Suzi – Izamar Saliha 180051629 Civil Engineering – Stage 2

ENERGY LOSS DUE TO FRICTION IN A CYLINDRICAL PIPE

- FULL THECNICAL REPORT –
_______________________________________________________________________________________

Introduction
The energy loss in a pipe is due to the interaction between the fluid and the wall of the pipe, which
generate a viscous stress because of the friction. In this experiment, the behaviour of a laminar flow
and a turbulent flow is observed. This experiment is focused on the energy losses due to friction in
the pipe, the relationship between head loss (loss in pressure), friction factor, flow rate and Reynolds
number for both laminar and turbulent flow in a smooth pipe is investigated. Some graphs are plotted
to better understand the results and some discussion is carried on with relative error analysis.

Aim and objective

This experiment is run to analyse the connection between energy loss and flow rate in a cylindrical
pipe. Two different diameters pipe and two different types of flows are observed in order to be to
analyse the difference between them and to analyse if the Reynold number and the friction factor is
influenced by these elements. Attention is put on the fact if the results can be generalised by using
dimensionless groups like the Reynold number or the friction factor.

Background and reading and theory

Energy loss is an important point to take in consideration when designing a pipe because a bad
understanding of it will lead to a loose inefficiency. A first initial difference to be made is between
laminar flow and turbulent flow (both observed in the experiment). Laminar flow is characterized by
a highly ordered motion, and smooth flowing speed with no disturbance and transversal motion; all
the particles move in straight lines parallel to the pipe walls with no virtual mixing between layers.
The velocity profile for this kind of flow is parabolic, and the maximum velocity is at the middle of
the pipe and the minimum is at the walls. The average velocity can be in considered like half of the
maximum velocity. Turbulent flow is characterized by chaotic motion with rapid variations in
pressure and velocity in space and time. Turbulent flow is diffusive; thus, it is rapidly mixing with an
increased rate of energy, mass transport, motion and variation streamline. The inter-exchange of the
molecules rises the sear stress whenever a relative motion exists. The velocity profile is fairly across
the centre of the pipe and drops drastically extremely close to the walls. The average velocity in the
pipe can be assumed approximately equal to the velocity at the centre of the pipe. One characteristic
of the two types of flow is the Reynold number which is the ratio of the inertia over the viscous force:
𝜌𝑉𝑐 𝐷 𝑉𝑐 𝐷
𝑅𝑒 = 𝜇
𝑜𝑟 𝑅𝑒 = 𝑣
(eq. 1)

Where: ρ = density
Vc = characteristic velocity (mean or average velocity)
D = critical length (diameter or distance between plates)
μ = dynamic viscosity of the fluid
v = kinematic viscosity of the fluid v = μ/ρ

When the Reynolds number is below 2100, the type of flow is considered as a laminar flow while
when the Reynolds number is above 4000, the type of flow is considered as turbulent flow.
The Reynold number is also related to the friction factor.
Friction is the force between two surfaces that are sliding, or trying to slide, across each other. In the
pipe case friction will occur between the different steam lines or between the fluid and the walls. A
friction factor can be taken into consideration in order to involve the friction force in the motion of

1
Suzi – Izamar Saliha 180051629 Civil Engineering – Stage 2

the fluid. The friction factor is related to the Reynold number, and different approaches are followed
in case of laminar flow or turbulent flow.
For laminar flow the relation between friction factor and Reynold numbers is determined theoretically
by the Hagen-Poiseuille formula:
64
𝑓= 𝑅𝑒
(eq. 2)

For turbulent flow, the relation between friction factor and Reynold number is demonstrated by
Blasius from experimental results (because of the chaotic movements in the turbulent flow, the only
way to approach it is through empirical equations obtained from observations and experiments), and
for smooth pipes only the friction factor is given by:
0.316
𝑓= (eq. 3)
𝑅𝑒 0.25

Due to the viscous effects, the laminar fluid flow will neglect the roughness of the pipe surface
while it is necessary for turbulent fluid flow to consider it and hence frictional factor is required to
be calculated using the equation below:
𝜀
𝑅𝑒𝑙𝑎𝑡𝑖𝑣𝑒 𝑟𝑜𝑢𝑔ℎ𝑛𝑒𝑠𝑠 = 𝐷 (eq. 4)

Where: ε = equivalent roughness

D = diameter of the pipe
In a turbulent flow, the friction factor, Moody’s chart attached in the relative section
The friction factor is also relater to the wall shear stress by the Fanning equation:
8𝜏𝑤𝑎𝑙𝑙
𝑓= ̅2
(eq. 5)
𝜌𝑢
Where: τwall = wall shear stress
ρ = density
𝑢̅2 = mean velocity ( 𝑢̅2 = 𝑄/𝐴 , where Q is the flow rate, and A is the cross-sectional area
of the pipe)
The relation between the head loss and the friction factor is given by the Darcy-Weisbach equation:

̅2
𝑓𝐿𝑢
ℎ𝑓 = (eq. 6)
2𝑔𝐷
Where: L=length pipe
𝑢̅2 = mean velocity
g = gravitational acceleration (9.81m/s2)
D = diameter of the pipe
The head loss is the measurement of energy dissipated in a system due to viscosity (hf: head loss due
to friction) or obstructions of the fluid from pipe elbows, valves or other elements (hm: minor head
loss).

ℎ𝑙𝑜𝑠𝑠 = ℎ𝑓 + ℎ𝑚 (eq. 7)
Where hf and hm are calculated from:

̅2
8𝑓𝐿𝑢 ̅2
𝑢
ℎ𝑓 = (eq. 8) ℎ𝑚 = 𝑘 (eq. 9)
2𝑔𝐷 2𝑔
Where: k = loss coefficient

2
Suzi – Izamar Saliha 180051629 Civil Engineering – Stage 2

Because there is no element such as valves, orifice or sensor in the pipe that may affect the flow, the
minor head loss will be neglected so the total head loss will be determined only from the frictional
head loss (eq. 8).
The last relation that will be discussed is the one between the head loss, the drop in pressure and the
wall shear stress. This is obtained by applying the momentum and energy equation to the flow in the
pipe.
The drop in pressure is the result of resistance to flow and a change in elevation between the start
and end of the pipe (not involved in the experiment).

∆𝑝 = 𝜌𝑔ℎ𝑓 (eq. 10)

Where: ∆𝑝 = drop in pressure
ρ = density
g = gravitational acceleration (9.81m/s2)
ℎ𝑓 = head loss due to friction
The momentum and energy equation gives:
∆𝑝 4𝜏𝑤𝑎𝑙𝑙
= (eq. 11)
𝐿 𝐷

And by substituting eq. 10 in eq. 11 the final relation between the head loss and the wall shear stress
is obtained:
4𝐿𝜏𝑤𝑎𝑙𝑙
ℎ𝑓 = (eq. 12)
𝜌𝑔𝐷

Description of the set up

Figure 1.
1. Armfield F1-10 Hydraulics Bench – is designed as a portable and self-contained service
module providing a controlled flow of water to a range of optional accessories
2. Armfield C6-MKII-10 Fluid Friction Apparatus – provides facilities for the detailed study of
fluid friction head losses, which occur when an incompressible fluid flows through pipes,
fittings and flow metering devices
3. Armfield F1-29 Fluid Statics and Manometer - is designed to demonstrate the properties of
Newtonian fluids and their behaviour under hydrostatic conditions
4. Volumetric tank – reservoir where the water is stored
3
Suzi – Izamar Saliha 180051629 Civil Engineering – Stage 2

5. Sight gauge and level scale - is used to measure the volume of the water inside the tank
6. Flow control – valve that control the flow of water and allows it to leave or return the stump
7. Power switch – switch that turns on/off the apparatus
8. Water is fed in from the hydraulics bench by a pipe – pipe that allow the water from the
volumetric tank into the Armfield C6-MKII-10 Fluid Friction Apparatus
9. Barbed connector - coupling for hose connection with continuous ridges on one or both ends
that grip the inside diameter of a hose to seal the connection
10. An in-line strainer - device to filter out grit and debris from a water line to protect appliances
from being damaged by contamination
11. Smooth bore pipe of 4.5mm diameter - pipe used for laminar flow
12. Smooth bore pipe of 12mm diameter – pipe used for turbulent flow
13. Pressure taping pipe connected at the start of the test section, which is connected to the
manometer - connection to existing piping or pressure vessels without the interrupting or
emptying of that section of pipe or vessel
14. Pressure taping pipe connected at the end of the test section, which is connected to the
manometer - connection to existing piping or pressure vessels without the interrupting or
emptying of that section of pipe or vessel
15. Isolating valves - system that stops the flow of process media to a given location
16. A pipe section made of clear acrylic with a Pitot static tube – element that allows water to
convey
17. A Venturi meter made of clear acrylic - element used to measure the flow rate through a tube
18. An orifice meter made of clear acrylic – element used to measure the rate of flow of a
substance
19. A globe valve – element used for throttling flow control
20. A gate valve – element used to completely shut off or open the fluid flow
21. Water is fed back into the volumetric tank via the exit tube - pipe that allow the water to flow
from the Armfield C6-MKII-10 Fluid Friction Apparatus back into the volumetric tank
22. Pressure tapings connected here to the manometer - connection to existing piping or pressure
vessels without the interrupting or emptying of that section of pipe or vessel
23. Measuring cylinder - piece of laboratory equipment used to measure the volume of a liquid

Experiment procedure and instrumentation

1. Take the hoses of the manometer and connect them to the pipe at the point where the
pressure drop has to be measured (13 and 14 in the apparatus).
2. Ensure all the valves are closed.
3. Open the line valves for the 4.5mm pipe and turn on the pump of the system to allow the
water to flow.
4. Ensure the system is free from any air bubble by opening and closing the valve several
times.
5. Check that the manometer level is the same.
6. Regulate the valves by opening them just slightly to ensure the flow is smooth and laminar.
7. Take the readings on the manometer.
8. Carry out a timed volumetric collection by closing the ball valve.
9. Read on the stopwatch how many litres were accumulated in the tank.
10. Open the ball valve again to drain the water off the system.
11. Increase slightly the flow of water by opening a little more the valves of the pipe.
12. Apply again steps from 7 to 10.
13. Increase again the flow for to times and apply steps from 6 to 9 to each. In total there are 4
readings for the 4.5 mm pipe.

4
Suzi – Izamar Saliha 180051629 Civil Engineering – Stage 2

14. To proceed with the 12 mm diameter pipe close the line valve of the 4.5mm pipe and switch
off the system.
15. Remove the hoses of the manometer from the 4.5mm pipe and connect them to the 12mm
pipe.
16. Open the line valves for the 12mm pipe and turn on the pump of the system to allow the
water to flow.
17. Ensure all the valves are closed.
18. Ensure the system is free from any air bubble by opening and closing the valve several
times.
19. Check that the manometer level is the same.
20. Repeat what was done for the 4.5mm but instead of slightly opening the pipe more every
time, full open it at the beginning and then slightly close it more every time a new reading
has to be taken.
21. After the 4 readings close the line valve of the 12mm pipe and switch off the system.
Error analysis
Sources of error in the experiment come from the fact that the valves are not perfectly close and tight.
This creates leakage from the pipe and bubbles forming inside the pressure taping pipes, which can
affect h values and the diy added to the water in the manometer for better reading, which affects the
water density. A parallax error and a zero error occurred while reading the reading from the
manometer and the stopwatch. The meniscus in the manometer might not be accurate as the water
level is not steady and consistent due to the unstable flow in the pipe. The viscosity of the fluid is
another parameter that may generate errors. The fluid which flows through the pipe might have
different viscosity due to the environment parameter such as pressure and temperature. Another error
may be a variation in the pipe diameter. An unmaintained pipe system might cause the inner side of
the pipe to be decayed, and hence the diameter would not be constant at every region. Deposition of
limescale or rust may cause a change in diameter. The velocity of the fluid might not be consistent
when the fluid is flowing inside the pipeline, and this can be considered an additional error factor.
The last error to take in consideration is the instrumental error from a bad calibration of the instrument.

Calculation procedure with one sample

Volume flow rate Q = Volume / time [m3/s]
Kinematic viscosity v = dynamic viscosity (μ) / density (ρ) [m2/s]
Velocity u = Volume flow rate (Q) / area (A) [m/s]
𝜌𝑉 𝐷
Reynold number 𝑅𝑒 = 𝜇𝑐 (eq. 1)
2𝑔𝐷∆𝐻
Experimental factor 𝑓= 𝐿𝑢̅2
(rearrange of eq.6)
64
Theoretical factor for 4.5 mm pipe 𝑓 = (eq. 2)
𝑅𝑒
0.316
Theoretical factor for 12 mm pipe 𝑓= (eq. 3)
𝑅𝑒 0.25
𝑓𝑡ℎ − 𝑓𝑒𝑥
% error %𝑒𝑟𝑟𝑜𝑟 = ∙ 100
𝑓𝑡ℎ

Sample calculation for the experimental factor

Pipe 1 case 1
𝑉𝑜𝑙𝑢𝑚𝑒 0.000145𝑚3
𝑄= = = 2.41667 ∙ 10−6 𝑚3 /𝑠
𝑇𝑖𝑚𝑒 60s
𝜇 0.00096𝑁𝑠/𝑚2
𝑣 = 𝜌 = 997𝑘𝑔/𝑚3 = 9.62889 ∙ 10−7 m2/s

5
Suzi – Izamar Saliha 180051629 Civil Engineering – Stage 2

𝑄 2.41667 ∙ 10−6 𝑚3 /𝑠
𝑢= = = 0.151953 m/s
A 0.000015904𝑚2
𝑘𝑔 m
𝜌𝑉𝑐 𝐷 997 3 ∙ 0.151953 ∙ 0.0045𝑚
𝑚 s
𝑅𝑒 = = = 710.1447
𝜇 0.00096𝑁𝑠/𝑚2
2𝑔𝐷∆𝐻 ⁄ 2
2 ∙ 9.81 𝑚 𝑠 ∙ 0.0045𝑚 ∙ 0.02𝑚
𝑓𝑒𝑥 = 2
= 2 = 0.076475
𝐿𝑢̅ 1𝑚 ∙ (0.151953 𝑚⁄𝑠)
64 64
𝑓𝑡ℎ = = 710.1447 = 0.090122
𝑅𝑒
𝑓𝑡ℎ − 𝑓𝑒𝑥 0.090122 − 0.076475
%𝑒𝑟𝑟𝑜𝑟 = ∙ 100 = ∙ 100 = 15. 14%
𝑓𝑡ℎ 0.090122
Pipe 2 case 1
𝑉𝑜𝑙𝑢𝑚𝑒 0.001772𝑚3
𝑄= = = 0.000180816 𝑚3 /𝑠
𝑇𝑖𝑚𝑒 9.8s
𝜇 0.00096𝑁𝑠/𝑚2
𝑣 = 𝜌 = 997𝑘𝑔/𝑚3 = 9.62889 ∙ 10−7 m2/s
𝑄 0.000180816 𝑚3 /𝑠
𝑢= = = 1.598772 m/s
A 0.000113097 𝑚2
𝑘𝑔 m
𝜌𝑉𝑐 𝐷 997 3 ∙ 1.598772 ∙ 0.012𝑚
𝑚 s
𝑅𝑒 = = = 19924.6971
𝜇 0.00096𝑁𝑠/𝑚2
2𝑔𝐷∆𝐻 ⁄ 2
2 ∙ 9.81 𝑚 𝑠 ∙ 0.012𝑚 ∙ 0.39𝑚
𝑓𝑒𝑥 = = 2 = 0.0359
𝐿𝑢̅2 1𝑚 ∙ (1.598772 𝑚⁄𝑠)
0.316 0.316
𝑓𝑡ℎ = 𝑅𝑒0.25 = 19924.69710.25 = 0.026597
𝑓𝑡ℎ − 𝑓𝑒𝑥 0.026597 − 0.0359
%𝑒𝑟𝑟𝑜𝑟 = ∙ 100 = ∙ 100 = 35.06%
𝑓𝑡ℎ 0.026597
Results

Figure 2.

Plot friction factor against the Reynold’s number

6
Suzi – Izamar Saliha 180051629 Civil Engineering – Stage 2

Figure 3.
7
Suzi – Izamar Saliha 180051629 Civil Engineering – Stage 2

Describe and discus your results and observations

It can be seen, at the end of the experiment, that for the laminar flow, the results are acceptable
because they prove the theory. For the turbulent flow instead only by merely looking at the graph and
the percentage error between the theoretical and the experimental value of the friction factor f, it can
be seen that the difference between the theoretical and the experimental values is way too hight.
If the two lines for pipe 1 in the graph are analysed, it can be seen that the theoretical line is straight
and matches the Hagen-Poiseuille formula. The experimental line seems to have more or less the
same gradient like the theoretical one except for the first value (pipe 1 case 1) that is slightly away
from the path of the straight line and its error is higher than the other values (15.14% versus an
average of 2.5%). The offset of the experimental line from the theoretical one probably is due to
errors that are repeated in every reading so probably parallax error or zero error. An inconsistency in
the flow and the presence of bubble air in the pipe may be an error as well, but this will give readings
more scattered than the actual ones. The first value can be considered affected by these errors.
Probably when the readings were taken, there was still air in the pump; thus, the flow was not
consistent. To reduce these errors, more readings can be done, and more accuracy can be put in
ensuring that the pipe is free from air. Parallax error and zero error can be reduced by paying more
attention during the readings.
For the two lines for the turbulent flow, it is tough to give a comparison. The theoretical line is straight
and matches the Blasius statement, but the experimental line is affected by an incredibly high
percentage error. The experimental values have irregular fluctuation and a completely different
pattern than the theoretical ones. These irregularities can be due to a flow more turbulent than
expected, and this can also be seen in the high values of the Reynold number, Re. If the table is
observed, it can be seen that between the theoretical values and the experimental values the percentage
error range is between 35% and 67%. These values are way too high if we consider a reasonable
engineering accuracy. The high turbulence may be due to obstruction like rust or limescale that have
been deposited during the years. To reduce these errors substituting the pipe every certain amount of
time can be the first approach. Considering a different fluid than water can be another solution to
reduce the friction and prevent damage in the apparatus. More readings can be taken in this case, as
well.
From the comparison between the Reynold numbers in the laminar flow and the Reynold numbers in
the turbulent flow, it can be seen that the transition Re region is between 1800 and 16000. A
theoretical range for the Re transition is between 2300 and 40005. The difference in the values may
be due to not enough readings taken during the experiment. Another reason can be due to the untrusted
values for the turbulent flow.
If the Reynold number Re and the friction factor f only are observed, it can be seen that even if they
are dimensionless elements, they still define the type of flow. If looking at the pipe 1 cases from 1 to
4, it can be observed that the theoretical and the experimental factors both decrease while Reynold
numbers increase. From this, it can be said that while the friction factor decreases, the Reynold
number Re, thus the parallel streamlines that represent the laminar Newtonian fluid start to be affected
by a slightly more random behaviour of the particle in the flux. For pipe 2 in cases from 1 to 4 it can
be observed that the theoretical value of the friction increase uniformly while the experimental values
of friction factor increase from case 1 to case 3 but with a minimum difference between case 2 and
case 3 and for case 4 the friction factor decrease compared with case 3. The decreasing difference
between case 4 and 3 is minimum, so it is not affecting the Reynold number Re, variation. The

8
Suzi – Izamar Saliha 180051629 Civil Engineering – Stage 2

Reynold numbers for pipe 2 decreases. Following the same reasoning made for pipe 1 while the
friction factor increases the Reynold number decrease and the chaotic motion inside the pipe slightly
begin to stabilise and create organised parallel streamlines. For pipe 1, the experiment is done on a
fluid that has initially laminate flow and change into a more turbulent or less laminar one. For pipe
two, the experiment is done on a fluid that is initially a turbulent flow and decrease to a less turbulent
or more laminate flow. From this, we can consider a flow with low Reynold number value and high
friction factor value a laminar flow and vice versa a flow with high Reynold number value and a low
friction factor value a turbulent flow.

Conclusions
For this experiment, the experimental values for friction factor are calculated and they are compared
to theoretical values. The values for pipe 1 can be compared and an overall analysis can be done
because the percentage error between the two is acceptable. For pipe 2 the percentage error between
the theoretical values and the experimental ones is more that 50% so a discussion between these
experimental values cannot be done accurately. Experimental errors play an important role in this
experiment and makes the reading for pipe 2 unusable.
Dimensionless factors like Reynold number and friction factor can define the type of flow. Flows
with low Reynold number value and high friction factor value are called laminar flows. Flows with
high Reynold number value and a low friction factor value are called turbulent flows.