1. JOMO KENYATTA UNIVERSITY OF AGRICULTURE AND TECHNOLOGY
2014
PERFORMANCE OF A
CENTRIFUGAL PUMP
EXPERIMENTAL ANALYSIS OF A ROTORDYNAMIC
PUMP
MIMISA DICKENS EN251-0305/2011
C I V I L , C O N S T R U C T I O N A N D E N V I R O N ME N T A L E N G I N E E R I N G D E P A R T ME N T
2. Abstract
The results of an experiment carried out to investigate the theory of a Rotor dynamic pump and to
determine the relationship between the head, discharge, the input power and the efficiency of a
centrifugal pump under the prescribed revolution speed are presents with much focus on the specific
aspects mentioned. This paper represents experimental study work carried out on centrifugal pump.
Vibrations and noise are both pre dominant due to hydraulic effects. The pump system used allows for
parametric variation of discharge. Data acquired are manually compiled and analyzed, reduced and
presented into a form that can be typically used to analyze pump characteristics. Reduced data is used
in determining the characteristic curve of the pump and to indicate the relationship between the
efficiency and flow rate and power.
Introduction
Centrifugal pumps are classified as rotary type of pumps in which a dynamic pressure developed enables
the lifting of water to great heights. The history of pumps dates back to the ancient day of technological
development in Egypt where the locals used water wheels with buckets mounted on them to move
water for the purposes of irrigation. It was not until the late 1600’s that true centrifugal pumps were
developed by Denis Papin, a French boy, who developed the hydraulic device though with straight
vanes. John G. Appold introduces the curved vane in 1851 thereby improving the efficiency of the
hydraulic device. It has been rapidly superseding the other types of pumps over the years and is
seemingly the most used kind of pump. It is most suited for situations requiring moderate to high flow
rates and modest increases in pressure. They are majorly used in municipal water supply systems,
circulating water heating and cooling systems applied in buildings, pump system in dish and cloth
washing machines and for pumping cooling water in automobile engines. Positive displacement pumps
are more suited for high pressure-low flow applications. Flow rate is a function of rotational speed and
has negligible dependence on pressure rise. They are also used to supply oil under very high pressure for
hydraulic actuators such as those on large earth moving machines.
Below is a sketch of a typical centrifugal pump.
Fluid which flows into the impeller within the
inner radius is given a significant momentum and kinetic energy thus enabling it to flow radially
3. outwards at a higher momentum and kinetic energy. As it leaves the outer radius of the impeller, it is
slowed down leading to a significant increase in pressure that was initially aimed for the system.
The actual head (H) produced by a centrifugal pump if dependant on the flow-rate (Q). The head-flow
relationship can be easily determined by selecting appropriate impeller geometry. Pumps are normally
designed in a way that head reduces with an increase in flow for the purpose of a stable flow rate when
the pump is connected to a piping system. A typical head-flow curve for a pump is as indicated below.
The result of the application of the mechanical energy equation applied on two sections of a piping
system proves that
퐻푃 +
푉2
1
2푔
+
푃1
훾
+ 푍1 = ℎ1 +
푉2
2
2푔
+
푃2
훾
+ 푍2
Where 퐻푃 is the pump head and ℎ1 = total head loss in the piping length under study. The others are the
pressure head and the velocity head of the system.
For any given pump operating at a given rotational speed, there is always only one operating point
where the geometry of the impeller blades is an optimum. When this is combined with the other forces
an efficiency of the system is obtained which is a function of the rate of flow. The efficiency is the ratio
of the fluid work (power outputted) to the shaft power input of the pump.
This relation can be shown as 휂 = 푌푄퐻푃
푃푠ℎ푎푓푡
The performance of a pump is highly dependent on the impeller and casing geometry, the
rotational speed, the size of the pump and the properties of the flowing speed. However, it is
not necessary to vary all these factors in order for one to be able to determine the performance
of a pump. Two geometrically identical pumps with flow rates adjusted so that the ratio of
tangential to radial fluid velocities is the same are said to be homologous. Homologous pumps
4. are known to have geometrical similarity and are also known to have the following
dimensionless parameters the same.
휋1 =
푄
푁퐷3
휋2 =
푔퐻
푁2퐷2
Where N is the angular speed and D the diameter. The Diameter is taken as a measure of the length
scale of the pump in question. A larger diameter indicates that all the other dimensions of the pump are
relatively larger. The outer diameter of the impeller is normally used. These relationships make it
possible to estimate the performance of a pump of known diameter by testing another pump with a
different diameter. It also becomes possible to determine the effects of a changing angular speed. This
scaling, however, is not perfect and a few errors are expected of it.
Materials Used
For the effectiveness of the process, a number of equipment and materials had to be availed. These are
as they have been listed below.
o A Centrifugal pump
o An Electric board comprised of an ammeter, a voltmeter and a power factor meter
o A V-notch with a hook gauge
o Pressure gauges on a suction pipe and a delivery pipe
o A thermometer
The apparatus were arranges in a set-up as shown below.
5. The set-up was arranged such that when the operation was started, all values are read simultaneously
for effectiveness of the process. The set-up was checked for correct layout with a few tests after which
the actual experiment was started and data collected.
Procedure
The temperature of the water was first measured after which the crest level of the v-notch was
measured using the hook-gauge. The operation of the pump was started with the gate valve closed after
which the gate valve was slowly opened and a small discharge set. The head above the v-notch was
measured using the hook gauge after it was clear that the flow had become steady. The readings of the
pressure gauges, voltmeter, ammeter and the power factor meter were recorded on the data sheet. The
procedure was repeated after the discharge was increased with the gate valve.
Theoretical Knowledge pertaining to the experiment
A pump is a device that supplies energy to a fluid. The effect of supplying energy can be s tudied via the
mechanical energy equation.
Δ퐸 + Δ
푝
휌
+
푔
푔푐
Δ푧 + Δ [
푉2
2훼푔푐
] + Σ퐹 = 푄ℎ − 푊푠
The equation neglects all shearing stresses. The power supplies in a system originate from a change in
pressure since pressure at section 2 is greater than the pressure at section 1. It also originates from the
change in level, change in kinetic energy and frictional changes.
In this experiment, it is assumed that there are no internal energy changes, no kinetic energy change,
zero heat generation and zero significant change in height. As such, the energy balance changes to
Δ
푝
휌
+ + Σ퐹 = − 푊푠
The actual shaft work done, therefore, is the total work done minus the frictional losses.
푊푠 = 푊푇 − Σ퐹
Definition of terms related to the study
1. Net positive suction Head (NPSH) – this is the difference between the static head at the suction
inlet and the head at the inlet at the vapour pressure.
2. Cavitation- this is the formation of bubbles around the impeller blades at low pressure areas
which move and collapse at high pressure areas. This collapse causes micro-jets orientated
towards the blade at extremely high pressure. This impact causes severe erosion of the impeller
blades in the presence of this phenomena, noise and great vibration will be detected.
6. 3. Efficiency – this is generally the ratio of the work done by the pump against the electrical energy
supplied by the pump.
Results and Tables:
Fundamental Data
Properties of water
Temperature 20°C
Density (ρ) 998.203 kg/m3
Specific weight (w) 9788.379 N/m3
Properties of centrifugal
pump
Revolution speed (N) 48.0 rev/s
Difference of the elevation of gauges (HG) 0.290 m
Properties of V-notch
Half angle of V-notch (θ) 45°
Coefficient of discharge (CdV) 0.576
Coefficient (KV) 1.360
Crest level (hook gauge) 0.224 m
Efficiency of motor (ηmo) 0.8
Operation Data
Stag
e
V-notch Electric board Pressure gauges
Gross
Head
(H)
m
Actual
power
(P)
× 103watts
Effici
ency
(ηo)
Specific
speed
(Ns)
Readi
ng
m
Head
(HV)
m
Discha
rge
(Q)
× 10−3m3
/s
Volta
ge
(V)
V
Curre
nt
(A)
A
Power
factor
(Pt)
(cos φ)
Input
powe
r
(PS)
× 103watt
Reading Pressure head
Head
differe
nce
(
p2 − p1
w
)
m
Gauge 1
(p'1)
cmHg
kg/cm2
Gauge
2
(p'2)
kg/cm2
Gauge 1
(p1/w)
m
Gauge 2
(p2/w)m
1 0.163 0.061 1.250 380 4.0 0.75 1.580 0.30 2.20 3.005 22.040 19.034 19.324 0.236
0.149
7
349.34
2 0.156 0.068 1.640 380 4.0 0.78 1.643 0.28 2.17 2.805 21.739 18.934 19.224 0.309
0.187
8
401.71
3 0.142 0.082 2.619 385 4.2 0.80 1.792 0.25 2.15 2.505 21.539 19.034 19.324 0.495
0.276
3
505.65
7. 4 0.137 0.087 3.036 385 4.3 0.81 1.858 0.24 2.12 2.404 21.238 18.834 19.124 0.568
0.305
9
548.76
5 0.130 0.094 3.684 385 4.4 0.82 1.925 0.21 2.10 2.104 21.038 18.934 19.224 0.693
0.360
2
602.13
6 0.123 0.101 4.409 385 4.6 0.83 2.037 0.17 2.09 1.703 20.938 19.235 19.525 0.843
0.413
7
651.07
7 0.116 0.108 5.213 385 4.8 0.84 2.151 0.12 2.04 1.202 20.437 19.235 19.525 0.996
0.463
2
707.95
8 0.109 0.115 6.099 380 4.9 0.85 2.193 0.02 2.00 0.200 20.036 19.836 20.126 1.202
0.547
9
748.55
9 0.098 0.126 7.664 380 5.3 0.85 2.372 0.00 1.85 0.000 18.533 18.533 18.823 1.412
0.595
3
882.27
10 0.065 0.159 13.710 385 6.5 0.86 2.982 0.00 1.11 0.000 11.120 11.120 11.410 1.531
0.513
5
1717.69
Calculations:
Discharge
5
2
Q = KVHV
KV =
8
15
CdV√2g tan θ
Where HV = head above V-notch,
CdV = coefficient of discharge of V-notch,
θ = half angle of V-notch,
KV = coefficient of V-notch.
Sample calculation
5
2
Q = KVHV
Q = 1.360 ∗ 0.061
5
2
Q=1.250
Input power
The motor in the hydraulics laboratory is a three-phase motor. The power supplied to the
shaft of the pump (Ps) is known as follows:
PS = √3AVPtηmo … . (13.3)
8. Where A = current (Ampere),
V = voltage (volt),
Pt = power factor(= cos φ),
ηmo = efficiency of motor.
Sample calculation
PS = √3 ∗ 4.0 ∗ 380 ∗ 0.75 ∗ 0.8
PS =1.580
Gross head
H =
p2
ρg
−
p1
ρg
+ HG
Sample calculation
H = 22.040 − 3.005 + 0.290
H = ퟏퟗ. ퟑퟐퟒ퐦
Overall efficiency
ηO =
ρQgH
PS
× 100(%)
Sample calculation
ηO =
998.203 ∗ 1.250 ∗ 9.81 ∗ 19.324
1.580
× 100(%)
ηO = ퟏퟒ. ퟗퟏ%
1) Specific speed
The specific speed of a pump is defined as
NS =
NQ
1
2
H
3
4
NS = ퟑퟒퟗ. ퟑퟒ
9. Discussions
From the results herein obtained and recorded, a graph relaying the characteristic curves of the
pump under study was developed. The curve indicated the peak capabilities of the pump in terms
of head and efficiency thereby indicating the performance properties.
Graph obtained from the results is as below.
150
Perfomance Curves of Pump
y = -0.1132x2 + 0.9751x + 21.842
y = -0.7885x2 + 15.476x - 1.3098
130
110
90
70
50
30
10
25
20
15
Head (m)
10
5
0
0 2 4 6 8 10 12 14
Efficiency (%)
Discharge (m3/s)
Head Input Power Efficiency Poly. (Head) Poly. (Efficiency)
From this graph, it is correct to indicate that the maximum efficiency of the pump is about76%.
The pump is also capable of producing a maximum head of about 24metres
During the design of the pump basis is placed on the power so that the power so that the
functionality of the pump produces the highest power output, which is usually not too far from
the range of acceptable efficiencies. This is quite evident in the graph obtained.
10. Similarly, specific speed for this pump was determined so as to estimate its peak values too. The
graph obtained was as below.
graph of 휂O vs. Ns
y = -401.91x2 + 456.35x - 48.108
90
80
70
60
Efficiency %
50
40
30
20
10
0
0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80
specific speed Ns
The specific speed obtained for the highest efficiency value is quite low for this size of pump.
This is an indication of the presence of some error in the values obtained.
The peak specific speed obtained from this graph is 0.56
Conclusion
The experiment was generally a success since characteristic curves of the pump were obtained.
The graph obtained indicates the peak capabilities of the pump.
However, the values obtained were not specifically accurate and. a such, the graphical
representation too.
Errors must have occurred in the system as a result of leakages in the system, friction and
erroneous and inaccurate recording and reading of values during the experiment. Possibly, errors
could also have been as a result of incorrect calculations and compilation of otherwise correctly
recorded data.
As such, a lot of care was taken during the collection of data and during the compilation of the
results so as to reduce the margin of error expected at the final issue of the report.
It is further recommended that the experiment be repeated if more accurate data is required for
the study of the performance characteristics of the centrifugal pump. This can be performed
under controlled conditions that ensure little margin of error. Such a condition would include
reducing the size of groups involved in the exercise.
11. References
1. Kumar, S., Gandhi, B. K., & Mohapatra, S. K. (2014). Performance Characteristics of
Centrifugal Slurry Pump with Multi-Sized Particulate Bottom and Fly Ash Mixtures.
Particulate Science and Technology, (just-accepted).
2. Marrero, T. R. Project-based Learning: Centrifugal Pump Operations.
3. Kumar, S., Gandhi, B. K., & Mohapatra, S. K. (2014). Performance Characteristics of
Centrifugal Slurry Pump with Multi-Sized Particulate Bottom and Fly Ash Mixtures.
Particulate Science and Technology, (just-accepted).