1

1
Centrifugal Pump Experiment
A. Amirfazli
Feb. 2009
Fluidedesign.com
Casing (aka volute)
Impeller
Eye of impeller
(where liquid
enters the pump)
2
Pump History
The ancient Egyptians used water wheels with buckets
mounted on them to move water for irrigation.
The true centrifugal pumps were developed in the late 1600`s
by DENIS PAPIN, a French -born inventor accredited with a
pumping device having straight vanes.
The British inventor-J OHN G. APPOLD, was responsible for
first introducing the CURVED VANE IN 1851.
2
3
Centrifugal Pumps
Centrifugal Pumps belong to wider group of
fluid machines called turbo machines.
Pumps add energy to a fluid stream. As
such they have a wide application:
Water distribution systems
Buildings; municipality; irrigation (farming).
Industrial processes
Oil pipelines; power plants; refineries
Machineries
Lubrication; fuel
4
OBJ ECTIVES
1. To measure the performance of a centrifugal pump
and compare the results with theoretical predictions.
2. To compare them with the manufacturers
specification.
3. To investigate affinity laws for pumps (MecE 330).
4. To study the performance characteristics of pumps
in parallel and series system configurations.
3
5
Centrifugal Pump
Fluidedesign.com
Casing (aka volute)
Impeller
Eye of impeller
(where liquid
enters the pump)
6
4
7
pump
inlet
outlet
motor
8
Experimental Setup
pumps
tank
Measurement tank (flow rate)
Return line Pressure transducer
5
9
Piping arrangement for conducting series or parallel
pump experiments
10
Experimental Measurements
The head produced by
pumps is measured
(pressure transducer)
The flow rate of the
pumps is measured
(gravitometry method
similar to Boiler Exp.)
6
11
Torque is measured by determining the force required to
hold the motor casing steady
The RPM (rotational speed) of pump is selected/set
The geometrical information for pump impeller (vane
angles, radius, etc.) is known/given
12
Head-Discharge Curve
0
1
2
3
4
5
6
7
8
9
10
0 20 40 60 80 100
Flow rate, Q (x10
-4
m
3
/s)
H
e
a
d
,

H

(
m
)
1500 RPM
1100 RPM
750 RPM
Shut-off head
7
13
Velocity Triangle and Adding Energy to the Fluid
U = .r
2
b
V
1
V
2
V
2
r
2
r
1
V
2
V
2r
V
2rel
Side view Top view
V
2

2
14
RED: high pressure BLUE: low pressure
impeller eye impeller tip
www.fortunecity.com
8
15
Euler Equation
. .
Angular momentum (T): T =mr
2
V
2
From velocity triangle we have: V
2
=U V
2r
cot
2
. .
therefore: T =m.r
2
[U V
2r
cot
2
]
.
From energy eqn.: T = mgH and U = r
2

Then: Euler Eqn.

=
2
2
2
cot 1
U
V
g
U
H
r
16
Non-dimensional Euler Eqn.
Where Head Coefficient (=Hg/U
2
) and
Flow Coefficient (= V
2r
/U)
2
cot 1 =
ideal
9
17
Experimental calculation for and
(= Hg/U
2
) is easy to calculate since we
directly measure H and we know RPM and
radius of the impeller which gives us U (=.r
2
)
To calculate (=V
2r
/U) again as above we
have U, but need V
2r
(=Q / A
exit
)
Q is directly measured, but A
exit
needs to
be calculated!
18
Vane thickness (t) Vane thickness (t)
Exit area (A
exit
) is equal to area at the impeller outlet
minus vane thickness, i.e. A
exit
= (2r
2
5t).b
There are 5 vanes
in the impeller
r
2
Impeller
thickness at exit

2
Outer
edge

2
on average
is ~28
10
19
Dimensionless Head and Flow Coefficients
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
Fl ow Coeffi ci ent
H
e
a
d

C
o
e
f
f
i
c
i
e
n
t

1500 RPM
1100 RPM
750 RPM
Manufacturer
(1750 RPM)
Ideal (Eul er's)
Shut-off head actual (head at zero flow rate) is ~50% of
the shut-off head ideal (= U
2
/g), why?
20
Deviations from Theory
Frictional losses in the impeller and casing is not
considered in the theory
These frictional losses increase with Q
Angle at which the fluid leaves the impeller is not

2
(it is always less due to slip)
The flow in the impeller is not uniform(1-
dimensional) as assumed in theory.
Some leakage from outlet to inlet of the pump
due to clearance between impeller and casing
11
21
Affinity Laws
Manufacturers to offer a wide range of
pumps use geometrical similarity to
enlarge or shrink a particular design.
If the Reynolds numbers are large, the
head and flow coefficient for geometrically geometrically
similar similar pumps will remain constant and
therefore:
2
2 2
1 1
2
1

=
D
D
H
H
3
2
1
2
1
2
1

=
D
D
Q
Q
22
Dimensionless head versus flow coefficient curves for
geometrically similar machines
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
0.00 0.05 0.10 0.15
Flow coeffi cient
Unit 454
Unit 453
Unit 452
Unit 451
Head
coefficient
12
23
Cut Impeller
Manufacturer to save on development costs, but
to offer wide range of pumps with various head-
flow characteristics often cut impellers and place
them in the same casing.
Cut impeller pumps are NOT geometrically
similar; therefore affinity laws will not apply!
24 www.fortunecity.com
Typical Pump Curve
13
25
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.00 0.05 0.10 0.15 0.20
Fl ow coeffi ci ent
H
e
a
d

c
o
e
f
f
i
c
i
e
n
t
Unit 454
Unit 453
Unit 452
Unit 451
Dimensionless head versus flow coefficient
curves for geometrically dissimilar machines
(cut impeller)
26
Flow Efficiency
Flow efficiency () is a measure of how much
of the input power (=T) is converted to the
fluid power (=pgQH).

=
T
gQH

14
27
Efficiency versus flow coefficient curves
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
80.0
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
Fl ow coeffi ci ent
E
f
f
i
c
i
e
n
c
y

(
%
)
1500 RPM
1100 RPM
750 RPM
Manufacturer
(1750 RPM)
28
Pumps in Series
Q = Q
1
= Q
2
H = H
1
+ H
2
15
29
Pumps in Series
0
5
10
15
20
25
30
0 20 40 60 80 100
Flow rate (x10
-4
) m
3
/s
H
e
a
d

m
2 pumps in series
A single pump
30
Pumps in Parallel
H = H
1
= H
2
Q = Q
1
+ Q
2
16
31
Final Note
Check the lab manual for details of what to
report and answer any questions posed.