1

Analysis of Axial Flow Pump Performance with Impeller

Blade Number and Impeller Diameter Variations

Michael Wijaya, Nazaruddin Sinaga

Departemen Teknik Mesin, Fakultas Teknik, Universitas Diponegoro
Jl. Prof. Sudharto, SH., Tembalang-Semarang 50275
Telp.

Abstract- A pump is a device used to move fluid from a lower method and singularity distribution method. CFD-based
pressure and/or position to a higher position. One example of a optimization design methods have developed rapidly with the
pump is an axial pump. Axial pumps are pumps that work by using rapid development of computer technology. Based on the
the rotational motion of the blades and have a flow direction numerical calculation results of the axial flow pump, adjust the
parallel to the axis of the shaft. Along with the development of the geometry parameters of the axial flow pump blades, so that the
times, methods to improve pump performance are quite a concern, flow pattern in the pump is better and the whirlpool, reflux and
especially to increase efficiency. One of them is the change in the secondary flow and other unstable flows can be avoided [Shi,
geometric aspect. Modification of the axial pump geometry 2016].
parameters can also change the performance of the pump itself, for
example making its efficiency higher. b. Literature Review
Kan Kan, et. al. discusses discusses the study of the large
Keywords- Pump; Axial; Geometry Variation; Performance; eddy simulation (LES) of the detailed flow structure around a
Efficiency rotating axial flow pump. It is shown that among the five different
flow rate working conditions, the integral turbulent kinetic energy
reaches its minimum value under the specified flow rate working
I. INTRODUCTION conditions. The larger the outflow from the designed flow rate
a. Introduction working conditions, the higher the total EPE.
Axial flow pumps are a common type of pump which Shi Lijian, et. al. study of backflow and disc friction loss
basically consists of a vane (axial impeller) in a pipe. Axial flow are the main reasons for the low efficiency of axial and tubular
pumps are most commonly used in industrial applications where flow pumps. Where the gap backflow is mainly affected by the
large flow rates are handled at lower pump head values. The blades impeller inlet and outlet pressures. In this study, eddy and bias
of the axial flow pump are mounted radially to the central hub flow phenomena were observed at the blade entrance of a full
portion of the impeller through which the fluid flows in the axial tubular pump.
direction. A casing such as a pipe is used in this pump, not a volute Zhang, et. al. discusses analytical and experimental
casing [Soe, 2015]. Axial flow pump which is a kind of high methods for unsteady flow that the time domain spectrum shows
specific speed pump with large flow and low head characteristics, that the static pressure fluctuation curves at the rotor inlet, outlet
widely used in agricultural irrigation, flood control and drainage, and stator outlet are periodic, and the amplitude of the pressure
municipal water supply, water environment, water cycle and water fluctuations increases from hub to end at the rotor inlet and outlet,
jet propulsion power generation [Yang, 2018] . but decreases at the shop stator. The amplitude of the pressure
With the increasing application of axial flow pumps, fluctuation is the largest at the rotor inlet, and the average
improving their efficiency continues to become increasingly amplitude decreases sharply from inlet to outlet.
important. One of the most useful methods for increasing its Yang, Shunseng, et. al. discussed that the pump BEP
efficiency is the installation of a guidevane behind the pump characteristics predicted by theoretical and CFD methods were
impeller. The guide vanes can expand the pump head by more accurate than the other two methods such as experimental
converting the kinetic energy of the rotating flow into pressure and analytical. The small difference between the experimental and
energy [Stepanoff, 1957]. The characteristic of the axial flow numerical results can be attributed to the neglect of the leakage
pump is that it has good hydraulic performance near the design loss through the balancing hole, the mechanical loss caused by the
flow rate; However, if the flow rate deviates, the hydraulic mechanical seal and bearing, as well as the surface roughness
performance of the axial flow pump will decrease [Cao, 2020]. values assigned to the machined surface.
The blade is the most important part of the axial flow Giosio, D.R., et. al. found that utilizing the pump
pump impeller, so the quality of the blade design directly impeller as a runner and incorporating inflow control. and found
determines the pump performance. The general methods of that the Unit's best efficiency was 79% and near-peak efficiency
designing axial flow pump blades are lifting method, circular arc under non-design conditions.
 2

Bozorgi A., et. al. discusses the use of an axial pump in consideration of the tip gap. The bigger the tip clearance, the
reverse mode instead of a routine turbine. In this mode, the flow bigger the vortex structure. By comparing the direction and value
and direction of rotation of the pump are reversed, and as a result of the leakage velocity at the tip, it is found that the leakage vortex
the hydraulic behavior of the pump is changed. Also, it was found strength can be suppressed by inhibiting the direction and velocity
that the axial pump can work well as a turbine in the Pico hydro of the leakage flow in the flow section.
power plant. Yabin Liu, et. al. discussed about the tip clearance
Bryan O., Sander Klos discusses the concept of running distance significantly affects the energy performance of axial flow
a centrifugal pump in reverse rotation mode has been recognized pumps. An increase in tip spacing of 1mm causes the pump head
by pump manufacturers for years under conditions of head below to decrease by 1.79 m, and the pump efficiency to decrease by
10 m and high flow rates, so axial flow pumps provide the most 5.49% at the design flow rate. At high flow rates, tip clearance
effective solution for converting turbines. results in a stronger negative impact on energy performance.
Bin Cheng, et. al. studied the numerical simulation of the
layout of the different guide separators analyzed by the Thus have been presented papers that examine topics similar to
momentum distribution of the inlet of the pumping source and the those discussed in this study. However, none of them have
uniformity of the pump flow distribution. The results showed that discussed the geometry variation, specifically the impeller
the guide splitter had a significant adjustment to the diversion diameter parameters and the number of impeller blades on the
channel momentum distribution and the inlet section velocity performance of the axial pump
distribution was influenced by the uniformity of the diversion
channel momentum distribution. The guide separator of suitable c. Problem Identification
size is applicable to reduce the influence of circulating flow and In this study, the study of the problem focused on the
make the flow in a normal distribution quickly after the flow design of the axial pump. On the rear side of the impeller, a
direction shift which can further improve the uniformity of flow guidevane is attached to guide the direction of flow through the
distribution. outlet. In the process of designing this axial pump, the model still
De you Li, et. al. studied to achieve the flow mechanism refers to international standards. However, there are several
during closing of the pump turbine guide vanes in pump mode, geometric parameters such as diameter and number of impeller
three-dimensional (3-D) incompressible simulation using a shear blades that will be varied, in order to see the effect on the pump
stress transition (SST) k-ω turbulence model. Analysis of the flow performance. So, some problems arise
characteristics revealed that severe fluctuations occurred due to
• How does variation of impeller diameter affect the
vortices in the retaining vane, and a high pressure circular ring in
performance of axial flow pump?
the chamberless chamber and a low pressure circular ring between
• How does the variation of the number of impeller blades
the outlet of the guide vane and the inlet of the retaining vane were
affect the performance of the axial flow pump?
identified.
Zhang, Desheng, et. al. discusses TLV paths simulated
under different flow rate conditions in axial flow pumps. The
d. Paper Objective
pressure difference between the two sides of the blade is the
The purpose of this report is to:
strength of the leakage flow. The fluid dynamics of pressure,
1. Knowing the Effect of Impeller Diameter Variations on
vorticity and turbulent kinetic energy within the TLV core are
the performance of axial flow pumps
revealed.
2. Knowing the effect of variations in the number of
Kim, Moon-Chan, et. al. discussing the effect of tip
impeller blades on the performance of axial flow pumps
clearance is clearly visible in the test results of this model. The
3. Knowing how geometry parameters affect the
efficiency difference according to the slit variation (1.5% diameter
performance of axial flow pumps in terms of head, as
and 0.7% diameter) is about 25% in overall efficiency. This
well as flow power and shaft power in relation to pump
difference is so large that the gap effect should be a very important
efficiency.
consideration in improving the performance of an axial flow
waterjet system.
e. Problem Limitations
Yang, Fan, et. al. studied that the internal flow field of
The limitations of the problem in this study are:
the guide vane is increased by adjusting the angle, and the flow
1. The pump under study is the type of Axial Flow Pump
separation of the tail inlet baluster and the guide vane is reduced
2. The analyzed part of the pump is its performance when
or eliminated, so that the hydraulic efficiency of the pumping
operated at a certain speed.
system will be improved.
3. Validation of the model is carried out on the reference
Kaya, Durmus stated that with the three- and four-blade
literature.
configurations, the guide vanes significantly increase efficiency.
The basic reason is the potential of the guide vane to convert the
f. Writing Systematics
tangential velocity energy back into pressure energy. Friction
The systematic arrangement of the completion of this
losses increase when guide vanes are used. The shaft moment
research is as follows:
increases for this reason. Some of the efficiency gain is spent on
1. Chapter 1 Introduction contains the background for
this friction, but the rest is still improving efficiency.
selecting the title, problem formulation/description,
Xu, Bin, et. al. studied that there are two obvious leakage
vortex structures in the end region of the axial flow pump with
 3

objectives to be achieved, problem boundaries,

systematic report writing. c. Flowchart
2. Chapter 2 Literature Review and Theoretical Basis
contains a literature review on solving this problem that Mulai

has been carried out by previous researchers. As well as Literature Study and Determination of
an explanation of the basic theories used for the problems Reference Literature
Run Calculation
to be solved. Determination of Variables and Simulation
Planning
3. Chapter 3 Methodology contains the author's way of
Convergence?
solving the problem.
Formulating Geometry
4. Chapter 4 Simulation contains the steps taken in making No
Yes
geometry using software. As well as the stages of the
Modeling and Making Named Selection Using
meshing process. At the same time doing grid ANSYS Storing Result Data
independence test
5. Chapter 5 Simulation results contains the results of the Making mesh with ANSYS Fluent with
Meshing Discussion of Simulation Results
validation and simulation results of geometric variations,
as well as a discussion of the simulation results that Perform setups such as energy equations, Conclusion
turbulence models, materials, and boundary
explain why a value has a certain trend or the value is low conditions on ANSYS Fluent
or high. Finish

6. Chapter 6 Closing contains Conclusions that write down Fill in the convergence conditions table

some of the main results that refer to the research Doing solution initialization
objectives. As well as suggestions containing several
suggestions that are considered important for further
research.
d. Fundamental Theory
Dynamic pumps are divided into several types, namely
II. METHODS AND MATERIALS axial pumps, centrifugal pumps and others. These pumps operate
a. Methodology by generating a fast fluid velocity and converting it into pressure
This study uses two methods, namely analysis and through changes in the fluid flow. These pumps generally have
simulation methods. The method of analysis is done by studying lower efficiency than positive displacement pump types, but
the reference literature. The simulation method is carried out to relatively low maintenance costs. Pumps of this type are capable
obtain the result value of the proposed variation. This simulation of operating at high flow rates and discharges [Korakianitis,
was performed using ANSYS Fluent. ANSYS Fluent itself is a 2016]. The following are the types of dynamic pumps:
software used for solving computational cases which are divided
into flow modeling, geometry, heat transfer, and others. ANSYS a. Centrifugal pump
is used to analyze various mathematical models, both A centrifugal pump is a mechanical device designed to
compressible and incompressible, steady or transient, and laminar move fluids by transferring rotational energy from one or more
or turbulent conditions. The work begins with conducting a driven rotors, called impellers. The fluid enters the impeller which
literature study and determining the literature that will be used as rotates rapidly along its axis and is discharged by centrifugal force
a reference, followed by determining the parameters that will be along its circumference through the ends of the impeller blades.
used as variables and planning the simulation. Then the geometry The action of the impeller increases the velocity and pressure of
and material determination of the study object and boundary the fluid and also directs it towards the pump outlet. The pump
conditions are formulated, followed by loading the model. In casing is specially designed to restrict fluid from the pump inlet,
ANSYS Fluent with Meshing, meshing is made and followed by direct it to the impeller and then slow and control the fluid before
setup on ANSYS Fluent, where energy equations, boundary discharge.
conditions and others are determined. Then the convergence
conditions are determined and initialization is carried out. Then do
a run calculation with ANSYS, and wait for the convergent
condition. If it is convergent, the result data such as contours and
graphs will be analyzed, then the final conclusion will be drawn.
b. In this study, the variables used for the review have been
determined based on the object of study contained in the reference
literature. The choice of geometry parameters to be simulated is
the variation of the impeller diameter and the number of impeller
blades. The number of diameter variations that will be given is 3
prices with a fixed number of impeller blades that are fixed for
each variation. Then for variations in the number of impeller
blades, 3 variations will also be given, with a fixed impeller Fig. 1 Centrifugal Pump
diameter for each variation. The simulation was carried out with a
constant angular speed for all variations, namely 1500 rpm. b. Axial pumps
 4

Axial pumps are also known as propeller pumps. This velocity of each unit of account from the outlet of the inlet. (m/s),
pump generates most of the pressure from the propeller and the lift and n is the number of units at the exit
force from the blades against the fluid. Single-stage vertical axial
pumps are the more commonly used type, but sometimes two- Pump Head
stage axial pumps are more economical to implement. Horizontal ∆𝑃
axial pumps are used to discharge large fluid flows with small H=
𝜌𝑔
pressures in the flow aliran. where ∆P is the pressure difference between the outlet pressure
and the inlet pressure. ρ is the density of the fluid, and g is the
acceleration due to gravity.

Torque
Torque value is a measure of the force that can cause an object to
rotate around an axis. Just as a force causes an object to accelerate
in linear kinematics, torque is what causes an object to acquire an
angular acceleration. In the pump calculation, the torque value is
taken from the magnitude of the moment on the hub and blade
impeller.

Fig. 2 Axial Flow Pump Efficiency

Pump efficiency is defined as the ratio of the flow power output
Axial pumps are used in many industrial and household from the pump to the pump shaft power input. The flow power is
applications, these types of pumps experience various kinds of determined by the flow rate and pressure delivered from the pump.
problems such as vibration and noise caused by complex pump Changing some geometry parameters for example on the impeller
geometries, flow conditions and cavitation phenomena. Due to this blade or hub can improve pump performance and save a lot of
problem, the performance and life of the pump will decrease. energy. Thus increasing the value of efficiency as well [Bellary,
Pump failure will increase further with high flow rate, high 2018]. Efficiency is expressed as:
reliability, large capability and high speed. However, an important
reason that causes the failure of the pump structure is the unstable 𝐹𝑙𝑜𝑤 𝑃𝑜𝑤𝑒𝑟
η= x 100
flow [Al-Obaidi, 2019]. 𝑆ℎ𝑎𝑓𝑡 𝑃𝑜𝑤𝑒𝑟

However, under conditions of small discharge there is also an Flow power is the amount of power in the resulting fluid flow, the
unstable operating area where severe vibration and noise will be value is obtained from the multiplication of the mass flow rate at
generated and the pump head will drop rapidly as the flow rate the outlet with the acceleration of gravity and the value of the head.
increases. In the unstable operating area, there is a strong While the Shaft Power is obtained by multiplying the torque value
interaction between the impeller structure and turbulent flow, by the rotational speed.
which will cause complex deformation of the impeller structure Flow Power = 𝑚̇ . 𝑔 . 𝐻𝑒𝑎𝑑
and even cracks. Therefore, in axial flow pumps, many scholars Shaft Power = Torque . RPM
devote themselves to studying unstable flow patterns [Meng,
2018]. III. CALCULATIONS
a. Pembuatan Geometri
Speed Distribution Uniformity The first stage, namely making geometry, is carried out on
form distribution of axial velocity on the outlet cross section. The solidworks 2017. The geometry created is the geometry that will
inlet shall be designed to provide a uniform flow distribution and be used for later validation, so that the geometry dimensions here
pressure distribution for the impeller. The uniformity (Vu) of the refer to the dimensions listed in the reference literature.
axial velocity distribution at the entrance reflects the quality of the
inlet design. The closer Vu to 100%, the more uniform the First, an extrude of the impeller hub is made with dimensions
distribution of axial flow velocity at the inlet outlet [Shi, 2017]. according to the reference literature.
The equation used is as follows

1 ∑𝑛 ⃗ 𝑎 )2
𝑖=1 (𝑣𝑎𝑖 𝑣
Vu = 1 − √ x 100%
𝑣⃗𝑎 𝑛

where Vu is the uniformity of the distribution of axial flow

velocity at the outlet of the inlet (%), va is the arithmetic mean of
the axial flow velocity at the outlet of the inlet, vai is the axial
 5

Then make a hub for the guidevane area, as well as the guidevane
with a total of 6.

Fig. 3 Extrude Impeller Hub

Followed by making one Impeller blade, then giving 4 circular

patterns according to the reference literature, then combine. Fig. 6 Guide Vanes

Finally, an outer casing is created which will later be given a fill

command for the fluid domain when it is done on ansys.

Fig. 4 Impeller Blade

Next create a hub for the intake section with a geometry that refers
to the reference literature. Fig. 7 Casing

Here is added a volume extract operation to fill the inner

cavity which will later become the domain of the fluid.

Fig. 5 Intake Hub

 6

results will be. Mesh making is done by adding local sizing for the
impeller and guidevane to refine the existing mesh in the area.
Then we describe the geometry and apply share topology to bring
it together. Furthermore, checking the existing areas and adding a
boundary layer with default conditions. Also check the quality of
the mesh, for example by looking at the value of Orthogonal
Quality. If the dominant is large, it means that the mesh is good
enough.

Fig. 8 Volume Extract

Followed by adding a plane and then cutting the new volume into
3 parts, namely the intake, impeller, and guide vane

Fig. 11 Meshing Results

c. Simulation with Fluent

After finishing meshing, then we click the switch to
solutions menu to move to Fluent. The next step is to do Setup to
prepare for calculations. Here you can also check the quality of the
mesh again.
In the general setup section, the solver setting selected is the
Pressure Based type, the Velocity formulation is Absolute, and for
the time it is selected steady.

Fig. 9 Split Body

Then a named selection is made according to the areas in the

volume, such as inlet, outlet, hub, impeller, and shroud..

Fig. 12 General Setup

Then select the Models menu, select the k-omega 2 turbulence

model, type STT and also select the Production Limiter option.

Fig. 10 Named Selection

b. Making of Mesh
The next stage is making meshing in ANSYS (Fluent
with Meshing), this is done to divide the existing geometry into
discrete elements so that it can be solved by simulation. The
smaller and neater the solution, the more precise the calculation
 7

Fig. 15 Cell Zone Conditions

Next is setup for Boundary Condition. Here, each part is

determined by its type, whether as a wall, interior, inlet, or outlet.
At the same time choose the type of inlet and outlet. The setting in
the Boundary Condition is left at the default condition, while the
changes are as follows:
 Shroud Impeller: Moving wall, Absolute Motion,
Rotational, Speed 0 rpm
Fig. 13 Setting Models
 Hub Impeller: Moving wall, Relative to Adjacent Cell
Zone, Rotational, Speed 0 rpm
Followed by selecting materials, which add materials from the
 Inlet: jenis mass flow inlet, massflow rate 0.046QBEP,
database that has been provided by ANSYS. Add h2o liquid.
which is 17.2 kg/s normal to boundary
 Outlet: Pressure Outlet, normal to boundary, 0 Pascal

Fig. 14 Materials

Then for Cell Zone Condition, the material in question is selected

for the existing zones. In this case the guidevane, impeller, and
intake are made of water liquid. And on the impeller, the Frame
Motion option is turned on because the impeller rotates. Rotation-
Axis Direction is selected Z=1 because it fits the Z axis, and for
Rotational Velocity it is filled according to the Reference
Literature.
Fig. 16 Boundary Condition

For reference values, values are assigned according to the default

state.
 8

Fig. 19 Monitors: Residual

Before performing calculations, initialization is done first. This is

done by initializing on the Solution Initialization menu. Selected
Standard Initialization, and Compute From Inlet. Reference and
Initial Value are left as conditions that have been filled.
Fig. 17 Reference Values

Solution method settings. The SIMPLE Algorithm scheme is

chosen, for the Gradient Least Squares Cell Based is flattened and
other parameters are paired to the initial state. And selected High
Order Term Relaxation.

Fig. 20 Solution Initialization

Next, Run Calculation is done, here it is chosen to do 700

iterations. Then wait until the simulation reaches the convergence
condition.
When the iteration stops, it means that the convergence condition
Fig. 18 Solution Method has been met. Here the convergent condition is reached at the
458th iteration
Then the contents of the Residual Monitor are determined, where
at the same time the convergence criteria are also determined.
 9

Fig. 24 Local Sizing with the First Element Number Var.

In the second Local Sizing, a torque of 266.99 Nm is obtained

Fig. 25 Torque from Second Local Sizing

Antara uji grid pertama dan kedua, didapat bahwa nilainya tidak
Fig. 21 Run Calculation and Convergence terlalu jauh. Perbandingan hasil menggunakan rumus:
d. Grid Independence Test ∆P = (Hn – Hn-1)/( Hn-1) x 100%
∆P0-1 = (231.75 – 226.99)/(226.99) x 100% = 2.097 %
After achieving convergent results, in this section it will
be determined whether the calculation results obtained depend on Since the difference in values between the two test grids is already
the number of grids/cells used. In this first calculation, the less than ±3%, it can be said that this calculation is independent of
calculation is carried out using the meshing option with a Target the number of grids. So that the mesh that will be used for
Mesh Size of 0.005. calculations is a finer one, namely the local sizing of the impeller
blade and guidevane of 0.004.

IV. CALCULATION RESULTS

a. Validation Resuts
Here it will be shown that the calculation results that have been
obtained previously are quite close to the results of calculations or
experimental results in the literature.
The geometry available in the reference literature is an axial flow
pump with an impeller diameter of 300 mm, a total of 4 impeller
Fig. 22 Local Sizing with the First Element Number Var. blades and a guidevane of 6.
In the reference literature, the author examines the effect of adding
The output value seen is torque. In the first Local Sizing, obtained an axial groove and a diagonal groove to the hydraulic
a torque of 231.75 Nm. performance and internal flow pattern of the axial flow pump. In
the reference literature there are experimental results, original
schemes, schemes with axial grooves, and schemes with diagonal
grooves. The parameters sought are also available for several mass
flow rate conditions with different values.

Fig. 23 Torque from First Local Sizing

Then for the Grid Independence test, we will compare the results
of the first calculation output with the results of calculations using
a larger number of elements. Done by making a finer mesh size.

Fig. 26 Head Result on Various Mass Flow Rate

The calculations carried out in the previous chapter took a simple

pipe sample with a mass flow rate of 17.2 kg/s. That is equivalent
 10

to the calculation of 0.046QBEP in the reference literature. It was 450 mm 16,34 16,48 17,33 18,21 17,52
found that the head size in this condition was 8.25m.
In the calculations that have been carried out with local sizing on
the impeller blade and guidevane of 0.004, the Total Pressure for
the inlet and outlet is obtained..

Fig. 27 Pressure Result for Validating Head Value

Fig. 28 Head Value on Diameter Variation
∆P
Head could be obtained from the following: H = It was found that as the impeller diameter increased, the head value
𝜌. 𝑔
at the same condition also increased. Whereas in contrast to mass
∆P: Pout-Pin = (36811,438 - -45594.329) = 82405,767 Pa flow, the increase in mass flow can increase or decrease the value
∆𝐏 𝟖𝟐𝟒𝟎𝟓,𝟕𝟔𝟕 of the Pump Head. The largest head value occurs at a mass flow
H (Head): H = = = 8.40 m of 70kg/s for diameters of 350 and 400 mm. As for the diameter
𝝆. 𝒈 𝟏𝟎𝟎𝟎 . 𝟗,𝟖𝟏
of 450 heads, the highest is 250 kg/s.
Compared with Head values in the reference literature for the
Tabel 2 Efficiency with Variation of Diameter at Various Mass
same conditions:
Flows
H calc − H ref
Error =
U ref
. 100% Mass
𝟖,𝟒𝟎 − 𝟖,𝟐𝟓 30 70 150 250 350
= . 𝟏𝟎𝟎% = 1,81% Flow
𝟖,𝟐𝟓 kg/s kg/s kg/s kg/s kg/s
Rate
Diameter
With this, validation has been successfully carried out by Variation
obtaining a simulation result value which is less than 10% with the 350 mm 3,98 10,53 25,08 37,83 46,66
reference literature, which is 1.81%. Therefore, we will proceed to 400 mm 2,67 6,89 16,17 27,25 37,66
the next stage, namely looking for simulation results with 450 mm 1,93 4,73 11,07 19,78 28,53
geometric variations.

c. Geometry Variation Simulation Results

After performing the simulation, the pressure values at
the inlet and outlet are obtained. To facilitate data acquisition,
expressions are used that can directly express the value of pump
head and pump efficiency.
1. Simulation Results of Variation in Diameter Ratio
Tabel 1 Head with Variation of Diameter at Various Mass Flows.

Mass
30 70 150 250 350
Flow
kg/s kg/s kg/s kg/s kg/s
Rate
Diameter
Variation
350 mm 12,12 12,74 11,16 9,65 8,54 Fig. 29 Efficiency Value on Diameter Variation
400 mm 14,17 15,25 15,20 13,96 12,62
 11

It was found that as the impeller diameter increased, the efficiency

value at the same condition decreased. In contrast to mass flow,
increasing mass flow can increase the value of Pump Efficiency.
The greatest efficiency value occurs at the largest mass flow,
which is 350 kg/s for each diameter variation

2. Simulation Results of Variation in Number of

Impeller Blades
Tabel 3 Heads with Variation in Number of Blades at various
Mass Flows
Mass
30 70 150 250 350
Flow
kg/s kg/s kg/s kg/s kg/s
Rate
Impeller
Number
3 8,57 9,52 8,96 7,64 6,91
5 14,88 15,72 13,52 11,20 9,73
Fig. 31 Graph of Efficiency against Blade Number Variations
7 17,54 18,42 16,32 13,47 13,27
It was found that with the increase in the impeller diameter value,
the pump efficiency value did not have a big difference. As for the
relation to mass flow, the increase in mass flow can increase the
value of Pump Efficiency. The greatest efficiency value occurs at
the largest mass flow, which is 350 kg/s for a variation of the
number of blades as much as 7.

3. Velocity Streamline with Variation of Impeller

Diameter

Fig. 30 Head Value on Blade Number Variation

It was found that with the increase in the number of impeller
blades, the head value under the same conditions also increased.
Whereas in contrast to mass flow, the increase in mass flow can
increase or decrease the value of the Pump Head. The largest head
value occurs at a mass flow of 70kg/s for all variations in the
number of blades:
Tabel 4 Efficiency with Variation in Number of Blades at Various (a)
Mass Flows
Mass
30 70 150 250 350
Flow
kg/s kg/s kg/s kg/s kg/s
Rate
Impeller
Number
3 4,10 10,78 25,14 38,75 46,89
5 4,21 11,13 24,82 38,76 47,27
7 4,20 10,93 25,08 40,93 49,24
 12

(a)

(b)
(b)

(c)
(c)
Fig. 33 Velocity Streamline on Diameter Variation
Fig. 32 Velocity Streamline on Diameter Variation
(a) 350 mm (b) 400 mm (c) 450 mm (a) 3 (b) 5 (c) 7
4. Velocity Streamline with Variation of Impeller
number
5. Pressure Contour with Variation of Impeller
Diameter
 13

(a)
(c)
Fig. 34 Pressure Contour at Variation of Pipe Diameter
(a) 350 mm (b) 400 mm (c) 450 mm

6. Kontur Tekanan dengan Variasi Jumlah Impeller

Blade

(b)

(a)
 14

7. GuideVane Velocity Vector at Variation of Impeller

Diameter

(a)

(b)

(b)

(c)
Fig. 35 Pressure Contour at Variation of Impeller Blade Number
(a) 3 (b) 5 (c) 7 (c)
Fig. 36 Velocity Vector on Variation of Impeller Diameter
(a) 300 mm (b) 400 mm (c) 450 mm
 15

8. GuideVane Velocity Vector on Variation in Number

of Impeller Blades

V. RESULTS AND DISCUSSION

From the simulation results of diameter variations seen in Table
5.1 and Fig. 5.3 that as the diameter increases, the head value will
increase. By simulating conditions at different mass flow rates, we
get a pattern that tends to be similar. Namely the increase in the
value of the head on the increase in the mass flow rate. However,
this only happens when the mass flow rate changes from 30 to
70kg/s. Starting from 70 kg/s, the increase in mass flow rate
actually reduces the head under the same diameter variation
conditions. This shows that to get a larger head can be done by
increasing the diameter of the impeller on the pump. At
sufficiently high operating speeds as simulated, the impeller
diameter is an important design parameter of the axial pump
because it affects its performance and internal flow characteristics
[Bai, 2017].
(a)
It was found that the largest head value that occurred for a
diameter of 350 mm was 12.74 m, namely at a mass flow rate of
70kg/s. For variations in diameter of 400 mm, the largest head is
obtained at 15.25 m at a mass flow rate of 70 kg/s and a value that
is almost similar to a flow rate of 150 kg/s. Then for the diameter
variation of 450 mm, it was found that the largest head was 18.21
m at a mass flow rate of 250 kg/s.
With this it can be stated that different pump impeller diameter
values have different ideal operating conditions. Judging from the
results, the diameters of 350 mm, 400 mm, 450 mm reached the
largest head conditions at mass flow rates of 50, 50-70, and 150
kg/s. This means that the larger the impeller diameter, the more
ideal the pump will be to operate at higher mass flow rates.
However, enlarging the impeller will make it less than ideal in
terms of power. Due to the increasing diameter of the impeller will
increase power consumption, which is then associated with a
(b) decrease in pump efficiency [Matlakala, 2019].
This is appropriate because we also perform simulations to find
pump efficiency at different mass flow rates. It was found that in
any given mass flow rate condition, the higher efficiency was the
one with the smallest diameter. And its value always increases
with increasing mass flow rate as well. The highest efficiency
values obtained for variations of 350 mm, 400 mm, and 450 mm
are 46.66%, 37.66%, 28.53%. It is hereby stated that to seek high
efficiency, we use variations with smaller diameters and operate
at high mass flow rates. However, with this we can not get a head
that is too big because the relationship is opposite. To get a large
head, we increase the diameter of the impeller. Meanwhile, to
increase the efficiency of the pump, we choose a small diameter.
Then also varied the number of impeller blades of the pump. It was
found that with increasing mass flow rate, the trend created by
each variation was similar, reaching the highest head at a mass
flow rate of 70 kg/s. The highest head values for variations of 3,
(c)
5, and 7 impellers are 9.52 m, 15.72 m, and 18.42 m. Under the
Fig. 37 Velocity Vector on Variation of Impeller Blade Number same conditions, the greatest head value is found in the variation
(a) 3 (b) 5 (c) 7 with the largest number of impellers. The cause of this is due to
the reduced loss that occurs by increasing the number of blades
 16

due to a reduction in secondary flow to a certain extent [Bai, VI. CONCLUSION

2017].
From the research results, the effect of variations in
Furthermore, the effect of the number of impeller blades on the
impeller diameter and the number of impeller blades on the
efficiency of the pump, it is found that the trend of each variation
performance of the axial pump that has been discussed previously
is increasing as the mass flow rate increases. With the highest
obtained the following results:
efficiency values obtained at a mass flow rate of 350 kg/s for each
variation, namely 46.89%, 47.27%, and 49.24% for the number of a. The validation of the geometry model that has been
impeller blades 3, 5, and 7. from the formed curve, we can see that carried out on the reference literature pursues results with
the difference in efficiency for each variation is not significant. errors below 10%. In the reference literature, it is found
although the number of blades was changed to more than double that with a mass flow rate of 0.046QBEP which is 17.2
from 3 to 7. kg/s, the head value obtained is 8.25 m. In the validation
geometry in this study, the head value is 8.40 m. The
With this it is found that the number of impellers has no significant
magnitude of the error is 1.81%.
effect on changes in pump efficiency. This is because, increasing
the number of impeller blades can cause an increase in loss due to b. It was found that the effect of the impeller diameter
friction with the increasing number of contact surfaces between variation on efficiency was inversely proportional. The
the fluid and the pump impeller. So, even though the secondary larger the impeller diameter actually makes the efficiency
flow phenomenon is compensated, there are other factors that smaller, but makes the head bigger. The largest head
reduce it again, causing the difference in efficiency values to be condition was achieved at a mass flow rate of 70 kg/s at
not large between variations in the number of existing impeller variations of 350 and 400 mm, and at conditions of 250
blades [Elyamin, 2019]. kg/s for variations of 450 mm. The highest efficiency
value is achieved at the largest head condition
Furthermore, if we look at the streamline that is formed from each
variation. It can be seen that the existing streamline conditions are c. It was found that the effect of variations in the number of
not much different except for the 450 mm diameter variation. Due impeller blades on efficiency was not very significant. As
to the asymmetry of the inlet longitudinal structure it affects the for the head, it is directly proportional to the increase in
flow pattern at the outlet. Especially in conditions of small flow the number of impeller blades, the larger the head. The
rates, the asymmetry of the structure will result in an uneven largest head is achieved at a mass flow rate of 70 kg/s for
velocity distribution at the inlet outlet [Xia, 2016]. all variations. The highest efficiency value is achieved at
the largest head condition.
Then there is a pressure contour that is formed on each variation
of the impeller diameter. It can be seen that the pressure
distribution tends to be similar at the intake, impeller, and guide VII. SUGGESTIONS
vanes. However, as the diameter of the impeller increases, a dark
blue zone begins to appear which indicates low pressure. And in In the simulation, it is found that there is a reverse flow
the largest diameter variation, there is a red zone at the ends of the phenomenon in several pump areas. When traced using CFD-
impeller which indicates high pressure. This is because as the Post, it can be seen that this reverse flow occurs only slightly and
diameter of the impeller increases, the same rotational speed will only in the periphery. However, this may lead to inaccurate
provide higher speeds at the ends of the impeller. This accelerates calculations that have been carried out. Suggestions for the
the fluid flow, which causes the low pressure colored zone to future should be a reassessment of the geometry and ensure that
appear. Variations in the number of impeller blades do not show a no parts are too thin and avoid asymmetrical inlets.
significant difference in the pressure contour. Because the only
change is the portion of the fluid that is pressed by the impeller,
the rotational speed and the flow dimensions are dominantly the REFERENCES
same. 1. Aung Kyaw Soe, Zin Ei Ei Win, Myat Myat Soe, (2015),
Then looking at the results of the velocity vector on the guide “Numerical Flow Analysis of an Axial Flow Pump”,
vanes, it can be seen that the direction of the velocity vector in International Journal of Engineering and Applied Sciences,
each variation is similar. But it can be seen in the largest diameter Vol.2
variation that more vectors are colored red, indicating a suddenly 2. Stepanoff, A.J. 1957. Centrifugal and Axial Flow Pumps,
higher velocity. Indicates that there is a higher flow velocity, thus New York, John Wiley and Sons
the largest impeller diameter variation gives the largest head.
3. Weidong Cao, Wei Li, (2020), “Study on the Performance
Then in the variation of the number of impeller blades, the velocity Improvement of Axial Flow Pump’s Saddle Zone by Using a
vector does not appear to have a significant difference. But the Double Inlet Nozzle”, Water, Vol.12
value of the head is getting bigger, because with the number of
blades more and more will make the fluid in a smaller portion 4. L J Shi, F P Tang, C Liu, R S Xie and W P Zhang, (2016),
depressed with the existing impeller blade. Makes the speed higher “Optimal design of multi-conditions for axial flow pump”,
when passing through the guide vanes zone and when exiting Earth and Environmental Science Vol.49
through the outlet. 5. K F Yang, J J Feng, G J Zhu, J L Lu and X Q Luo, (2018),
“Study on improvement of hump characteristic of an axial
 17

flow pump by grooving inlet wall”, Earth and Environmental 19. Bin Xu, Xi Shen, Zhang Desheng, Weibin Zhang, (2019),
Science, Vol.163 “Experimental and Numerical Investigation on the Tip
Leakage Vortex Cavitation in an Axial Flow Pump with
6. Kan Kan, Zixuan Yang, Pin Lyu, Yuan Zheng, Lian Shen,
Different Tip Clearances”, Processes, Vol.7
(2020), “Numerical study of turbulent flow past a rotating
axial-flow pump based on a level-set immersed boundary 20. Yabin Liu, Lei Tan, (2019), “Investigation of tip leakage
method”, Renewable Energy, Vol.168 vortex characteristics around tip clearance in a mixed flow
pump”, Earth and Environmental Science, Vol.240
7. Lijian Shi, Wenpeng Zhang, Haifeng Jiao, Fangping Tang,
Li Wang, Dandan Sun, Wei Shi, Soheil Soleimani kutanaeia, 21. Theodosios Korakianitis, Mohammad A. Rezaienia, Gordon
C.X. Lina, Dexin Wang, (2020), “Numerical simulation and M. Paul, Akbar Rahideh, Martin T. Rothman, Sahand
experimental study on the comparison ofthe hydraulic Mozafari, (2016), “Optimization of Centrifugal Pump
characteristics of an axial-flow pump and a full tubular Characteristic Dimensions for Mechanical Circulatory
pump”, Renewable Energy, Vol.153 Support Devices”, Journal of American Society for Artificial
Internal Organs, Vol. 62.
8. Zhang De-sheng, SHI Wei-dong, CHEN Bin, GUAN Xing-
fan, (2010), “Unsteady flow analysis and experimental 22. Ahmed Ramadhan Al-Obaidi, Amer A. Mohammed, (2019),
investigation of axial-flow pump”, Journal of “Numerical Investigations of Transient Flow Characteristic
Hydrodynamics, Vol.22 in Axial Flow Pump and Pressure Fluctuation Analysis
Based on the CFD Technique”, Journal of Engineering
9. Sun-Sheng YangShahram DerakhshanFan-Yu Kong, (2012),
Science and Technology, Vol.12
“Theoretical, numerical and experimental prediction of
pump as turbineperformance”, Renewable Energy, Vol.48 23. Fan Meng, Shouqi Yuan and Yanjun Li, (2009). “Fluid–
structure coupling analysis of impeller in unstable region for
10. D.R.GiosioaA.D.HendersonaJ.M.WalkerbP.A.BrandnerbJ.E.
a reversible axial-flow pump device”, Advances in
SargisoncP.Gautam, (2015), Renewable Energy, Vol.78
Mechanical Engineering, Vol.10
11. A. Bozorgi, E. Javidpour, A. Riasi, A. Nourbakhsh, (2012),
24. Lijian Shi, Fangping Tang, Rongsheng Xie, Weipeng Zhang,
“Numerical and experimental study of using axial pump as
(2017), “Numerical and experimental investigation of tank-
turbine in Pico hydropower plants” Renewable Energy,
type axial-flow pump device”, Advances in Mechanical
Vol.53
Engineering, Vol.9
12. Bryan Orchard,Sander Klos, (2009), “Pumps as turbines for
25. S.A.I. Bellarya, A. Husain, A. Samad, R.A. Kanai, (2018),
water industry”, World Pumps, Vol.8
“Performance Optimization of Centrifugal Pump for Crude
13. Bin Chen, Yonghai Yu., (2012), “CFD Simulation and Oil Delivery”, The Journal of Engineering Research, Vol.15
Optimization for Lateral Diversion and Intake Pumping
26. Yuxing Bai, Fanyu Kong, Bin Xia, Fei Zhao, Yingying Liu,
Stations”, Procedia Engineering, Vol.28
(2017), “Effects of Impeller Diameter on High-Speed
14. Deyou Li, Hongjie Wang, Zhenggui Lia ., Torbjørn Kristian, Rescue Pump”, Mathematical Problems in Engineering
Nielsen., Rahul Goyal, XianzhuWeiae Daqing Qin, (2018),
27. M E Matlakala, D. V. V. Kallon, K. E. Mogapi , I. M.
“Transient characteristics during the closure of guide vanes
Mabelane, D. M. Makgopa, (2019), “Influence of Impeller
in a pump-turbine in pump mode”, Renewable Energy,
Diameter on the Performance of Centrifugal pumps.”,
Vol.118
Materials Science and Engineering, Vol.665
15. Desheng Zhang, Weidong, B.P.M.(Bart) van Esch, Lei Shi,
28. Yuxing Bai, Fanyu Kong, Bin Xia and Yingying Liu,
Michael Dubuisson (2015), “Numerical and experimental
(2017), “Effect of blade number matching of impeller and
investigation of tip leakage vortex trajectory and dynamics
diffuser in high-speed rescue pump”, Advances in
in an axial flow pump”, Computers & Fluids, Vol.112
Mechanical Engineering, Vol.9
16. Moon-Chan, KimHo-HwanChun, (2007), “Experimental
29. Gamal R.H. Abo Elyamin, Magdy A. Bassily, Khalil Y.
Investigation into the performance of the Axial-Flow-Type
Khalil, Mohamed Sh. Gomaa, (2019), “Effect of impeller
Waterjet according to the Variation of Impeller Tip
blades number on the performance of a centrifugal pump”,
Clearance”, Ocean Engineering, Vol.34
Alexandria Engineering Journal, Vol.58
17. Fan Yang, Hao-ru Zhao, and Chao Liu, (2015),
30. Chenzhi Xia, Li Cheng, Chao Liu, Jiren Zhou, Fangping
“Improvement of the Efficiency of the Axial-Flow Pump at
Tang, Yan Jin, (2016), “Research on the performance of
Part Loads due to Installing Outlet Guide Vanes
low-lift diving tubular pumping system by CFD and Test”,
Mechanism” Mathematical Problems in Engineering,
Earth and Environmental Science, Vol.49.
Vol.2016
18. Durmus, Kaya, (2002), “Experimental study on regaining
the tangentialvelocity energy of axial flow pump” Energy
Conversion and Management, Vol.44