EXPERIMENTAL ANALYSIS AND CFD

SIMULATION OF MINOR LOSSES IN PIPES

A thesis submitted in partial fulfilment of the requirement for
The award of the degree of
BACHELOR OF ENGINEERING

IN

MECHANICAL ENGINEERING

Submitted by
R. MOUNIKA (314126520134)
REETHU.P (314126520117)
P. MURALI KRISHNA (314126520122)
M.SRAVAN KUMAR (314126520101)
SAHIL ALAM KHAN (314126520140)
Under the guidance of

Dr. RAJESH GHOSH, B.E., M.Tech., Ph.D.

Associate Professor

DEPARTMENT OF MECHANICAL ENGINEERING

ANIL NEERUKONDA INSTITUTE OF TECHNOLOGY & SCIENCES

(Permanently affiliated to Andhra University, Approved by AICTE, and Accredited by

NBA & NAAC with ‘A’ grade)

Sangivalasa - 531162, Bheemunipatnam (Mandal), Visakhapatnam (Dist.),

Andhra Pradesh, India.

2018
 ANIL NEERUKONDA INSTITUTE OF TECHNOLOGY & SCIENCES
(Ariliated to Andhra University, Approved by AlCTE, Aceredited by NBA & NAAC with A grade)
SANGIVALASA, VISAKHAPATNAM (District)- 531162

RAGNANAMAMBRAW

ANITS
CERTIFICATE

This is to certify that the Project Report entitled EXPERIMENTAL

ANALYSIS & CFD SIMULATION OF MINOR LOSSES IN PIPES" being
submitted by RAPETI MOUNIKA (314126520134), PARAVADA REETHU
(314126520117), PICHIKA MURALI KRISHNA (314126520122), SAHIL ALAM
KHAN (314126520140), MULAKA SRAVAN KUMAR (314126520101) n

partial fulfillments for the award of degree of BACHELOR OF TECHNOLOGY in

MECHANICAL ENGINEERING of ANDHRA UNIVERSITY. It is the work of
bona-fide, carried out under the guidance and supervision of Dr. Rajesh Ghosh,
Associate Professor, Department Of Mechanical Engineering, ANITS during the
academic year of 2014-2018.

Approved By
PROJECT GUIDE HEAD OF THE DEPARTMENT

(Dr. Rajesh Ghosh) (Dr. B. Naga Raju)

Associate Professor Head of the Department
Mechanical Engineering Department Mechanical Engineering Department
ANITS, Visakhapatnam. ANITS, Visakhapatnam...

PROFESSOR & HEAD

Department of Mechanical Eng1neerng
ANIU NEEPA A INSTITUTEOf TECHNOLOGY&SCENCE
Sangivalása-531 162 VISAKHAPATNAM Dist A
 BOARD OF EXAMINERS
THIS PR0JECT IS APPROVED BY THE

INTERNAL EXAMINER:
kaju
Dr. B.MNaga
.Tech,M.E.,Ph.d
Professor & HOD
Engineering9
Mechanical

Dept of
ANITS.Sangivalasa.
isakhapatnam-53 1 162.

EXTERNAL EXAMINER:
 ACKNOWLEDGEMENT

On the submission of our project report entitled “Experimental Analysis and CFD
Simulation of Minor Losses in Pipes” we would like to give our heartiest thanks and
gratitude to Dr.Rajesh Ghosh, Associate Professor, Department of Mechanical
Engineering, Anil Neerukonda Institute of Technology & Sciences, for his continuous
motivation, constant support and guidance throughout the past year.

We are very thankful to Prof.T.V.HanumanthaRao, Principal and

Prof.B.Nagaraju, Head of the Department, Mechanical Engineering, Anil Neerukonda
Institute of Technology & Sciences for their valuable support and facilities.

We would like to thank the technical staff of fluid machinery lab for their continuous
cooperation and their guidance in helping us to understand the technical details of hydraulic
machines in the lab.

Finally, we would like to convey our thanks to everyone, who have contributed
directly or indirectly for the completion of this project work.

PROJECT ASSOCIATES

R. MOUNIKA (314126520134)
REETHU.P (314126520117)
P. MURALI KRISHNA (314126520122)
M.SRAVAN KUMAR (314126520101)
SAHIL ALAM KHAN (314126520140)
 ABSTRACT

This project “Experimental Analysis and CFD Simulation of Minor Losses in Pipes”,
deals with the frictional losses produced in a pipe due to shear stress and viscosity of fluids.
This paper contains flow analysis of fluid in different pipe geometry. It focusses on the losses
in piping systems, as working fluid through pipes plays an important role in functionality of
industries like chemical industries, petroleum industries etc. Whenever there is need of
transferring fluids in major piping systems we come across many obstacles such as elbow-
junctions, bends, contractions, expansions. All this together affects the overall efficiency by
causing major and minor losses in pipes.

The experimental analysis is done by considering the continuity equation. By varying the
pipe geometry and flow parameters, the velocity at inlet and outlet of the pipe is calculated.
This velocity is used to calculate the coefficient of loss for pipe.

The purpose of this project work is to investigate the steady, incompressible fluid flow and
to get familiarize with CFD. The simulations were done using ANSYS FLUENT CFD 14.5
software to observe the effect of changes in velocity of flow, drop in pressure and effect of
static pressure, dynamic pressure and stream flow due to change in geometry.

In this project work, analysis of results was done and the results obtained in experiment on
different pipe geometry were compared to be closely conforming to the results of ANSYS.
 LIST OF CONTENTS

Page No

LIST OF FIGURES I
LIST OF TABLES II
LIST OF GRAPHS III
NOMENCLATURE IV

CHAPTER-1 INTRODUCTION 1
1.1 GENERAL 2
1.2 OBJECTIVE 3
1.3 METHODOLOGY 3

CHAPTER-2 LITERATURE REVIEW 4

CHAPTER-3 THEORETICAL ANALYSIS 10

3.1 LOSSES IN PIPE 11
3.2 MAJOR LOSS IN PIPE 12
3.3 MINOR LOSSES IN PIPE 14
3.3.1 Loss of head due to sudden expansion 15
3.3.2 Loss of head due to sudden contraction 16
3.3.3 Loss of head due to bend in pipe 17
3.4 INTRODUCTION TO ANSYS 18
3.4.1 Generic Steps for Solving Any Problem in ANSYS 18
3.4.2 ANSYS Fluent Features 19
3.5 COMPUTATIONAL FLUID DYNAMICS
3.5.1 Definition and history 20
3.5.2 Governing equations 20
3.5.3 Applications of CFD 21
3.5.4 Advantages of CFD 21
3.5.5 Disadvantages of CFD 21
3.4.6 CFD analysis procedure 22
CHAPTER-4 PROCEDURE OF EXPERIMENT 24
4.1 STEPS 25

CHAPTER-5 OBSERVATIONS AND CALCULATIONS 26

5.1 SUDDEN EXPANSION 27
5.1.1 Experimental observation and calculation 27
5.1.2 Results obtained from ANSYS 28
a. Modelling 28
b. Meshing 28
c. Stream function 29
d. Absolute pressure 29
e. Velocity 30
f. Dynamic pressure 30
g. Velocity variation 31
h. Result 31
5.2 SUDDEN CONTRACTION 33
5.2.2 Experimental observation and calculation 33
5.2.3 Results obtained from ANSYS 34
a. Modelling 34
b. Meshing 34
c. Stream function 35
d. Absolute pressure 35
e. Velocity 36
f. Dynamic pressure 36
g. Static pressure 37
h. Result 37
5.3 BEND IN PIPE 38
5.1.1 Experimental observation and calculation 38
5.1.2 Results obtained from ANSYS 39
a. Modelling 39
b. Meshing 39
c. Velocity 40
d. Dynamic pressure 40
 e. Static pressure 41
f. Result 41

CHAPTER-6 ANALYSIS OF RESULTS 42

6.1 COMPARISION BETWEEN ANSYS AND
EXPERIMENTAL RESULT S 43

CHAPTER-7 CONCLUSIONS 44

REFERENCES
 I

LIST OF FIGURES

FIG 3.1 SUDEEN EXPANSION IN PIPE 14

FIG 3.2 VENNA CONTRACTA FORMED IN SUDDEN CONTRACTION 16
FIG 3.3 BEND PIPE 17
FIG 4 MINOR LOSSES APPARATUS 25
FIG 5.1 GEOMETRY OF SUDDEN EXPANSION OF PIPE 28
FIG 5.2 MESH OF SUDDEN EXPANSION OF PIPE 28
FIG 5.3 STREAM FUNCTION 29
FIG 5.4 ABSOLUTE PRESSURE 29
FIG 5.5 CONTOURS OF VELOCITY 30
FIG 5.6 CONTOURS OF DYNAMIC PRESSURE 30
FIG 5.7 GEOMETRY OF SUDDEN CONTRACTION 34
FIG 5.8 MESH OF SUDDEN CONTRACTION 34
FIG 5.9 CONTOURS OF STREAM FUNCTION 35
FIG 5.10 CONTOURS OF ABSOLUTE PRESSURE 35
FIG 5.11 CONTOURS OF VELOCITY 36
FIG 5.12 CONTOURS OF DYNAMIC PRESSURE 36
FIG 5.13 CONTOURS OF STATIC PRESSURE 37
FIG 5.14 GEOMETRY OF ELBOW 39
FIG 5.15 MESH OF ELBOW 39
FIG 5.16 CONTOURS OF VELOCITY 40
FIG 5.17 CONTOURS OF DYNAMICS PRESSURE 40
FIG 5.18 CONTOURS OF STATIC PRESSURE 41
 II

LIST OF TABLES

TABLE 5.1 OBSERVATION OF SUDDEN EXPANSION PIPE 27

TABLE 5.2 CHANGES IN PRESSURE VS VELOCITY 32
TABLE 5.3 OBSERVATIONS FOR SUDDEN CONTRACTION 33
TABLE 5.4 OBSERVATIONS FOR BEND 38
TABLE 6.1 RESULTS OBTAINED FROM ANSYS
AND EXPERIMENT 43
 III

LIST OF GRAPHS

GRAPH 5.1 VELOCITY VARIATIONS AT THE OUTLET

CROSS SECTION 31
GRAPH 5.2 VARIATION OF PRESSURE DIFFERENCE VS
VELOCITY 32
GRAPH 6.1 RESULTS IN ANSYS AND EXPERIMENTS 43
 IV

NOMENCLATURE

hf = head loss due to friction in unit of length

f = friction factor

D = Pipe Diameter

V = Flow velocity

C= Chezy’s constant

m=hydraulic mean depth

𝑡 = time taken

K = Loss coefficient

ℎe = Loss in head due to expansion

ℎ =Manometer Difference

Q = Discharge

V1=velocity at D1

V2=velocity at smaller diameter

𝑘C=co-efficient of loss due to contraction

𝑘 b=co-efficient of loss due to expansion

𝑘 e=co-efficient of loss due to expansion

 1

CHAPTER 1

INTRODUCTION
 2

1.1 GENERAL:

Pipe network is very common in industries throughout the country, where fluid and gases
are transported from one point to another. The pressure loss depends on the type of flow of
the fluid in the network, pipe material, and the fluid flowing through the pipe. When any
fluid flows through a pipe, the velocity adjacent to the pipe wall is zero and the velocity
gradually increases from the wall. Maximum velocity is observed at the centre of the pipe.
Due to increase in the velocity gradient, shear stresses are produced in the fluid due to its
viscosity. This viscous action attributes to loss of energy which is commonly known as
loss due friction or frictional loss.
William Froude stated the following laws of fluid friction under turbulent flow.

For a turbulent flow, frictional resistance is:

1. Directly proportional to 𝑉𝑛, where n varies between 1.5 and 2.

2. Proportional to fluid density.

3. Proportional to surface area in contact.

4. Independent of the pressure

5. Dependent on the nature of the surface in contact.

If losses are minute in a pipe network then the efficiency is higher. Moreover, all networks
should be designed to undergo minimum loss.
 3

1.2 OBJECTIVES:
 Calculate the minor losses (due to sudden expansion, sudden contraction and bend)
in lab and find the co-efficient of loss for their geometry.
 Modelling of different pipe geometry like elbow, sudden enlarge, sudden contract
pipe etc. in ANSYS software.
 Simulation of fluid flow through these pipes.
 Calculation of minor losses with the help of ANSYS.
 Comparison of ANSYS obtained results with experimental obtained results.

1.3 METHODOLOGY
This project can broadly be divided into the following stages.
1) Identifying the problem statement and formulating objectives.
2) Preparation for project:
a. This includes all preparatory things like literature review, data
collection from laboratory etc.
b. Laboratory practical that are to be undertaken for this project are
frictional losses in pipes of different geometry.
c. Various models of pipes are to be modelled in ANSYS Software for
the analysis and comparison of the results from laboratory and
ANSYS.
3) Optimization of result:
a. Flow analysis for fluid flowing through different pipe geometry using
data obtained from practical, theoretical and ANSYS methods.
b.Comparison between ANSYS and experimental results.
 4

CHAPTER 2

LITERATURE REVIEW
 5

Ackeret et al [1] discussed special features of internal flow. He concluded that there is a
predominant role played by the equation of continuity, especially if compressibility is
involved. If the width of the duct is not growing too fast along its length, separation is
followed by reattachment. He observed that in case of internal flow also, three-dimensional
boundary layers can appear as in external flow. We have applied equation of continuity to
pipes of different geometry when fluid is flowing through it.

Celata et al [2] investigated the possibility of wall roughness effects and geometric
deviations for micro tubes ranging from 31 to 326 micro meters. The intent was to model
how accurately fluid flow behaved in accordance with the classical Hagen-Poiseuille flow
for different diameter micro tubes, and to possibly see around what size deviation from this
accepted flow model occurred. An uncertainty analysis was carried out for the Darcy
equation, and a slip parameter was incorporated into the laminar velocity profile equation to
extrapolate a modified Darcy equation. By this we have simulated the pipes having different
geometry.

Hager and Dupraz et al (1985) [3] derived a theoretical equation for obtaining the
coefficient of contraction in terms of the contraction ratio, the inlet angle of the contraction
and the length ratio of the contracted reach. The flow conditions were those of transitional
flow from subcritical to supercritical passing through critical at the minimum depth point
through the contraction length. They verified their expression experimentally. Based on this,
we have calculated loss coefficient by conducting experiment on different pipes and
compared the results with ANSYS results.

Laursen et al (1970) [4] studied the contraction coefficient at sudden expansion at bridge
locations. Four distinct flow zones (accretion, contraction, expansion and abstraction) were
identified and discussed. It was found that the contraction coefficient varies between 0.7 for
about 30% contraction ratio and 1.0 for no contraction. The use of different constrictions for
peak discharge measurement by indirect methods was discussed by Matthi (1976) and was
outlined in French (1986). We have calculated the loss coefficient and had observed the
variations by considering different pipe geometry.
 6

Kindsvater, Carter and Lacy et al (1953) and Kindsvater and Carter et al (1955) [5]
carried out an experimental investigation to address the effects of different types of
contractions on discharge characteristics. Formica (1955) tested experimentally the various
design for channel transition (contraction and expansion). The main results of Formica work
are reported in Chow (1959). Basing on this we have conducted the experiment for different
discharge for a certain volume.

Rathakrishnan et al and Sreekanth et al [6] studied flows in pipe with sudden

enlargement. They concluded that the non-dimensional base pressure is a strong function of
the expansion area ratios, the overall pressure ratios and the duct length-to-diameter ratios.
They showed that for a given overall pressure ratio and a given area ratio, it is possible to
identify an optimal length-to-diameter ratio of the enlargement that will result in maximum
exit plane total pressure at the nozzle exit on the symmetry axis (i.e. minimum pressure loss
in the nozzle) and in a minimum base pressure at the sudden enlargement plane. By this
study, we have come to know that as the diameter of the pipe changes for a given length,
pressure variation is observed.

Wick et al [7] has studied the effect of boundary layer on sonic flow through an abrupt cross-
sectional area. He observed experimentally that the pressure in the corner of expansion was
related to the boundary layer type and thickness upstream of the expansion. He considered
boundary layer as a source of fluid for the corner flow. Based on this concept of boundary
layer, we have observed the variations in velocity from the centre of the pipe to the extreme
walls. At the centre, the velocity is found to be maximum. Due to the relative motion between
the fluid molecules, a decrease in the velocity is observed from the centre to pipe walls. At
the pipe wall, the fluid molecules come to rest due to the direct contact between fluid
molecules and pipe wall. The fluid layer next to this has a velocity nearer to zero and it
thereby varies from layer to layer.
 7

Mandal et al. (2008) [8] examined the shape and stability of Taylor bubbles and Taylor
drops in liquid-liquid systems. They noted the effect of tube diameter and inclination on the
shape and velocity of the Taylor bubble and drop. They have reported that the velocity of
both Taylor bubble and drop increases with increase in tube diameter. From this journal we
have concluded to note the variation of velocity at different diameters.

Rodriguez et al. (2009) [9] studied the frictional pressure drop encountered during
horizontal and vertical core flow. They have used both viscous (ρ = 925 kg/m3 and µ = 0.5
Pa-s) as well as ultra-viscous crude oil (ρ = 972.1 kg/m3 and µ = 36.95 Pa-s) in two pipes of
diameter 0.0284 m and 0.077m for their study. They have noticed a reduction in frictional
pressure with addition of water in all the cases. Further they have modified the model
proposed by Parda and Bannwart (2001) to predict frictional pressure gradient. Based on this
we have determined pressure drop at different pipe sections. And using this we have
calculated the loss coefficient.

Vallentine et al [10] Sixth International Water Technology Conference, IWTC 2001,

Alexandria, Egypt (1958) investigated the effect of pipe contraction placed normal to
channel axis. His observations covered data that include different diameters of flow. In
practical applications we come across different pipes. Basing on this journal we have taken
different pipe geometry and found out the geometry of pipe which gives minimum frictional
loss.

Delhaye et al (1981), Wadle et al (1989) Schmidt and Friedel et al (1997), Guglielmini

et al (1997), Fossa and Guglielmini et al (2002) [11] Researchers have proposed different
models for predicting pressure drop in expansion. Some studies are also devoted on the effect
of area change on flow regimes (Fossa et al. (2006), Ahmed et al. (2007), Ahmed et al.
(2008), Chen et al. (2009). Extracting from this article we have determined pressure drop at
different sections when fluid is flowing through pipe.
 8

Timothy J. Rennie et al [12] studied the heat transfer characteristics of a double pipe helical
heat exchanger for both counter and parallel flow. Both the boundary conditions of constant
heat flux and constant wall temperature were taken. The study showed that the results from
the simulations were within the range of the pre-obtained results. For dean numbers ranging
from 38 to 350 the overall heat transfer coefficients were determined. The results showed
that the overall heat transfer coefficients varied directly with the inner dean number but the
fluid flow conditions in the outer pipe had a major contribution on the overall heat transfer
coefficient. The study showed that during the design of a double pipe helical heat exchanger
the design of the outré pipe should get the highest priority in order to get a higher overall
heat transfer coefficient. We have conducted experiment on different pipe geometry to find
the coefficient of loss and compared the results with ANSYS results.

J.S. Jayakumar et al [13] observed that the use of constant values for the transfer and
thermal properties of the fluid resulted in inaccurate heat transfer coefficients. Based on the
CFD analysis results a correlation was developed in order to evaluate the heat transfer
coefficient of the coil. In this study, analysis was done for both the constant wall temperature
and constant wall heat flux boundary conditions. The Nusselt numbers that were obtained
were found to be highest on the outer coil and lowest in the inner side. Various numerical
analyses were done so as to relate the coil parameters to heat transfer. The coil parameters
like the diameters of the pipes, the Pitch Circle Diameters have significant effect on the heat
transfer and the effect of the pitch is negligible.

G.Satish, K.Ashok Kuma, V. Vara Prasad, S k.M.Pasha [14]studied in their paper about
the flow through sudden and gradual change of pipe diameter (enlargement and contraction)
was numerically simulated with water by unsteady flow in k-epsilon scheme. The major
observations made related to the pressure and velocity contours in the process of flow
through these pipes. Sudden enlargement creates more severe formation of flow eddies than
sudden contraction. Also, the losses are more at the point where the enlargement in the pipe
begins. In the sudden contraction, vena-contracta’s are formed at the point of contraction
and effect of viscosity is negligible on the pressure drop through sudden contraction.
 9

Wan Kai, Wang Ping [15] studied in their paper about Using standard k-ε model with
FLUENT software on large diameter CFD numerical simulation of air flow in a 90 ° bent
tube. The standard k-ε model belongs to the eddy viscosity model, which adopts closed
RANS equations to solve the model. Assuming the air flow rate of 15 m/s the continuous
and stable manner flows through the elbow. As the flow rate is small, it can be considered
incompressible fluid. By homogenization of the continuity equation and instantaneous
Navier-Stokes equations, the Cartesian coordinate system under adiabatic, steady,
incompressible fluid flow is governed by the control equation. Basing on this paper we have
simulated to 90 degrees bent pipe.
 10

CHAPTER 3

THEORETICAL ANALYSIS
 11

3.1 LOSSES IN PIPES:

When a fluid flows through a pipe, the fluid experiences some resistance, due to which there
are some losses in the energy of fluid.

Energy Losses

Major Losses
Minor Losses
(Frictional Losses)
Sudden
Expansion
Sudden
Contraction

Bend

This loss of energy is classified as:

1. Major Energy losses (Friction loss)

2. Minor losses
 Sudden expansion
 Sudden contraction
 Bend in Pipe
 12

3.2 MAJOR LOSS IN PIPE (ENERGY LOSS DUE TO FRICTION):

Friction loss is the loss of energy or “head” that occurs in pipe flow due to viscous effects
generated by the surface of the pipe. Friction Loss is considered as a "major loss" and it is
not to be confused with “minor loss” which includes energy lost due to obstructions. In
mechanical systems such as internal combustion engines, it refers to the power lost in
overcoming the friction between two moving surfaces.

This energy drop is dependent on the wall shear stress (τ) between the fluid and pipe surface.
The shear stress of a flow is also dependent on whether the flow is turbulent or laminar. For
turbulent flow, the pressure drop is dependent on the roughness of the surface, while in
laminar flow, the roughness effects of the wall are negligible. This is due to the fact that in
turbulent flow, a thin viscous layer is formed near the pipe surface which causes a loss in
energy, while in laminar flow, this viscous layer is non-existent.

Friction loss has a few reasons, including:

 Frictional losses rely on upon the states of flow and the physical properties of the
system.
 Movement of fluid atoms against one another.
 Movement of fluid atoms against within surface of a channel.
 Bends wrinkles, and other sharp turns in hose or channelling.

In channel flows the losses because of contact are of two types: skin-rubbing and structure
grinding. The former is because of the roughness of the inward a piece of the channel where
the fluid interacts with the pipe material, while the latter is because of obstructions present
in the line of flow maybe a curve, control valve, or anything that changes the course of
movement of the flowing fluid.
 13

1. Darcy-Weisbach Formula:
In many practical engineering applications, the fluid flow is more rapid, therefore turbulent
rather than laminar. Under turbulent flow, the friction loss is found to be roughly
proportional to the square of the flow velocity and inversely proportional to the pipe
diameter, that is, the friction loss follows the phenomenological Darcy–Weisbach equation.
It is a method to calculate friction loss resulting from fluid motion in pipes is by using the
Darcy-Weisbach Equation. For a circular pipe:

Where,
=Head loss due to friction in unit of length
f= friction factor
D = Pipe Diameter
V = Flow velocity

2. Chezy’s formula:

In fluid dynamics, the Chezy’s formula describes the mean flow

velocity of steady, turbulent open channel flow:

𝑣 = 𝐶√𝑅𝑖

Where,
𝑣 is average velocity [m/s],
C is Chezy's coefficient [m½/s],
R is the hydraulic radius (~ water depth) [m], and
i is the bottom slope
 14

3.3 MINOR LOSSES IN PIPE:

Minor losses in pipes come from changes and components in a pipe system. This is different
from major losses because those come from friction in pipes over long spans. If the pipe is
long enough the minor losses can usually be neglected as they are much smaller than the
major losses. Even though they are termed “minor”, the losses can be greater than the major
losses, for example, when a valve is almost closed the loss can be almost infinite or when
there is a short pipe with many bends in it. There are three types of forces that contribute to
the total head in a pipe, which are elevation head, pressure head, and velocity head. Minor
losses are directly related to the velocity head of a pipe, meaning that the higher the velocity
head there is, the greater the losses will be. Units for minor losses are in length, such as feet
or meters, the same as any of the three types of head. A separate head loss coefficient, k, can
be determined for every element leading to minor losses. K is a dimensionless parameter to
help determine head loss. The coefficient is then multiplied by the velocity head to get the
head loss as shown below,
Head loss = head loss coefficient × velocity head

𝑣
ℎ=𝑘×
2𝑔
Where,
ℎ is the head loss
𝑘 is the loss coefficient.
𝑣 is the velocity
𝑔 is the acceleration due to gravity
Each, geometry of pipe entrance has an associated loss coefficient.
The minor loss of energy (or head) happens in the following cases:
1. Loss of head due to bend in the pipe.

2. Loss of head due to sudden expansion.

3. Loss of head due to contraction.

4. Loss of head due to different pipe fitting.

5. Loss of head due to entrance of a pipe.

 15

3.3.1 Loss of head due to sudden expansion:

Expansions are defined when the flow in a pipe goes from a small area to a larger area and
the velocity slows down. It is the exact opposite for contractions, the flow goes from a larger
pipe to a smaller one and the velocity increases. The loss or energy is due to turbulence, or
eddies, formed at the point where the pipe sizes change.
Because of sudden change in diameter across the pipe from D1 to D2, the fluid flowing
through the pipe is not fit to the unexpected change of the boundary. Thus, the flow separates
from the boundary and turbulent eddies are formed as indicated in fig 3.1. The loss of head
happens because of the creation of these eddies.

Fig. 3.1 Sudden expansion

Loss of energy is a result of turbulence. Measure of turbulence relies upon the difference in
the pipe diameters.
Head loss,

Where,
he = loss in head due to expansion.
V1=velocity at D1
 16

3.3.2 Loss due to sudden contraction:

Sudden contractions are defined when the area of the pipe diameter reduces suddenly along
the length of the channel (at the 90-degree plot). The downstream velocity will be higher
than the upstream velocity. The streamlines cannot follow the abrupt change of geometry
and hence gradually converge from an upstream section of the larger tube. However,
immediately downstream of the junction of contraction of area, the cross-sectional area of
the stream tube becomes the minimum and less than that of the smaller pipe. This section of
the stream tube is known as vena-contracta, after which the stream widens again to fill the
pipe. The flow pattern after the vena-contracta is similar to that after an abrupt enlargement,
and the loss of head is confined between section 1-1 and section 2-2. Therefore, we can say
that the loss due to contraction is not for the contraction itself, but due to the expansion
followed by the contraction.

Fig. 3.2 Vena-contracta formed in sudden contraction

Head loss due to contraction is expressed as,

,
Where,
𝑣 2=velocity at smaller diameter
𝑘 c =co-efficient of loss due to contraction
 17

3.3.3 Loss of head due to bend in pipe:

Bends are provided in pipes to change the direction of flow through it. An additional loss of
head, apart from that due to fluid friction, takes place in the course of flow through pipe
bend. The fluid takes a curved path while flowing through the pipe bend as shown in fig.3.3.

Fig. 3.3 Bends in pipe

Whenever a fluid flows in a curved path, there must be a force acting radially inwards on the
fluid to provide the inward acceleration, known as centripetal acceleration. Fluid particles in
this region, because of their close proximity to the wall, have low velocities and cannot
overcome the adverse pressure gradient and this leads to a separation of flow from the
boundary and consequent losses of energy in generating local eddies. Losses also take place
due to a secondary flow in the radial plane of the pipe because of a change in pressure in the
radial depth of the pipe.
Loss in head due to Bend is expressed as:

Where,
𝑘 b= co-efficient of loss
𝑣 = velocity of flow in pipe
 18

3.4 INTRODUCTION TO ANSYS

ANSYS is a general-purpose finite element modelling package for numerically solving a
wide variety of mechanical problems. These problems include static/dynamic, structural
analysis (both linear and nonlinear), heat transfer, and fluid problems, as well as acoustic and
electromagnetic problems.
ANSYS is the standard FEA technique in Mechanical Engineering Department also used in
Civil and Electrical Engineering, as well as in the Physics and Chemistry departments.
ANSYS provides a cost-effective way to explore the performance of products or processes
in a virtual environment. This type of product development is termed virtual prototyping.
With virtual prototyping techniques, users can iterate various scenarios to optimize the
product long before the manufacturing is started. This enables a reduction in the level of risk,
and in the cost of ineffective designs. The multifaceted nature of ANSYS also provides a
means to ensure that users are able to see the effect of a design on the whole behaviour of
the product, be it electromagnetic, thermal, mechanical etc.

3.4.1 Generic Steps for Solving Any Problem in ANSYS

Like solving any problem analytically, we need to define our solution domain,
physical model, boundary conditions and the physical properties in ANSYS. You then solve
the problem and present the results, compare to numerical methods, the main difference is
an extra step called mesh generation. This is the step that

 Build Geometry
 Define Material Properties
 Generate Mesh
 Apply Loads, and boundary conditions
 Obtain Solution
 Present the Results
Build Geometry: In this stage construct a two or three-dimensional representation of the
object to be modelled and tested using the work plane coordinate system within ANSYS.

Define Material Properties: Now that the part exists, define a library of the necessary
materials and material properties that compose the object (or project) being modelled.
 19

This includes thermal and mechanical properties of the object.

Generate Mesh: At this point ANSYS understands the makeup of the part. Now define how
the modelled system should be broken down into finite pieces.

Apply Loads: Once the system is fully designed, the last task is to apply the system with
constraints, such as physical loadings or boundary conditions.

Obtain Solution: In this step we obtain the solution. In this step we need to understand
within what state (steady state, transient… etc.) the problem must be solved.

Present the Results: After the solution has been obtained, there are many ways to present
ANSYS results, choose from many options such as tables, graphs, and contour plots.

3.4.2 ANSYS Fluent Features

a) Efficient and Flexible Workflow

Fluent is fully integrated into the ANSYS Workbench environment, a platform designed for
efficient and flexible workflows, CAD associatively and powerful capabilities in geometry
modelling and meshing. The built-in parameter manager makes it easy to rapidly explore
multiple design options.
b) Go Faster with High Performance Computing (HPC)
With HPC, ANSYS Fluent delivers CFD simulation solutions faster so that engineers and
designers can make better decisions sooner in the design cycle. While ANSYS HPC provides
linear scalability on systems with tens of thousands of processors, there is more to HPC than
just the number of cores. ANSYS also optimizes processor architecture, algorithms for
model partitioning, optimized communications and load balancing between processors to
deliver results in breath-taking speed on a wide variety of simulation models.
c) Turbulence Modelling
ANSYS Fluent software places special emphasis on providing a wide range of turbulence
models to capture the effects of turbulence accurately and efficiently. Several innovative
models such as the Menter–Langtry γ–θ laminar–turbulent transition model™ are available
only in Fluent.
d) Fluid-Structure Interaction
Fluent models the effects of solid motion on fluid flow by coupling with ANSYS structural
mechanics solutions through the Workbench unified user environment. Fluent users enjoy
 20

robust and accurate two-way FSI without the need to purchase, administer or configure third-
party coupling and pre- and post-processing software.
e) Heat Transfer & Radiation
Fluent handles all types of radiative heat exchange in and between fluids and solids, from
fully and semi-transparent to radiation, or opaque. You can choose from a variety of spectral
models to account for wavelength dependencies in a simulation and to account for scattering
effects.

3.5 COMPUTATIONAL FLUID DYNAMICS:

3.5.1 DEFINITION AND HISTORY:

Computational Fluid Dynamics (CFD) is the use of computer-based simulation to analyse
systems involving fluid flow, heat transfer and associated phenomena such as chemical
reaction. A numerical model is first constructed using a set of mathematical equations that
describe the flow. These equations are then solved using a computer programme in order to
obtain the flow variables throughout the flow domain.

3.5.2 GOVERNING EQUATIONS

The governing equations of fluid flow represent mathematical statements of the conservation
laws of physics. Each individual governing equation represents a conservation principle. The
fundamental equations of fluid dynamics are based on the following universal laws of
conservation. They are,
Unsteady state 3-D equation of continuity: -
𝜕 ∂⍴
+ 𝑑𝑖𝑣 (⍴𝑢) = 0
𝜕𝜕𝑡
Momentum equation: -
𝜕(⍴u) 𝜕𝑃
+ 𝑑𝑖𝑣 (⍴𝑢μ) = − + 𝑑𝑖𝑣(μ𝑔𝑟𝑎𝑑𝑢)
𝜕𝑡 𝜕𝑥
Energy equation: -
∂(⍴i)
+ div(⍴iu) = −Pdiv u + div(k grad T) + ∅ + 𝑆
∂t
 21

3.5.3 APPLICATIONS OF CFD:

The earliest adopters of CFD were the aerospace, automotive and nuclear industries. Further
growth and development in CFD and its ability to model complex phenomena along with
the rapid increase in computer power have constantly widened the range of application of
CFD. CFD is applied in a wide range of industries including mechanical, process, petroleum,
power, metallurgical, biomedical, and pharmaceutical and food industries.
CFD techniques have been applied on a broad scale in the process industry to gain insight
into various flow phenomena, examine different equipment designs or compare performance
under different operating conditions.
Examples of CFD applications in the chemical process industry include drying, combustion,
separation, heat exchange, mass transfer, pipeline flow, reaction, mixing, multiphase
systems and material processing.

3.5.4 ADVANTAGES OF CFD

 No restriction to linearity.
 Complicated physics can be treated.
 Time evaluation of flow can be obtained.
 It has the potential of providing information not available by other means.
 Computational investigation can be performed with remarkable speed. Designer can
study the implications of hundreds of different configurations in minimum time and
choose the optimum design.
 It gives detailed and complete information. It can provide the values of all the
relevant variables (pressure, velocity, temperature, concentration, turbulence)
throughout the domain of interest.
3.5.5 DISADVANTAGES
 Truncation errors
 Boundary condition problems
 Computer costs
 Computer storage & speed
 22

3.5.6 CFD ANALYSIS PROCEDURE

Computational fluid dynamics (CFD) study of the system starts with the construction of
desired geometry and mesh for modelling the dominion. Generally, geometry is simplified
for the CFD studies. Meshing is the discrete process of the domain into small volumes where
the equations are solved by the help of iterative methods. Modelling starts with the
describing of the boundary and initial conditions for the dominion and leads to modelling of
the entire system. Finally, it is followed by the analysis of the results, discussions and
conclusions.
The complete CFD analysis procedure can be divided into the following six stages.
a) Initial thinking
It is very important to understand as much as possible about the problem being simulated in
order to accurately define it. This stage involves collecting all the necessary data required
for the simulation including geometry details, fluid properties, flow specifications, and
boundary and initial conditions.
b) Geometry creation
The geometry of the flow domain is created using specialised drawing software.
Usually, 2-D sketches are first drawn and 3-D tools are then used to generate the full
geometry.
c) Mesh generation
In this stage the continuous space of the flow domain is divided into sufficiently small
discrete cells, the distribution of which determines the positions where the flow variables are
to be calculated and stored. Variable gradients are generally more accurately calculated on
a fine mesh than on a coarse one. A fine mesh is therefore particularly important in regions
where large variations in the flow variables are expected. A fine mesh, however, requires
more computational power and time. The mesh size is optimised by conducting a mesh-
independence test whereby, starting with a coarse mesh, the mesh size is refined until the
simulation results are no longer affected by any further refinement.
d) Flow specification
Flow specification involves defining the fluid physical properties, flow models, boundary
conditions, and initial flow conditions, as determined in the initial thinking stage.
 23

e) Calculation of the numerical solution

When all the information required for the simulation has been specified, the CFD software
performs iterative calculations to arrive at a solution to the numerical equations representing
the flow. The user needs also to provide the information that will control the numerical
solution process such as the advection scheme and convergence criteria.
f) Results analysis
Having obtained the solution, the user can then analyse the results in order to check that the
solution is satisfactory and to determine the required flow data. If the results obtained are
unsatisfactory, the possible source of error needs to be identified, which can be an incorrect
flow specification, a poor mesh quality, or a conceptual mistake in the formulation of the
problem.

Start

Initial thinking Conceptual

mistake?

Geometry and mesh Yes

Poor mesh?
generation

No
Yes
Flow specification Incorrect flow
specifications?

Calculate numerical
solution

Results not acceptable

Analyse the results

Results acceptable

Stop
 24

CHAPTER 4

PROCEDURE OF EXPERIMENT
 25

4.1 STEPS:
1. First of all, bench valve, gate valve and the flow control valves are opened and after
that the pump is started to fill the test rig with water.
2. Air, if present is bleed from the pressure tap points and the manometers by adjusting
the bench and flow control valves and air bleed screw.
3. The scales of all the manometer levels are checked when all the valves are fully
opened. The level is adjusted with the help of air bleed screw.
4. The reading is recorded for a selected flow from all the manometers after the water
levels have steadied.
5. The flow rate is determined by collecting some fixed volume of water (15000 cm 3)
in volumetric storage tank with the help of stopper. A digital stopwatch is used to
record time. This is used to calculate the discharge of water.
6. Step 4 and 5 are repeated in different pipes for two more flows.
7. The flow rate is adjusted by the control valve and pressure drop across the gate valve
is measured from the pressure gauge.
8. Step 7 is repeated for more two flow rates.

Fig.4.1 Apparatus of Minor Loss

 26

CHAPTER 5

OBSERVATIONS
&
CALCULATIONS
 27

5.1 SUDDEN EXPANSION:

5.1.1 EXPERIMENTAL OBSERVATION AND CALCULATION:

OBSERVATION:
Diameter of the smaller pipe before expansion is 19 mm and after expansion is 25.4mm.
Diameter of larger pipe before expansion is 25.4mm and after expansion is 38.1mm.
Table 5.1: Observation table of sudden expansion

Volume Discharge Manometer

Rise in Time Head
Sl. No Q Difference
level(cm) (sec) he(m) (m/sec)
(m3/sec) h(mm)

Pipe 1 10 15000 10 1.5×10-3 25 0.315 5.29 0.22

Pipe 1 10 15000 11 1.3×10-3 19 0.24 4.8 0.2

Pipe 2 10 15000 13 1.1×10-3 7 0.084 2.17 0.35

Pipe 2 10 15000 23 6.43×10-4 2.5 0.03 1.26 0.33

CALCULATION:
For pipe 1,
Q= volume/ time
= 0.015/10
= 1.5× m3/sec
.
A= ᴨ = ᴨ× = 0.28× m2

Q = A×v1
Q=1.5×10-3 = 0.28×10-3 ×v1
V1 =5.29 m/s
.
he =h( − 1) = 25×( − 1) = 25×12.6 mm

he = 0.315 m

he = ke×
𝟐𝒈𝒉𝒆 𝟐×𝟗.𝟖𝟏×𝟎.𝟑𝟏𝟓
𝒌𝒆 = 𝒗𝟏 𝟐
= 𝟓.𝟐𝟗𝟐
= 0.22
 28

5.1.2 FROM ANSYS:

a) Modelling

Fig.5.1. Modelling of Sudden expansion of pipe

b) Meshing

Fig.5.2. Meshing of sudden expansion pipe

 29

c)Stream function

Fig. 5.3: Stream function of flow

Just after the junction, stream line at the surface is disconnected, formation of eddies
occurs and further touches the surface.

d)Absolute Pressure

Fig. 5.4: Contours of absolute pressure in pipe

At the inlet, pressure is constant and near the junction, variation in pressure can be shown.
And at outlet, pressure is much more than inlet.
 30

e) Velocity

Fig 5.5: Contours of velocity magnitude

There is decrease in velocity of fluid when it reaches the junction. At the right of the junction,
velocity decreases in right direction and after some distance, velocity reaches steady state
and velocity at outlet is less than inlet.
f) Dynamic Pressure

Fig 5.6: Contour of dynamic pressure

It is nearly same as velocity contour because dynamic pressure is directly proportional to
square of the velocity.
 31

g) Velocity variation

Graph 5.1: velocity variation at the outlet surface

At the pipe surface, velocity is zero and it is increased as we go away from surface to towards
the axis and velocity is maximum at the axis.

h) Result:
Pressure at outlet =1.01×105 pa.
Pressure at inlet=9.73×104 pa.
Difference in pressure= (1.01×105) - (9.73×104 )= 3.7×103 pa.
. × . ×
Head difference = = = 0.377 m
× .

he = ke×
× . × .
𝑘 = = = 0.26
.
 32

Table 5.2. Change in pressure and corresponding head between inlet

and outlet at different velocity
Velocity(m/s) change in pressure(Pa) head loss(m)

1 133.7 0.013655

2 516 0.0527

3 1411.3 0.144137

4 2365.9 0.241631

5 3410 0.348266

6 4727.2 0.482793

7 6517.65 0.665653

8 8270.7 0.844694
9 9863.4 1.007357

10 12016.4 1.227245

Graph 5.2: Variation of pressure difference at different velocity

 33

5.2 SUDDEN CONTRACTION:

5.2.1 EXPERIMENTAL OBSERVATION AND CALCULATION:
OBSERVATION:
Table 5.3: Observation table of sudden contraction
Sl No Rise Volume Time Discharge Manometer Head V1
in (sec ( Difference (m) (m/sec) (m/sec)
level ) h(mm)
(cm)

Pipe 1 10 15000 10 1.5×10-3 24 0.315 2.97 5.29 0.22

Pipe 1 10 15000 18 8.18×10-4 7.1 0.106 1.62 2.89 0.4

Pipe 2 10 15000 16 9.0×10-4 4.05 0.051 0.995 1.77 0.31

Pipe 2 10 15000 18 8.18×10-4 3.25 0.054 0.905 1.61 0.41

CALCULATION:
For pipe 1
Q= volume/ time
= 0.015/10
= 1.5×10-3 m3/sec
.
A= ᴨ = ᴨ× = 0.28× m2

Q = A×v2
1.5×10-3 = 0.28×10-3 ×v2
V2=5.29 m/s
.
hc = h( − 1) = 25×( − 1) = 25×12.6 mm

hc = 0.315 m

hc = kc×
𝟐𝒈𝒉𝒄 𝟐×𝟗.𝟖𝟏×𝟎.𝟑𝟏𝟓
𝒌𝒄 = = = 0.22
𝟓.𝟐𝟗𝟐
 34

5.2.2 FROM ANSYS

a) Modelling

Fig 5.7: Geometry of sudden contraction

b) Meshing

Fig 5.8: Mesh of sudden contraction

 35

c) Stream Function

Fig 5.9: Contour of Stream function

In the above fig 5.9, it clearly shows that stream lines change their path at junction and
minimum area of flow occurs at just right of the junction which is known as vena-contracta.

d) Absolute Pressure

Fig 5.10: Contour of absolute pressure

Pressure decreases when flow is contracted at the junction of the pipe surface having very
high pressure, and at outlet (smaller diameter) pressure is lower.
 36

e) Velocity

Fig 5.11: Contour of velocity

Velocity increases when stream enters in smaller diameter and velocity is maximum at vena-
contracta because area of flow is minimum at that place.

f) Dynamic Pressure

Fig 5.12: Contour of Dynamic pressure

It is nearly same as velocity profile because, it is directly proportional to square of the

velocity.
 37

g) Static Pressure

Fig 5.13: Contour of static pressure

From above Fig 5.13, it clearly shows that there is loss in static pressure between inlet and
outlet. Pressure at inlet (pipe having large diameter) is more than the outlet (pipe having
smaller diameter).

h) Result:
Pressure at inlet=9.35× 10 Pa.
At outlet=6. 0 × 10 Pa.
Difference in pressure=3.35× 10 Pa.
. × . ×
Head difference = = = 0.342 m
× .

For pipe 1 and V2 = 5.29 m/s.

hc = kc×

𝟐𝒈𝒉𝒄 𝟐×𝟗.𝟖𝟏×𝟎.𝟑𝟒𝟐
𝒌𝒄 = = = 0.24
𝒗𝟐 𝟐 𝟓.𝟐𝟗𝟐
 38

5.3 BEND
5.3.1 EXPERIMENTAL OBSERVATION AND CALCULATION:
OBSERVATION:

Table 5.4 Observation table for bend

Sl No Rise in Volume Time Discharge Manometer Head

level(cm) ( ) (sec) ( Difference(mm) (m) (m/sec)

Pipe 1 10 15000 10 1.5×10-3 14.2 0.1786 2.96 0.4

Pipe 1 10 15000 15 1×10-3 7 0.09 1.97 0.45

Pipe 2 10 15000 16 9×10-4 16 0.1984 3.16 0.39

Pipe 2 10 15000 24 6.21×10-4 8 0.1 2.18 0.41

CALCULATION:
For pipe 1,
Q= volume/ time
= 0.015/10
= 1.5× m3/sec
.
A= ᴨ = ᴨ× = 0.645× m2

Q = A×v
Q=1.5×10-3 = 0.645×10-3 ×v
V =2.96 m/s

hb = kb×

𝟐𝒈𝒉𝒃 𝟐×𝟗.𝟖𝟏×𝟎.𝟏𝟕𝟖𝟔
𝒌𝒃 = = = 0.4
𝒗𝟐 𝟐.𝟗𝟔𝟐
 39

5.3.2 FROM ANSYS

a) Modelling

Fig 5.14: Geometry of elbow

b) Meshing

Fig 5.15: Mesh of elbow

 40

c)Velocity

Fig 5.16: contour of velocity

Velocity is maximum at the bend because of the change in direction of flow.

d)Dynamic Pressure

Fig 5.17: Contour of dynamic pressure

 41

e) Static Pressure

Fig 5.18: contour of static pressure

Static pressure increases at the bend and after the bend static pressure decreases gradually.

g) RESULT:

For bend pipe,

Inlet pressure= 2.23 103 Pa
Outlet pressure= -5.6×10-1 Pa
Difference in Pressure = 2.286x103 Pa
pressure difference 2.286 × 1000
head loss = = m
ρg 1000 × 9.81
=0.233m

Now, kb = Coefficient of bend

𝟐𝒈𝒉𝒃 𝟐×𝟗.𝟖𝟏×𝟎.𝟏𝟕𝟖𝟔
𝒌𝒃 = 𝒗𝟐
= 𝟐.𝟗𝟔𝟐
= 0.52
 42

CHAPTER 6

ANALYSIS OF RESULTS
 43

6.1 COMPARISON BETWEEN ANSYS AND EXPERIMENTAL

RESULTS

Table 6.1: Results in ANSYS and Experiments

RESULT Kb Ke Kc

From experiment 0.4 0.22 0.22

From ANSYS 0.52 0.26 0.24

Graph 6.1: Results in ANSYS and Experiment

0.6

0.5

0.4

0.3

0.2

0.1

0
Kb Ke Kc

Experiment ANSYS
 44

CHAPTER 7

CONCLUSION
 45

The experimental study is conducted and the results obtained from experiment and
CFD simulation are compared with each other. The graph shows the variation of head loss
with respect to change in the velocity.
From the entire project, it can be concluded that the results obtained from CFD
simulation are slightly higher than the results obtained from the experiment.
The variation in the results is because of various parameters acting on the fluid such
as resistance between layers, errors in taking the readings, and air acting on the surface of
the fluid.
Loss co-efficient of bends is 3% more in ANSYS results than result obtained from
experiments. The difference between results is less which means values are approximately
same. When the bend curvature is high, the velocity profiles at the bend inlet are shifted
towards the inner pipe wall, whereas at low curvature the velocity profiles remain symmetric .
Loss co-efficient of contraction is 1.8% more in ANSYS results than the result
obtained from experiments. The values of result are nearly same.
Loss co-efficient of expansion is 0.9% more in ANSYS results than the result
obtained from experiments. In this geometry, the results obtained from both ANSYS and
experiments are almost same.
 REFERENCES
[1] Sovran G., Ackeret, Fluid mechanics of internal flow, Elsevier Publishing Company
(1967).

[2] G. P. Celata, M. Cumo, S. McPhail, and G. Zummo, "Characterization of fluid dynamic

behaviour and channel wall effects in microtube," International Journal of Heat and Fluid
Flow, vol. 27, pp. 135-143, 2006.

[3] Hager, W.H., “Cavity Outflow from a Nearly Horizontal Pipe,” Int. J. Multiphase Flow,
25, 349 (1999).

[4] Laursen (1970), “Contraction coefficient at sudden expansion at bridge locations”.

[5] Kindsvater, Carter and Lacy et al (1953) and Kindsvater and Carter et al (1955) Effects
of different types of contractions on discharge characteristics.

[6] Rathakrishnan, E. and Sreekanth, A.K., Flow in pipes with sudden enlargement,
Proceedings of the 14th International Symposium on Space Technology and Science, Tokyo,
Japan, p.491 (1984).

[7] Wick, R. S., The effect of boundary layer on sonic flow through an abrupt cross-sectional
area change, Journal of the Aeronautical Science, Vol 20, p.675 (1953).

[8] Tapas K. Mandal, “Motion of Taylor Bubbles and Taylor Drops in Liquid−Liquid
Systems”, Ind. Eng. Chem. Res., 2008, 47 (18), pp 7048–7057.

[9] Rodriguez (2009)International Journal of Multiphase Flow Volume 36, Issues 11–
12, November–December 2010

[10] Vallentine Sixth International Water Technology Conference, IWTC 2001, Alexandria,
Egypt (1958), Effect of pipe contraction placed normal to channel axis.
[11] Delhaye et al (1981), Wadle et al (1989) Schmidt and Friedel et al (1997), Guglielmini
et al (1997), Fossa and Guglielmini et al (2002) “Oil–water flows through sudden contraction
and expansion in a horizontal pipe–Phase distribution and pressure drop”.

[12] Timothy J Rennie, “Comparison of heat transfer rates between a straight tube heat
exchanger and a helically coiled heat exchanger”, International Communications in Heat and
Mass Transfer 29 (2), 185-191.

[13] J.S. Jayakumar, “Experimental and CFD investigation of convective heat transfer in
helically coiled tube heat exchanger”.

[14] “Comparison of flow analysis of a sudden and gradual change Of pipe diameter using
fluent software” G.Satish, K.Ashok Kuma, V.Vara Prasad, Sk.M.Pasha IJSRD -
International Journal for Scientific Research & Development| Vol. 2, Issue 05, 2014|

[15] “CFD numerical simulation analysis of small and medium caliber 90 ° circular bend”
Wan Kai,Wang Ping Proceedings of the 2nd International Conference on Computer Science
and Electronics Engineering (ICCSEE 2013) “Evaluation of head losses in fluid”.
de