Home

Search

Collections

Journals

About

Contact us

My IOPscience

Analysis of the flow field into a two stages and double entry storage pump taking into account
two geometries of stator blades

This content has been downloaded from IOPscience. Please scroll down to see the full text.
2010 IOP Conf. Ser.: Earth Environ. Sci. 12 012016
(http://iopscience.iop.org/1755-1315/12/1/012016)
View the table of contents for this issue, or go to the journal homepage for more

Download details:
IP Address: 181.64.209.27
This content was downloaded on 07/06/2015 at 19:14

Please note that terms and conditions apply.

25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

Analysis of the flow field into a two stages and double entry
storage pump taking into account two geometries of stator
blades
G Dunca1, S Muntean2 and E C Isbasoiu1
1

Hydraulics and Hydraulic Machinery Department, Politehnica University of Bucharest,

313 Splaiul Independentei, sector 5, Bucharest, Romania, Tel.: (+40) 723284815, Fax:
(+40) 2140298.65,
2
Centre of Advanced Research in Engineering Sciences, Romanian Academy - Timisoara
Branch, Bv. Mihai Viteazu 24, RO-300223, Timioara, Romania
E-mail: [email protected]
Abstract. The paper presents the 3D numerical analysis of the flow into a hydraulic passage of
the two stages and double entry storage pump. One of the reasons for choosing this machinery
was that, even from the beginning of its operation, high levels of noise and vibration were
recorded. According to the literature, these can be considered as effects of the impeller-stator
phenomenon. After only 100 hours of operation, the pump first stator blades was bend and the
second stator blades was broken. As a rehabilitation solution, 100 mm of the chord were cut
from the stator blades, near the leading edge. After the rehabilitation, a decrease of the noise
and vibration levels during pump operation was observed. In order to analyse the pump
behaviour, three measurements campaigns were conducted, after the rehabilitation. Yet, the
experimental results were not very conclusive. A more detailed experimental analysis on a real
turbo machine is very difficult and expensive. Thus, in order to obtain more detailed
information regarding the impeller-stator phenomenon inside the analysed pump, a numerical
analysis was realized. The impeller-stator (between the first impeller and first stator as well as
between second impeller and second stator) and stator-impeller (between the first stator and
second impeller) interactions are taken into account with mixing interface method. The
hydrodynamic field from the inlet to the outlet is obtained. As a result, the pressure rise and
hydraulic efficiency are computed at best efficiency point. These values are validated against
experimental data measured into the storage pump. Comparing the numerical results obtained
for the two geometries of the stators, it can be seen that they have different behaviour during
the pumps operation. It can be considered that, although the same geometry modification was
realized for both the stators, the effects on the flow parameters are different, only for the
second stator being possible to observe a net improvement of the blade load.

1. Introduction
The centrifugal impeller of a pump interferes with the stator blades which produces pressure fluctuations
downstream the impeller. The pressure fluctuations amplitude can be as high as the total pressure rise inside the
pump in case of small radial gaps between the impeller and stator blades, Arndt et al. [1, 2] Brennen [3] and
Tsukamoto et al. [4]. This leads to noises and vibrations causing unacceptable levels of stress that can affect
machines structure. Barrio et al [5] have studied the influence of the radial gap between the impeller and the tongue
on the dynamic forces generated inside a simple entry and single stage centrifugal pump.
The turbo machines flow is essentially unsteady due to the rotor-stator interaction [6]. On the other hand,
rigorously speaking, the geometrical periodicity of the stator/rotor blade rows cannot be used since there are
differences in flow from one inter-blade channel to another. Spence et al. [9] presented a numerical investigation
of the time variation of pressure within a complete centrifugal pump and the focus is on the variation of the
pressure pulsation close to the impeller outlet in function of the discharge. Computing the real flow (threedimensional, turbulent and unsteady) through a complex storage pump still requires large computer memory and

c 2010 IOP Publishing Ltd



25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

CPU time. However, a simplified simulation technique must be employed to obtain useful results for storage
pump analysis, using currently available computing resources [7, 8]. As a result, with carefully chosen
assumptions, one can devise a methodology for computing the pump flow, such that good and engineering useful
results are obtained.
In this paper, numerical simulations into the storage pump with double entry and two stages centrifugal pump
were carried out in order to investigate the hydrodynamic field for two different geometries of stator blades. The
test case problem is presented in the next section. Next, computational domains, boundary conditions and
numerical method are presented. Velocity and pressure coefficients are plotted for impellers as well as on
original and modified geometry of the first and second stators. The conclusions are drawn in last section.

2. Test case: Galceag pumping station

The Galceag pumping station consists in two units with two stages and double entry. It takes the water from
the Cugir lake and leads it into the hydro power plant Galceag penstock. The pumps discharge is brought to the
two pumps through a headrace with 6500 m length and 2800 mm diameter. The headrace branch in two ducts,
one for each pump, and two slide gates are mounted on each of these ducts, upstream the pumps. Before the
branch, a surge tank is installed on the headrace. The discharge ducts have 338 m length and 1.3 m diameter and
for each pump a sphere vane of 700 mm diameter is mounted.
Centrifugal
pump
sphere valve
Dn = 700

slide gate
Dn = 1000
pn = 6

discharge
duct

Fig. 1 Centrifugal pump with double entry and two stages installed in the Galceag storage pump
The analyzed pump is a centrifugal pump with double entry and two stages [10]. The pump is a specific
speed of nq = 28.

Fig. 2 Centrifugal pump with double entry and two stages installed in the Galceag storage pump
During the experimental investigation, the pressure was measured using two transducers, one set at the
pumps inlet and the second one set at outlet. The acquired pressure mean values were compared with the
pressure values read by means of two pressure manometers. In order to identify the possible operating disorders,
as well as their causes, the signal for each measurement was processed. A representative sample of data was used

25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

and the continuous component of the signal was extracted in order to compute the FFT. This leaded to
indications regarding vibrations and distortions sources.
The sampling period being 0.003 seconds resulted a cutting frequency of 166.5 Hz (the Nyquist frequency)
and a resolution of 0.645 Hz. The results of spectral analyze of the outlet pressure are synthetically shown in
Figure 3, and in Figure 4 the results of spectral analyze of the inlet pressure are shown. For the outlet pressure,
the results show that in the frequency domain of 0163.5 Hz the spectrum is quasi-symmetric around the value
of 100 Hz. The main pressure fluctuations are around 24.914 Hz, 43.12 Hz, 86.722 Hz, 99.258 Hz and
115.371Hz.
The frequency at 1000 rot/min motor rotational speed is:

fm =

1000
= 16.666 Hz,
60

(1)

and the fundamental frequency is 100 Hz which corresponds to the above frequency multiplied by the impellers
number of blades (7 blades):

f =

N
1000
n =
7 = 116.62 Hz.
60
60

Fig. 3 Pressure spectrum in outlet duct

(2)

Fig. 4 Pressure spectrum in the inlet duct

The pressure spectrum in the inlet duct (Fig. 4) has higher amplitude at 26.426 Hz, 29.648 Hz, 43.184
Hz and 48.984 Hz in the range 0163.5 Hz. The pumps inlet hydraulic circuit generates these frequencies and
they cannot be correlated with the pumps specific frequencies. It can be considered that these frequencies
(24.914 Hz and 43.12 Hz), which are present also in the outlet pressure spectrum, derive from the inlet hydraulic
circuit and are sent through the pump in the outlet hydraulic circuit.
Because a detailed experimental analysis inside the pump would have been difficult, in order to obtain
distinct results regarding the possible sources of these vibrations, a numerical analysis of the flow inside the
centrifugal pump was realized, taking into account the two different geometries of the stators blades.

3. Computational domains, equations and boundary conditions

In our case, three-dimensional flow investigation through half of the pump is considered due to the plane of
symmetry, Fig. 5. Even so, the hydraulic circuit is complex and includes: suction elbow, impeller 1 (R1) with
seven blades, stator 1 (St1) with two rows each with nine blades, impeller 2 (R2) with same geometry with
impeller 1, stator 2 (St2) with nine blades and volute with three stay vanes and one tongue, Fig. 6. The
computational domains are the following: entire suction elbow, one blade channel for both identical impellers,
one channel for first stator, and half of the volute with all blades of the second stator and all stay vanes.
Two different geometries of the stator blades are investigated in order to evaluate the impeller-stator
interaction. The original geometry of the stator blades (marked with grey in Fig. 7) is reshaped. As a result, a
new geometry of the stator blades (coloured with light blue) is obtained, Figure 7.
An unstructured mesh with 565231 cells is generated for the suction elbow, a structured mesh with 475680
cells is generated on the computational domain of the two impellers, a structured mesh with 859380 cells is used
on the stator computation domain, and an unstructured mesh with 1471360 cells is generated for the volute,
using Gambit.

25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

Fig. 5 Cross view through the storage pump. The half

of hydraulic passage is marked with colours: suction
elbow (blue), first impeller (red), stator (yellow),
second impeller (red)
and the volute (light blue).

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

Fig. 6 Exploded axonometric view of the half

hydraulic passage of Galceag storage pump.

Original geometry
Modified geometry
Cutwater diameter
modified geometry
1132

Cutwater diameter
1104

Impeller diameter
1100

Fig. 7 Two geometries for the stators blades: original (with gray) and modified (light blue).
In the present investigation a full 3D, steady, turbulent and incompressible fluid flow is considered. Thus the
upper equations can be written
G
V = 0
G
G G p
G
+ V
V V = f

(3a)
(3b)

These equations are applied in this form in the suction elbow, stators and volute. For impellers analysis, a
steady relative flow is assumed, in a rotating frame of reference with angular speed = k :

W = V r
The following boundary conditions are considered in our computation for each domain:

(4)

25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

Table 1 Boundary conditions imposed on each computational domain

Surface
inlet
outlet
solid
periodic

Boundary condition
velocity components and turbulent quantities
pressure field
wall
periodicity for all quantities

Mixing interface method for coupling velocity and pressure fields from consecutive components is used. In
the mixing plane approach, each fluid zone is treated as a steady-state problem. Flow-field data from adjacent
zones are passed as boundary conditions that are spatially averaged or mixed" at the mixing plane interface.
This mixing removes any unsteadiness that would arise due to circumferential variations in the passage-topassage flow field, thus yielding a steady-state result. Despite the simplifications inherent in the mixing plane
model, the resulting solutions can provide reasonable approximations of the time-averaged flow field [12, 13].
Four mixing interface planes (two annular sections between suction elbow and first impeller and first stator
and second impeller as well as two cylindrical sections between impellers and stators, labeled MI in Figure 5) are
used in order to compute the flow through the entire storage pump. An iterative process is employed to obtain the
absolute velocity and pressure continuity across all mixing interfaces. It starts by computing the suction elbow,
with a constant suction elbow outflow pressure. The resulting outflow velocity components are circumferentially
averaged and corrected to preserve the prescribed discharge, then used as inflow condition on the impeller
computational domain. The same procedure is applied for coupling the first impeller domain with the first stator
domain (R1-St1), the stator domain with the second impeller domain (St1-R2) and the second impeller domain
with volute domain that includes the second stator (R2-St2). The volute computation is performed with inflow
velocity profile on the inlet section and constant pressure on the outlet section displaced in the downstream duct.
After coupling all components downstream, a new pressure profile is obtained on the inlet section of the volute
domain. Consequently, the flow in the second impeller is computed taking into account the new pressure profile
from volute domain. After the coupling between the volute and second impeller, other computational domains
are connected upstream in order to obtain continuous pressure field. Then, computing the suction elbow with a
new outflow pressure is performed, and so on. The iterating process is stopped when no significant changes in
pressure field occur from the previous iteration for all mixing planes. The convergence of the iterative scheme is
investigated. It is found that relaxation must be employed in order for the iterative process to converge. Excellent
convergence of the iterative mixing interface algorithm is achieved, with no more than 5-6 iterations. Threedimensional turbulent computation is performed using the FLUENT 6.3 code and SST k turbulence model
[14, 15].
The numerical investigations are performed for one operating regime near to the best efficiency point with
parameters from Tab. 2:
Table 2 Parameters of the operating point investigated
Parameter

Value

Rotational speed n [rpm]

Discharge Q [m3/s]
Pumping head H [m]

1000
3
247

Hydraulic power Ph [kW]

7269.2

4. Numerical results
Hydrodynamic field into the storage pump is computed. As a result, the pressure coefficient and the velocity
coefficient distributions are obtained. A correct evaluation of this minimum pressure level is crucial in evaluating
the storage pump cavitational behavior.
The pressure coefficient is defined according to the eq. (5) and absolute velocity coefficient in eq. (6):

25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

cp =

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

p p ref

gH
v
cv =
2 gH

trailing edge

(5)
(6)

z = 10%

shroud

z = 50%
z = 90%

suction side

inlet
section

hub

pressure
side
leading edge
Fig. 8 Shroud-to-hub sections at 10%, 50% and 90% selected for plotting the numerical results.
Figure 8 presents the shroud-to-hub sections selected in order to plot the numerical results. The absolute
velocity coefficient distribution on the mixing plan section at the first impeller outlet (corresponding to a radius
of R=0.555 m) is presented in Fig. 9. On this section three analysing section were made, corresponding to three
values of the axial coordinate, one next to the shroud, z = 10% , one in the middle of the mixing section,
z = 50% and one next to the hub, z = 90% . The distribution of this velocity coefficient is plotted in terms of
the angular coordinate of the impeller outlet section. One can observe that the non-uniformity is more significant
near the shroud due to the meridian shape of the channel, Fig. 9(right). However, these circumferential nonuniformities are not passed downstream in the stator due to mixing interface technique [12]. It can be seen that
there is a pronounced non-uniformity of the velocity field at the impeller outlet corresponding to the wake jet
phenomenon [13], as Prof. Brennen has shown [3].

z10%

suction side

cv [-]

pressure
side

z50%
cv [-]

z90%

impeller
blade
suction side

Fig. 9 Absolute velocity coefficient distribution on the outlet section of the first impeller (R1)

25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

0.9

1.8

0.8

1.6

0.7

1.4

0.6

1.2

0.5

PS

Cp [-]

Cp [-]

The pressure coefficient distribution in seven shroud-to-hub sections on the first impeller is presented in Fig.
10. One can observe a uniform pressure coefficient distribution on the first impeller blade, excepting the leading
edge. That means the impact flow is on the blade pressure side (PS) near the leading edge. The same conclusion
can be drawn for the second impeller from Fig.11.

0.4

10%

SS

SS

0.8

Shroud

0.3

PS

Shroud

0.6

10%

30%
50%

0.2

30%

0.4

50%

70%
0.1

90%

90%

Hub
0.2

0.4

0.6

0.8

1.2

1.4

Hub

0
0

70%

0.2

1.6

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

l/D2 [-]

l/D2 [-]

Fig. 10 Pressure coefficient distribution on the blade

of the first impeller (R1)

Fig. 11 Pressure coefficient distribution on the

blade of the second impeller (R2).

Figure 12 shows the pressure coefficient distribution on the first blade row of the first stator with original
geometry. One can see in Fig. 12 the loading near to the leading edge on the first blade of the first stator (St1).
Moreover, it can be seen that the loading near to the shroud is less than the loading to the hub that means the
different flow impact along to the leading edge. Indeed the flow to the outlet is leaned according to the trailing
edge of the first impeller generating different flow impact on the straight first stator blade. It can be observed in
Fig. 13 that the pressure has lower values near the leading edge in the middle section (50%) on the modified
blade geometry of the first stator (St1_mod), which indicates a different attack angle on the stator inlet. However,
this difference is not so clear near the hub. Nevertheless, the new geometry of the first stator blades increases the
gap between the first impeller outlet (R1) and the first stator inlet (St1). Consequently, the impeller-stator
interaction (R1-St1) is mitigated.
The same analysis is performed for impeller-stator interaction between second impeller (R2) and the second
stator (St2) presented in Figs. 14 and 15. In this case the benefit is clearer than for the first one. One can observe
an improved pressure coefficient distribution near to the leading edge of the second stator new geometry
(St2_mod plotted in Fig. 15) than the same distribution computed on the original geometry (St2 from Fig. 14).
Consequently, it can be stated that the flow incidence on the reshaped blade of the second stator is improved.
1.4

1.4

1.2

1.2

PS

PS

0.8

Cp [-]

SS

Cp [-]

SS

0.8

0.6

Shroud
0.6

Shroud

10%

10%

0.4

30%

30%

0.4
50%

50%
0.2

70%

70%

0.2

90%

90%
Hub

0
0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

Hub

0
0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

l/D2 [-]

l/D2 [-]

Fig. 12 Pressure coefficient distribution on the first

blade (original geometry) of the first stator (St1)

Fig. 13 Pressure coefficient distribution on the

first blade

25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

(modified geometry) of the first stator (St1_mod

).

PS

2.0

1.8

SS

1.6

SS

1.6

1.4

1.4

1.2

1.2
Cp [-]

Cp [-]

PS

2.0

1.8

1.0
0.8

1.0
0.8

Shroud

0.6

10%

10%
30%

30%

0.4

Shroud

0.6

50%

0.4

50%

70%

70%

0.2

0.2

90%

90%

Hub

0.0

Hub
0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

0.1

0.2

0.3

0.4

l/D2 [-]

0.5

0.6

0.7

0.8

0.9

l/D2 [-]

Fig. 14 Pressure coefficient distribution on the blade

(original geometry) of the second stator (St2)

Fig. 15 Pressure coefficient distribution on the blade

(modified geometry) of the second stator (St2_mod)

PS

1.8

SS

1.6

PS

1.4

PS

Cp [-]

1.2

PS

SS

1
0.8

SS

PS

0.6

SS

R1
St1_1
St1_1_mod
St1_2

SS

0.4

R2

0.2

St2
St2_mod

0
0

l/D2 [-]

Fig. 16 Pressure coefficient distribution along the mean streamline of the entire storage pump hydraulic
passage
Pressure coefficient distribution along the mean streamline for the entire analysed hydraulic passage is
presented in Fig. 16. For the first impeller, the minimum pressure coefficient is above critical value avoiding any
cavitational risk into the storage pump.

5. Conclusion
The paper presents a computing methodology as a first step in the three-dimensional numerical analysis of
the flow into a hydraulic passage of the two stages and double entry storage pump. First, the 3D geometry of the
hydraulic passage (suction elbow, first impeller, first stator with two cascades, second impeller, second stator and
volute) is reconstructed. Consequently, the 3D computational domains for each part are generated. Second, the
3D turbulent steady flow is performed in the hydraulic passage of storage pump using the commercial FLUENT
code. The impeller-stator (between the first impeller and first stator as well as between second impeller and
second stator) and stator-impeller (between the first stator and second impeller) interactions are taken into
account with mixing interface method. The hydrodynamic flow field from the inlet to the outlet is obtained. Thus,

25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

the blade geometries of the first and second stator cascades are modified. As a result, the 3D flow around
modified geometry is computed. The hydrodynamic field for modified geometry is analyzed against original one.
Comparing the results obtained for the two geometries of the stators, it can be seen that they have different
behaviour during the pumps operation, due to the different conditions at the outlet of the two impellers situated
upstream each of the analysed stators. It can be considered that, although the same geometry modification was
realized for both the stators, the effects on the flow parameters are different, only for the second stator being
possible to observe a net improvement of the blade load. The cavitational behaviour was not studied. After this
methodology is validated, an unsteady analysis will be the subject of a further study, in order to properly study
the impeller-stator interaction phenomenon inside the double entry and two stages pump.

Acknowledgments
Dr. Sebastian Muntean was supported by the Romanian Academy program Hydrodynamics Optimization
and Flow Control of the Hydraulic Turbomachinery in order to Improve the Energetic and Cavitational
Performances. Numerical computations have been performed at the Numerical Simulation and Parallel
Computing Laboratory from the Politehnica University of Timisoara, National Center for Engineering of
Systems with Complex Fluids.

G
V
G
w

Nomenclature
p
cp
cv
g

Absolute velocity [m/s]

Relative velocity [m/s]
Fluid density [kg/m3]
Pump suction head [m]

Static pressure [Pa]

Pressure coefficient [-]
Absolute velocity coefficient [m/s]
Gravity [m/s2]

References
[1]

Arndt N, Acosta A J, Brennen C E and Caughey T K 1990 Experimental investigation of rotor-stator

interaction in a centrifugal pump with several vaned diffusers J. of Turbomachinery 112(1) 98 108
[2] Arndt N, Acosta A J, Brennen C E and Caughey T K 1989 Rotor-stator interaction in a diffuser pump J. of
Turbomachinery 111 213 -221
[3] Brennen C E 1994 Hydrodynamics of Pumps (Oxford University Press)
[4] Qin W and Tsukamoto H 1997 Theoretical Study of Pressure Fluctuations Downstream of a Diffuser
Pump Impeller J of Fluid Engineering 119 647
[5] Barrio R, Blanco E, Parrondo J, Gonzlez J and Fernndez J 2008 The Effect of Impeller Cutback on the
Fluid-Dynamic Pulsations and Load at the Blade-Passing Frequency in a Centrifugal Pump J of Fluids
Engineering 130 111102
[6] Marx Y P 1997 Unsteady Simulations of Impeller - Diffuser Interactions in a Centrifugal Pump EPFLSCR No 9.
[7] Stuparu A, Muntean S, Balint D, Anton L E and Baya A 2007 Numerical Investigation of a Storage Pump at
Constant Speed Proc. of the 3rd Romanian German Workshop on Turbomachinery Hydrodynamics
(Timisoara, Romania) pp 73-84
[8] Stuparu A, Muntean S, Baya A and Anton L E 2008 3D Numerical Investigation of Flow through the
Centrifugal Pump with Double Flux Proc. of the 3rd Workshop on Vortex Dominated Flows
(Timisoara, Romania) pp 75-80
[9] Spence R Amaral-Teixeira J 2008 Investigation into pressure pulsations in a centrifugal pump using
numerical methods supported by industrial tests Computers & Fluids 37 690704
[10] Dunca G, Isboiu E C, Clinoiu C, Bucur D M and Ghergu C M 2008 Vibrations level analyse during
pumping station Galceag operation Scientific Bulletin of Politehnica University of Bucuresti Series D
70(4) 181-189
[11] Dunca G, Bucur D M, Isboiu E C and Clinoiu C 2007 Transient behavior analysis. Case study Galceag
pumping station Scientific Bulletin of Politehnica University of Bucuresti Series C 69(4)
[12] Dunca G, Muntean S and Isboiu E C 2009 3D numerical flow analysis into a double stage and double
flux storage pump Proc. of the Intrenational Conf. on Modelling Fluid Flow (CMFF'09) (Budapest,
Hungary) pp 811-817

25th IAHR Symposium on Hydraulic Machinery and Systems

IOP Conf. Series: Earth and Environmental Science 12 (2010) 012016

IOP Publishing
doi:10.1088/1755-1315/12/1/012016

[13] Dunca G, Muntean S and Isboiu E C 2009 3D Numerical Analysis of the Impeller - Stator Interaction
into a Storage Pump The International Conf. on Energy and Environment, CIEM2009
[14] Menter F R 1994 Two-equation eddy-viscosity turbulence models for engineering applications AIAA
Journal 32(8) 269 289
[15] Menter F R, Kuntz M and Langtry R 2003 Ten Years of industrial Experience with the SST Turbulence
Model Proc. of the Turbulence, Heat and Mass Transfer 4 625-632

10