See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/314143790

Effects of Blade Discharge Angle, Blade Number and Splitter Blade Length on
Deep Well Pump Performance

Article  in  Journal of Applied Fluid Mechanics · March 2017

DOI: 10.18869/acadpub.jafm.73.239.26056

CITATIONS READS

10 894

3 authors, including:

Ergun Korkmaz
Isparta University of Applied Sciences
9 PUBLICATIONS   11 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Mechanical Representation of Damage in Continuum Damage Mechanics View project

All content following this page was uploaded by Ergun Korkmaz on 24 May 2017.

The user has requested enhancement of the downloaded file.

Journal of Applied Fluid Mechanics, Vol. 10, No. 2, pp. 529-540, 2017.
Available online at www.jafmonline.net, ISSN 1735-3572, EISSN 1735-3645.
DOI: 10.18869/acadpub.jafm.73.238.26056

Effects of Blade Discharge Angle, Blade Number and

Splitter Blade Length on Deep Well Pump
Performance
E. Korkmaz1†, M. Gölcü2 and C. Kurbanoğlu3
1 Suleyman Demirel University, Energy Systems Engineering Department, 32260 Cunur, Isparta, Turkey.
2 Pamukkale University, Automotive Engineering Department, 20017 Kinikli, Denizli, Turkey.
3 Istanbul Medeniyet University, Mechanical Engineering Department, 34730, Goztepe, Istanbul, Turkey.

†Corresponding Author Email: ergunkorkmaz@sdu.edu.tr

(Received December 29, 2015; accepted November 4, 2016)

ĄBSTRACT

Impellers with splitter blades are used for pumps and compressors in the design of turbomachines. Design
parameters such as the number of blades, blade discharge angle and impeller discharge diameter impact affect
pump performance and energy consumption. In this study, the effect of the number of blades (z=5, 6, and 7),
blade discharge angles (β2b=25, and β2b=35) and splitter blade lengths (40, 55, 70, and 85% of the main
blade length) on Deep Well Pump (DWP) performance has been studied experimentally. In the experiments,
pump casing, blade inlet angle, blade thickness, blade width and impeller inlet and discharge diameters have
been kept fixed while other parameters such as the number of blades, blade discharge angles and splitter blade
lengths have been allowed to vary. As a result of the experimental study, the highest efficiency of all the
impellers for best efficiency point (b.e.p) has been obtained on the impeller with the number of blades z=6,
blade discharge angle 2b=25 and 85% splitter blade addition compared to impellers without splitter blades.

Keywords: Deep well pump; Blade discharge angle; Pump performance; Impeller; Splitter blade.

NOMENCLATURE

b.e.p best efficiency point Pe brake horsepower

b1 blade inlet width Q flow rate
b2 blade outlet width Wosb without splitter blade
D1 impeller inlet diameter z number of blades
D2 impeller outlet diameter
e blade thickness 1b blade inlet angle
Hm head
2b blade discharge angle
L length of main blade
Ls length of splitter blade g general efficiency
L non-dimensional splitter blade length

1. INTRODUCTION water distribution, industrial water networks,

thermal water networks, agricultural irrigation
Most of the electricity consumed in industrialized systems and water supply grids. Especially in
countries is used in electric motor applications, countries whose economies largely depend on
which are typically pump, fan and compressor agriculture, underground waters have a broad field
drives (De Almeida et al. 2003; Ahonen et al. of application in agricultural irrigation. Considering
2010). Pumps of the centrifugal type are broadly technologies used in transporting underground
used in the transfer of liquids due to the simplicity waters to the surface, it can be seen that DWPs with
of their structure, low cost and low space demands. low specific speed are used. One of the largest
Deep well pumps (DWPs) are one of the most disadvantages of such pumps is low efficiency. In
widespread types of such pumps and are currently the studies that have been conducted for energy
used in a number of fields such as underground saving, it has been seen that one of the areas of high
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

potential energy saving is pumping systems (Kaya have a great effect on the efficiency and a
et al. 2008). suggestion was made to study the blade discharge
angle in a more detailed manner in order to increase
A variety of principles have been developed to the fan efficiency.
solve the hydraulic problems of pumps with low
specific speed. One of those is the splitter blade Kaya (2003) experimentally investigated regaining
principle. In order to increase efficiency, the the tangential velocity energy of axial flow pump
diameter of the impeller of a pump with low by using two different axial flow pump impellers
specific speed needs to be decreased. Moreover, in with and without splitter blades. As a result of the
addition to decreasing the diameter, a larger blade investigation, it is found that using splitter blades
discharge angle (β2b) and a greater number of blades and regaining the tangential velocity energy of the
need to be used. This will ensure that the desired pump improves the total pump efficiency about 3%.
head is achieved. Thus, losses occurring as a result
of the increase of the blade discharge angle can be Miyamoto et al. (1992) have studied the effect of
decreased by placing splitter blades between two splitter blades on the flow and performance by
main blades. The reason for using splitter blades measuring speed and pressure in closed and half-
instead of main blades is to decrease the clogging at open impeller channels. Flows and characteristics in
the impeller inlet caused by a large number of the channel of impellers with splitter blades have
blades. The use of splitter blades plays an important been compared with the flows and characteristics in
role in obtaining optimum design and identifying the channel of impellers without splitter blades. It
the best area among the design points (Korkmaz, has been claimed that in impellers with splitter
2008). blades, the blade loads tended to decrease and
emphasized that absolute peripheral speeds and
A variety of studies have been previously conducted total pressures of such impellers was significantly
on the use of splitter blades to solve the hydraulic higher than that of impellers without splitter blades.
problems of pumps with low specific speeds. In the
study where they considered the calculation of a When splitter blades are used in impellers, it is
three-dimensional flow in a centrifugal impeller considerably important for the pump efficiency to
with splitter blades, Kui and Jian (1988) have used determine the optimum length and position of the
a method they named “flow-surface coordinates splitter blade. Splitter blades are generally
iteration” to calculate the three-dimensional flow in positioned at the geometric center between two
turbo machines. Empirical data obtained from main blades and have a suitable size. Yuan (1997)
centrifugal pump impellers with different splitter wrote that the radial length of a splitter blade was
blades have been provided to provide the validity of generally 2/3 of the length of the main blade or
the calculation method offered. It was indicated in 0.5~0.75 times the impeller outlet diameter while
the said study that the aerodynamic load on the the inlet radius of the splitter blade was
radial part of the blades could be higher in order to approximately 0.4~0.6 times the impeller outlet
increase impeller efficiency where the number of diameter.
blades is deceased and that splitter blades could be Zha and Yang (1986) used super-short blades that
used in the channel in order to solve this problem. had a very small radial length or whose inlet radius
Gui et al. (1989) closely examined the effects of the was 0.75~0.85 times the impeller outlet diameter in
splitter blades on the performance of a forward- the improvement of pump performance. The
curved centrifugal fan. The study consists of two optimum length and position of splitter blades in the
parts that involve experimental work and numerical improvement of pump efficiency should be
calculations. The experiments have been conducted supported experimentally.
on a special impeller with adjustable splitter blades Blade discharge angle has very important role in the
to observe performance under difference performance of the centrifugal pump (Li, 2012). It
circumstances. The results of the experiments is well-known that the Euler’s head is inversely
conducted have shown that the change in the proportional to flow rate Q with discharge angle of
circumferential position of the splitter blades blade β2  90 and the smaller the β2, the more rapid
considerably affects fan performance and that the decrease of Euler’s head with flow rate. Former
splitter blades with a proper length could increase experience indicates that with appropriate other
the total pressure coefficient. As a general result of geometrical parameters the head-flow rate curve is
the study, it has been explicitly determined that the basically stable when β2  30. In some cases β2 can
circumferential position of splitter blades affects the be as high as 40 without droop (Yuan, 1997).
performance of forward leaning fans. It has been
emphasized that when the splitter blades are In this study, the effect of the number of blades
positioned close to the pressure surface, the total (z=5, 6 and 7), blade discharge angles (β2b=25 and
pressure coefficient increased whereas when they β2b=35) and splitter blade lengths (40, 55, 70, and
were positioned close to the suction surface of the 85% of the main blade length) on DWP
main blade, the efficiency increased somewhat. It performance has been studied experimentally. In the
has been claimed that increasing the length of the experiments, pump casing, blade inlet angle, blade
splitter blade might increase the total pressure thickness, blade width and impeller inlet and
coefficient but that there was no certain rule relating discharge diameters have been kept fixed while
to the effect of the splitter blade length on the other parameters such as the number of blades,
efficiency curve. It has been further emphasized blade discharge angles and splitter blade lengths
that the discharge angle of the splitter blade did not have been allowed to vary.

530
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

Fig. 1. Deep Well Pump test rig (not scale).

Fig. 2. An example of the pump impellers designed (2b=25, z=5, impeller without splitter blades).

Table 1 Deep Well Pump design point values

Q 36 m /h3
1b 18 e 4 mm D1 72 mm

Hm 13 m 2b 25 and 35 dm 20 mm D1i 62 mm

n 2850 rpm b1 25 mm dh 30 mm D1d 82 mm

z 6 b2 14 mm D0 78 mm D2 132 mm

2. TEST RIG literature (Stepanoff, 1957; Dicmas, 1987; Lobanoff

and Ross, 1992; Tuzson, 2000; Karassik et al.
A DWP with low specific speed has been designed 2001). Design point values of the DWP have been
in this study. Geometric factors other than the provided in Table 1 and an example of the designed
number of blades (z=5, 6 and 7), blade discharge impellers is shown in Fig. 2.
angle (β2b=25 and β2b=35) and length of splitter
blades have been kept constant and splitter blades 3. DETERMINATION OF PUMP
were added to the geometric center of two main CHARACTERISTICS
blades at the size of 40, 55, 70 and 85% of the main
blade length. The purpose was to experimentally Unlike most volumetric pumps, rotodynamic pumps
study the effect of splitter blade length on the pump are capable of pumping fluids at flow rates that vary
performance for different numbers of blades and depending on project values and suction conditions.
blade discharge angles and to obtain pump Here, the most important factor affecting the flow
characteristics. A DWP test unit was assembled as rate is the manometric head of the pump.
shown in Fig. 1 for this purpose. Depending on the manometric head, a change in the
The DWP design values were used as Hm=13 m, flow rate will lead to changes in the brake
Q=36 m3/h, n=2850 rpm. Impellers for DWPs are horsepower and efficiency values. These curves that
generally designed using empirical equations in the indicate the change of manometric head, brake

531
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

horsepower and efficiency depending on a variety blades. To the number of blades z=5, 6 and 7 and an
of flow rate values is called characteristic curves. impeller with 24 splitter blades with a blade
Drawing the characteristic curves of a pump is discharge angle of 2b=25 and 2b=35, splitter
particularly important in the determination of the blades were added to the geometric center of two
pump’s b.e.p. EN ISO 9906 (1999) has been blades that had a size of 40, 55, 70 and 85% of the
referred to in determining the pump characteristics. main blade length. Due to instability at very low
flow rates, the pump characteristics have been
Manometers are generally used to measure pressure
considered in the range of 27 – 63 m3/h.
in classical pump test systems. However, pressure
transducers or pressure transmitters have recently 3.1 Characteristics of Impellers without
started to be broadly used due to measurement Splitter Blades
precision demands. High-precision pressure
transmitters with 24 V DC supply, 2-cable In order to determine the effects of the number of
connection and 4~20 mA output have been used in blades and blade discharge angle on pump
pressure measurements. The vacuum pressure performance, 6 impellers were tested with numbers
transmitter has a measurement range of -1~0 bar of blades z=5, 6 and 7 and blade discharge angles
while the positive pressure transmitter has a 2b=25 and 2b=35. Figure 3a, Fig. 3b and Fig. 3c
measurement range of 0~10 bar (Wika, 2015). A indicate Hm-f(Q), Pe-f(Q) and g-f(Q)
magnetic-type flow-meter has been used for flow characteristics, respectively.
rate measurements (Krohne, 2005) and a network
analyzer has been used for brake horsepower
measurements (Entes, 2006). The network voltage 16

has an important effect on the motor rotation speed 15

(and therefore on the characteristic values of the 14

pump). To avoid a possible negative effect of this 13

12
situation during the comparison of impeller
Hm (m)

11
characteristics, a three-phase servo voltage 10
B2b=25, z=5, wosb
B2b=25, z=6, wosb
regulator has been used in the test settings with a 9 B2b=25, z=7, wosb
precision of ±2% of network voltage. A 8
B2b=35, z=5, wosb

submersible motor with a speed of 2850 rpm has 7

B2b=35, z=6, wosb
B2b=35, z=7, wosb
been used in the experiments. A frequency inverter 6

has been used to obtain the motor rotation speed of 20 25 30 35 40 45

Q (m 3/h)
50 55 60 65 70

2850 rpm and a PLC unit has been used for its (a)
control. Accuracy of the used devices and 4800
uncertainty of the calculated results have been B2b=25, z=5, wosb
B2b=25, z=6, wosb
provided in Table 2. 4400
B2b=25, z=7, wosb
B2b=35, z=5, wosb
4000
B2b=35, z=6, wosb

Table 2 Accuracies of the measurements and the

P e (W)

B2b=35, z=7, wosb

3600

uncertainties in the calculated results

3200
Measurements Accuracy
2800
Pressure transmitter  0.025
2400
Magnetic flow-meter  0.015 20 25 30 35 40 45 50
Q (m 3/h)
55 60 65 70

Network analyzer  0.01 (b)

47
Three-phase servo voltage regulator ±%2
45
Calculated results Uncertainty
43

Efficiency ±%1.2 41
 g (%)

39 B2b=25, z=5, wosb

One of the most significant problems of low B2b=25, z=6, wosb

B2b=25, z=7, wosb
37
specific speed pumps is their low efficiency. In B2b=35, z=5, wosb
order to increase efficiency, splitter blades can be 35 B2b=35, z=6, wosb
B2b=35, z=7, wosb
positioned between two main blades with suitably 33

selected number of blades and blade discharge 20 25 30 35 40 45

Q (m 3/h)
50 55 60 65 70

angle. For this purpose, splitter blades were (c)

positioned on an impeller with no splitter blades at Fig. 3. a) Hm-f(Q) b) Pe-f(Q) c) ηg-f(Q)
the geometric center of two main blades with characteristics of impellers without splitter
various splitter blade lengths (at 40, 55, 70 and 85% blades for different blade discharge angles and
of the main blade length). This experimentation has numbers of blades.
made it possible to study the effects of splitter blade
length and changes in the blade discharge angle on
DWP performance. As seen on Fig. 3a, concerning b.e.p, the head
increases with an increase in the number of blades
A total of 30 impellers have been used in the and blade discharge angle. The results obtained
experiments, 6 of which were without splitter from the experiments are in conformity with the

532
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

results provided in the literature (Yuan, 1997; Yuan blades z=5 and blade discharge angle 2b=35,
et al. 1993; Gölcü, 2001; Gölcü, 2005; Gölcü, respectively. As seen on Figs. 5a and 5b, splitter
2006a,b; Gölcü and Pancar, 2006; Gölcü et al. blade addition has increased the head and power
2007; Li, 2012). Although an increase in the blade consumed by the pump. The highest head and
discharge angle leads to an increase in the head, it efficiency increases have been obtained on the
also increases the power consumed by the pump impeller with 85% splitter blade addition (Figs. 5a
(Fig. 3b). Moreover, considering the b.e.p, a and 5c).
decrease is observed in the general efficiency as the
blade discharge angle is increased (Fig. 3c).
Among the impellers without splitter blades, the
highest efficiency was obtained using the impeller
with the blade discharge angle 2b=25 and number
of blades z=7. For the same blade discharge angle,
as the number of blades increased, the head also
grew. Additionally, increase was also observed in
the power consumed by the pump and efficiency.
Increases of 7.51% in head, 4.77% in power and
2.56% in efficiency were obtained with an impeller
with 2b=25 and z=7 compared to an impeller with (a)
2b=25 and z=5.
If the number of blades is kept constant, as the
blade discharge angle is increased, a decrease is
observed in efficiency as head and power consumed
by the pump increase. An increase of the blade
discharge angle has increased head by 8.41% and
power consumed by the pump by 14.84% in an
impeller with 2b=35 and z=5 compared to an
impeller with 2b=25 and z=5. However, a 5.60%
decrease has been observed in efficiency with the
increase of the blade discharge angle.
(b)
3.2 Characteristics of Impellers with
Splitter Blades
In order to observe the effects of splitter blade
length on the pump performance, a total of 24
impellers have been tested with the numbers of
blades z=5, 6 and 7 and blade discharge angles
2b=25 and 2b=35. Effects of the splitter blade
length has been observed on DWP performance by
adding to the geometric center of main blades
splitter blades with a size of 40, 55, 70 and 85% of
the main blade length on impellers with different
blade numbers and blade discharge angles. On the (c)
graphs, L is used to indicate main blade length Fig. 4. a) Hm-f(Q) b) Pe-f(Q) c) ηg-f(Q)
whereas the splitter blade length is indicated by Ls. characteristics of impellers with and without
For instance, Ls/L=0.70 indicates that splitter blades splitter blades for z=5 and β2b=25.
have been added that are 70% of the main blade
length. Literature sources indicate that the optimum
efficiency can be obtained with the number of
3.2.1 Characteristics of impellers with splitter blades range z=5-8 (Schweiger and Gregori, 1987)
blades and a number of blades z=5 further stating that the critical number of blades for
Figures 4a, 4b and 4c indicate Hm-f(Q), Pe-f(Q) and splitter blade applications would be z=5 (Gölcü et
g-f(Q) characteristics of impellers with a number al. 2007). In the study conducted, the researchers
of blades z=5 and blade discharge angle 2b=25, emphasized that for the blade discharge angle
respectively. As seen from Figs. 4a and 4b, addition β2b=15, a blade number of 5 and over would not be
of splitter blades has led to an increase in head and suitable for the use of splitter blades and that for
power consumed by the pump. The greatest blade numbers z=5 and higher, the blade discharge
increase in the head and power has been obtained angle would need to be increased.
on the impeller with 70% splitter blade addition
while the greatest efficiency increase has been In a study by Gölcü et al. (2007), it has been
obtained on the impeller with 85% splitter blade detected that splitter blade addition for β2b=15 and
addition (Fig. 4c). z=5, as seen on Figs. 6a, 6b and 6c, would lead to
an increase in power and would decrease both head
Figures 5a, 5b and 5c show Hm-f(Q), Pe-f(Q) and and pump efficiency at points following the best
g-f(Q) characteristics of impellers with number of efficiency point. In our study, however, head and

533
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

efficiency increases have been obtained at all the impeller without splitter blades, the impeller
operation points with the addition of splitter blades with 85% splitter blade addition had a 5.45%
to an impeller with the number of blades z=5 after increase in head, 0.36% increase in brake
the blade discharge angle was increased (2b=25 horsepower and 5.06% in efficiency. Apparently,
and 2b=35) (Figs. 4a, 4c, 5a, 5c). while 5.45% increase in head and 5.06% increase in
efficiency have been observed on the impeller with
85% splitter blade addition, the increase in brake
16 horsepower is at a negligible level.
15
14

13
12 16 z=5 (a) without splitter blade
Hm (m)

Ls=0.25L
11 Ls=0.35L
14 Ls=0.50L
10 B2b=35, z=5, wosb Ls=0.60L
Ls=0.80L
9 B2b=35, z=5, Ls/L=0.40
12
B2b=35, z=5, Ls/L=0.55

Hm (m)
8
B2b=35, z=5, Ls/L=0.70
7 10
B2b=35, z=5, Ls/L=0.85
6
20 25 30 35 40 45 50 55 60 65 70 8
Q (m 3/h)
(a) 6
5000
4
4600 0 2.5 5 7.5 10 12.5 15 17.5
Q (l/s)
4200 1.9
z=5 (b)
Pe (W)

3800 1.8
B2b=35, z=5, wosb
3400
B2b=35, z=5, Ls/L=0.40 1.7
B2b=35, z=5, Ls/L=0.55
Pe (kW)

3000
B2b=35, z=5, Ls/L=0.70 1.6
B2b=35, z=5, Ls/L=0.85
2600 without splitter blade
Ls=0.25L
20 25 30 35 40 45 50 55 60 65 70 1.5 Ls=0.35L
Q (m 3/h)
Ls=0.50L
(b) 1.4
Ls=0.60L
Ls=0.80L
46

44 1.3
42 0 2.5 5 7.5 10 12.5 15 17.5
40
Q (l/s)

38 60
 g (%)

z=5 (c)
36
34 B2b=35, z=5, wosb
55
32 B2b=35, z=5, Ls/L=0.40

30 B2b=35, z=5, Ls/L=0.55

(%)

B2b=35, z=5, Ls/L=0.70

28
B2b=35, z=5, Ls/L=0.85 50
g

26
20 25 30 35 40 45 50 55 60 65 70
without splitter blade
Q (m 3/h)
Ls=0.25L
(c) 45 Ls=0.35L
Ls=0.50L
Fig. 5. a) Hm-f(Q) b) Pe-f(Q) c) ηg-f(Q) Ls=0.60L
Ls=0.80L
characteristics of impellers with and without 40
splitter blades for z=5 and β2b=35. 0 2.5 5 7.5
Q (l/s)
10 12.5 15 17.5

Fig. 6. (a) Hm-Q, (b) Pe-Q and (c) ηg -Q

3.2.2 Characteristics of Impellers with characteristics of the impeller (z=5, β2b=15)
Splitter Blades and a Number of Blades z=6 with different lengths of splitter blades
(Gölcü et al. 2007).
Figures 7a, 7b and 7c indicate Hm-f(Q), Pe-f(Q) and
g-f(Q) characteristics of impellers with a number
Figures 8a, 8b and 8c indicate Hm-f(Q), Pe-f(Q) and
of blades z=6 and blade discharge angle 2b=25,
g-f(Q) characteristics of impellers with a number
respectively. As seen from Fig. 7a, addition of
of blades z=6 and blade discharge angle 2b=35,
splitter blades has led to an increase in head of the
respectively. As seen from Figs. 8a and 8b, addition
pump. The greatest head increase has been obtained
of splitter blades has led to an increase in head and
on the impeller with 85% splitter blade addition.
power consumed by the pump. The greatest head
Among the impellers with splitter blades, the lowest
and efficiency increase has been obtained on the
power has been obtained on the impeller with 85%
impeller with 70% splitter blade addition (Figs. 8a
splitter blade addition. This is indicated on Fig. 7b.
and 8c). Compared to impellers without splitter
Among the impellers with splitter blades, the
blades, the impeller with 70% splitter blade addition
highest efficiency has been obtained on the impeller
yielded 10.62% head increase, 5.25% brake
with 85% splitter blade addition (Fig. 7c).
horsepower increase and 5.06% efficiency increase.
Compared to the impeller without splitter blades, Compared to impellers without splitter blades, the
the impeller with 70% splitter blade addition had a impeller with 85% splitter blade addition yielded
4.57% increase in head, 1.96% increase in brake 8.58% head increase, 7.65% brake horsepower
horsepower and 2.53% in efficiency. Compared to increase and 0.83% efficiency increase.

534
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

16 17
B2b=25, z=6, wosb B2b=35, z=6, wosb
15 16 B2b=35, z=6, Ls/L=0.40
B2b=25, z=6, Ls/L=0.40
14 B2b=25, z=6, Ls/L=0.55 15 B2b=35, z=6, Ls/L=0.55
13 B2b=25, z=6, Ls/L=0.70 B2b=35, z=6, Ls/L=0.70
14
B2b=25, z=6, Ls/L=0.85 B2b=35, z=6, Ls/L=0.85
12

Hm (m)
Hm (m) 13
11
12
10
11
9
8 10

7 9

6 8
20 25 30 35 40 45 50 55 60 65 70 20 25 30 35 40 45 50 55 60 65 70
Q (m 3/h) Q (m 3/h)

(a) (a)
3800 5000

3600
4600

3400
4200
3200

Pe (W)
P e (W)

3800
3000
B2b=25, z=6, wosb B2b=35, z=6, wosb
3400
B2b=25, z=6, Ls/L=0.40 B2b=35, z=6, Ls/L=0.40
2800
B2b=25, z=6, Ls/L=0.55 B2b=35, z=6, Ls/L=0.55
3000
2600 B2b=25, z=6, Ls/L=0.70 B2b=35, z=6, Ls/L=0.70
B2b=25, z=6, Ls/L=0.85 B2b=35, z=6, Ls/L=0.85
2600
2400
20 25 30 35 40 45 50 55 60 65 70
20 25 30 35 40 45 50 55 60 65 70 Q (m 3/h)
Q (m3/h)

(b) (b)
46
48

46 44

44 42

42 40
 g (%)
 g (%)

40
38
B2b=25, z=6, wosb B2b=35, z=6, wosb
38
B2b=25, z=6, Ls/L=0.40 36 B2b=35, z=6, Ls/L=0.40
36 B2b=25, z=6, Ls/L=0.55 B2b=35, z=6, Ls/L=0.55
B2b=25, z=6, Ls/L=0.70 34 B2b=35, z=6, Ls/L=0.70
34
B2b=25, z=6, Ls/L=0.85 B2b=35, z=6, Ls/L=0.85
32
32
20 25 30 35 40 45 50 55 60 65 70
20 25 30 35 40 45 50 55 60 65 70 Q (m 3/h)
Q (m 3/h)
(c) (c)
Fig. 7. a) Hm-f(Q) b) Pe-f(Q) c) ηg-f(Q) Fig. 8. a) Hm-f(Q) b) Pe-f(Q) c) ηg-f(Q)
characteristics of impellers with and without characteristics of impellers with and without
splitter blades for z=6 and β2b=25. splitter blades for z=6 and β2b=35.

3.2.3 Characteristics of Impellers with Considering the results at the best efficiency point,
Splitter Blades and a Number of Blades z=7 all additions of splitter blades decrease the
efficiency, as seen on Fig. 9c. For instance, while an
Figures 9a, 9b and 9c indicate Hm-f(Q), Pe-f(Q) and impeller with 70% splitter blade addition
g-f(Q) characteristics of impellers with a number demonstrates a head decrease of 2.60% and
of blades z=7 and blade discharge angle 2b=25, efficiency decrease of 4.32% compared to the
respectively. As seen from Fig. 9a, the greatest head impeller without splitter blades, a 1.80% increase in
increase has been obtained on the impeller with brake horsepower has been observed. On the other
85% splitter blade increase. hand, on an impeller with 85% splitter blade
addition, a 2.60% head increase and 7.32% brake
Among the impellers with splitter blades, the lowest horsepower increase have been obtained compared
power has been obtained on the impeller with 55% to the impeller without splitter blades, but its
splitter blade addition (Fig. 9b). However, the head efficiency decreased by 4.37%. The reason why
value of the impeller with 55% splitter blade splitter blade additions negatively affect efficiency
addition is smaller than that of the impeller without at z=7 and β2b=25 is the clogging and contraction
splitter blades (Fig. 9a). that occur on the inlet and outlet cross sections due
Considering efficiency values of impellers with the to the too high number of blades on the outlet (z=14
number of blades z=7 and blade discharge angle for the outlet cross section) and the too small blade
2b=25, an interesting situation has been observed. discharge angle (2b=25). As the blade discharge
While concerning b.e.p, the efficiency values of all angle increases, the outlet contraction coefficient
impellers with splitter blade additions are very close also grows and is inversely proportional to the
to each other, they are lower than the efficiency of number of blades. Figures 10a, 10b and 10c indicate
the value of the impeller without splitter blades (Fig. Hm-f(Q), Pe-f(Q) and g-f(Q) characteristics of
9c). This leads us to conclude that the addition of impellers with a number of blades z=7 and blade
splitter blades to impellers with the number of blades discharge angle 2b=35, respectively. As seen from
z=7 and blade discharge angle 2b=25 is unsuitable. Figs. 10a and 10b, the addition of splitter blades

535
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

leads to an increase in the head and power addition at z=7 and 2b=25 would negatively
consumption of the pump. The greatest head impact efficiency. However, such addition at z=7
increase has been obtained on the impeller with and 2b=35 has been observed to yield an
70% splitter blade increase (Fig. 10a) while the efficiency increase. Given this fact, it can be
greatest efficiency increase is yielded by the claimed that a 10 increase in the blade discharge
impeller with 85% splitter blade increase up to the angle in this situation eliminates the effect of
best efficiency point and by the impeller with 70% contraction. While efficiency decreased at =7 and
splitter blade increase after that point (Fig. 10c). 2b=25 for all splitter blade lengths, it has been
Considering the best efficiency point, while observed to decrease for all splitter blade lengths at
efficiencies of both impellers are close to each z=7 and 2b=35 up to b.e.p.
other, the efficiency value of the impeller with 85%
splitter blade addition is greater than that of the
impeller with 70% splitter blade addition. 17
B2b=35, z=7, wosb
16 B2b=35, z=7, Ls/L=0.40
15 B2b=35, z=7, Ls/L=0.55
16 B2b=35, z=7, Ls/L=0.70
B2b=25, z=7, wosb 14
B2b=35, z=7, Ls/L=0.85
B2b=25, z=7, Ls/L=0.40

Hm (m)
14 13
B2b=25, z=7, Ls/L=0.55
12
B2b=25, z=7, Ls/L=0.70
12
B2b=25, z=7, Ls/L=0.85 11
Hm (m)

10 10

9
8
8
20 25 30 35 40 45 50 55 60 65 70
6 Q (m3/h)

(a)
4 5000
20 25 30 35 40 45 50 55 60 65 70
Q (m 3/h)
(a) 4600

3800 4200
Pe (W)

3600
3800
3400 B2b=35, z=7,wosb
3400 B2b=35, z=7, Ls/L=0.40
3200 B2b=35, z=7, Ls/L=0.55
P e (W)

3000 B2b=35, z=7, Ls/L=0.70

3000
B2b=25, z=7, wosb B2b=35, z=7, Ls/L=0.85
B2b=25, z=7, Ls/L=0.40 2600
2800
B2b=25, z=7, Ls/L=0.55 20 25 30 35 40 45 50 55 60 65 70
Q (m3/h)
2600 B2b=25, z=7, Ls/L=0.70
(b)
B2b=25, z=7, Ls/L=0.85
2400 46
20 25 30 35 40 45 50 55 60 65 70
Q (m 3/h) 44
(b)
42
46

44 40
 g (%)

42 B2b=35, z=7,wosb
38
40 B2b=35, z=7, Ls/L=0.40
38 36 B2b=35, z=7, Ls/L=0.55
 g (%)

B2b=35, z=7, Ls/L=0.70

36
34 B2b=35, z=7, Ls/L=0.85
34 B2b=25, z=7, wosb
B2b=25, z=7, Ls/L=0.40 32
32
B2b=25, z=7, Ls/L=0.55 20 25 30 35 40 45 50 55 60 65 70
30 Q (m 3/h)
B2b=25, z=7, Ls/L=0.70
28 B2b=25, z=7, Ls/L=0.85 (c)
26 Fig. 10. a) Hm-f(Q) b) Pe-f(Q) c) ηg-f(Q)
20 25 30 35 40 45
Q (m 3/h)
50 55 60 65 70
characteristics of impellers with and without
(c) splitter blades for z=7 and β2b=35.
Fig. 9. a) Hm-f(Q) b) Pe-f(Q) c) ηg-f(Q)
characteristics of impellers with and without
splitter blades for z=7 and β2b=25. Head (Hm), break horsepower (Pe) and general
efficiency (g) values depending on the number of
blades (z), blade discharge angle (2b) and
Apparently, compared to the impeller without
splitter blades, the impeller with 70% splitter blade dimensionless splitter blade length ( L  L s ) have
increase yielded a 5.32% head increase, 3.33%
L
Impellers with and without splitter blades used in
brake horsepower increase and 1.92% efficiency
the experiments have been assigned model numbers
increase. Compared to the impeller without splitter
based on their number of blades, blade discharge
blades, the impeller with 85% splitter blade increase
yielded a 4.58% head increase, 2.12% brake angle and splitter blade lengths and Hm, Pe, and g
horsepower increase and 2.39% efficiency increase values at b.e.p of the impellers according to their
model numbers are presented in Figs. 11a, 11b, and
We had already concluded that splitter blade 11c.

536
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

(a)

(b)

(c)
Fig. 11. a) Hm, b) Pe and c) ηg values of impellers with and without splitter blades at b. e. p.

As seen from Fig. 11a, the addition of splitter of blades z=7 and blade discharge angle 2b=25
blades has ensures a head increase except for some has been obtained without splitter blades.
of the impellers with splitter blades with the number
of blades z=7 and blade discharge angle 2b =25 4. CONCLUSIONS
(except for model numbers 23 and 24). Figure 11c
shows that the splitter blade addition to impellers
with number of blades z=7 and blade discharge As a results of the experimental study; an increase
in head, brake horsepower and general efficiency
angle 2b =25 negatively affects the efficiency.
with impellers without splitter blades is observed
As seen from Fig. 11b, splitter blade addition has with increasing the blade number as similar with
generally caused an increase in the power consumed results of previous studies (Gölcü and Pancar, 2005;
by the pump compared to impellers without splitter Gölcü et al. 2006). Moreover, it has been concluded
blades. that the blade discharge angle has greater effect on
pump characteristics compared to the number of
As seen on Table 3 and Fig. 11c, the highest blades. As the blade discharge angle increased, head
efficiency of all the impellers for b.e.p has been and brake horsepower have been observed to
obtained on the impeller with the number of blades increase while a decrease has been observed in the
z=6, blade discharge angle 2b=25 and 85% splitter efficiency until b.e.p. Although head value
blade addition compared to impellers without increased with higher blade discharge angle, general
splitter blades. Since the increase in the brake efficiency of the pump decreased due to a much
horsepower is negligible, the 85% splitter blade greater increase in the brake horsepower. Due to
addition is suitable economically. On the impeller the increase of the absolute velocity with the
with the number of blades z=7 and blade discharge increase of the blade discharge angle, losses grew
angle 2b=25, the addition of splitter blades has and pump efficiency decreased. At the same time,
negatively impacted the efficiency, yielding an increase of the blade discharge angle led to an
efficiency value lower than that of the impeller increase in separation losses. Considering the
without splitter blades for all splitter blade lengths. results at the best efficiency point;
The best efficiency on the impeller with the number

537
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

Table 3 Characteristic values of impellers with and without splitter blades at b.e.p
Model Factors Test results (b.e.p.)
Number
z β2b (º) L Hm (m) Pe (W) ηg (%)
s
1 5 25 wosb 11.05 3101 43.71
2 5 25 0.40 11.31 3143 44.12
3 5 25 0.55 11.57 3186 44.52
4 5 25 0.70 11.98 3327 44.16
5 5 25 0.85 11.88 3245 44.89
6 5 35 wosb 11.98 3561 41.26
7 5 35 0.40 12.91 3764 42.07
8 5 35 0.55 12.86 3791 41.60
9 5 35 0.70 13.69 3935 42.65
10 5 35 0.85 13.84 3865 43.91
11 6 25 wosb 11.36 3179 43.83
12 6 25 0.40 11.52 3260 43.31
13 6 25 0.55 11.67 3260 43.90
14 6 25 0.70 11.88 3241 44.94
15 6 25 0.85 11.98 3190 46.05
16 6 35 wosb 12.70 3721 41.87
17 6 35 0.40 13.07 3888 41.21
18 6 35 0.55 13.53 3857 43.02
19 6 35 0.70 14.05 3916 43.99
20 6 35 0.85 13.79 4005 42.22
21 7 25 wosb 11.88 3249 44.83
22 7 25 0.40 12.13 3471 42.88
23 7 25 0.55 11.26 3221 42.86
24 7 25 0.70 11.57 3307 42.89
25 7 25 0.85 12.19 3487 42.87
26 7 35 wosb 13.53 3857 43.02
27 7 35 0.40 13.84 3923 43.26
28 7 35 0.55 14.05 3939 43.73
29 7 35 0.70 14.25 3986 43.85
30 7 35 0.85 14.15 3939 44.05

On impellers without splitter blades blade discharge angle 2b=25, an 85% splitter
blade addition yielded 7.51% head increase, 4.64%
While the highest head and brake horsepower are
power increase and 2.69% efficiency increase
obtained with the blade discharge angle 2b=35 compared to impellers without splitter blades. On
and number of blades z=7, the highest efficiency impellers with the number of blades z=5 and blade
has been obtained with the blade discharge angle
discharge angle 2b=35, an 85% splitter blade
2b=25 and number of blades z=7. Shojaeefard et addition yielded 15.52% head increase, 8.53%
al. (2012) have been investigated both numerically power increase and 6.42% efficiency increase
and experimentally the effects of blade outlet angle compared to impellers without splitter blades. On
and passage width on the performance of a impellers with the number of blades z=6 and blade
centrifugal pump. In their study, the optimum
discharge angle 2b=25, an 85% splitter blade
centrifugal pump performance has been obtained
addition yielded 5.45% head increase, 0.34% power
with the blade discharge angle of 30 for 21 mm increase and approx. 5% efficiency increase
blade exit width. Babayigit et al. (2015) conducted compared to impellers without splitter blades. Since
a numerical study to analyze the effect of blade exit the increase in the brake horsepower is negligible,
angle on the performance of multistage centrifugal the 85% splitter blade addition is suitable
pump impeller for the blade exit angle values of economically. Unlike on other impellers, on
18°, 20°, 25°, 30° and 35°. In their study, the most impellers with the number of blades z=6 and blade
convenient blade exit angle of 18° was obtained.
discharge angle 2b=35, the highest efficiency has
On impellers with splitter blades been obtained with the 70% splitter blade addition.
Such addition yielded 10.63% head increase, 5.24%
Splitter blade addition has been observed to yield an power increase and approx. 5% efficiency increase
increase in general efficiency and head on impellers compared to impellers without splitter blades. On
with all numbers of blades and blade discharge impellers with the number of blades z=7 and blade
angles other than impellers with the number of discharge angle 2b=25, none of the splitter blade
blades z=7 and blade discharge angle 2b=25. addition yielded an increase in efficiency. On the
On impellers with the number of blades z=5 and contrary, the efficiency decreased by an average of

538
E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

4.37% for all splitter blade sizes. Therefore, splitter Pamukkale University, Denizli, Turkey.
blade addition is not suitable for impellers with the
number of blades z=7 and blade discharge angle Gölcü, M. (2006a). Neural network analysis of
head-flow curves in deep well pumps. Energy
2b=25. This is caused in splitter blade applications
Convers Manage 47(7-8), 992-1003.
by the high number of blades (z=7) combined with
the small blade discharge number (2b=25). A Gölcü, M. (2006b). Artificial neural network based
decrease is also observed in the outlet contraction modeling of performance characteristics of
coefficient compared to 2b=35. On impellers with deep well pumps with splitter blade. Energy
the number of blades z=7 and blade discharge angle Convers Manage 47(18-19), 3333-43.
2b=35, an 85% splitter blade addition yielded
Gölcü, M. and Y. Pancar (2005). Investigation of
4.58% head increase, 2.12% power increase and
performance characteristics in a pump impeller
2.39% efficiency increase compared to impellers
with low blade discharge angle. World pumps
without splitter blades.
468, 32-40.
Considering the results that have been obtained, it
Gölcü, M., N. Usta and Y. Pancar (2007). Effects of
has been concluded that for splitter blade addition,
splitter blades on deep well pump
the optimal number of blades would be z=6,
performance. Journal of Energy Resources
optimal blade discharge angle 2b=25 and optimal Technology 129, 169-76.
splitter blade length in the range of Ls=0.70L ~
Ls=0.85L. Finally, it has been concluded that low Gölcü, M., Y. Pancar and Y. Sekmen (2006).
efficiency, which is one of the most significant Energy saving in a deep well pump with
problems of pumps with low specific speed, could splitter blade. Energy Convers Manage 47,
be solved by properly selecting the number of 638-51.
blades, blade discharge angle and adding splitter
Gui, L., C. Gu and H. Chang (1989). Influences of
blades.
splitter blades on the centrifugal fan
performances. In Proceeding of ASME
ACKNOWLEDGEMENTS International Gas Turbine and Aeroengine
Congress and Exposition.
We thank the Administrative Unit of Scientific
Research Projects at Süleyman Demirel University Karassik, I. J., J. P. Messina, P. Cooper and C. C.
(SDUBAP) and the Scientific and Technological Heald, (2001). Pump Handbook. 3rd ed.,
Research Council of Turkey (TÜBİTAK) for their McGraw-Hill, New York.
contributions in this study. Kaya, D. (2003). An experimental study on
regaining the tangential velocity energy of
REFERENCES axial flow pump. Energy Convers Manage 44,
1817-29.
Ahonen, T., J. Tamminen, J. Ahola, J. Viholainen, Kaya, D., E. A. Yagmur, K S. Yigit, F. C. Kilic, A.
N. Aranto, and J. Kestilä (2010). Estimation of S. Eren and C. Celik (2008). Energy efficiency
pump operational state with model-based in pumps. Energy Convers Manage 49, 1662-
methods. Energy Convers Manage 51, 1319- 73.
25.
Korkmaz, E. (2008). Analysis of the Effect of
Babayigit, O., O. Kocaaslan, M. H. Aksoy, K. M. Splitter Blade Length and Circumferential
Guleren and M. Ozgoren (2015). Numerical Position at Different Blade Discharge Angles
identification of blade exit angle effect on the on the Deep Well Pump Performance. Ph. D.
performance for a multistage centrifugal pump thesis (in Turkish), Süleyman Demirel
impeller. EPJ Web of Conferences. University, Isparta, Turkey.
De Almeida, A. T., P. Fonseca, H. Falkner and P. Krohne, (2005). Krohne Optiflux 1000/5000
Bertoldi (2003). Market transformation of Electromagnetic Flow Sensor, Sandwich
energy-efficient motor technologies in the EU. Versions, for Volumetric Flow Rate
Energy Policy 31, 563-75. Measurement Electrically Conductive Liquids.
Dicmas, J. L. (1987). Vertical Turbine, Mixed Flow, Quick Start, Germany.
and Propeller Pumps. McGraw-Hill, New Kui, L. D. and J. L. Jian (1988). Calculation of
York. complete three-dimensional flow in a
EN ISO 9906/AC. (1999). Rotodynamic Pumps - centrifugal rotor with splitter blades. In
Hydraulic Performance Acceptance Tests - Proceeding of ASME International Gas
Grades 1 and 2. Turbine and Aeroengine Congress and
Exposition, Amsterdam.
Entes, (2006). Entes MPR-60S digital network
analyzers user guide (in Turkish), 37, Istanbul, Li, W. G. (2011). Blade exit angle effects on
Turkey. performance of a standard industrial
centrifugal oil pump. Journal of Applied Fluid
Gölcü, M. (2001). Analysis of Effects of Adding Mechanics 4(2), 105-119.
Splitter Blade to Impeller on Efficiency in
Deep Well Pumps. Ph. D. thesis (in Turkish), Lobanoff, V. S. and R. R. Ross (1992). Centrifugal

539
 E. Korkmaz et al. / JAFM, Vol. 10, No. 2, pp. 529-540, 2017.

Pumps: Design and Application. 2nd ed, Gulf Stepanoff, A. J. (1957). Centrifugal and Axial Flow
Publishing Company, Houston. Pumps: Theory, Design, and Application. John
Wiley and Sons Inc., New York.
Miyamoto, H., Y. Nakashima and H. Ohba (1992).
Effects of splitter blades on the flows and Tuzson, J. (2000). Centrifugal Pump Design. John
characteristics in centrifugal impellers. JSME Wiley and Sons Inc., New York.
International Journal Series II: Fluids
Engineering, Heat Transfer, Power, Wika (2015).Wika instrument corporation web site.
Combustion, Thermophysical Properties 35(2), http://www.wika.us/upload/OI_S_10_S_11_en
238-46. _us_18255.pdf, Access date: 24.11.2015.

Schweiger, F. and J. Gregori (1987). Design effects Yuan, S. (1997). Advances in hydraulic design of
on performance characteristics of centrifugal centrifugal pumps. In Proceeding of ASME,
pumps. In Proceeding of ASME, Applied Fluid engineering division. Summer meeting,
Mechanics, Biomechanics, and Fluids Vancouver, British Col., Canada.
Engineering Conference, Cincinnati, Ohio. Yuan, S., C. Chen, W. Cao, S. Li and S. Jin (1993).
Shojaeefard, M. H., M. Tahani, M. B. Ehghaghi, M. Design method of obtaining stable head-flow
A. Fallahian, M. Beglari (2012). Numerical curves of centrifugal pumps. ASME, Pumping
study of the effects of some geometric machinery, FED 154, 171-75.
characteristics of a centrifugal pump impeller Zha, S. and M. G. Yang (1986). Research on the
that pumps a viscous fluid. Computers and improvement of efficiency of low specific
Fluids 60, 61–70. speed centrifugal pump. J. Jiangsu Inst. of
Tech 7(4), 1-12.

540

View publication stats