Computers & Fluids 60 (2012) 6170

Contents lists available at SciVerse ScienceDirect

Computers & Fluids

j o u r n a l h o m e p a g e : w w w . e l s e v i e r . c o m / l o c a t e / c o m p fl u i d

Numerical study of the effects of some geometric characteristics

of a centrifugal pump impeller that pumps a viscous uid
M.H. Shojaeefard a, M. Tahani a,, M.B. Ehghaghi b, M.A. Fallahian a, M. Beglari a
a
b

School of Mechanical Engineering, Iran University of Science and Technology, Tehran, Iran
Department of Mechanical Engineering, Tabriz University, Tabriz, Iran

a r t i c l e

i n f o

Article history:
Received 17 July 2011
Received in revised form 30 November 2011
Accepted 27 February 2012
Available online 7 March 2012
Keywords:
CFD
Turbulence modeling
Hydraulic oil
Centrifugal pump
Geometry

a b s t r a c t
The performance of centrifugal pumps drops sharply during the pumping of viscous uids. Changing
some geometric characteristics of the impeller in these types of pumps improve their performance. In this
investigation, the 3-D ow in centrifugal pump along with the volute has been numerically simulated.
This numerical solution has been carried out for different cases of primary geometry, and for the changes
made to the outlet angle and passage width of the impeller, and also for simultaneous modications of
these parameters. The nite volume method has been used for the discretization of the governing equations, and the High Resolution algorithm has been employed to solve the equations. Also, the k  x SST
has been adopted as the turbulence model in the simulation. In the steady state, this simulation is dened
by means of the multi-reference frame technique, in which the impeller is situated in the rotating reference frame, and the volute is in the xed reference frame, and they are related to each other through the
Frozen Rotor. The obtained numerical results are compared with the experimental ones, and the outcome shows acceptable agreement between the two. The ow analysis indicates that with the modication of the original geometry of the pump, at the 30 outlet angle and the passage width of 21 mm, the
pump head and efciency increases compared to the other cases; this improvement is due the reduction
of losses arising from the generation of eddies in the passage and outlet of the impeller.
2012 Elsevier Ltd. All rights reserved.

1. Introduction
Due to the large complexity of ow and geometry in the radial
ow pumps, there are still many unknown issues associated with
the complete ow pattern in these pumps, which need to be
investigated. Moreover, the conduction of experimental studies
on samples with different volute and impeller geometries is
time-consuming and costly, and because of the complicated geometry, it is not possible to carry out a thorough investigation of the
ow eld for a vast number of operating conditions. Therefore,
the numerical ow analysis has recently become an appropriate
method of investigation of the ow patterns and losses.
Based on the main pioneering researches on the centrifugal
pump handling viscous uid (since 1926), the performance as a
function of oil viscosity was investigated, and the obtained results
were used for the design and selection of these pumps [14]. Today, because of drastic changes in the design and structure of newer models of hydraulic pumps, the previously obtained results on
older models cannot be used with high condence.
Li [510] investigated the effects of uids viscosity on the
performance of centrifugal oil pumps experimentally and
Corresponding author. Tel.: +98 2177240360.
E-mail address: Tahani@iust.ac.ir (M. Tahani).
0045-7930/$ - see front matter 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.compuid.2012.02.028

numerically; he presented the ow pattern inside the impeller.

According to these researches, the high viscosity results in rapid increases in the disc friction losses over outside of the impeller
shroud and hub as well as the hydraulic losses in ow channels
of the pump. The viscosity of uid not only affects the slip coefcient, but also causes the reduction of ow in the impeller and volute. Furthermore, he presented that there is a wide wake near the
blade suction side of the centrifugal pump impeller. Also, there is
not a jet near the blade pressure side, and the ow pattern is essentially different from the well-known jet/wake model. He obtained
the optimum number of blades for the impeller when uids with
different viscosities are pumped.
In the year 2004 Asuaje et al. [11] conducted 3D ow solution
by CFD tools. In this research, a design procedure was established.
This method was based on the geometrical design and the performance analysis. Their design tool took into account models and
correlations resulting from experimental data dealing with many
ranges of industrial centrifugal pumps which constitute a signicant database.
In 2007, Kergourlay et al. [12] investigated the effects of separated blades on the ow eld of water in centrifugal pumps.
According to this research, adding the splitters has negative and
positive effects on the pump behavior. It increases the head rise
compared to the original impeller that is mainly due to the

62

M.H. Shojaeefard et al. / Computers & Fluids 60 (2012) 6170

Nomenclature
b2
BEP
CFD
DH
g
H
k
MRF
OL
P
Pd
Ps
PL
Q
r
RNG
S
SDUSs
SST
t
u
V

passage width of Impeller (mm)

best efciency point
computational uid dynamics
hydraulic diameter (m)
gravity acceleration
head (m)
turbulent kinetic energy (m2/s2)
multiple reference frames
over load performance
pressure (Pa)
discharge pressure (Pa)
suction pressure (Pa)
part load performance
ow rate (m3/h)
position vector (m)
re-normalization group
source term
Skew Upwind Differencing Schemes
shear stress transport model
time (s)
relative velocity of uid (m/s)
capacity (l)

increase of the impeller slip factor which helps conduction of the

ow.
Shojaeefard et al. [13,14] performed experimental and numerical investigations to obtain the effect of the impellers outlet angle
during the pumping of oil. As a result of these researches, when the
blade outlet angle increases, the width of wake at the outlet of
impeller decreases, this phenomenon causes the improvement of
centrifugal pump performance when handling viscous uids.
In 2008, Grepsas et al. [15] conducted the numerical study and
optimized design of blades in centrifugal pumps by means of the
evolutionary strategy (dynamic algorithm). In this research, commercial software for the analysis of ow was used to conduct parametric studies of the effect of some geometric parameters of a
centrifugal pump impeller, and the results revealed that their modications could have a signicant impact on its performance.
In 2009, Anagnostopoulos [16] presented a quick numerical approach for the analysis of ow and the design of impeller blades. In
this research, a numerical method for the two-dimensional and
turbulent ows in the impeller of centrifugal pumps was developed. Spence and Amaral-Teixeira [17] studied the geometrical
variations on the pressure pulsations and performance characteristics of a centrifugal pump by CFD method. The results of this paper
presented by concentrating on the selected locations around the
pump and provided the detailed information regarding the pressure pulsation close to the impeller outlet, in the volute and in
the leakage ow region. There are many other valuable references
covering the ongoing research and review of the computational
uid dynamics to study the ow in the impeller of a centrifugal
pump [1821].
In this article, the procedure of 3-D investigation of ow in centrifugal pumps includes the sections of geometry denition, mesh
generation, analysis of equations, and processing of results. The nite volume method is used for the discretization of the governing
equations of ow. The turbulence model used in the simulation is
the k  x SST model. At the steady state case, this simulation
has been analyzed by using the multi-reference frame technique,
in which the impeller is situated in the rotating reference frame
and the volute is in the xed reference frame. The obtained numer-

altitude (m)

Greek symbols
b2
outlet angle of blade(degree)
q
density of the uid (kg/m3)
l
viscosity (Pa s)
lt
eddy viscosity (Pa s)
t
kinematic viscosity (mm2/s)
X
rotational speed (rpm)
s
stress tensor
g
efciency (%)
Subscripts and superscripts
Avg
average
d
discharge
elect
electrical
i, j
components
in
inlet
out
outlet
s
suction
t
total

time-averaged value

ical results are compared with the experimental ones, and acceptable correlation is found between the two sets of results.
Numerically solving the complete 3-D geometry of the centrifugal pump, offering an impeller for improving the efciency, simulating and comparing various possible geometries, and plotting of
Static pressure contours and velocity vectors inside the pump are
some of the signicant features of this article.
2. Geometry of the centrifugal pump and the generation of
mesh
In this report, for the numerical investigation of the ow eld of
centrifugal pump, the geometry of the pump model: 65200 (made
by the Pump Iran Co.) is used. This centrifugal pump has single axial suction and vane less volute casing; equipped with an impeller
of 209 mm in outside diameter and six backwards curved blades.
The blade outlet and wrapping angles of the impeller are 27.5
and 140 respectively. The shroud of the impeller made of metal
is machined. The roughness of the impeller and volute is 100 lm.
The pump tested is driven by a three-phase AC electric motor,
whose rated power is 5.5 kW and speed is 1450 rpm. First, the initial geometry of this centrifugal pump (which has been designed
for the pumping of water) is simulated by using the available technical specications. In the next steps, in order to analyze the effect
of changing the uid viscosity, the outlet angle and the ow passage width are modied. In this analysis, the denition of geometry
covers the three pump sections of volute, impeller, and outlet pipe,
which are connected together for the analysis of the whole pump
(Fig. 1).
Then, in the mesh generation part of this code, mesh conguration is produced based on the type of physics that is considered for
the problem. For better conformity of the geometry with the computational domain, at the near-wall regions, the structured mesh is
used for the boundary layer, and at regions away from the wall, the
unstructured mesh conguration is employed to correctly cover
the complex geometry. For producing the unstructured mesh conguration, six-sided, pyramid, and wedge-shaped elements are
used in appropriate situations, which are shown in Fig. 1.

M.H. Shojaeefard et al. / Computers & Fluids 60 (2012) 6170

63

Fig. 1. General view of the centrifugal pump model and the mesh conguration used for ow analysis.

values or the time-averaging approach; however, the most common method of modeling turbulent ows is the time-averaging
method. Using this approach for the case of incompressible ows,
the general forms of the governing equations could be expressed as
relations (1)(4). In these equations, the effect of terms resulting
from ow rotation, including the Coriolis and centrifugal forces
can be modeled by adding the source term to the equations of
momentum.
Since the pumped uid is incompressible and the ow is in a
steady state, the continuity equation has the following form:

@ui
0
@xj

Also, the equation of conservation of momentum together with

the denition of the source term, and the shear stress is expressed
as relation (2):
Fig. 2. A view of the modeled impeller.

2.1. Change of geometry of the initial pump

@
@P
@
qui uj

sij  qu0i u0j Sui

@xj
@xi
@xj

where the source term (including the centrifugal and Coriolis force
terms) is written as Eq. (3) and the average shear stress is obtained
from relation (4):

In Figs. 2 and 3, the geometry of the impeller and the outlet angle of blades have been shown, respectively. It can be observed that
the changes made to the blade are applied on the outlet angle b2.
As is clear from these gures, the passage width of the impeller
can be modied on the meridian plane; and the changes applied
to the original geometry with the outlet angle of 27.5 and passage
width of 17 mm are outlined in Table 1. To increase the passage
width, the distance between the front and back plates in the
meridian plane are increased. It should be mentioned that the applied change in the outlet angle of the blade is not accompanied by
any change in its inlet angle.

In order to take advantage of mesh compaction near the rigid

walls, the wall rule function is applied to the turbulence model
equations. This function is selected in such a way that all the mesh
points fall outside the viscous sub-layer.

2.2. Governing equations and the selection of a turbulence model

2.3. Flow in the rotating reference frame

In the averaging of steady-state incompressible ows, the conservation equations can be solved based on the average Reynolds

The analysis of ow in centrifugal pumps (because of the ows

rotational motion) leads to the examination of two types ows:

! ! !
! !
Sui q2 X  u X  X  r 

sij l



@ui @uj

@xj @xi

3
4

64

M.H. Shojaeefard et al. / Computers & Fluids 60 (2012) 6170

Fig. 3. Meridian plane with two passage widths of 17 and 21 mm.

uated with high precision using the k  x model and the SST function, and the obtained results enjoy better accuracy than those of
the k  e model; therefore, the SST turbulence model is used for
the numerical investigation of ow inside the centrifugal pump.

Table 1
Specications of the modied geometry.
Passage width
of Impeller(b2) (mm)

Outlet angle of blade (b2) ()

17
21

27.5
27.5

30
30

32.5
32.5

steady, and unsteady. If the volute has xed blades (diffuser), because of the alternating motion of blades and the confrontation between the impeller and diffuser blades, the ow is of the unsteady
type, relative to the xed external reference frame. However, when
the diffuser doesnt have xed blades, the ow can be studied in a
steady state by dening a rotating reference frame.
This simulation is dened by means of the multi-reference
frame technique, in which the impeller is situated in the rotating
reference frame, and the volute is in the xed reference frame,
and they are related to each other through the Frozen Rotor.
The frozen rotor method employs a quasi-steady algorithm, where
the rotor and stator are modeled at a xed (frozen) position relative to each other. Rotational terms are included in the moving
frames, but transient effects are neglected. This provides an efcient method for the calculation of interactions between impellers
and casings (volutes), and is also a viable option for compact machines with small distances between rotor and stator. In this way
the solution provides a snapshot of the ow regime [22].
When the NavierStokes equations of motion are solved in a
rotating reference frame, the uid acceleration appears as an additional term in the momentum equations, which was shown in relation (3).
2.4. Turbulence model
With the help of the mixing function value, the SST model automatically uses the k  x model in the near-wall regions and the
k  e model in the regions away from the wall. This model rst
modies the energy production term in the kinetic energy transfer
equation [23,24].
Considering the studies conducted on the two models of k  e
and RNG k  e, it is concluded that the near-wall ow can be eval-

2.5. Discretization of equations and the solution method

In the developed computer code, the nite volume method has
been used for the discretization of equations, but the analysis of
the geometry has been based on the nite element approach. So,
the geometrical exibility of the nite element method can be
used; however, the equations are dealt with in the form of nite
volume [22]. As is presented in the following equations, the nite
volume method is implemented by integrating the equations over
a selected control volume, with the help of the Gaussian theory. In
this approach, the computational space is divided into small elements and the surfaces of the control volume are dened on the
mid plane of each of those elements. The numerical scheme employed in this case, involves generating nite control volumes from
the mesh, which is shown in Fig. 4.
In this gure, each node is surrounded by a set of surfaces which
comprise the control volume. The control volume is constructed
around each mesh node using the median-dual discretization
scheme. The boundary of the control volume is dened by lines
joining the centers of the element edges with the element centroids surrounding the node, as shown in right hand side of this
gure.
The governing equations in integral form are applied to each nite control volume such that the relevant quantity (mass or
momentum) is conserved in a discrete sense for each. Considering
an isolated mesh element for simplicity, such as the one shown in
left hand side of this gure, the surface uxes of the continuous
equations must be discretely represented at the three integration
points to complete the conversion of the continuous equations into
their discrete forms. The integration points (ipn) are located at the
center of each boundary segment surrounding the control volume
as shown in Fig. 4. The surface ow rates are presented separately
at the integration points, and nally, the discredited forms of the
equations are obtained as follows:

@
@t

q dv
v

Z
s

qui dnj 0

M.H. Shojaeefard et al. / Computers & Fluids 60 (2012) 6170

65

Fig. 4. Typical mesh element with nodes (n1, n2 and n3) and integration points (ip1, ip2 and ip3).

@
@t

qui dv

quj ui dnj
s

P dni

@
@t

qu dv

Z
s



Z
@ui @uj
dnj Sui dv

@xj @xi
v

quj u dnj
 s
Z
@u
dnj
C
@xj
s


Z
@ui @uj
dnj Su dv

@xj @xi
v

In this computer code, all the velocity vector components for

surfaces are determined in the Cartesian coordinates, and all the
ow rates are obtained at the integration points. The integration
points for each volume are shared by the adjacent volumes, and
the output ow rate of each control volume is equal to the input
ow rate of the adjacent volume. One of the strong points of this
numerical method is that in this way, the conservation laws are
completely satised for the equations, as follows:

qV

qV

qV

q  q

Dt

X
quj Dnj ip 0


 X
ui  ui
_ ip uj ip
m

Dt
ip
X
X  @ui @uj  

PDni ip
l

Dnj
Su i V
@xj @xi
ip
ip
ip


u  u
Dt

ip

X
ip

_ ip uip
m


X  @u
C
Dnj
Su V
@xj
ip
ip

10

The mass ow rate is obtained from the following equation:

_ ip quj Dnj ip

m

11

The diffusion terms are estimated with the help of the weight
functions of the nite element method, and in the momentum
equations, the surface integral of the pressure gradient terms is
determined at the integration points. The advection terms are analyzed by the Skew Upwind Differencing Schemes (SUDSs) method,
which in this approach, the advection terms are in the streamwise
direction.
In the segregated solution strategy, the momentum equation is
solved by using an initial solution for pressure, and then a relation
is obtained for pressure modication. Because the guess correct
scheme is a linear system, a large number of iterations together
with relaxation coefcients are needed for the variables.

In this computer program, the coupled solver has been used,

so that the hydrodynamic equations (u, v, w, and p) are solved in
a system. This type of solution is used in the totally implicit discretization of equations in the given time step. In solving the stable
problems, the time step acts like an accelerator parameter, and
directs the approximate solution on the basis of the stable solution,
which, for the reduction of the number of required iterations for
the convergence of the stable problem or the reduction of the volume of computations for each time step in the analysis, is independent of time.
2.6. Boundary conditions and physical properties of working uids
The static pressure of the reservoir is determined at the entrance boundary of the pump (entrance to the impeller). Since
the inow to the pump from the suction reservoir occurs at a high
volume, and this ow has already been smoothed by bafe plates,
the average turbulence intensity is considered as 5%, which, in light
of the existing system and the ow at the entrance to the pump, is
a totally empirical value. The mass ow rate or the exit velocity is
dened at the exit plane of the pipe attached to the volute. In addition, the impeller and the rotating section are modeled in the rotating frame, and the volute and the pipe attached to it are modeled in
the xed reference frame. Since the whole pump is modeled in this
analysis, at the solid boundaries (including the pump casing,
impeller, and the volute), the no-slip condition with relative roughness of 100 lm is applied. This roughness value is obtained
through the Reynolds stresses and the shear stress terms in the
equations. The average tangential velocity between the xed and
rotating sections is determined by simultaneously solving these
two regions.
The uids used in the numerical simulation and experimental
investigation, are water and oil with the density of 998 and
875 kg/m3 and kinematic viscosity of 1 and 43 mm2/s respectively
at 25 C.
2.7. Evaluation of the dependency of results from mesh conguration
For the reduction of computation time and the improvement of
accuracy, the optimum number of mesh elements in the simulation
has been investigated. Also, the total pressure rise inside the pump
and mean differences of circumferential velocities values for the
best efciency point performance (BEP, with maximum efciency)
has been used as the evaluation parameter for the effect of mesh
size on the solution. Finally, the least number of dependent
mesh elements for which the output pressure is obtained with negligible change, has been calculated. Table 2 shows the study of the

66

M.H. Shojaeefard et al. / Computers & Fluids 60 (2012) 6170

Table 2
Evaluation of the dependency of mesh and summary of CPU time for original impeller.
Element
number

Total pressure rise

inside the pump at
BEP (kPa)

Mean differences of
circumferential
velocities values (m/s)

CPU time (for

Core2Due CPU of
2.5 GHz) (h)

628146
1045668
1842482
2173604
3277226

197.983
186.274
180.169
178.888
178.261

0.226
0.304
0.371
0.398
0.402

14.8
25.1
47.4
58.3
90.4

As the numbers in the above table demonstrate, at 3277226 elements, the output pressure does not change much. The number of
mesh elements includes the sum of elements in the impeller, volute, and the outlet pipe.
2.8. Summary of experimental setup

dependency of results from mesh conguration and summary of

CPU time for the centrifugal pump for pumping viscous uids at
the blade outlet angle (b2) of 27.5 and passage width (b2) of
17 mm.

A centrifugal pump test setup, shown in Fig. 5, was used to compare the impeller geometries effects on the performance of centrifugal pump.
The pipe of the rig was made of stainless steel with inner diameter of 80 mm. The tank net volume was 2400 l. The operation condition was controlled by a gate valve on the discharge pipe. A digital
pressure transmitter gage was used at the pump inlet and outlet
pipes to measure the inlet and outlet pressures accurately. Bafe
plates were used for damping the disturbance of discharged uid.

Fig. 5. Centrifugal pump test setup.

Fig. 6. Centrifugal pump performance diagrams at different working conditions for original Impeller geometry.

67

M.H. Shojaeefard et al. / Computers & Fluids 60 (2012) 6170

Fig. 7. HQ curve of the centrifugal pump at different geometries.

Fig. 8. Comparison of the numerical and experimental results.

The following parameters were measured under steady conditions to draw the main performance diagrams.
The pump head at all stages was computed using Bernoulli
equation:

The centrifugal pump efciency at all stages was dened as the

ratio between the pump output power and the input electrical
power consumption.

g
Ps

V 21
2g

Z1 H

Pd

V 22
2g

Z 2 Hl

12

For the same inlet and outlet pipe diameter and velocity
(V1 = V2):

Pout;pump
cQH

;
Pin;pump
Pin;pump

Pin;pump P out;elect gelect  Pin;elct

14

The efciency of this electromotor was almost constant (83%).

Input electrical power was measured using a three-phase
wattmeter.
3. Results and discussions



Pd  Ps

Z 2  Z 1 Hl

where Hl was the head of losses in the outlet pipe.

13

3.1. Effects of viscosity on performance

The mentioned parameters were measured under steady conditions to draw the main performance diagrams (PQ, gQ and HQ),

68

M.H. Shojaeefard et al. / Computers & Fluids 60 (2012) 6170

Fig. 9. Variations of efciency vs. ow rate for different uids and impeller geometries.

Fig. 10. Static pressure contours for impellers at BEP at 50% span.

at different working conditions, i.e. part load (PL), best efciency

point (BEP), and overload (OL). These diagrams are shown in
Fig. 6 for different uids with original impeller without improving

in geometry. This gure shown that, Frictional losses reduce the

pressure inside the impeller, diffuser, and volute resulting in
reduction of the head and efciency. On the other hand, the friction

M.H. Shojaeefard et al. / Computers & Fluids 60 (2012) 6170

on the discs including the wheel in the case of oil increases the
absorption power compared with the case of water.
Based on these results, at the best efciency points (that located
at QBEP(water) = 47.3 and QBEP(oil) = 48.2 m3/h) the efciency and
head values decrease about 20.5% and 1.4 m respectively, and the
power consumption increases about 1.03 kW during the pumping
of oil. The mentioned factors result in a remarkable reduction in
the performance.
3.2. The numerical head-ow rate (HQ) curve
For obtaining the HQ curves of centrifugal pumps in the cases
of steady state, turbulent ows and incompressible uids, the
amounts of total pressure at the entrance and exit of the simulated
geometry (such as experimental test points) are determined, and
by using relation (15), the pump head is calculated:

Pt2  Pt1
qg

15

The head-ow rate (HQ) curve resulting from the numerical

solution for the pumping of viscous uids has been illustrated in
Fig. 7.
The numerical results indicate that the centrifugal pump with
b2 = 30 and b2 = 21 mm has the highest pressure head in oil pumping; therefore, the noted pump has been built, and its results have
been compared with those of the simulation.
Because of laboratory limitations, the experimental results have
been exclusively obtained for three different kinds of impellers
with: b2 = 27.5 and b2 = 17 mm; b2 = 32.5 and b2 = 17 mm; and

69

b2 = 30 and b2 = 21 mm. Even with this few numbers of experiments, the numerical simulation can be veried for other impeller
blade angles and uid types, and the obtained results can be accepted without further experimentation.
Fig. 8 shows the comparison between the numerical and experimental HQ results during the pumping of water and oil by the
centrifugal pump. Comparing the experimental and numerical
values of head show that the average difference percentage in
the different cases (27.517 mm-water), (30.021 mm-oil),
(32.517 mm-oil) and (27.517 mm-oil) was 4.06%, 3.35%,
4.08% and 4% respectively. According to this gure, the numerical
and experimental results show satisfactory correlation.
3.3. The efciency-ow rate (gQ) curve
Fig. 9 shows the variations of efciency with the ow rate for
water and oil. It was depicted that increasing the outlet angle in
PL performance condition increases the efciency more than
increasing the passage width. The Results shown in this gure reveal that by improving the impeller geometry (outlet blade angle
of 30 and impeller passage width of 21 mm), the increment
magnitude of efciency around the BEP and OL operating conditions was 7.93% for mentioned viscous uid.
3.4. Static pressure contours and velocity vectors on the blade-to-blade
plane
When the centrifugal pump is pumping a highly viscous uid,
due to the increase of hydraulic losses and skin friction, the pump

Fig. 11. Velocity vectors for impellers at BEP at 50% span.

70

M.H. Shojaeefard et al. / Computers & Fluids 60 (2012) 6170

head diminishes, compared to the case in which water is pumped;

and this fact should also be demonstrated by the static pressure
contours.
The static pressure distribution over the suction and pressure
side on the middle-span plane of the impellers for the best efciency point condition for the viscous uid ow are illustrated in
Fig. 10. As this gure the pressure increases gradually along
streamwise direction within impeller blade-to-blade passage and
has higher pressure on pressure side than suction side for each
impeller. It is demonstrated that the average of pressure in the outlet area of impeller (b2 = 30 and b2 = 21 mm) is higher than other
impellers. This results cause to increasing the head as well as
shown in Fig. 7.
Fig. 11 illustrates the velocity vectors on the middle-span plane
of the impellers for the viscous uid ow in the numerical simulations performed at the best efciency point condition. It is noticed
that a big zone with low velocity exists there in the original impeller. The center region of the blade-to-blade passage near the suction side has a low velocity region and is considered as stall ow.
This phenomenon could be considered as jet-wake structure
development phase. The improvement of performance can be seen
in the ow separation in the passage of the impeller. As can be
seen, at the impeller blade outlet angle of 30 and passage width
of 21 mm, there is uniform ow with less number of vortices.
4. Conclusion
In this study, the effects of blade outlet angle and passage width
on the performance of a centrifugal pump have been investigated
numerically and experimentally. Considering the comparison between the performances of centrifugal pumps with different
impellers, during the pumping of water and oil, the following conclusions can be made:
 The friction on the discs including the wheel in the case of oil
decreases the head and efciency and increases the power consumption compared with the case of water. On the other hand,
the performance of centrifugal pumps drops sharply during the
pumping of viscous uids (Fig. 6). Changing the original geometry of the impeller improves the centrifugal pump
performance.
 Numerical results show that the impeller blade with the angle
of 30 and passage width of 21 mm produces a higher head relative to the other ve blade settings (Fig. 7).
 The results obtained from the numerical and experimental
investigations on a 65200 centrifugal pump performance have
satisfactory agreement, and demonstrate that increasing the
impeller passage width from 17 to 21 mm increases the head
and hydraulic efciency due to reduction of the friction losses.
Also, the centrifugal pump performance with the impeller passage width equal to 21 mm, at outlet blade angle of 30
improves in comparison with 27.5 and 32.5. This is due to
reduction of the dissipation arising by vortex formation in
impeller passage when the pump handles viscose liquid (Figs.
8 and 9).
 However, by increasing the blade outlet angle to 30 and the
passage width to 21 mm, the pump efciency decreases at part
load (in comparison with b2 = 32.5 and b2 = 17 mm), but
increases to the highest level at the best efciency point and
the overload performances, compared with other blade
congurations.

 The static pressure contours for impellers at BEP show that the
pressure increases gradually along streamwise direction within
impeller blade-to-blade passage and has higher pressure on
pressure surface than suction surface for each conguration. It
is demonstrated that the average of pressure in the outlet area
of impeller (b2 = 30 and b2 = 21 mm) is higher than the others.
 Velocity vectors for impellers at BEP show that in the improved
impeller, the width of wake at the outlet of impeller and the
hydraulic losses decrease, these phenomena cause the improvement of centrifugal pump performance when handling viscous
uids.

References
[1] Dauherty RL. A further investigation of performance of centrifugal pump when
pumping oils, Bulletin 130. Gould Pumps, Inc; 1926.
[2] Stepanoff AJ. Pumping viscous oils which centrifugal pumps. Oil Gas J 1940(4).
[3] Moore J. Calculations of three-dimensional viscous ow and wake
development in a centrifugal impeller. AICHE Symp Ser 1979:617.
[4] Stoffel B. Tests on centrifugal pumps for handling viscous liquids. Institute of
Chemical Engineerings; paper no. 3, 1980.
[5] Li WG, Hu ZM. An experimental study on performance of centrifugal oil pump.
Fluids Mach 1997;25(2):37.
[6] Li WG, Deng DL. Reynolds number of centrifugal oil pumps and its effects on
the performance. Pump Technol 1999(3):103.
[7] Li WG. Effects of viscosity of uids on centrifugal pump performance and ow
pattern in the impeller. Int J Heat Fluid Flow 2000(21):20712.
[8] Li WG. Numerical study on behavior of a centrifugal pump when delivering
viscous oils Part 1: Performance. Int J Turbo Jet Engines 2008;25(2):6179.
[9] Li WG. The sudden-rising head effect in centrifugal oil pumps. World Pumps;
2000 [October].
[10] Li W-G. Inuence of the number of impeller blades on the performance of
centrifugal oil pumps. World Pumps 2002:325.
[11] Asuaje M, Bakir F, Kouidri S, Rey R. Inverse design method for centrifugal
impellers and comparison with numerical simulation tools. Int J Comput Fluid
Dyn 2004;18(2):10110.
[12] Kergourlay G, Younsi M, Bakir F, Rey R. Inuence of splitter blades on the ow
eld of a centrifugal pump: test-analysis comparison. Int J Rotat Mach 2007:13
[Article ID 85024].
[13] Shojaeefard MH, Boyaghchi FA, Ehghaghi MB. Experimental study and threedimensional numerical ow simulation in a centrifugal pump when handling
viscous uids. IUST Int J Eng Sci 2006;17(34):5360.
[14] Shojaeefard MH, Ehghaghi MB, Boyaghchi FA. Studies on the inuence of
various blade outlet angles in a centrifugal pump when handling viscous. Am J
Appl Sci 2007;4(9):71824.
[15] Grapsas V, Stamatelos F, Anagnostopoulos JS, Papantonis D. Numerical study
and optimal blade design of a centrifugal pump by evolutionary algorithms. In:
12th International conference on knowledge-based intelligent information
and engineering systems, KES 2008, September 5, 2008. Zagreb
(Croatia): Springer Verlag; 2008. p. 2633.
[16] Anagnostopoulos JS. A fast numerical method for ow analysis and blade
design in centrifugal pump impellers. Comput Fluids 2009;38:2849.
[17] Spence R, Amaral-Teixeira J. A CFD parametric study of geometrical variations
on the pressure pulsations and performance characteristics of a centrifugal
pump. Comput Fluids 2009;38:124357.
[18] Gugau M. Transient impeller-volute interaction in a centrifugal pump,
technich universitat darmstadt, fg tubomaschine und uidantriebstechnik,
Petersentrasse 30, 64287 Darmstadt; 2003.
[19] Zhou W, Zhao Z, Lee TS, Winoto SH. Investigation of ow through centrifugal
pump impellers using computational uid dynamics. Int J Rotat Mach
2003;9:4961.
[20] Cheah KW, Lee TS, Winoto SH, Zhao ZM. Numerical ow simulation in a
centrifugal pump at design and off-design conditions. Int J Rotat Mach 2007:8
[Article ID 83641].
[21] Klimovskii KK. Improving the efciency of a centrifugal pump. Thermal Eng
(English Translation of Teploenergetika) 2008;55:2534.
[22] Elder R, Tourlidakis A, Yates M. Advances of CFD in uid machinery design.
IMechE Seminar Publications; 2002, ISBN:978-1-86058-353-7.
[23] Menter FR. Two-equation eddy-viscosity turbulence models for engineering
applications. AIAA J 1994;32:1598605.
[24] El-Behery SM, Hamed MH. A comparative study of turbulence models
performance for separating ow in a planar asymmetric diffuser. Comput
Fluids 2011;44:24857.