Engineering Applications of Computational Fluid

Mechanics

ISSN: (Print) (Online) Journal homepage: https://www.tandfonline.com/loi/tcfm20

Influence of blade inlet angle on the performance

of a single blade centrifugal pump

Jianfang Chen, Weidong Shi & Desheng Zhang

To cite this article: Jianfang Chen, Weidong Shi & Desheng Zhang (2021) Influence of blade
inlet angle on the performance of a single blade centrifugal pump, Engineering Applications of
Computational Fluid Mechanics, 15:1, 462-475, DOI: 10.1080/19942060.2020.1868341

To link to this article: https://doi.org/10.1080/19942060.2020.1868341

© 2021 The Author(s). Published by Informa View supplementary material

UK Limited, trading as Taylor & Francis
Group

Published online: 17 Feb 2021. Submit your article to this journal

Article views: 2060 View related articles

View Crossmark data Citing articles: 8 View citing articles

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=tcfm20
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS
2021, VOL. 15, NO. 1, 462–475
https://doi.org/10.1080/19942060.2020.1868341

Influence of blade inlet angle on the performance of a single blade centrifugal

pump
Jianfang Chena,b , Weidong Shia,c and Desheng Zhanga
a Research Center of Fluid Machinery Engineering and Technology, Jiangsu University, Zhenjiang, People’s Republic of China; b School of
Petroleum Engineering, Changzhou University, Changzhou, People’s Republic of China; c School of Mechanical Engineering, Nantong
University, Nantong, People’s Republic of China

ABSTRACT ARTICLE HISTORY

Single blade centrifugal pump is suitable for transporting sewage water with large obstacles in diam- Received 10 September 2020
eter, which is required to achieve high performance. The blade inlet angle is a parameter that affects Accepted 15 December 2020
the performance of the single blade centrifugal pump. However, the selection of blade inlet angle is KEYWORDS
frequently based on experiences and demonstrates certain arbitrariness. Therefore, six impellers with Single blade centrifugal
different blade inlet angles are designed using ANSYS Bladegen and NX software. These impellers pump; blade inlet angle;
are categorized into Groups 1 and 2. The blade inlet angle of the impellers in Group1/Group2 numerical calculation;
increases/decreases from the front shroud to the rear one, respectively. Subsequently, the numerical external characteristics
simulation is carried out to investigate the effect of the blade inlet angle on the performance of the
single blade pump. The blade inlet angle has more obvious influence on the head, efficiency, and
shaft power for the impellers in Group 1 than those in Group 2. The lowest pressure is negative for
one of the impellers in Group 1. Moreover, an optimum blade inlet angle improves the pump perfor-
mance. In addition, the alternating load is small on the blades of the impellers in Group 2. The present
results are helpful in designing the single blade centrifugal pump and improving its performance.

1. Introduction
experimental, and numerical methods. Nishi et al. (2006)
Environmental protection has been paid increasing and Nishi et al. (2009) applied the CFD analysis to pre-
attention with the sustained and rapid economic devel- dict the performance of single blade sewage pump to
opment, together with the continuous advancement reduce radial force fluctuation components and proposed
of urbanization. Building and improving various types a method for designing the impeller of single blade cen-
of sewage treatment facilities is necessary. Thus, the trifugal pumps. Benra et al. (2005) investigated the effect
research on sewage pumps becomes very urgent. For of the impeller position on flow parameters through
example, the single blade centrifugal pump is investigated numerical and experimental methods. Moreover, De
as a good case. This type of pump is a special category Souza et al. (2008) and De Souza et al. (2010) optimized
of centrifugal pumps. It has a single flow channel that is the single blade impeller through design of experiments
formed by a large wrap angle envelope (Tan et al., 2017). and computational fluid dynamics (CFD), proposed a
Given that the single blade pump only has one blade, it novel volute design method with the constant speed
provides a large channel cross section for the fluid flow method, and applied the proposed method to a single
(Michael et al., 2009). Thus, this type of pump is com- blade impeller. Benra and Dohmen (2008) investigated
monly used as the sewage pump to transport substances the radial force of a single blade pump through experi-
that contain solids to avoid clogging disturbance. How- ments and CFD software. Subsequently, they employed
ever, the structure with low blade number faces great numerical and PIV schemes to verify the reliability of a
challenges, such as low efficiency, pressure fluctuation, Navier–Stokes solver that predicts the transient flow in a
and hydrodynamic unbalance (Aoki, 1984; Pei, Yuan, & single blade pump. Pei, Dohmen & Yuan (2012a) studied
Dohmen, 2012b). the vibration and periodic unstable flow that originated
To improve the sewage pump performance, many from the single blade pump under non-design conditions
researchers have performed studies using theoretical, by using numerical and experimental methods. Pei et al.

CONTACT Weidong Shi wdshi@ntu.edu.cn

Supplemental data for this article can be accessed here. https://doi.org/10.1080/19942060.2020.1868341
© 2021 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use,
distribution, and reproduction in any medium, provided the original work is properly cited.
 ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 463

(2013) investigated the periodic unsteadiness in the rotor

and volute domains for the normal operation quantita-
tively. Tan et al. (2018) obtained the characteristics of the
fluid force that induced the vibration of a single blade
pump under different operating conditions. Wang, Long
& Yang (2020) and Wang, Long & Wang (2020) stud-
ied the influence of impeller blade structure on pump
performance.
The aforementioned review indicates that most of the
studies are focused on design methods and analyzing the
unsteady and turbulence flows, and only few research has
carried out investigations on the impact of the blade inlet
angle on the performance of the single blade centrifugal Figure 1. View of the single blade centrifugal pump.
pump. As a matter of fact, the blade inlet angle affects
the degree of blade bending, which further influences the
blade area of the inlet flow. Meanwhile, the position of Table 1. Single blade centrifugal pump specifications.
the stagnation point and the appearance of cavitation are Parameters Value Parameters Value
affected (Guan, 2011). Therefore, investigating the influ- Head (H) 11 m Impeller inlet 45 mm
ence of the blade inlet angle on the pump performance is diameter (D1 )
Rated flow rate (Qd ) 20 m3 /h Impeller outlet 125 mm
very necessary. diameter (D2 )
For the present study, the 3D modeling software Rotating speed (n) 2940 rpm Impeller outlet width 31 mm
(b2 )
is used to determine the blade inlet angle, which has Volute inlet diameter 135 mm Blade wrap angle (ϕ) 360°
not been conducted before. Six impellers with differ- (D3 )
Volute inlet width 46 mm Blade outlet angle 18°
ent blade inlet angles are designed to reveal the impact (b3 ) (β 2 )
of the blade inlet angle on the behavior of the sin- Specific speed (ns ) 132 Blade inlet angle (β 1 ) To be determined
gle blade centrifugal pump. Then, the CFD simula-
tion is carried out so that the performance of the six
impellers is investigated. We hope to obtain some valu- streamlines divide the blade axial projection into four
able results to evaluate the relationship between the equal parts. For the present study, six impellers with dif-
blade inlet angle and the single blade centrifugal pump ferent blade inlet angles are designed. The blade inlet
performance. angle β 1 , which is the inlet angle at the blade suction side,
is shown in Figure 2(b). Moreover, the corresponding
inlet angles are presented in Table 2, where six impellers
2. Computational model are divided into two groups.
In Group 1, the streamline inlet angle of each impeller
2.1. Studied pump
increases 3° from s1 to s5 , whereas the streamline inlet
Figure 1 shows the single blade centrifugal pump (the angle of each impeller in Group 2 decreases 3°. Designed
front chamber is removed) investigated in the present impellers are called impellers 11, 12, 13, 21, 22, and 23.
study. The volute is stationary, and the impeller rotates at The inlet angles for the same streamlines of impellers
a rotating speed of 2940 rpm. The single blade centrifugal 11, 12, and 13 differ sequentially by 3°, whereas those of
main specifications are listed in Table 1. For the present impellers 21, 22, and 23 are the same. The blade inlet
computation, the water at 25?? is selected as the operating angles at different streamlines for the six impellers are
fluid. listed in Table 2.
According to design parameters, six impellers with
2.2. Blade inlet angle and impeller model different blade inlet angles and the same other parameters
are designed with the Ansys Bladegen and NX software.
When the impeller model is established in accordance An isometric view of one of the impellers is shown in
with the geometric parameters, the profile of the blade Figure 3. The present blade is a twisted one.
leading edge is not a complete surface for the inlet angle Figure 4(a) shows the evolution of blade cross section
higher than 30°. Therefore, the maximum blade inlet for different streamlines. Moreover, the blade thickness
angle is set to 30°. gradually increases from streamline s1 to s5 regardless of
Moreover, Figure 2 shows the axial and plan projec- the value of the streamline inlet angle. Figure 4(b) shows
tion of the single blade impeller. Hence, the three axial the shapes of the blade leading edges for different blade
464 J. CHEN ET AL.

Figure 2. Schematic presentation of the impeller. (a) Axial projection (b) Plan projection (removed front shroud).

Table 2. Blade inlet angles of six impellers. be a steady state to simplify the study. Note that CFX
Streamline is a coupled solver, and governing equations are solved
Impeller s1 s2 s3 s4 s5 Remarks
simultaneously.
Impeller 11 12° 15° 18° 21° 24°
Impeller 12 15° 18° 21° 24° 27° Group1
Impeller 13 18° 21° 24° 27° 30° 3.1. Numerical model
Impeller 21 24° 21° 18° 15° 12°
Impeller 22 27° 24° 21° 18° 15° Group2 Figure 5 indicates that the computational domain
Impeller 23 30° 27° 24° 21° 18°
includes the impeller inlet extension, front chamber,
impeller, rear chamber, volute, and volute outlet exten-
inlet angles. When the inlet angle studied in this paper sion. Studies (Peng et al., 2020; Zhou et al., 2020) show
is relatively small, the pressure side of the blade leading that extending the impeller inlet and volute outlet in
edge is symmetrical to its suction side. However, with the length can improve the stability of inlet and outlet fluid
increase in angle, the blade leading edge gradually shows flow, respectively.
asymmetry.

3. Numerical analyses 3.2. Grid generation

Computational fluid dynamics (CFD) is a modern sim- For the present study, the ICEM CFD software is applied
ulation technology that is widely used in various fields to generate the unstructured grids for impeller and volute
(Abadi et al., 2020; Ghalandari et al., 2019; Salih et al., regions, whereas structured grids are used for other
2019). For the present investigation, CFD software regions. The head and efficiency under the design con-
ANSYS CFX is used to simulate the flow in the single dition are studied as the criteria to perform the grid
blade centrifugal pump. Here, the flow is assumed to independency analysis for impeller 12. Figure 6 shows

Figure 3. Isometric view of a designed single blade impeller. (a) Impeller (b) Blade.
 ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 465

Figure 4. Blade shapes in different cases. (a) Shapes of cross section of blade, (b) Shapes of blade leading edges.

the calculated results. As the grid number increases, the

stability and efficiency of the head also rise. However,
when the grid number is more than 2.5×106 , both of
them vary by less than 0.5%. Thus, based on the results
of the grid independency test, the number of grids for
all cases is approximately 2.84×106 . y+ near the walls of
the impeller and volute are approximately 80. Figure 7
illustrates the generated grids for the impeller, the front
chamber, and the rear chamber.

3.3. Turbulence model

In this steady state case, the Reynolds Average Navier–
Stokes equations are used to express the fluid flow in
Figure 5. Computational domain.
the pump. The turbulence model should be selected to
solve the RANS equations. The standard k–ε, RNG k–ε,
466 J. CHEN ET AL.

the wall surface of the impeller. The wall surface rough-

ness of the volute is defined as 50 μm. The predefined total
pressure and the mass flow rate are selected as the inlet
and outlet boundary conditions, respectively. The inlet
pressure is set to a standard atmospheric pressure, and
the inlet turbulence intensity is considered 5%. A region
between rotational and stationary ones is connected by
a frozen rotor interface, whereas two stationary regions
are connected by a general connection interface. To opti-
mize the calculation time and accuracy, the calculation
stops when the convergence residual is less than 10−4 .
In addition, models, such as LES, DES and SAS, belong
Figure 6. Grid independency. to the scale resolving turbulence model. The charac-
teristics of these models have very high demands for
computational resource and cost. Therefore, they are not
very suitable for the real engineering question because
standard k–ω, and SST k–ω are four commonly used tur- of a huge computational cost. Limited to the finite com-
bulence models. To select a reasonable turbulence model, putational resource, combining the above analyses, the
the calculated results for impeller 12 under the design standard k–ε is chosen here. In addition, although the
condition for the four turbulence models are listed in fluid flow in a single blade centrifugal pump may be
Table 3. As shown in Table 3, the standard k–ε turbu- unsteady, the present computation shows that some valu-
lence model can obtain a result similar to the experimen- able results can be obtained by means of the present
tal one (the related introduction can be seen in section condition. Therefore, unsteady computation, unsteady
5), and the computational time is the shortest among boundary conditions, and the scale resolving turbu-
the other models. Hence, in this research, the standard lence model are not applied here because of complex
k–ε turbulence model is used to discriminate the turbu- computations.
lence flow. The equations are discretized by finite volume
technique. High resolution and first order are adopted
for the advection scheme and the turbulence numerics,
respectively. 4. Results and discussions
4.1. Analyses of the external characteristics
The numerical simulations of the six impellers with dif-
3.4. Boundary conditions
ferent blade inlet angles have been carried out in the
Scalable wall function is applied to the near wall region. foregoing section. The obtained heads of the six impellers
Moreover, the no-slip boundary condition is imposed on are presented in Table 4. The head decreases as the flow

Figure 7. Computational domain and grids of the impeller and chamber.

Table 3. Comparison for different turbulence models.

Head (m) Efficiency (%) Relative error in Head (%) Relative error in Efficiency (%) Consuming time (minutes)
Experimental results 13.11 57.0
Standard k-ε 13.87 59.6 5.50 4.56 111
RNG k-ε 13.87 58.9 5.50 3.33 209
Standard k-ω 14.04 60.8 7.09 6.67 410
SST k-ω 13.76 59.7 4.96 4.73 217
 ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 467

rate increases. At each operating point, no fixed mean- Table 5. Obtained efficiencies for different impellers.
ingful relationship is observed between the head and Q (m3 /h) 12 16 20 24 28
the blade inlet angle. When the flow rate is 16, 20, and Impeller 11 50.2 56.0 60.5 61.5 60.1
28 m3 /h, impeller 11 in Group 1 has the highest head Impeller 12 49.5 55.0 61.0 60.2 59.5
Impeller 13 50.2 57.3 58.5 62.3 59.3
whereas the head of impeller 13 is the lowest. Further- Impeller 21 50.5 55.2 60.8 62.8 59.7
more, impeller 22 in Group 2 has the highest head, except Impeller 22 50.1 55.3 61.3 62.5 60.0
for the flow rate of 24m3 /h. Table 4 lists the difference Impeller 23 50.1 55.5 60.8 62.3 59.4

between the maximum and the minimum heads of the

three impellers in Groups 1 and 2 at each operating point,
impeller 13 is the lowest. The difference between them is
where the maximum difference is 0.37 m. The influence
2.8%. At the design condition, the impellers in Group 2
of the blade inlet angle on the head of Group 1 is more
can improve the pump efficiency well.
obvious than that of Group 2. Further analyses show
Figure 8 illustrates the histogram of the shaft power
that impeller 21 has the lowest head at low flow rates,
and the power loss for different impellers. The variation
whereas that of impeller 23 is the smallest at high flow
of the shaft power with the flow rate of six impellers is
rates. Moreover, the produced head of impeller 21 is
similar in the whole operating range. The shaft power
the same as that of impeller 23 when the pump works
increases with the flow rate. When the blade inlet angle
at the design condition. The comparison of the heads
varies, the shaft powers of the three impellers in Group
between impellers 11 and 21, impellers 12 and 22, and
1 are significantly affected. Notably, the consumed shaft
impellers 13 and 23 at different working conditions shows
power of impellers11, 12, and 13 at the design condi-
that the maximum head difference is 0.27, 0.32, and
tion are higher than those of impellers 21, 22, and 23,
0.18 m, respectively. Furthermore, the produced heads of
respectively. The corresponding difference between them
impellers 11, 22, and 23 are higher for most of work-
is 24, 9, and 33 W, respectively. Moreover, the power
ing conditions. In other words, when the inlet angle is
loss against flow rate has almost a parabolic distribution.
small, the impeller head in Group 1 is high. Moreover,
When the inlet angle changes, the power losses of the
as the inlet angle increases, the head of the impeller in
impellers in Group 1 change dramatically, whereas those
Group 2 rises.
of Group 2 change relatively gently. At the design condi-
Table 5 shows the efficiency of each impeller. The sin-
tion, the power loss of the impeller in Group 2 is higher
gle blade pump cannot operate at low flow rates because
than that of Group 1. Power losses of impellers 11, 12 and
the efficiency abruptly decreases with the flow rate. The
13 are 15, 12, and 43 W higher than those of impellers
influence of the blade inlet angle on the efficiency of
21, 22, and 23, respectively. This finding is consistent
Group 1 is more obvious than that of Group 2. When
with the results of the efficiency analysis in the foregoing
the flow rate is higher than 20 m3 /h, the efficiency for the
section.
impeller in Group 2 is higher than those of Group 1. The
The relationship between the external characteristics
range with a high efficiency of the impeller in Group 2 is
of a single blade centrifugal pump and its blade inlet angle
wider than that of Group 1. At each operating point, the
difference among impeller efficiencies is small. Therefore,
the blade inlet angle has a small impact on the efficiency
of the single blade centrifugal pump. The comparison of
the efficiencies between impellers 11 and 21, impellers
12 and 22, and impellers 13 and 23 indicate that the lat-
ter impeller efficiency is higher than that of the former
impeller at the design condition, respectively. Moreover,
the efficiency impeller 22 is the highest, whereas that of

Table 4. Obtained heads for different impellers.

Q (m3 /h) 12 16 20 24 28
Impeller 11 16.33 15.18 13.92 12.18 10.96
Impeller 12 16.69 15.00 13.87 12.08 10.89
Impeller 13 16.52 14.96 13.55 12.30 10.70
Hmax -Hmin 0.36 0.22 0.37 0.22 0.26
Impeller 21 16.32 14.95 13.73 12.45 10.94
Impeller 22 16.70 15.02 13.75 12.40 10.98
Impeller 23 16.58 14.97 13.73 12.35 10.78 Figure 8. Histogram of shaft power and power loss for six
Hmax -Hmin 0.38 0.07 0.02 0.1 0.2 impellers.
468 J. CHEN ET AL.

may be interpreted as follows. When the blade inlet angle lowest efficiency under the design condition. Figure 10
increases, the blade bending is reduced, thereby expand- shows that wp suddenly decreases at about 160°, and the
ing the blade inlet flow area. Therefore, pump head and region from 0° to 160° is where fluid separation occurs
efficiency are improved. On the other hand, given that on the blade pressure surface in Figure 9. Further anal-
a single blade centrifugal pump has only one blade, the ysis shows that wp of impellers in Group 2 is larger than
capability of the pump on the fluid delivery is weakened, those of in Group 1 at 0°. Moreover, wp of the impellers
and thus pump head and efficiency decreases. The ulti- in the two groups approximately decreases to the same
mate increase or decrease in pump head and efficiency value at the same position (80°). That is, the velocity gra-
depends on the strength of the foregoing two condi- dient of the impellers in Group 2 is larger than that of in
tions. First, the blade streamline inlet angles of impellers Group 1. Therefore, as shown in Figure 9, the flow sep-
11, 12, and 13 increase from the front shroud to the aration of impellers in Group 2 at the front of the blade
rear one, thereby causing the impeller channel to dif- pressure surface is more serious than that of impellers in
fuse and generate vortices. Thus, the efficiency of the Group 1.
three impellers is slightly lower than that of impellers In Figure 10, very high velocities are generated on the
21, 22, and 23, respectively. This finding can be con- suction surface of the blade leading edge because of the
firmed in Figure 10(a’,b’). That is, the fluctuation of the flow around the blade leading edge. ws is the relative
relative velocity on the streamline of the pressure sur- velocity on the streamline of the blade suction surface.
face of impellers in Group 1 is more serious than that of In the range of 0° to 20°, ws of impeller 21 is smaller than
Group 2. that of impeller 11. Thus, the pressure of the former is
higher than that of in the latter (Figure 9). The same is
true for impellers 23 and 13. Similarly, for impellers 12
4.2. Flow field analyses
and 22, the difference of ws between them is very small;
Figure 9 depicts the relative velocity contour distribution hence, the pressure difference is also very minimal. Fur-
for six impellers at mid-span under the design condition thermore, because of the thickness of the impeller blade
(Qd = 20 m3 /h). The relative velocity distributions of the trailing edge, the wake flow is observed at the blade out-
six impellers are similar. Low velocity exists at the par- let. The wake at the blade outlet is related to the velocity
tial blade pressure/suction surface, the blade inlet, and difference between the blade pressure surface and the
near the tongue, and it reaches the maximum at the blade suction surface here. The greater the velocity difference
outlet. For the six impellers, a wake flow at the blade is, the more significant the wake is. In Group 1, impeller
outlet. 11 has the smallest velocity difference, and impeller 22
Figure 10 shows the relative velocity distribution is the smallest one in Group 2. In summary, the relative
curves of the blade pressure surface and suction surface velocity distributions of impellers 11 and 22 are the most
streamlines at mid-span under the design condition. In uniform in Groups 1 and 2, respectively.
Figure 10, the circumferential angle is plotted on the Figure 11 shows the pressure distribution of six
abscissa. Considering that the blade wrap angle is 360°, impellers at mid-span under the design condition.
the blade inlet is set to 0°, and the blade outlet is set According to Figures 9 and 10, the pressure distribution
to 360°. Figure 10(a) shows the changes in the relative of the impeller corresponds to its velocity distribution.
velocity from 0° to 11.25° for the middle streamline on For instance, when the relative velocity is high, the pres-
pressure surface and suction surface of the impellers in sure is low, and vice versa. For the impellers in Group
Group 1. Figure 10(b) presents the impellers in Group 1, the high-pressure range decreases with the increase
2. The range from 0° to 11.25° is exactly around the in blade inlet angle. Moreover, the low pressure at the
blade leading edge. Figure 10(a’,b’) shows the changes of blade inlet first increases and then decreases as the blade
the two above variables from 11.25° to 360°. After the inlet angle rises. Although the low-pressure region in
fluid enters into the impeller, it circumvents the blade impeller 13 decreases, negative pressure appears at the
leading edge. After a short distance, the relative veloc- blade inlet. For clarification, the negative pressures are
ity on the streamline of the blade pressure surface (wp ) lower than the water vapor pressure (3.17kPa) at the stan-
reaches the maximum value (the relative velocity can be dard atmospheric condition. Therefore, cavitation and
considered to reach its maximum value at 0°) and then corrosion easily occur at the suction surface near blade
decreases to the minimum value at approximately 80°. inlet. Meanwhile, for the impellers in Group 2, the mini-
Subsequently, although the relative velocity fluctuates, it mum pressure first decreases and then increases, and no
generally increases until the impeller outlet is reached. negative pressure occurs. Compared with impeller 13, the
wp of impeller 13 fluctuates the most seriously. This phe- high-pressure range in impeller 23 rises. More specifi-
nomenon may be the reason why impeller 13 has the cally, the minimum pressure increases from −1.08 kPa to
 ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 469

Figure 9. Relative velocity contour distributions for six impellers under the design condition.

9.83 kPa. Similarly, for impellers 11 and 21, the minimum 2, the possibility of cavitation is low at the lowest pressure
pressure value increases from 4.02–10.2 kPa, whereas the area.
minimum pressure value has a minimal difference for Figure 12 shows the positions of points O1–O8 at the
impellers 12 and 22. Therefore, for the impellers in Group impeller outlet. Points O1–O8 divide the impeller outlet
470 J. CHEN ET AL.

Figure 10. Relative velocity distribution curves of the middle streamline on the pressure/suction surface

into eight equal parts. The circumferential angle of point of impeller 12 is the minimum, whereas that of impeller
O1 is set to 0°. Consistent with the direction of rotation 13 is the maximum. Moreover, the pressure standard
of the impeller, the circumferential angle of point O2 is deviation of impellers 21, 22, and 23 are 3.75, 3.79, and
set to 45°. Therefore, the circumferential angles of points 3.69 kP, respectively. Meanwhile, the difference in their
O3–O8 increase by 45° in turn. mixing collision losses is negligible. The mixed colli-
The pressure circumferential distribution of the six sion loss of impellers in Group 2 is lower than those of
impellers at the impeller outlet is presented in Figure 13. Group 1.
The pressure distributions of all impellers along the cir- Figure 14 shows the relative velocity distribution at the
cumference are extremely non-uniform. Moreover, min- impeller outlet. The relative velocity distributions of the
imum and peak pressure values are found downstream six impellers are not remarkably uniform. As the non-
and near the tongue, respectively. The non-uniformity uniformity degree increases, the flow becomes more dis-
of the impeller outlet pressure represents the mixing organized at the outlet, and the energy exchange becomes
loss that occurred here. As the non-uniformity degree more intense. Furthermore, the greater the relative veloc-
increases, the outlet flow becomes more disordered, and ity is, the lower the converted energy will be. The stan-
the loss of the mixing collision between the fluid in volute dard deviations of the relative velocity of impellers 11,
and the impeller outflow rises. The standard deviation of 12, and 13 are 0.77, 0.78, and 0.88 m/s, respectively. On
pressure is used to evaluate the corresponding fluctua- the other hand, the average relative velocities of the three
tions. The pressure standard deviations of impellers 11, impellers are 11.18, 11.40, and 13.25 m/s, respectively.
12, and 13 are 4.41, 4.23, and 6.02 kP, respectively. These Therefore, for the three impellers in Group 1, the adop-
standard deviations indicate that the mixing collision loss tion of a smaller blade inlet angle is beneficial to reduce
 ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 471

Figure 11. Pressure distributions in six impellers under the design condition.

the non-uniformity of the fluid flow at the impeller out- impellers in Group 2 is lower than that of Group 1, and
let, thereby improving the effect that impellers trans- the corresponding energy conversion degree of the for-
form the fluid kinetic energy into pressure energy. The mer is higher than the latter. Meanwhile, the sudden
standard deviations of the relative velocity of impellers expansion loss from the impeller to the volute can be
21, 22, and 23 are 0.76, 0.89, and 1.00 m/s, respectively. reduced, whereas the relative velocity non-uniformity of
The average relative velocities of the three impellers impellers in Group 2 is slightly higher than that of Group
are 11.19, 11.07, and 11.26 m/s. The relative velocity of 1. For the two groups, the relative velocity fluctuates
472 J. CHEN ET AL.

more seriously at the impeller outlet as the blade inlet

angle increases. Obviously, the outlet relative velocity
of impeller 13 is much greater than those of the other
impellers.
Figure 15 shows the blade load distribution for six
impellers on the streamline at mid-span. The blade sur-
face load refers to the pressure difference between the
pressure surface and the suction surface at the same
radius for the same blade. For Group 1, the blade load
of impeller 11 changes relatively flat, whereas that of
impeller 13 changes abruptly at approximately 200°. Flow
separation is the reason for the sudden change in blade
load. For Group 2, the blade load distribution of the
three impellers is very similar, and no sudden change is
Figure 12. Corresponding point to the circumferential angle of
the impeller outlet. observed in the blade load. For the two groups, Figure
12 shows that the blade load distribution of the impellers
in Group 2 is flatter than those of Group 1. Therefore, the

Figure 13. Pressure distribution along the circumferential direction. (a) Impeller 11, 12 and 13 (b) Impeller 21, 22 and 23.

Figure 14. Relative velocity distribution along the circumferential direction. (a)Impellers 11, 12 and 13 (b) Impellers 21, 22 and 23.
 ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 473

Figure 15. Blade load distribution on the middle streamline.(a) Impellers 11, 12 and 13 (b) Impellers 21, 22 and 23.

internal flow of the impellers in Group 2 is more uniform,

and the alternating load is small on the blades, which does
not cause fatigue failure easily.

5. Experimental validation
An open pump test rig is set up in the Jiangsu University
(China) to verify the accuracy of the calculation (Figure
15). The WT2000 pressure transmitters are applied to
measure the inlet and outlet pressures with the accuracy
of ±0.1%. Moreover, the LWGY-50A turbine flow meter
is applied to obtain the amount of fluid through the pump
with the precision of ±0.5%. Figure 17. Performance curves of the model pump.
Impeller 12 is selected as the test impeller. The perfor-
mance curves of the model pump are shown in Figure 16.
The efficiencies of the numerical computation are higher similar trends. Compared with the experimental data,
than those for the experiment. The difference between the maximum relative errors of the pump head is 6.6%,
the numerical and experimental results may come from except for the flow rate of 28m3 /h, and the maximum rel-
the ignored various losses in the numerical simulation. ative errors of efficiency is 7.8%. Therefore, the numerical
However, the numerical and experimental results show results have an acceptable accuracy (Figure 17).

6. Conclusion
The relationship between the blade inlet angle and the
behavior of the single blade centrifugal pump is inves-
tigated on the basis of CFD simulation and experiment
results. The following conclusions are drawn on the basis
of the aforementioned analyses:

(1) In comparison with the three impellers of Group 2,

the head and shaft power of the three impellers in
Group 1 are more sensitive to the blade inlet angle.
The distribution of the efficiency curve versus the
Figure 16. Open pump test rig. flow rate for Group 2 impellers is almost flat. Blade
474 J. CHEN ET AL.

inlet angle has an evident influence on the pressure Fluids Engineering Division Summer Meeting. Houston,USA,
and velocity distribution. The present results show 1, 1223–1231.
that impeller 11 in Group 1 and impeller 22 in Group De Souza, B., Niven, A., & Daly, J. (2008). Single-blade
impeller development using the design of experiments
2 have the best performance. method in combination with numerical simulation. Pro-
(2) The blade inlet angle has a great influence on the ceedings of the Institution of Mechanical Engineers, Part E:
first quarter of the blade pressure surface and the Journal of Process Mechanical Engineering, 222(3), 135–142.
fluid recirculation at the blade inlet. For the three https://doi.org/10.1243/09544089JPME218
impellers in Group 1, the relative velocity on the last De Souza, B., Niven, A., & Mcevoy, R. (2010). A numer-
ical investigation of the constant- velocity volute design
of streamline of the blade pressure surface fluctuates
approach as applied to the single blade impeller pump. Jour-
seriously. Moreover, the alternating load is large on nal of Fluids Engineering, 132(6), 1–7. https://doi.org/10.11
the blades of the three impellers in Group 1. 15/1.4001773
(3) The analysis of the distributions of pressure and Ghalandari, M., et al. (2019). Investigation of Submerged Struc-
velocity suggests that the energy conversion ability of tures’ Flexibility on Sloshing Frequency using a bound-
impellers in Group 2 are higher than that of Group ary element method and finite element analysis. Engineer-
ing Applications of Computational Fluid Mechanics, 13(1),
1. For the studied impellers, the use of an impeller 519–528. https://doi.org/10.1080/19942060.2019.1619197
with a decreasing inlet angle from the front shroud Guan, X. F. (2011). Theory and design of modern pump. China
to the rear one is recommended. Astronautic Publishing House.
(4) Considering that the present blade is a twisted blade, Michael, H., Kuhimann, J., & Benra, F. K. (2009). Designing
the blade manufacturing accuracy is not in agree- high-power sewage water pumps. World Pumps, 514, 20–25.
Nishi, Y., Fujiwara, R., & Fukutomi, J. (2009). Design method
ment with the expected one. Therefore, the error
for single blade centrifugal pump impeller. Journal of Fluid
between the calculated and the experimental results Science and Technology, 4(3), 786–800. https://doi.org/10.
seems to be slightly huge. The inner flow mecha- 1299/jfst.4.786
nism, the appearance of radial force, and the noise Nishi, Y., Makita, K., & Fukutomi, J. (2006). A study on a
of the single blade centrifugal pump are promising new type of sewage pump. Nippon Kikai Gakkai Ronbun-
research directions in the future. shu B Hen(Transactions of the Japan Society of Mechanical
Engineers Part B (Japan), 72(720), 1984–1992.
Pei, J., Dohmen, H. J., & Yuan, S. Q. (2012a). Investigation
of unsteady flow-induced impeller oscillations of a single-
Disclosure statement blade pump under off-design conditions. Journal of Fluids
No potential conflict of interest was reported by the authors. and Structures, 35, 89–104. https://doi.org/10.1016/j.jfluidstr
ucts.2012.08.005
Pei, J., Yuan, S. Q., & Dohmen, H. J. (2012b). Numerical predic-
tion of unsteady pressure field within the whole flow passage
Funding of a radial single-blade pump. Journal of Fluids Engineering,
This work was supported by the National Natural Science 134(10), 1–11. https://doi.org 10.1115/1.4007382
Foundation of China [Nos 51979138 and 51776087]. Pei, J., Yuan, S. Q., & Yuan, J. P. (2013). Numerical analysis
of periodic flow unsteadiness in a single-blade centrifugal
pump. Science China Technological Sciences, 56(1), 212–221.
https://doi.org/10.1007/s 11431-012-544-x
References Peng, G., Huang, X., Zhou, L., Zhou, G., & Zhou, H. (2020).
Abadi, A. M., Sadi, M., Farzaneh-Gord, M., Ahmadi, M.H., Solid-liquid two-phase flow and wear analysis in a large-
Kumar, R., &Chou, K.W. (2020). A numerical and exper- scale centrifugal slurry pump. Engineering Failure Analy-
imental study on the energy efficiency of a regenerative sis, 114, 104602. https://doi.org/10.1016/j.engfailanal.2020.
heat and mass exchanger utilizing the counter-flow Maisot- 104602
senko cycle. Engineering Applications of Computational Fluid Salih, S. Q., Suliman, M., Rasani, M.R., Ariffin, A.K., &Yaseen,
Mechanics, 14(1), 1–12. https://doi.org/10.1080/19942060. Z.M. (2019). Thin and sharp edges bodies-fluid interac-
2019.1617193 tion simulation using cut-cell immersed boundary method.
Aoki, M. (1984). Instantaneous inter blade pressure distri- Engineering Applications of Computational Fluid Mechan-
butions and fluctuating radial thrust in a single-blade ics, 13(1), 860–877. https://doi.org/10.1080/19942060.2019.
centrifugal pump. Bulletin of JSME, 27(233), 2413–2420. 1652209
https://doi.org/10.1299/jsme1958.27.2413 Tan, L. W., Shi, W. W., & Zhang, D. S. (2018). Experiment on
Benra, F. K., & Dohmen, H. J. (2008). Investigation on the fluid force induced vibration for centrifugal pumps. Transac-
time-variant flow in a single blade centrifugal pump. Pro- tions of the Chinese Society for Agricultural Machinery, 49(5),
ceedings of the 5th WSEAS International Conference on 179–185. https://doi.org/10.6041/j.issn.1000.1298.2018.05.
Fluids Mechanics (Fluids 08), 629–634. 020
Benra, F. K., Dohmen, H. J., & Sommer, M. (2005). Peri- Tan, L. W., Shi, W. D., Zhang, D. S., Zhang, Y., & Zhang,
odically unsteady flow in a single-blade centrifugal pump W. Q. (2017). Effects of blade outlet angle on per-
numerical and experimental results. Proceedings of ASME formance for single blade pumps. Journal of Drainage
 ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS 475

and Irrigation Machinery Engineering, 35(10), 835–855. performance of submersible well pumps. International Jour-
https://doi.org/10.3969/j.issn.1674-8530.16.0235 nal of Simulation Modelling, 19(1), 100–111. https://doi.org/
Wang, H., Long, B., Wang, C., Han, C., & Li, L. (2020). 10.2507/IJSIMM19-1-506
Effects of the impeller blade with a Slot structure on Zhou, L., Han, C., Bai, L., Li, W., El-Emam, M., & Shi,
the centrifugal pump performance. Energies, 13(7), 1628. W. (2020). CFD-DEM bidirectional coupling simulation
https://doi.org/10.3390/en13071628 and experimental investigation of particle ejections and
Wang, H., Long, B., Yang, Y., Xiao, Y., & Wang, C. (2020). energy conversion in a spouted bed. Energy, 118672.
Modelling the influence of inlet angle change on the doi:10.1016/j.energy.2020.118672