Project Report 2018-2019

CHAPTER 1
INTRODUCTION

1.1 CENTRIFUGAL PUMP

Centrifugal pumps basically consist of a stationary pump casing and an impeller

mounted on a rotating shaft. The pump casing provides a pressure boundary for the
pump and contains channels to properly direct the suction and discharge flow. The
pump casing has suction and discharge penetrations for the main flow path of the
pump and normally has small drain and vent fittings to remove gases trapped in the
pump casing or to drain the pump casing for maintenance.
The illustration directly below, is a simplified diagram of a typical centrifugal pump
that shows the relative locations of the pump suction, impeller, volute, and discharge.
The centrifugal pump casing guides the liquid from the suction connection to the
center, or eye, of the impeller. The vanes of the rotating impeller impart a radial and
rotary motion to the liquid, forcing it to the outer periphery of the pump casing where
it is collected in the outer part of the pump casing called the volute. The volute is a
region that expands in cross-sectional area as it wraps around the pump casing. The
purpose of the volute is to collect the liquid discharged from the periphery of the
impeller at high velocity and gradually cause a reduction in fluid velocity by
increasing the flow area. This converts the velocity head to static pressure. The fluid is
then discharged from the centrifugal pump through the discharge connection.

Fig no. 1.1 Centrifugal pump

1
 Project Report 2018-2019
Centrifugal pumps can also be constructed in a manner that results in two distinct
volutes, each receiving the liquid that is discharged from a 180 degrees region of the
impeller at any given time. Pumps of this type are called double volute pumps (they
may also be referred to a split volute pumps). In some applications the double volute
minimizes radial forces imparted to the shaft and bearings due to imbalances in the
pressure around the impeller. A comparison of single and double volute centrifugal
pumps is shown in fig no 2

Fig no.1. 2 single and double volute Fig no. 1.3 Pump characteristic curve
For a given centrifugal pump operating at a constant speed, the flow rate
through the pump is dependent upon the differential pressure or head developed by the
pump. The lower the pump head, the higher the flow rate. A vendor manual for a
specific pump usually contains a curve of pump flow rate versus pump head called a
pump characteristic curve. After a pump is installed in a system, it is usually tested to
ensure that the flow rate and head of the pump are within the required specifications.
A typical centrifugal pump characteristic curve is shown in the figure 3.
There are several terms associated with the pump characteristic curve that must be
defined. Shutoff head is the maximum head that can be developed by a centrifugal
pump operating at a set speed. Pump runout is the maximum flow that can be
developed by a centrifugal pump without damaging the pump. Centrifugal pumps
must be designed and operated to be protected from the conditions of pump runout or
operating at shutoff head.

2
 Project Report 2018-2019
1.2 CAVITATION

The vapour bubbles that form in a pot of boiling water will eventually rise to the
surface and burst. The energy release (explosion) that occurs upon burst is quite small.
But when that same bubble forms in the vanes of an impeller and begins its outward
migration, an extremely powerful force can be released.

Cavitation (from cavity) is defined as the rapid formation and collapse of vapour
bubbles or pockets in a liquid, due to dynamic action, and resulting in the formation of
cavities on the surfaces of solid boundaries. These solid boundaries can exist in any
number of structures including hydrofoils, pipes, and fittings but, in our industry, the
primary victims are impellers and propellers. Cavitation was first explained in 1917
by the English physicist, Lord Rayleigh, as he was studying a mysterious metal
erosion on ship propellers. Although its effects upon the surfaces of an impeller have
been witnessed by many of us in the pump industry cavitation, itself, remains a largely
misunderstood phenomenon. Actually, this is not too surprising when one considers
the complex physical behaviour of the relatively, simple water molecule and the
dynamic forces it encounters within the centrifugal pump.

The physics and thermodynamics of water vapour bubble formation are quite complex
but well understood. The complexities of cavitation have also been well researched
and a wealth of data on its causes and effects exist in both printed and on-line
literature. Unfortunately, most of the articles on cavitation that I have read tend to
circumvent what I call the "intuitive approach" to its understanding. The good ones
are typically too academic and mathematically complex for the majority of us and, the
"not so good" ones just gloss over the phenomenon itself and move on to its rather
vague causes and effects. Some even depict cavitation occurring on the wrong side of
the vane.

The goal of this thesis is to helps to gain an intuitive understanding of the cavitation
process and its causes. I will leave its prevention to the many articles that already
exist. Since boiling is the precursor to cavitation, we will begin by taking a look at
some of the properties of water and their role in the cavitation process. We will then

3
 Project Report 2018-2019
put NPSH into perspective and review some of the dynamics of centrifugal pump
operation that also directly affect the process. We will conclude with a discussion of
the different types of centrifugal pump cavitation and the conditions that can cause
them to occur.

1.3 NPSHa & NPSHr

The boiling point of water, in a pot, is a function of atmospheric pressure and the
temperature of the water. When a centrifugal pump is involved in the boiling process,
three additional components must be factored into the equation.
The NPSH available to a centrifugal pump combines the effects of atmospheric
pressure, water temperature, supply elevation, and the dynamics of the suction piping.
The following equation illustrates this relationship. All values are in feet of water.
NPSHa = Ha +/- Hz - Hf + Hv - Hvp
Where: Ha is the atmospheric pressure
Hz is the vertical distance from the surface of the water to the centerline of the
pump's suction (positive or negative)
Hi is the friction formed in the suction piping Hv is the velocity head at the
pump's suction
Hvp is the vapor pressure of the water at its ambient temperature
With the possible exception of Hv, the components of NPSHa are pretty self
explanatory. Unlike Hz which represents potential energy of a water column. Hv or
velocity head is the kinetic energy of a mass of water moving at some velocity V. It
can be easily calculated by determining the velocity m the suction piping from a
velocity table and substituting that value for V in the equation h =V2/ 2g (where g is
the force of gravity S 32 ft/sec2). It is usually small (at a velocity of 7 feet per second,
h = 0.765') but can be important when overall NPSHa is small or when suction lift
conditions exist.
At first glance, the equation for NPSHa looks pretty cut & dried but. it is actually
quite dynamic All of file variables can be in a continuous state of change.
Atmospheric pressure, for example, can vary by several feet depending upon weather
conditions. Likewise, water temperature and supply elevation can be seasonal. And,
velocity head and suction line friction vary as a function of flow. Usually the "worst
case" values for each of its components are used when calculating NPSHa.

4
 Project Report 2018-2019
A centrifugal pump begins to cavitate when file NP5H available falls below the vapor
pressure of the liquid in a pump's flow passages. Most published definitions of the Net
Positive Suction Head required by a pump (NPSHr) state that it is file head required to
prevent cavitation. Recently a knowledgeable engineer, and member of the Hydraulic
Institute, told me that file definition has changed over time. It now defines NPSHr as
the suction head at which a particular pump's hydraulic performance is degraded by
3%. This tidbit of knowledge is quite alarming because the degradation, itself, is due
to cavitation and. at the 3% level, it is almost certainly damaging cavitation. It is also
reasonable to expect that damaging cavitation could occur well before degradation
reaches the 3% level. The Hydraulic Institute's standards stipulate that the points on a
pump manufacturer's NPSHr curves must reflect these values. This same engineer said
that reputable manufacturers often inflate the NFS Hr points on published curves in
order to keep those of us unfamiliar with this definition out of trouble. Watch for the
HI to change the name of NPSHr to NPSHi
It is reasonable to expect that a small amount of cavitation, due to turbulence from the
impeller, can occur much of the time. In recent years a new term, NFSHi (inception),
was developed to define the pressure required at the suction that will suppress all
cavitation. The cavitation that occurs between NFSHi and the point where damage
occurs is called incipient cavitation. This form of cavitation appears to cause little, if
any, damage in normal pumping applications, however, there are instances where
excessive margins of NFSHa can increase the severity of incipient cavitation to a
point that it can cause damage, out, this is the stuff of pump design (pump and suction
specific speed).
NFS Hr cannot be calculated and must be determined by actual testing of each and
every pump model. Why does a pump require a positive suction head? Quite simply, it
is impossible to design a centrifugal pump that exhibits absolutely no pressure drop
between the suction inlet and its minimum pressure point, which normally occurs at
file entrance to the impeller vanes. And, even if we could its application would be
limited because it would be unable to take water from any source below its inlet.
Therefore, all pump systems must maintain a positive suction pressure that is
sufficient to overcome this pressure drop. If the pressure is not sufficient, the water
will boil.

5
 Project Report 2018-2019
Cavitation - The Process, Its Types,Causes
Suction cavitation, file most common and predictable form of cavitation, occurs when
the NPSH available at the suction inlet falls below the minimum NPSH required by a
particular pump. The areas most sensitive to this type of cavitation are the low
pressure sides of the impeller vanes near the inlet and file front shroud, where vane
curvature is greatest. When water flows over the surfaces of these convex curvatures,
the pressure near the surface is lowered and the flow can separate from the surface. If
the pressure is low enough, water vapor pockets or bubbles can form in the separated
area and then collapse or implode) upon entering an area of higher pressure. This
process of formation and collapse can occur thousands of times over a very short
linear distance of travel. Unlike an air bubble, which simply dissolves into the
surrounding water as pressure increases, the vapor bubble actually changes state from,
gas to liquid. This collapse or implosion of a bubble can be difficult to comprehend,
so let’s compare its action to that of a bubble bursting on the surface.
There are two distinct differences between the burst of a vapor bubble on the surface
of a boiling pot of water and its implosion or collapse beneath the surface. First, there
is a very large difference in the concentration of energy or the amount of energy
released per unit of volume. An exploding bubble releases an extremely small amount
of energy ion the order of 1+ atmospheres) while an imploding bubble, releases a very
large amount of energy (on the order of 10* atmospheres). The reason for the large
difference has to do with the change of state we discussed earlier. When water boils
and changes state to a vapor bubble, it gains energy in the process. But, it also gains
volume - - an increase of about 1,673 times that of its liquid state at 2120 F. (When
that same vapor bubble forms at 68°F, a typical cavitation temperature, its volume
increase is approximately 55,000 times that of its liquid state) The pressure in the
bubble is relatively small - - just slightly more than that of the atmosphere. When that
bubble rises to the surface and bursts, its contents remain in the vapor state and its
energy is simply dispersed into the air above. But, when that same vapor bubble
begins to collapse, it loses volume and the pressure (energy) within increases
accordingly. At total collapse, when its volume is almost zero, it changes back to a
liquid and all of the energy, gamed m the boiling process, is released into the water

6
 Project Report 2018-2019
around it in tire form of heat. Additionally, a shock wave is generated as the
surrounding water rushes in to fill the void left by the bubble.
The other difference is the manner in which that energy is dispersed. When a water
vapor bubble bursts its energy is dispersed into the atmosphere from a nearly perfect
circle on the surface of the water. It is not a violent explosion since its pressure is only
slightly greater than that of the atmosphere. In other words a relatively small amount
of energy is released over a relatively large area. But when a vapor bubble collapses,
all of its energy is released into an extremely small volume due to the shrinkage of the
bubble. This energy release can also be directed m a single direction, due to the
formation of something known as a re-entrant micro jet. If the implosion is near the
impeller surface, it has been shown that the direction of the micro jet will likely be
towards its surface. If so, the entire energy content of the bubble (both heat and shock
wave) will strike the impeller in an almost microscopic area.
So. it is a combination of the energy concentration and its focused direction, in the
form of a micro jet. that differentiates an explosion from an implosion and the
potential damage the latter can cause. Remember, a single cavitation bubble will not
cause detectable damage. Pitting is an erosion process that occurs, over time, due to
the continuous formation and collapse of vapor bubbles over a particular area. Severe
cavitation can cause significant damage in just a few days while minor cavitation may
not be noticed for years. Research has shown that the life of a cavitation bubble is on
the older of three milliseconds. When translated into "normal speak", this means that a
single bubble can form, collapse, and reform 333 times in a single second. During
severe cavitation tens, if not hundreds, of thousands of bubbles may be forming at any
point in time!
Even if the bubbles implode well above the surface of the impeller and pitting is
avoided, vibration and noise are likely to be present. Vibration is also due to the shock
waves emitted by the imploding bubbles and can lead to other forms of mechanical or
structural damage. The noise created by pump cavitation is a primary indication of its
presence and is similar to the sounds emitted by a pot of water just before it reaches
the boiling point. This "crackling" sound is caused by vapor bubbles that form at the
bottom of the pot. where the temperature is highest, but implode as they reach cooler
temperatures just a fraction of an inch above the bottom. (If water boiled in a pot at

7
 Project Report 2018-2019
5
6S F instead of 212°F, the bottom of the pot would not last very long. Why? Because
the bubble would be almost 55 times larger and so would the shock wave it generates.)
Current evidence also points to the presence of extremely high "pin point"
temperatures during bubble implosion but their damaging effect, if any, is secondary
to the mechanical shock wave that is released.
The figure no:1.4 is that of a water vapor bubble, in various stages of collapse, near
the surface of some solid boundary such as the vane of an impeller. The solid lines are
those that were predicted by a model and the dotted lines are those obtained through
experimentation. In stages 1 through 3 compression is fairly spherical, but starting in
stage 4 there is a small depression on the side of the bubble directly opposite of the
solid boundary. This depression becomes more pronounced in the latter stages and the
bubble becomes toroidal in its three dimensional shape.
Upon implosion, the re-entrant micro jet would normally form at the bottom of the
bubble and direct its energy towards the solid boundary.

fig.1.4 water vapour bubble- stages of collapse fig 1.5 series of bubble collapse

8
 Project Report 2018-2019

Fig. 1.6 Cavitation pattern during flow Fig 1.7 Damage at the leading edge

The fig 2 is a series of actual photographs of bubble collapse below a solid boundary.
It also shows the formation of the re-entrant micro jet. In stage 7 there is a small dot at
the bottom of the bubble. It grows in size in the following stages and appears to
penetrate the upper surface in stage 14. Stage 18 is just prior to total collapse and
generation of the shock wave.
Fig 3 shows the cavitation pattern during flow, bubbles are forming to the left and
imploding at the impeller’s surface in the upper right. Fig 4 shows the damage at the
low pressure side of the leading edge of an impeller vane due to suction cavitation.
1.4 CAVITATION PARAMETER
This is non-dimensional ratio of forces suppressing cavitation to the forces causing
cavitation. It is given by
σ = (hu – hv)/(v2/2g)
Because of scale effects cavitation parameter does not have a unique value even for
geometrically similar components tested with same fluid.
1.5 TYPES OF CAVITATION
According to the conditions under which, cavitation takes place and according to the
principle physical characteristic the appearance of cavitation may be classified into the
following types.

1.5.1 TRAVELLING CAVITATION

This type is composed of individual transit cavities for bubbles to form in the liquid
and move with the liquid, as they expand, shrink & then collapse. Such traveling
transit bubbles may appear at the low pressure points along the solid boundary or in
the liquid interior either at the cores of moving vortices or in the high turbulent region

9
 Project Report 2018-2019
in a turbulent shear field. The traveling of such cavities is the distinguishing from
other transit cavities.

1.5.2 FIXED CAVITATION

The term fixed cavitation refers to the situation that sometimes develops after
inceptions, in which the liquid flow detaches from the rigid boundary of an immersed,
body or a flow passage to form a pocket or cavity attached to the boundary. The
attached or fixed cavity is stable in a quasisteady sense. Fixed cavities sometimes have
the appearance of a highly turbulent boiling surface. In other cases the surface
between the liquid and the large cavity may be so smooth as to be transparent.

1.5.3 VORTEX CAVITATION

In vortex cavitation cavities are found in the cores of vortices, which form in zones of
high shear. The cavitation may appear as traveling cavities. Vortex cavitation is one of
the earliest observed types, since it often occurs on the blade tips of ships propellers.
In fact, this type of cavitation is often referred to as cavitation. Tip cavitation is not
restricted to open propellers but also occurs in ducted propellers such as are found in
propeller pumps at hydrofoil tips. Other examples of vortex cavitation are the
cavitation in the wake caused by the boundary layer separation from sphere.

1.5.4 VIBRATORY CAVITATION

In vibratory cavitation the forces causing cavities to form and collapse are due to
continuous series of high amplitude, high frequency pressure pulsation in the liquid.
These pressure pulsations are generated by a submerged, surface, which vibrates,
normal to it’s face and sets up pressure waves in the liquid. The measure difference
between this type of cavitation and the earlier once that the same liquid is used for
cavitation study in the former case where as the liquid element passes through the
cavitation zone only once.

1.6 EFFECTS OF CAVITATION

Generally the effects of cavitations are harmful in nature but there are situations where
effects have some useful applications. The harmful effects are:
1. Loss in efficiency of the performance of hydraulic machines like turbines, pumps &
Propellers.

10
 Project Report 2018-2019
2. Production of the noise, which is undesirable in the case of submarines.
3. Increase in the drag force of cavitating parts
4. Undesirable vibrations due to collapse of the bubbles are due to shock waves set up.
5. Material damage due to erosion.
1.7 THE BENIFICIAL EFFECTS OF CAVITATION:
1. Cavitation can be used for agitation and mixing.

2. A cavitation noise boomer can be used as sound source for an echo ranging survey
of ocean bottom conditions
3. Jet cavitations can be used very effectively for tunneling through rock.
1.7.1 Cavitation descriptors and similarity :
It is well known that in a flow system the pressure difference between any two points
is proportional to the difference between the squares of the velocities. When applied to
two geometrically similar system under similar flow conditions, these means that the
difference between the heads at two points is proportional to the square of
representative velocity, the values of the pressure head and velocity in the two system
being taken at corresponding points.

In caviling pump, two convenient points to consider are the point in the suction pipe
ahead of the impeller eye and the point of minimum pressure in the impeller. The
pressure head at the former point can be denoted by Hs .The pressure at the latter
points is in general not known. However, just when cavitation is about to begin, this
pressure should be equal to the saturated vapour pressure. Thus when cavitation is
imminent, the head difference between these two points is Hs-Hvp, where Hvp is
saturated vapour pressure head .If “v” is the velocity at some point in the system, then
the ratio
(Hs-Hvp) / (V2/2g)

will remain constant in all geometrically similar systems when Hs and V are measured
at corresponding points, and cavitation is imminent in all the systems (assuming that
the point of minimum pressure occurs at corresponding points in all the systems).
Now Vs is the velocity at the points where Hs is measured, the ratio Vs/V is constant
for all geometrically similar systems, and therefore the following relation is also
satisfied for all the systems:

11
 Project Report 2018-2019
(Hs + V s/ 2g- Hvp) / (V / 2g) = constant
2 2
…………….………(1)
Under similar conditions of operation, the total head produced by two geometrically
similar pumps is also proportional to the velocity head:
H / (V2/2g) = constant …………………………………………(2)
Eliminating v between equation (1) and (2),
(Hs + V2s /2g – Hvp) / H =constant …………………………...(3)
For all geometrically similar pumps under similar flow conditions. This constant is
usually denoted by ‘σ’, and the numerator on the left hand side of equation (3) by H sv,
so that
σ = Hsv /H …………………………………………(4)
Where, Hsv =Hs + V2s/2g – Hvp …………………………………… (5)
Hsv is called Net positive suction head, NPSH for short since it is a measure of the
margin of pressure available against cavitation inception. The use of σ, not only to
indicate cavitation inception, but also to denote the stage of cavitation occurring in the
machine, was first proposed by Thomas, and is named after him .Its use in describing
cavitation in turbines is unquestioned .For pumps, however, sigma is not entirely
suitable, because it involves H which depends on the impeller outlet diameter,
whereas it is the inlet geometry which largely determines pump cavitation. For similar
conditions of operation in geometrically similar pumps, the specific speed is constant:
Ns = n  (Q / H) = constant ………………………(6)
Where, n = pump speed.
Q = discharge,
Eliminating H from between equations (4) and (6),
Ns = n  (Q)*σ / Hsv = constant……………………….. (7)

The expression on the left hand side of equation (7) does not now contain any
term, which depends on the outlet diameter, and has found wide acceptance as a
cavitation descriptor for pumps.

12
 Project Report 2018-2019
CHAPTER 2
AIM AND OBJECTIVES
2.1 AIM
 The aim of this thesis is to accurately predict the cavitation of the centrifugal
pump with the help of Computational Fluid Dynamics (CFD)

2.2 OBJECTIVES
 The main objective of this work is to present the cavitation modelling in a
centrifugal pump impeller without the use of experimental works that in the
most case showing high cost.

 How the operating conditions and input parameters affect the impeller
geometry.

13
 Project Report 2018-2019
CHAPTER 3
LITERATURE REVIEW

Shalin Marathe and Rishi Saxena , Numerical analysis on effect of exit blade
angle on cavitation in centrifugal pump: June 2013 This paper presents the effect
of outlet blade angle on cavitation in centrifugal pump. The experiment is performed
on a centrifugal pump test rig consisting of backward bladed impeller at different
operating conditions and characteristics of the pump are predicted. Modeling of the
centrifugal pump along with the different configuration of the impeller having
different exit blade angles is carried out using Creo Parametric. Numerical simulation
is carried out using ANSYS CFX and standard k-_ turbulence model is implemented
for the analysis purpose. Cavitation is clearly predicted in the form of water vapor
formation inside the centrifugal pump from the simulation results. Analytical analysis
is carried out in order to find out NPSHr of the pump and Cavitation number (_c)
which indicates the cavitation phenomenon in the centrifugal pump. From the results
it has been found that the pump having low value of the blade exit angle will have less
chances of getting affected by the cavitation phenomenon.
LIU Houlin*, WANG Yong, YUAN Shouqi, TAN Minggao, and WANG Kai
“Effects of Blade Number on Characteristics of Centrifugal Pumps” chinese
journal of mechanical engineering Vol. 23,aNo. *,a2010 . The blade number of
impeller is an important design parameter of pumps, which affects the characteristics
of pump heavily. At present, the investigation focuses mostly on the performance
characteristics of axis flow pumps, the influence of blade number on inner flow filed
and characteristics of centrifugal pump has not been understood completely.
Therefore, the methods of numerical simulation and experimental verification are used
to investigate the effects of blade number on flow field and characteristics of a
centrifugal pump. The model pump has a design specific speed of 92.7 and an
impeller with 5 blades. The blade number is varied to 4, 6, 7 with the casing and other
geometric parameters keep constant. The inner flow fields and characteristics of the
centrifugal pumps with different blade number are simulated and predicted in
noncavitation and caviation conditions by using commercial code FLUENT. The
impellers with different blade number are made by using rapid prototyping, and their
characteristics are tested in an open loop. The comparison between prediction values

14
 Project Report 2018-2019
and experimental results indicates that the prediction results are satisfied. The
maximum discrepancy of prediction results for head, efficiency and required net
positive suction head are 4.83%, 3.9% and 0.36 m, respectively. With the increase of
blade number, the head of the model pumps increases too, the variable regulation of
efficiency and cavitation characteristics are complicated, but there are optimum values
of blade number for each one. The research results are helpful for hydraulic design of
centrifugal pump.
Myung Jin Kim, Hyun Bae Jin, and Wui Jun Chung: “A Study on Prediction of
Cavitation for Centrifugal Pump”, World Academy of Science, Engineering and
Technology Vol:6 2012-12-22. In this study, to accurately predict cavitation of a
centrifugal pump, numerical analysis was compared with experimental results
modelled on a small industrial centrifugal pump. In this study, numerical analysis was
compared with experimental results modelled on a small industrial centrifugal pump
for reliable prediction on cavitation of a centrifugal pump. To improve validity of the
numerical analysis, transient analysis was conducted on the calculated domain of full-
type geometry, such as an experimental apparatus. The numerical analysis from the
results was considered to be a reliable prediction of cavitaion.
Sujoy chakraborty, Kishan choudhary, Pransenjit datta, Bishop Debbarma;
“Performance prediction of centrifugal pumps with variations of blade number”;
Journal scientific & industrial research. Carried out two dimensional study of
steady, static pressure given out and incompressible flow characteristics inside the
passage with different numbers of blades of centrifugal pump impeller. The
investigation focuses mainly on the effective efficiency of the pump. Centrifugal
pump with impeller blades 5, 6 and 7 have been prepared and its efficiency at 3000
rpm is assess by FLUENT 6.3 software. The numerical analysis displays that with the
increase of the head, blade number, the head and static pressure of the model become
greater, but the efficiency of centrifugal pump varies with number of blades.

15
 Project Report 2018-2019

CHAPTER 4
MATHEMATICAL MODELLING

The fluid in the cavitating flow field for the pumps is considered as a homogeneous
and compressible mixed medium of vapour and liquid. The continuity and momentum
equations in the Cartesian coordinates are as follows:

where ⍴m and µm are the mixture density and dynamic viscosity, calculated by
weighted average of each phase volume fraction, u is the velocity, p is the pressure,
and µt is the turbulent viscosity, respectively. Subscripts i, j, k denote the axes
directions. The phases change vigorously in the cavitating flow, leading to a great
gradient of density in cavitation. Here, the full cavitation model proposed by Singhal
is employed, and the mixture density is defined as

where fv, fg, fl are the component mass fractions of the vapour, gas, liquid, ⍴ v,⍴ g,⍴ 1
are the component densities, and αv.αg,α1, are the component volume fractions. The
vapour generation and condensation rates are modelled as follows:

16
 Project Report 2018-2019
where the empirical factors are Ce=0.02 and Cc=0.01, the gas mass fractions
fg=1.5×10 5, the surface tension coefficient λ =0.0717 N/m, Pv denotes vaporization
pressure, and k denotes turbulent kinetic energy. The RNG k-Ԑ turbulence model,
which makes some corrections on the model coefficients in the Ԑ equation of standard
k-Ԑ turbulence model and can predict the rotating and curvature flow more accurately,
is widely applied in fluid machinery. For RNG k-Ԑ turbulence model, the k and Ԑ
equations can be written as:

where Ԑ and Gk denote the turbulent kinetic energy dissipation rate and turbulent
kinetic energy generation term, respectively. The empirical constants are αk=1.39
αs=1.39, and Cµ=0.09. The compressibility should be considered for the cavitating
flow consisting of liquid and vapour. To take into account the effects of the
compressibility, a filter function can be introduced to the turbulence model

where F is the filter function of the filter size (Δ), the turbulent kinetic energy and
dissipation rate, as shown below

In order to ensure the filtering process, the filter size in the present study is chosen to
be larger than the largest grid scale employed in the computation. According to eq.
(10), the turbulent structure smaller than the filter size will not be resolved. In the
calculation, a commercial CFD code Fluent coupled with the RNG k-Ԑ turbulence
model modified by a user defined file was used. The SIMPLEC algorithm was used to
solve the above steady flow governing equations. The boundary conditions are as
follows: the constant flow velocity at the inlet and the pressure at the outlet, the

17
 Project Report 2018-2019
standard wall functions were imposed over the impeller blades and sidewalls, the
volute casing and the inlet and outlet pipe walls in order to work with the filter
function. The pressure at pipe outlet was decreased step by step with a convergence at
every given pressure value. The multiple rotating reference frame (MRF) was applied
to couple the rotation and station domains.

18
 Project Report 2018-2019

CHAPTER 5
NUMERICAL ANALYSIS

Computational fluid dynamics investigation is very useful tool in designing of

machinery in turbo commercial enterprise. With the aid of the CFD proceed towards,
the compounded internal flows in water pump impellers, which are not fully
comprehend yet, can be well augur, to speed up the pump design procedure.
Computational fluid dynamics (CFD) analysis is being progressively applied in the
design of centrifugal pumps.

Computational Fluid Dynamics (CFD) is a powerful tool used to model the real life
behavior of fluids. It allows the optimization of design parameters without the need
for the costly testing of multiple prototypes. What is more, it is also a powerful
graphical tool for visualizing flow patterns that can give insight into flow physics that
otherwise would be very difficult and costly to discover experimentally, if possible at
all. Governing equations exist to model fluid behavior, but it is not always possible to
apply them to many of the complex flow patterns we see in the real world directly as
there would be too many unknown variables. However, CFD involves creating a
computational mesh to divide up real world continuous fluids into more manageable
discrete sections. The governing equations for fluid flow can then be applied to each
section individually, but as the properties of each section are inevitably linked to its
neighbouring sections, all the sections can be solved simultaneously until a full
solution for the entire flow field can be found. This method obviously requires a huge
amount of computational power, nevertheless with the advancement of modern
computing, solutions that would take months to compute by hand can now be found in
seconds using nothing more than an ordinary desktop or laptop computer.

As with everything, CFD is not without its limitations. Its accuracy or validity are
dependent on a multitude of different factors: the quality and appropriateness of the
mesh, the degree to which the chosen equations match the type of flow to be
modelled, the interpretation of the results, the accuracy of the boundary conditions
entered by the user or the level of convergence of the solution, to name but a few.
Often it comes down to the skill of the user, as each flow problem will be slightly
19
 Project Report 2018-2019
different and as a result, will require a slightly different modelling approach.
However, experimental data can provide a valuable reference point with which to
check the validity of CFD models.

Recent advances in computing power, in junction with powerful graphics and

interactive 3D influence of models have made the process of creating a CFD model
and analyzing results much less labour intensive, lesser time and, hence, cost. Thus,
CFD is any dominant tool for pump designers. The application of CFD tools in
industry which uses turbo machinery is quite common today. Many tasks can
arithmetically be solved much faster and cheaper than by means of experiments.
Nevertheless the highly wobbly flow in turbo machinery raises the question of the
most appropriate method for modeling the rotation of the impeller.
5.1 METHODOLOGY OF CFD ANALYSIS

Following figure describes methodology of CFD analysis.

Fig no:5.1 Flow chart of CFD

5.1.1 Initial Design:

Initial design of model is the planning decision and the geometry is generated
depending on this initial design considerations, using CFD model tool or other design,

20
 Project Report 2018-2019
which is used to input the problem geometry, generates the grid; define the flow
parameter and the boundary condition to code.

5.2.2 Geometry generation:

The geometry of design needs to be created from the initial design. Any modeling
software can be used for modeling and then shifted to some other simulation software
for the analysis purposes.

5.2.3 Mesh Design Generation:

Mesh generation (Gridding) is the process for subdividing a region to be modeled in to

set of small control volume. Associated with each control volume there will be one or
more value of the dependent flow variable (e.g. velocity, pressure and temperature
etc.) usually they represents some local type of averaged values. Numerical algorithms
constituting representing approximations to the conservations of law of mass,
momentum and energy are then used to compute these variables in each control
volume.

5.2.4 Pre processor

Pre processor is used to input the problem geometry, generates the grid; define the
flow parameter and the boundary conditions to code.

5.2.5 Solver

Solver is used to get governing equations of the flow subject to the conditions
provided there are three different methods to use as given below:

Finite difference method: Finite difference method utilized Taylor series

expansion to write the derivatives of the variables as the differences between the
values of variable at various points in space or time.

Finite elements method: In the finite element method, the fluid domain under the
consideration is divided in to small number of sub domains, known as element. The
simple function is assumed for the variation of each variable in each element. The
summation of the variations of variable in each element is used to describe the whole
fluid flow.
21
 Project Report 2018-2019
Finite volume method: Finite volume method is currently the most popular in
CFD. The main reason is that it can resolve all difficulties that the other two have.
Generally the finite volume method is the special case of finite element method.

5.2.6 Post processor

Post processor is used to message the data and show the results graphically and easy
to read format.

22
 Project Report 2018-2019

CHAPTER 6
RESEARCH METHODOLOGY

The blade number of impeller is an important design parameter of pumps, which

affects the characteristics of pump heavily. At present, the investigation focuses
mostly on the performance characteristics of axis flow pumps, the influence of blade
number on inner flow filed and characteristics of centrifugal pump has not been
understood completely. Therefore, the methods of numerical simulation and
experimental verification are used to investigate the effects of blade number on flow
field and characteristics of a centrifugal pump. The model pump has a design specific
speed of 92.7 and an impeller with 5 blades. The blade number is varied to 4, 6, 7 with
the casing and other geometric parameters keep constant. The inner flow fields and
characteristics of the centrifugal pumps with different blade number are simulated and
predicted in noncavitation and caviation conditions by using commercial code
FLUENT.
Here in this investigation we select various parameters for analysing the cavitation
behaviour such as blade number, inlet and outlet pressures, blade angles and blade
leading edge angle. The parameters are analysed by Ansys FLUENT.

6.1 CREATING IMPELLER USING BLADEGEN

ANSYS BladeGen is a geometry creation tool that is specialized for turbomachinery

blades. BladeGen is a component of ANSYS BladeModeler. The Blade Modeler
software is a specialized, easy-to-use tool for the rapid 3-D design of rotating
machinery components. Incorporating extensive turbo machinery expertise into a user-
friendly graphical environment, the software can be used to design axial, mixed-flow
and radial blade components in applications such as pumps, compressors, fans,
blowers, turbines, expanders, turbochargers, inducers and others.

Blade Modeler provides the essential link between blade design and advanced
simulation including computational fluid dynamics and stress analyses. Blade Modeler
contains a rich set of tools and functions for designing a turbo machinery blade from
scratch, using industry-specific tools, workflow, and language that the blade designer
expects.
23
 Project Report 2018-2019
With BladeGen, the user can re-design existing blades to achieve new design goals or
create completely new blade designs from scratch. When either re-designing or
evaluating an existing blade design, BladeGen facilitates the import of blade geometry
interactively or through user supplied files. BladeGen allows sculpted or ruled element
blades with linear or compound lean leading or trailing edges. Over/Under-Filing can
be applied and leading and trailing edge shapes are easily specified as a full radius, an
ellipse ratio, or a simple cutoff.

BladeModeler represents a pivotal link between blade design, advanced analysis and
manufacturing. Used in combination with ANSYS analysis software, users can rapidly
evaluate the performance of a component. BladeGen model files can be imported into
DesignModeler using the BladeEditor feature. BladeEditor provides a seamless path
to both structural and fluid analysis, which enables the user to efficiently transition
from preliminary blade design, to full 3-D viscous flow analysis, and finally to the
users native CAD system.

In analyzing the performance of pump systems, the presence of cavitation can have a
significant impact on the performance of the pump. Cavitation bubbles emerge when
the local static pressure of the working fluid (a liquid) falls below the vapour pressure.
Vigorous cavitation can lead to loss of head rise and, in many cases, structural damage
due to erosion from collapsing vapour bubbles striking the pump blade passages. This
involves the use of FLUENTs moving reference frame model and the multiphase
mixture model (with the cavitation sub‐model enabled).Upon using blade gen tool we
have to create an impeller and casing of a centrifugal pump and the certain operating
conditions and geometry parameters are required to create the impeller geometry. The
figure 6.1shows the dialogue boxes of operating conditions and geometry parameters.

24
 Project Report 2018-2019

Fig 6.1 operating conditions and geometric parameters

The values given in the operating conditions and geometrical parameters are available
from the paper “Effects of blade number on characteristics of centrifugal pumps”, by
LIU Houlin, WANG Young, YUAN Shouqi, TAN Minggao and WANG
Kai(Research Centre of Fluid Machinery Engineering and Technology, Jiangsu
University, Zhenjiang 212013, China). In this paper they create an experimental setup
and study the performance of the pump at different number of impellers. Here the

25
 Project Report 2018-2019
validation of the numerical analysis of centrifugal pump charecteristics is also done
through comparing the values with this paper.

6.1.1 EFFECT OF BLADE NUMBER ON PUMP CHARACTERISTICS

The pump under consideration is a five blade centrifugal pump operating at 2900 rpm.
The flow enters through a circular inlet pipe, passes through the impeller, and is
discharged radially through a uniform cylindrical exit boundary. The working fluid is
water and the flow is assumed to be steady and incompressible. The fluid properties
are as follows:
liquid density = 1000 kg/m3
liquid viscosity = 0.001 Pa‐s
vapor density = 0.01927 kg/m3
vapor viscosity = 8.8E‐6 Pa‐s
vapor pressure of liquid = 3540 Pa
The simulation will assume a total fixed volume flow rate of 50 m3/hr at the inlet.
This will be imposed using a fixed inlet velocity boundary condition. The flow will be
assumed fully turbulent, with an entrance turbulence intensity of 5%. A uniform
pressure outlet boundary condition is imposed at the outlet of the vaneless space. The
absolute inlet static pressure depends on the prescribed outlet absolute pressure.
Hence, the outlet pressure will implicitly define the level of cavitation that occurs
(relative to the minimum absolute pressure that occurs on the suction side of the
impeller), based on the specified vapour pressure for the liquid. In the present case,
the outlet pressure is set to 295000 Pa.

26
 Project Report 2018-2019
CASE 1: BLADE NUMBER,( Say Z=5)

Fig 6.2 number of blades 5

Fig 6.3 The geometry of the model

The figure6.3 shows the geometry of a five blade impeller. Next step is to meshing the
geometry.

27
 Project Report 2018-2019

Fig 6.4 Mesh model

Fig 6.5 results in blade gen

By putting the blade number as 5 and other parameters as like the experimental values
we get the Net Positive Suction Head (NPSHR) as 3.63m. And this value is similar to
the experimental analysis(NPSHR=3.68m). Thus from these observations we can say
that our numerical analysis is validated

28
 Project Report 2018-2019

CHAPTER 7
RESULTS AND DISCUSSION

A total of 4 parameters are used to find the cavitation parameters of the centrifugal
pump. They are
 Number of blades, Z
 Inlet and outlet pressures
 Inlet flow angle
 Trailing edge angle
7.1 NUMBER OF BLADES
The blade number of impeller is an important design parameter of pumps, which
affects the characteristics of pump heavily. At present, the investigation focuses
mostly on the performance characteristics of axis flow pumps, the influence of blade
number on inner flow filed and characteristics of centrifugal pump has not been
understood completely. Therefore, the methods of numerical simulation and
experimental verification are used to investigate the effects of blade number on flow
field and characteristics of a centrifugal pump. The model pump has a design specific
speed of 92.7 and an impeller with 5 blades. The blade number is varied to 4, 5,6, 7.
7.1.1 NUMBER OF BLADES, Z=4

Fig no.7.1 Profile of static pressure

29
 Project Report 2018-2019

Fig no.7.2 Contours of volume fraction

7.1.2 NUMBER OF BLADES ,Z =5

Fig no.7.3 Contours of static pressure

30
 Project Report 2018-2019

Fig no. 7.4 Contours of volume fraction

7.1.3 NUMBER OF BLADES ,Z =6

Fig no.7.5 Contours of static pressure

31
 Project Report 2018-2019

Fig no. 7.6 Contours of volume fraction

7.1.4 NUMBER OF BLADES ,Z =7

Fig no.7.7 Contours of static pressure

32
 Project Report 2018-2019

Fig no.7.8 Contours of volume fraction

7.1.5 EXPERIMENTAL AND NUMERICAL NPSHR FOR DIFFERENT

NUMBER OF IMPELLERS
Table no 7.1 Experimental and numerical NPSHr

Blade number(Z) Experimental NPSHR(m) Numerical NPSHR(m) % error

4 4.04 4.07 1.4
5 3.68 3.63 1.3
6 4.66 4.22 9.4
7 4.95 5.02 7

Chart Title contours of volume fraction

volume fraction

4
Experimenta
NPSHR (m)

l NPSHR(m) 5
Numerical 6
NPSHR(m)
7

Blade number Z blade number

Fig 7.9 results

7.1.6 OBSERVATIONS:

33
 Project Report 2018-2019
From the graph and table we conclude that the lower value of NPSHR is at when
impeller number, Z=5 and all the other values the value of NPSHR is larger. So we
conclude that the chance of cavitation parameter is less when the required net positive
suction head (NPSHR) minimum and it is at blade number , Z=5. Thus the default
value of number of blade is set as 5. Further analysis are done on this default value.

7.2 INLET AND OUTLET PRESSURES

The second parameter selected for the study is suction and delivery gauge pressures.
From the published papers and Ansys customers port the normal pressure range is lies
between 250000 Pa and 300000Pa. So in this tests fix the number of impeller is fix as
5 from the previous observations. In this study we set the input pressure value
(283000 Pa) is setting as a default value and outlet pressure is varying.

7.2.1 INLET PRESSURE 283000Pa AND OUT LET PRESSURE 290000Pa

Fig no.7.10 Contours of static pressure

34
 Project Report 2018-2019

Fig no. 7.11 Contours of volume fraction

7.2.2 INLET PRESSURE 283000Pa AND OUT LET PRESSURE 295000Pa

Fig no.7.12 Contours of static pressure

35
 Project Report 2018-2019

Fig no.7.13 Contours of volume fraction

7.2.3 INLET PRESSURE 283000Pa AND OUT LET PRESSURE 300000Pa

Fig no.7.14 Contours of static pressure

36
 Project Report 2018-2019

Fig no.7.15 Contours of volume fraction

7.2.4 INLET PRESSURE 283000Pa AND OUT LET PRESSURE 305000Pa

Fig no.7.16 Contours of static pressure

37
 Project Report 2018-2019

Fig no.7.17 Contours of volume fraction

7.2.5 INLET PRESSURE 283000Pa AND OUT LET PRESSURE 310000Pa

Fig no.7.18 Contours of static pressure

38
 Project Report 2018-2019

Fig no.7.19 Contours of volume fraction

7.2.6 OBSERVATIONS

volume volume inlet Outlet volume

volume fraction vs pressure variation
fraction, volume fraction,
290000, fraction, 305000,
pressure(Pa) pressure (Pa) fraction
volume
0.995 295000, 0.994
0.991 fraction, 290000
volume fraction

300000, 295000 283000 290000 0.995

0.982
300000 283000 295000 0.991
305000 283000 300000 0.982
310000 283000 305000 0.994
volume
fraction,
310000,
283000 310000 0.945
Outlet pressure
0.945
Fig no. 20. Volume fraction vs pressure Table no. 2 . Volume fraction vs pressure

From the observations we can find out that while increasing the outlet gauge pressure
with a constant value of inlet gauge pressure the amount of volume fraction is in an
decreasing manner except in one case. That means the amount of vapour content is
decreasing in the impeller blade region with increasing in pressure difference between
the inlet and out let respectively. So we conclude that the amount of cavitation is
decreasing with increasing the outlet pressure gauge reading.

39
 Project Report 2018-2019

7.3 INLET FLOW ANGLE

Inlet flow angle is the angle between direction of fluid enters into the eye of the pump
and the impeller. In normal working conditions the inlet flow angle is 900. In this
study we have to change the angle as 800, 1000 and 1100 and check how these
parameters affect the pump performances.

7.3.1 INLET FLOW ANGLE 800

Fig no.7.21 Contours of static pressure

40
 Project Report 2018-2019

Fig no.7.22 Contours of volume fraction

7.3.2 INLET FLOW ANGLE 1000

Fig no.7.23 Contours of static pressure

41
 Project Report 2018-2019

Fig no.7.24 Contours of volume fraction

7.3.3 INLET FLOW ANGLE 1100

Fig no.7.25 Contours of static pressure

42
 Project Report 2018-2019

Fig no.7.26 Contours of volume fraction

7.3.4 OBSERVATIONS
In normal working conditions the inlet flow angle is 900 that means radial impeller. In
this study we have to change the inlet flow angle as above and below the default
value.
From the results a graph is plotted as inlet flow angle vs volume fraction. From the
graph we can see that at 800 inlet flow angle the volume fraction is negligible while
increasing the angle above the 900 the amount of volume fraction suddenly increases
to a high value. Hence there is a large amount of cavitation is occurred in the impeller
while increasing the inlet flow angle. But in the case of backward flow there is a
chance of reverse flow of fluid. Hence in normal cases radial impellers are used.

inlet flow angle vs volume fraction

volume
fraction, 110,
0.984
volume fraction

80
100
110

inlet flow angle

Fig no. 7.27 inlet flow angle vs volume fraction

43
 Project Report 2018-2019
7.4 TRAILING EDGE BLADE ANGLE

The present numerical study and passive flow control are concerned with the effect of
the blade leading edge angle on the cavitation and performance of a centrifugal pump.
There is a limited number of works regarding the effect of the blade leading edge
angle on the pump cavitation.
7.4.1 TRAILING EDGE BLADE ANGLE 150

Fig no.7.28 Contours of static pressure

Fig no. 7.29 Contours of volume fraction

44
 Project Report 2018-2019
0
7.4.2 TRAILING EDGE BLADE ANGLE 27

Fig no.7.30 Contours of static pressure

Fig no.7.31 Contours of volume fraction

45
 Project Report 2018-2019
0
7.4.3 TRAILING EDGE BLADE ANGLE 40

Fig no.7.32 Contours of static pressure

Fig no.7.33 Contours of volume fraction

7.4.4 OBSERVATIONS
Trailing edge angle is the outlet blade angle, by changing the corresponding angle
there is a sudden change in cavitation phenomena. From fig 7.33we can see that at a
particular value of trailing edge angle (270) value of volume fraction is zero. If any
deviation from this value in upward or downward there is a sudden increment of
volume fraction. So the mixture vapour bubble is large enough to cause the cavitation.

46
 Project Report 2018-2019
Trailing edge angle vs volume fraction

volume fraction 15
27
40

Trailing edge angle

Fig 7.34 Trailing edge vs volume fraction

47
 Project Report 2018-2019
CHAPTER 8
CONCLUSION

The effects of blade number, inlet and outlet pressures, inlet flow angle and trailing
edge angle are on inner flow and characteristics of centrifugal pump were resea
rched by using the methods of numerical simulation. The main research conclusions
are below.

With the increase of blade number, the head of centrifugal pump grows all the time,
the change regulations of efficiency and NPSHR are complex, but there is an optimum
value for the best efficiency and cavitation characteristics, the optimum blade number
of the model pump in this paper for efficiency is 7, 5 blades secondly and for
cavitation characteristic is 5, 4 blades secondly, so the model pump with the 5 blades
has the best general characteristics.

With increasing the outlet gauge pressure with respect to the inlet pressure the volume
fraction amount is in decreasing manner. But in the case of inlet flow angle the
cavitation phenomenon is increases with blade angle. For trailing edge angle there is
an optimum value (270) for minimum value of cavitation and any deviation will cause
a large increment of vapour particles.

So in this paper we check different parameters how they affect or influence in the
phenomenon of cavitation on impeller of a centrifugal pump and in each case there is
an optimum value for minimum of that.

48
 Project Report 2018-2019

CHAPTER 9
REFERENCES

1.Shalin Marathe and Rishi Saxena ,” Numerical analysis on effect of exit blade angle
on cavitation in centrifugal pump”: June 2013

2 .LIU Houlin*, WANG Yong, YUAN Shouqi, TAN Minggao, and WANG Kai
“Effects of Blade Number on Characteristics of Centrifugal Pumps” chinese journal of
mechanical engineering Vol. 23,2010

3.Myung Jin Kim, Hyun Bae Jin, and Wui Jun Chung: “A Study on Prediction of
Cavitation for Centrifugal Pump”, World Academy of Science, Engineering and
Technology Vol:6 2012-12-22

4. Sujoy chakraborty, Kishan choudhary, Pransenjit datta, Bishop Debbarma;

“Performance prediction of centrifugal pumps with variations of blade number”;
Journal scientific & industrial research.

5. John s. Anagnostopoulos, “CFD Analysis and Design Effects in a Radial Pump

Impeller”, Wear transactions on fluid mechanics Issue 7, Vol. 1, July 2006.

6. M. Hofmann, B. Stoffel, O. Coutter-delgosha, R. Fortes- patellia, JL. Rebound,

”Experimental and numerical studies on a centrifugal pump with 2D-curved blades in
cavitating motion”, CAV 2001, Session B7.005

7.Sujoy Chakraborty a*, K.M. Pandey a, Bidesh Roya, “Numerical Analysis on

Effects of Blade Number Variations on Performance of Centrifugal Pumps with
Various Rotational Speeds” International Journal of Current Engineering and
Technology, 2012

49
 Project Report 2018-2019

50