetric flow rate through the pipe.

7 Head Loss 2 28
4) Assume a value for the volumetric flow rate in each pipe such that the sum of the
volumetric flow rates into each junction is equal to the sum of the volumetric flow rates
out of the junction. Note that the fluid tends to follow the path of least resistance
through the pipe network, so that the pipes having high values of K have relatively low
volumetric flow rate.
5) Determine the algebraic sum of total head losses, 𝒉𝒇 , for each closed loop by
using the following sign convention. If the flow is clockwise, then the total head loss
and the volumetric flow rate are positive. However, if the flow is counterclockwise, then
the total head loss and the volumetric flow rate are negative.
6) Determine the arithmetic sum of 𝟐𝑲𝑸, that is 𝟐𝑲𝑸 for each closed loop. During
this summation, consider all 𝟐𝑲𝑸 values as positive.
7) Calculate the correction for the volumetric flow rate, ΔQ, for each loop from
ℎ𝑓
∆𝑄 =
(2𝑄𝐾)

8) Calculate the improved values of the volumetric flow rate, Q', by using
𝑄′ = 𝑄 − ∆𝑄
9) Repeat calculations in steps (5) through (8) until ΔQ from step (7) becomes negligibly
small. The Q' value is used for the next cycle of iteration.
10) Calculate the average velocity, V = 4Q/(πd2) for each pipe.
11) Calculate the Reynolds number, Re = ρVd/μ for each pipe.
12) Determine the improved value of the friction factor, f, for each pipe from the
Moody diagram in Figure 10.5 by using the Reynolds number, Re, and the relative
roughness, ɛs/d.
13) Compare the assumed and the improved values of the friction factor for each pipe
and repeat steps (lii) through (xii) as many times as needed in order to obtain the
desired accuracy in the friction factor.

7 Head Loss 2 29
Example
Water at 20°C is flowing through the pipe network, as shown in Figure. All pipes are 1"
cast iron, and the elbows are standard. Determine the volumetric flow rate of water in
pipes a, b, c, d and e.

Solution:

To be completed in class

7 Head Loss 2 30
7 Head Loss 2 31
7 Head Loss 2 32
7 Head Loss 2 33
Interconnected Reservoir Systems

Another example of pipe systems, which has practical importance in water supply
systems is when three or more reservoirs at various elevations are interconnected at a
joint.

Figure shows three interconnected reservoirs. Reservoir A is at the highest elevation,

reservoir B is at the intermediate elevation and reservoir C is at the lowest elevation.
Three pipes 1, 2 and 3 meet at a common junction J, and Q1 . Q2 and Q3 are the
volumetric flow rates in the respective pipes.

It is obvious that the flow takes place from the highest reservoir, that is reservoir A,
towards the junction, and then from the junction to the lowermost reservoir, that is
reservoir C. However, the direction of the flow in the second pipe is still uncertain. If
the total head at the free surface of reservoir B is greater than the total head at
junction J, then the flow takes place from reservoir B towards junction J. In this case,
the continuity equation at junction J becomes

𝑄1 + 𝑄2 = 𝑄3 𝑤ℎ𝑒𝑛 ℎ𝑡𝐵 > ℎ𝑡𝐽

However, when the total head at the free surface of the reservoir B is smaller than the
total head at junction J, then the flow takes place from junction J towards reservoir B.
Hence, the continuity equation at junction J is

𝑄1 = 𝑄2 + 𝑄3 𝑤ℎ𝑒𝑛 ℎ𝑡𝐵 < ℎ𝑡𝐽

7 Head Loss 2 34
The procedure for determining the volumetric flow rate for each pipe may be given as
below:

1) Calculate the total head, ht, at the free surface of each reservoir.
2) Determine the relative roughness, ɛs/d, for each pipe by using the pipe diameter, d,
and the pipe material.
3) Assume a friction factor, f, for each pipe from the Moody diagram by using the
relative roughness, ɛs/d, in the fully rough flow region. The reason for the
assumption of the flow in the fully rough region is that no iteration is necessary if
this assumption is true.
4) Express the total head loss, hf, for each pipe in the form of hf = KQ2, where K is the
equivalent resistance to the flow and Q is the unknown volumetric flow rate
through the pipe.
5) Assume the total head for the junction, and determine the direction of flow in
pipes which are connected to intermediate reservoirs. Write the continuity
equation for the junction.
6) Write the extended Bernoulli equation along the streamlines between the free
surface of each reservoir and the junction.
7) Determine the volumetric flow rate, Q, for each pipe,
8) Determine whether the continuity equation is satisfied at the junction or not. If the
continuity equation is not satisfied, then repeat the calculations in steps (5)
through (7) as many times as necessary.
9) Calculate the average velocity, V = 4Q/(πd2) for each pipe,
10) Calculate the Reynolds number, Re = ρVd/μ for each pipe,
11) Determine the improved value of the friction factor, f, for each pipe from the
Moody diagram in Figure 10.5 by using the Reynolds number, Re, and the relative
roughness, ɛs/d.
12) Compare the assumed and the improved values of the friction factor for each pipe
and repeat steps (4) through (9) as many times as needed in order to obtain the
desired accuracy in the friction factor.

7 Head Loss 2 35
Example:
Water at 20°C is flowing through the three reservoir system, which is shown in Figure.
All pipes are cast iron. Neglecting all minor head losses, determine the volumetric flow
rate for each pipe.

Solution

To be completed in class

7 Head Loss 2 36
7 Head Loss 2 37
7 Head Loss 2 38
7 Head Loss 2 39