Vol. 42, No. 3 (2020) 382-399, DOI: 10.24874/ti.803.11.19.

06

Tribology in Industry

RESEARCH
www.tribology.rs

Computational Investigation of Erosion Wear on

Industrial Centrifugal Pump Handling Solid-Water
Flows

J. Kumara, G. Tiwaria, A. Rawata,*, V.K. Patela

a Department of Applied Mechanics, MNNIT Allahabad, Prayagraj, 211004, Uttar Pradesh, India.

Keywords: ABSTRACT
Erosion
The Centrifugal pumps handling solid-water mixture are heavily afflicted
Solid-water flows
by erosion occurring due to the transportation of solid particles.
Industrial Pumps
Eventually, the erosion leads to degradation in performance and
CFD
mechanical properties of the pump materials. The objective of present
Stainless steel
work is twofold; a comprehensive erosion study on three of the common
pump materials (Carbon Steel, SS 304 and SS 316) by considering solid-
* Corresponding author: water mixture with two different erodent materials (Silicon Carbide and
Silicon dioxide), and to find the critical locations of erosion wear on the
Anubhav Rawat
pump by utilizing commercial CFD code ANSYS R16.0 Fluent. Numerical
E-mail: Anubhav-r@mnnit.ac.in
simulations on a 3-D model of the pump have been conducted and the
Anubhav_1982@yahoo.com erosion rates caused due to erodent (SiC and SiO 2) in various parts of the
pump namely spiral casing, front Shroud, back shroud and vanes have
Received: 29 November 2019 been calculated for pump materials carbon steel, SS 316 and SS 304 with
Revised: 5 April 2020 the concentrations of erodent varying from 5-25 % and at fixed speed of
Accepted: 1 June 2020 1450 rpm. Erosion on carbon steel is found to be 1.15-4.16 times higher
than the other two steel for all parts at 25 % concentration for SiC
erodent material. The corresponding figure for SiO 2 is 1.25-2. Further, it
is observed that Casing receives 7-13 times higher erosion than Shrouds
and 4-7 times higher than vanes for SiC. Whereas, the wear due to SiO 2 is
found to be around 12-30 times higher than SiC at 25 % concentration.
Thus, Carbon steel receives more wear and SiO 2 causes more wear over
the pump parts. Erosion is not only found to be critically dependent upon
pump and erodent materials but also on the interaction of the erodent
material with pump material surface. In general, Erosion is found to be
increasing linearly with concentration and causing unequal wear at each
part of the pump. It is also found that the critical parts for erosion wear
of the industrial pumps are where flow suddenly changes direction.
© 2020 Published by Faculty of Engineering

382
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

1. INTRODUCTION considering the effect of particle size, velocity

and concentration. Aribo et al. [4] did a
Transportation of solid-liquid mixtures through comprehensive erosion-corrosion combined
pipelines is a vastly used method in many study in a pot tester due to slurry. Adnan et al.
industries. Few industries do it on purpose to [5] predicted that the erosion damage in lime
handle solid- liquid flows viz. Transportation of slurry pumps obtained using three-dimensional
minerals and coal ash through pipelines. Many numerical analysis is directly correlated to
industries design their pump systems to impact velocity, concentration by weight and
transport single-phase liquids only but the diameter of solid particles along with the
dispersed phase manifests itself as an impurity. critical role played by temperature. It is also
Hydraulic power plants are one such example, reported that the tongue and belly of spiral
where the pumps are designed to handle water casing are highly erosive. Rahul et al. [6] found
flows only. Still, the presence of sand solid out the ways to estimate the pump efficiency
particles not only changes the flow and the erosion of centrifugal pumps used for
characteristics across them but also cause a hydraulic transportation of solids based on
substantial amount of erosion wear to them. optimization of pump design, selection of pump
This will deplete the wall thickness of pump at material, concentration and flow characteristics
various locations and ultimately may cause of slurry and ways to optimize both
failure of pumps many years before their performance and erosion and further used
designed life. The failure of operational pumps is Eulerian-Eulerian model to predict the effect of
severely associated with heavy losses in terms of solids on the performance of centrifugal pumps
money incurred both in maintenance or and the head and efficiency show a decline with
replacement of the pumps and delays in the increase in the particle size and
operations at hydraulic power plants. Mainly, concentration. Whereas, high specific gravity
two mechanisms by which solid-liquid flow particles mitigate the effect of a decrease in
causes material depletion to the pumps are head and efficiency of pumps to an extent.
cavitation and slurry erosion due to the solid
particles present in the liquid flow. The type of Gautam et al. [7] studied the slurry erosive
solid-liquid mixture (slurry) determines the behavior of brass with sand particle slurry, and
characteristics of solid particles present, such as according to them, the rate of material loss was
their diameter and concentration. found to be dependent on r concentration of
sand and also impact velocity. They also
A peculiar and prevalent problem in the soda predicted that the wear rate aggravates with
ash-making industry is the erosion caused by impingement angles up to 600 and declines
lime slurry, and complications are compounded thereafter. W.K. Chan [8] based on their
due to properties of the surface, erodent, and experimental work on a centrifugal impeller in a
slurry involved. Extensive experimental and closed circuit found that erosion rate increases
numerical analysis have been carried out over below the critical NPSH (net positive suction
the years in the field of pump and pump casing head), then decreases and ultimately increases
erosion, specifically into the erosion of surfaces for a declining NPSH. Gandhi et al. [9] have
Kishore et al. [1] studied the effect of thermo studied erosion wear for the parallel flow of
mechanical processing of steels on the sand solid-liquid mixture flows in a slurry pot tester
slurry erosion behavior using a pot tester. They and have found an increase in parallel flow wear
studied slurry erosion resistance of 13Cr4Ni with the increase in solid concentration, particle
stainless steel after thermo-mechanical size, and velocity. However, dependency on
processing and predicted that the thermo- velocity is greater compared to concentration.
mechanically processed specimens show higher Contrary to this, Rawat et al. [10] found the
slurry erosion resistance as compared to the as- dependency of erosion wear on concentration to
received stainless steel. Gupta et al. [2] studied be more pronounced than velocity at high
sand-water slurry erosion behavior on white concentrations. Yuan et al. [11] employed
cast iron under crayon conditions. More et al. Bitter’s model to ascertain wear coefficients
[3] studied the erosion wear characteristics of associated with erosion by particle
SS304 using Taguchi technique using a pot impingement. Pagolthivarthi et al. [12] carried
tester. They did a detailed parametric study out a predictive analysis of the slurry flow in

383
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

centrifugal pump casings with respect to facilitates this visualization and critical analysis
conditions pertaining to geometry and other of three-dimensional complex flows in the
working conditions. They found that a gradient pumps and also aids in its hydraulic design.
exists with respect to the solid wall shear and Thus, simulation results are employed in the
Solid concentration with a gradual increase performance determination of the pump and to
from the upstream tongue to the downstream alter or eliminate the experiments required in
belly region. Kaushal et al. [13] estimated the pump design process. In particular, this
pressure drop, concentration profile at various becomes important for pumps in hydraulic
flow conditions, and solid concentration inside power plants dealing with sand as an impurity. A
a horizontal pipeline. Mesa et al. [14] reported 3-D model of a centrifugal pump provided by
that slurry of an acidic solution consisting of Kirloskar Pvt. Ltd. is used in the present work to
hard particles causes extensive loss in the investigate the rate of erosion at the pump
material on interaction with commercial casing, impeller blades, and hub. The rate of
stainless steel 410 and AISI 420. Satish [15] erosion depends on various parameters like
carried out a study to understand the erosion velocity, angle of impact of the solid particle,
wear phenomenon in the pump at varied speed particle diameter, concentration of the erodent
settings with different particle size distribution particles, type of erodent particle etc. Different
and concentration with the help of ANSYS-CFX materials respond differently to a change in
computational fluid dynamics software. Xiao et these parameters. The present work, therefore,
al. [16] used a stainless-steel prototype to study aims to investigate the rate of erosion at various
the change in the flow characteristics and components of the pump for commonly used
erosion mechanism due to material loss and pump materials such as carbon steel, stainless
subsequent geometry distortion. Two models steel 304, and stainless steel 316 at a designed
were used for the study of the matter in the speed of 1450 rpm. A comparison of erosion
Eulerian frame of reference, and the results rates at different components of the pump is
indicated a great reliance on flow imperative for the aforementioned pump
characteristics and erosion patterns on materials to determine the suitability of these
geometric variations. Kumar et al. [17] studied materials from an erosion resistance standpoint.
the erosion behavior of AISI 316 pipe bend with The change in erosion owing to the nature of the
the implication of a swirl of different vane erodent particles is investigated by calculating
angles, and Baghel et al. [18] studied the erosion rates with Silicon Dioxide (SiO2) and
erosion wear on hot forged materials. Silicon Carbide (SiC) as erodent particles. The
effect of particle concentration on these
Few more important wear related [19-21] materials has also been studied by calculating
studies are worth mentioning but as far as erosion rates for particle concentrations of Cw =
erosion wear studies on pumps are concerned 5 %, 10 %, 15 %, 20 % and 25% with respect to
many useful studies [22-32] are previously done the continuous phase for all pump materials at
on various experimental and numerical aspects the fixed speed.
related to erosion wear on pumps but in general
for a single material and there is no detailed
comparative study on special pump materials 2. MATHEMATICAL MODELLING AND
namely Carbon Steel, SS304 and SS316 that too COMPUTATIONAL METHODOLOGY
considering the flow of solid-liquid mixture with
the erodent materials of Silicon Carbide (SiC) 2.1 Geometry of the industrial centrifugal
and Silicon dioxide (SiO2) as far as centrifugal pump
pumps handling sands in hydraulic power plants
is concerned. The study is also essential, and The centrifugal pump geometry utilized in the
rare in terms of observing the patterns of present work is a 3-D model conceived in ANSYS
erosion wear at different locations and parts of R16.0 design modeler and is comprised of three
the pump viz. casing, vanes, shrouds etc. Real- parts, namely the inlet pipe with diameter 69.51
time experimentations are costly, and there is no mm, impeller, and a spiral casing as shown in
established numerical strategy in pump design Fig. 1. The impeller consists of six vanes, front
for wear considering the above mentioned shroud, back shroud, vanes, and impeller eye.
scope. Numerical simulation using CFD The eye diameter is 31.991 mm. The impeller at

384
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

the inlet has a diameter of 69.51 mm and a width orthogonal quality is desired for cells. The
of 13 mm. The outlet at the impeller has a orthogonal quality of a cell is considered to be
diameter of 139 mm. The casing outlet has a good if it is farther from 0 and the closer it is to
diameter of 20 mm. Carbon steel, stainless steels 1. The mesh for the model had quality
SS 304 and SS 316 are used as existing industrial parameters within the acceptable limits with a
pump materials. mean aspect ratio of 4.634, mean orthogonal
quality of 0.85 and a mean skewness of 0.69.
2.2 Grid generation of 3-D model
2.3 Grid independency test (GIT)
Meshing has been done using the ICEM CFD
module of ANSYS R16.0, as shown in Fig. 2 Mesh independence for the simulation was
below. The inlet pipe, spiral casing, and impeller checked by comparing the efficiency of the pump
consist of tetrahedral, pyramid, and prism at various grid sizes shown in Fig. 3 and Table.1.
elements. The quality of a mesh significantly The efficiency was not found to vary much
influences the stability of numerical after11379287 number of mesh elements.
computation. In order to obtain a high quality
grid, mesh quality measures like aspect ratio, Therefore, the grid with 11379287 elements is
orthogonal quality, and skewness are kept selected.
within permissible limits. The aspect ratio of an
element is the degree of squishiness of an Table 1. Grid independency test.
element. It is the ratio of the maximum and No. of elements Efficiency (%)
minimum distance between cell centroid and 2.02751E6 58.42
face centroid. Skewness is an indication of how 5.70477E6 77.246
9.04676E6 84.568
close the cell shapes are and with that of the
1.13793E7 87.874
equilateral cell of equivalent volume. High 1.42686E7 87.876
values of skewness may lead to difficulties in the 1.82686E7 87.877
convergence of the solution. A higher 2.82686E7 87.878

Outlet

Delivery pipe Front

shroud
Tongue Inlet
Spiral casing Impeller
eye

Impeller Vanes Back

outlet shroud

(a) (b) (c)

Fig. 1. 3-D model of the centrifugal pump (a) Full Geometry, (b) Vanes assembly, (c) Shroud Assembly.

(a) (b) (c) (d)

Fig. 2. Mesh (a) Isometric view of model, (b) Front view of spiral casing, (c) Isometric view of impeller, (d)
Front view of impeller.

385
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

100 which is a two-equation eddy viscosity model is

used as a closure model. RANS equation is given
90 in Eq. 2.
𝜕(𝜌𝑢𝑖 ) 𝜕 𝜕𝑝
+ 𝜕𝑥 (𝜌𝑢̅𝑖 𝑢̅𝑗 + 𝜌𝑢𝑖 ′ 𝑢′ 𝑗 ) = − 𝜕𝑥 +
Efficiency (%)

80
𝜕𝑡 𝑗 𝑖
𝜕 𝜕𝑢 ̅𝑗
̅̅̅ 𝜕𝑢
70 𝜕𝑥𝑗
[ 𝜇 ( 𝜕𝑥 𝑖 𝜕𝑥 )]. (2)
𝑗 𝑖

60 (𝜌𝑢𝑖 ′ 𝑢′ 𝑗 ) are the Reynolds’s stresses. The

Reynold’s stresses as per Boussinesq’s
50 approximation is given in Eq. 3.
5.0M 10.0M 15.0M 20.0M 25.0M 30.0M
Number of Elements 2 ̅̅̅𝑖
𝜕𝑢 ̅𝑗
𝜕𝑢
− 𝜌𝑢𝑖 𝑢𝑗 = − 𝜌 𝑘𝛿𝑖𝑗 + 𝜇𝑡 ( + ). (3)
3 𝜕𝑥𝑗 𝜕𝑥𝑖
Fig. 3. Grid independency test.
Where 𝜇𝑡 is turbulent viscosity given by:-
2.4 Computational methodology 𝑘
𝜇𝑡 = 𝜌 .
𝜔
Numerical simulation is done using the Euler- The transport equations for k and ω as per the
Lagrange approach, wherein the fluid flow is SST k-ω model are given in Eq. 4 and Eq. 5.
the solid erodent particles are treated as
𝜕(𝜌𝜔) 𝜕(𝜌𝑈𝑗 𝜔) 𝛾 𝜕
resolved by treating it as a continuous phase + = 𝑃𝑘 − 𝛽𝜌𝜔2 +
𝜕𝑡 𝜕𝑥𝑗 𝜐𝑡 𝜕𝑥𝑗
and discrete particles on account of them
having a low volumetric fraction when 𝜕𝜔 1 𝜕𝑘 𝜕𝜔
(𝛤𝜔 ) + (1 − 𝐹1 )2 𝜌𝜎𝜔2 𝜔 . (4)
compared with the continuous phase. The 𝜕𝑥𝑗 𝜕𝑥𝑗 𝜕𝑥𝑗
interaction between the continuous phase and 𝜕(𝜌𝑘) 𝜕(𝜌𝑈𝑗 𝑘) 𝜕 𝜕𝑘
the solid phase is modelled using the one-way + ̃𝑘 − 𝛽 ∗ 𝜌𝜔𝑘 +
=𝑃 (𝛤 ). (5)
𝜕𝑡 𝜕𝑥𝑗 𝜕𝑥𝑗 𝑘 𝜕𝑥𝑗
coupling approach. Unlike the two-way 𝜇 𝜇
coupling approach wherein the change in with, Γk = 𝜇 + 𝜎 𝑡 , Γω = 𝜇 + 𝜎 𝑡 ,
𝑘 𝜔
momentum of the solid particles is taken as
∂(Ui )
source terms in the momentum equation of the 𝑃𝑘 = 𝜏𝑖𝑗 ̃𝑘 = 𝑚𝑖𝑛(𝑃𝑘 ; 𝑐1 𝜀).
,𝑃
∂xj
continuous phase, the one-way coupling
approach neglects the effect that the particles where, 𝛤𝑘 and 𝛤𝜔 are effective diffusivities and
have on the continuous phase on account of taken C1=10, β*=0.09. F1 is the blending function
them being low in volume when compared to which helps in resolving the near wall flow with
the continuous phase. The erosion calculation the k-ω model and the rest of the flow by k-ɛ and
at the wall boundaries is calculated using the is given in Eq. 6.
erosion model proposed by Oka et al. [33].
𝐹1 = tanh(𝑎𝑟𝑔14 ). (6)

2.5 Governing equations for continuous √𝑘 500𝜈 4𝜌𝑘𝜎𝜔

Where, 𝑎𝑟𝑔1 min (𝑚𝑎𝑥 ( 𝜔𝑦 𝛽∗ ; 𝑦2 𝜔 ) ; 𝑦2 𝐶𝐷 2 ) .
phase 𝑘𝜔

σω1 = 2, σω2 = 1.168.

The continuous phase is resolved using the
continuum approach, and the standard continuity 1 𝜕𝑘 𝜕𝜔
𝐶𝐷𝑘𝜔 = max (2𝜌𝜎𝜔2 ; 1.0𝑒 −10 ).
equation is used for mass conservation. The 𝜔 𝜕𝑥𝑗 𝜕𝑥𝑗
continuity equation is given in Eq. 1. The SST k-ω model introduces an upper limit for
𝜕𝜌
+ ∇ ∙ (𝜌𝜐⃗) = 𝑆𝑚 . (1) turbulent stress in boundary layers and as a
𝜕𝑡 result µt is given in Eq. 7.
Where 𝑆𝑚 is the term associated with mass 𝑎1 𝑘
addition from a source term and 𝜌 and 𝜐 are 𝜇𝑡 = 𝜌 . (7)
𝑚𝑎𝑥(𝑎1 𝜔;√2 𝑆∙𝐹2 )
density and velocity of the continuous phase.
Where a1=0.31, and 𝐹2 is blending function given
by:
The momentum for the continuous phase is
resolved using the Reynolds-Averaged Navier- 𝐹2 = tanh(𝑎𝑟𝑔24 ) , 𝑎𝑟𝑔2 = ( 2
√𝑘 500𝜐
; 𝑦2 𝜔 ).
Stokes equation (RANS). The SST k-ω model, 𝜔𝑦 𝛽 ∗

386
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

2.6 Governing equations for dispersed phase material hardness(Hv) and shape of erodent
particles. v and d stand for the velocity of impact
The dispersed phase is resolved using the and particle diameter, whereas v’ and d’ indicate
Lagrangian approach. The Eq. 8 for the solid the standard values for the same, which are
particle therefore are: utilized for erosion damage correlations in the
⃗⃗⃗⃗⃗⃗
𝑑𝑢 ⃗⃗− ⃗⃗⃗⃗⃗⃗
𝑢 𝑢𝑝 𝑔⃗⃗(𝜌𝑝 −𝜌) experiments. Experimentally obtained exponents
𝑑𝑡
𝑝
= 𝜏𝑟
+ 𝜌𝑝
+ 𝐹⃗ . (8) of v and d can also be used for dimesionless terms
of (v/v’) and (d/d’). All constants are reproduced
Where 𝐹⃗ is an additional acceleration accounting from the Oka et al. [33] and given in Table 2.
for forces such as gravity forces, virtual mass
⃗⃗− ⃗⃗⃗⃗⃗⃗
𝑢 𝑢𝑝 2.8 Boundary conditions, solution control
force etc. 𝜏𝑟
is the drag force per particle mass
and convergence
and 𝜏𝑟 is the particle relaxation time given by: -
2
𝜌𝑝 𝑑𝑝 24 ANSYS fluent R16.0 was used for the numerical
𝜏𝑟 = .
18𝜇 𝐶𝑑 𝑅𝑒 simulation. The semi-implicit method for
Where 𝑅𝑒 is the Reynolds number relating to the pressure linked equations was utilized for solving
relative velocity of the particle with respect to the pressure-velocity coupling. The convective
the flow velocity. terms and diffusion terms are solved using the
⃗⃗𝑝 −𝑢
𝜌𝑑𝑝 |𝑢 ⃗⃗| second order upwind scheme and second order
𝑅𝑒 = 𝜇
. central differencing scheme, respectively. For the
continuous phase, the mass flow rate was
2.7 Erosion modelling specified at the inlet, and the outlet was
considered a pressure outlet. The impeller was
The model proposed by Oka et al. [34] was used set at a speed of 1450 rpm, which was the design
for calculating the erosion rate at the walls. The condition for the pump. The casing was set as a
erosion rate at the walls is given in Eq. 9. stationary wall with no-slip boundary condition,
1
and the blades were set as moving walls moving
𝛦(𝛼) = 109
𝑚̇𝜌𝑚 (𝑠𝑖𝑛 𝛼)𝑛1 at the speed of the impeller. Interfaces were
specified, firstly between the inlet pipe and the
(1 + 𝐻𝑣 (1 − 𝑠𝑖𝑛 𝛼))𝑛2 (9)
impeller inlet and secondly between the impeller
𝑏𝑘1
𝑣 𝑘2 𝑑 𝑘3 outlet and the casing inlet. The SiO2 and SiC
𝐾(𝑎𝐻𝑣 ) ( ′) ( ′) .
𝑣 𝑑 erodent particles were treated as a discrete
Where , Ε(𝛼) is the rate of erosion in kg/s for a phase. These particles were assumed to be of a
particle impact angle of α, K1 is a function of uniform diameter of 106 µm and were released
particle properties like angularity, K2 is the from the pipe inlet and escaped the flow domain
velocity constant which is independent of the from the outlet. The particles were set to have
particle diameter but is dependent on the erodent perfectly elastic collisions with the surface of the
particle and the material hardness, K3 is the casing and the blades. The erosion values at the
particle size constant independent of the particle casing, vane, front shroud, and back shroud were
velocity but dependent on the erodent particle calculated using a user defined function, which
and material hardness, a and b are the load utilized the model given by Oka et al. [33] and as
relaxation properties of the material and n1 and mentioned in the equation above.
n2 are constants which are a function of the

Table 2. Different constants for the erosion modeling (Oka et al. [33,34]).
Pump Hv v’ d’ Density
K K1 K2 K3 n1 n2 a b Erodent
material (GPa) (m/s) (µm) (kg/m3
CS 65 -0.12 2.32 0.19 0.74 1.82 0.30 0.604 1.34 104 326 )
7700 SiO2
CS 45 -0.05 3.07 0.19 0.74 2.09 0.30 0.604 1.34 99 326 7700 SiC
SS 304 65 -0.12 2.31 0.19 0.82 0.88 1.18 0.028 2.9 104 326 7890 SiO2
SS 304 45 -0.05 3.28 0.19 0.82 0.96 1.18 0.028 2.9 99 326 7890 SiC
SS 316 65 -0.12 2.32 0.19 0.74 1.78 1.19 0.056 1.37 104 326 7730 SiO2
SS 316 45 -0.05 3.08 0.19 0.74 2.04 1.19 0.056 1.37 99 326 7730 SiC
CS= Carbon Steel, SS= Stainless Steel

387
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

Table 3. Various hydraulic parameters of the pump at different regimes.

Inlet total Outlet total Hydraulic
Discharge Torque Input power Output power Head
pressure pressure efficiency
(cumec) (N-m) (Watt) (Watt) (m)
(Pa) (Pa) (%)
0.0035 101793 162657 1.64 248.72 213.02 6.21 85.64
0.0036 101818 161939 1.66 253.33 216.43 6.12 85.43
0.0037 101847 161378 1.68 255.79 220.26 6.06 86.11
0.0038 101886 160047 1.70 258.48 221.01 5.92 85.50
0.0039 101903 159946 1.73 262.66 226.36 5.91 86.17
0.0040 101924 159879 1.75 267.11 231.82 5.90 86.78
0.0042 101980 159399 1.80 274.42 241.15 5.85 87.87
0.0043 102033 158638 1.83 278.58 243.40 5.77 87.37
0.0044 102084 156567 1.85 281.82 239.72 5.55 85.06
0.0045 102149 156076 1.89 286.98 242.67 5.49 84.55

2.9 Validation of computational methodology These heads were compared against the results
obtained by the numerical strategy of the pump,
For the validation of the computational as mentioned above using ANSYS fluent R16.0.
methodology explained above the experimental The results were found to be in good agreement
heads developed at different discharge by water (±2 % deviation) with the experimental results
handling industrial centrifugal pump at designed provided, as shown in Fig. 4b. Corresponding
speed of 1450 rpm were chosen which were pressure contours have also been shown in Fig.
provided by Kirloskar Pvt. Ltd., Pune, Maharashtra, 4a for the sake of clarity.
India. Details of all the data are given in Table 3.

3. RESULTS AND DISCUSSIONS

The erosion values for pump casing, blades, and

hub are determined using computational
methodology as explained in section 2.0 above at a
designed speed of 1450 rpm and is summarized in
Fig. 5. Analysis of the results so obtained has been
explained in the subsequent sections.

Model of centrifugal pump

Grid generation
a)
7.0
Experimental Turbulence model
6.5 CFD
Run user define function
6.0 For erosion modelling
Total Head (m)

5.5

5.0 Boundary
conditions
4.5

4.0 Solutions
0.0032 0.0036 0.0040 0.0044 0.0048
3 -1
Discharge (m - s )
Visualization and analysis of results
b)
Fig. 4. (a) Pressure contours for 3d model, (b) Fig. 5. Flow Chart of the adopted computational
Validation of head. methodology.

388
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

3.1 Erosion density distribution along & Kim [5], Gandhi et al. [9] and Rawat et al.
various components of the pump [10], Gupta et al. [35] for the various
concentrations of solids.
Before analyzing the pattern of erosion wear due
-7
to solid particles, it is inevitable to see how the 4.0x10

Erosion Rate (kg-s-1m-2)

particles are traversing inside the pump along Carbon steel_SiC
with the fluid. Thus, particle tracks for each of -7
SS316_SiC
3.0x10
the cases studied in the current work have been SS304_SiC
shown in Figs. 6a and 6b.
-7
2.0x10

-7
1.0x10

0.0
0 5 10 15 20 25 30
Concentration(%)
a)
-6
(a) (b) 6.0x10

Erosion Rate (kg-s-1m-2)

Carbon steel_ SiO2
Fig. 6. Solid particle density distribution for SiC at (a) 5.0x10
-6

Cw = 5 %, (b) Cw = 25 %. SS 316_SiO2
4.0x10
-6
SS 304_SiO2
These shall be analyzed in subsequent sections 3.0x10
-6

for each component of the pump. The erosion

rate has been calculated using the one-way 2.0x10
-6

coupling method. Since the continuous phase -6

1.0x10
has been accurately resolved, the erosion rate
calculated in the following sections would only 0.0
be dependent on erosion parameters such as 0 5 10 15 20 25 30
type of pump material and the type of erodent Concentration(%)
material used and the concentration of the b)
-6
erodent. The particle track as shown in Fig. 6 7.0x10
Erosion Rate (kg-s-1m-2)

Carbon steel_ SiC

depicts that the number of particles impacting 6.0x10
-6
Carbon steel_ SiO2
with the pump increases whenever there is a
5.0x10
-6 SS 316_SiC
change in the direction of the flow of fluid.
-6
SS 316_SiO2
4.0x10
3.2 Erosion at pump casing SS 304_SiC
-6
3.0x10 SS 304_SiO2
The average values of erosion wear at pump 2.0x10
-6

casing have been calculated separately from the -6

1.0x10
computational methodology at various
concentrations ranging from 5 % to 25 % for 0.0
0 5 10 15 20 25 30
both of the erodent material (SiO 2 and SiC) on
Carbon steel, Stainless steel 304 and Stainless Concentration(%)
steel 316 pump materials separately. The c)
results so obtained have been drawn in Fig. 7. Fig. 7. Erosion rate in spiral casing of carbon steel,
The figure shows that as the concentration of SS316 and SS304, (a) For erodent particle SiC, (b) For
particles increases, the erosion rate increases erodent particle SiO2, (c) Comparison of erosion rate
almost linearly. This is obvious since with the for erodent particle SiC and SiO2.
increase in a number of particles due to an
increase in the concentration of solid particles, Further, the change of erodent material and
the overall contact surface area of particles target material (casing material) affects the
available for wear increases. These results are amount of erosion wear. For the same erodent
consistent with the results mentioned in Noon material, the rate of increase in erosion wear is

389
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

higher in the case of Carbon steel than SS 316 reveals that the erosion rate values for the
and SS 316 has a higher erosion rate than SS casing. For a detailed analysis of erosion on the
304. It is essential to mention that the pump casing at different highest at the
toughness of Carbon steel is lower than SS316 narrowest part of the casing regardless of the
and SS304 both, whereas the toughness of erodent used or the concentration of the
SS316 is lesser than SS304. Thus, the behaviors erodent. This is due to the fact that due to the
of erosion wear on Carbon steel can be easily narrower flow passage area, the particle
explained, but in the case of SS316 and SS304, velocities may be highest causing, more wear.
not only the toughness but also the way the The particles tracks are shown in Fig. 6 also
particles interact with the surfaces also play an suggest and depict the concentration of
important role. Values of n1 and n2 are higher in particles in the narrower zone of the casing.
the case of SS316 than SS304, Table 1, Oka et al. From Figs. 8 and 9, the contours for the spiral
[33,34]. Figure 7c reveals that the erosion wear casing, the erosion rate is found to be nearly
for SiO2 is higher than SiC on all three target uniform in the rest of the flow passage except
materials since the hardness of SiO 2 is higher for the regions where the flow direction is
than SiC. The experimental values shown in the changed suddenly. The sudden change of
above graphs are average values for the casing. direction causes the secondary flow of particles
For a detailed analysis of erosion on the pump and a centrifugal action on the particles
casing at different locations, the erosion rate providing relatively higher energies to the
contours have been plotted in Figs. 8 and 9. particles, which results in relatively more wear
Careful analysis of the figures (Figs. 8 and 9) than the rest of the part of the spiral casing.

(a) (i) Cw = 5 % (ii) Cw = 15 % (iii) Cw = 25 %

(b) (i) carbon steel (ii) SS316 (iii) SS304

Fig. 8. Erosion rate contours at spiral casing with SiC as erodent for (a) Carbon Steel at different concentrations,
(b) Erosion contour for carbon steel, SS316 and SS304 at Cw = 25 % .

390
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

(a) (i) Cw = 5 % (ii) Cw = 15 % (iii) Cw = 25 %

(b) (i) carbon steel (ii) SS316 (iii) SS304

Fig. 9. Erosion rate contours at spiral casing with SiO2 as erodent (a) For carbon steel at different concentrations,
(b) Erosion contour for carbon steel, SS316 and SS304 at Cw = 25 %.

3.3 Erosion wear on impeller vanes 1.8x10

-6
Erosion Rate (kg-s-1m-2)

Carbon steel_ SiO2

-6
1.5x10 SS316_SiO2
The erosion wear variation rate at the vanes of
1.2x10
-6 SS304_SiO2
the pump has the same patterns as that for the
pump casing i.e., erosion wear increases as the 9.0x10
-7

concentration increases, and SiO2 creates more -7

6.0x10
wear than SiC (Fig. 10c). Also, Carbon Steel, SS
316 and 304 observe erosion wear due to
-7
3.0x10

erodent materials in increasing fashion (Fig. 10). 0.0

Whereas, the erosion in the blades occurs 0 5 10 15 20 25 30
primarily at the tip of the blades and at the top Concentration(%)
b)
of the blades (Figs. 11 and 12), where the axial 2.4x10
-6
Erosion Rate (kg-s-1m-2)

flow from the inlet pipe is turned to the radial Carbon steel_ SiC
-6 Carbon steel_ SiO2
outward flow in the blade passage. 2.0x10
SS316_SiC
-6
1.6x10 SS316_SiO2
-7
1.0x10
Erosion Rate (kg-s-1m-2)

Carbon steel_SiC 1.2x10

-6 SS304_SiC
-8 SS316_SiC SS304_SiO2
8.0x10
SS304_SiC 8.0x10
-7

-8
6.0x10 4.0x10
-7

4.0x10
-8 0.0
0 5 10 15 20 25 30
-8
Concentration(%)
2.0x10
c)
0.0
Fig. 10. Erosion rate in vane of carbon steel, SS316
0 5 10 15 20 25 30 and SS304 (a) For erodent particle SiC, (b) For
Concentration(%) erodent particle SiO2, (c) Comparison of erosion rate
a) for erodent particle SiC and SiO2.

391
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

To visualize this aspect better, an enlarged view than the erosion rate on the pressure side of the
of vanes has been shown in Fig. 11d. The erosion blade as the velocities of the fluid is higher on
rate on the suction side is also found to be more the suction side.

(a) (i) Cw = 5% (ii) Cw = 15% (iii) Cw = 25%

(b) (i) carbon steel (ii) SS316 (iii) SS304

Suction side

Pressure Side

Leading edge

Tailing edge

(c) Enlarged view of vane depicting wear incurred at Cw=25 % for carbon steel for SiC erodent particle

Fig. 11. Erosion rate contours at the vanes with SiC as erodent (a) For carbon steel at different concentrations,
(b) Erosion contour for carbon steel, SS316 and SS304 at C w = 25 %, (c) Enlarge view of vanes.

392
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

(a) (i) Cw = 5 % (ii) Cw = 15 % (iii) Cw = 25 %

(b) (i) carbon steel (ii) SS316 (iii) SS304

Fig. 12. Erosion rate contours at the vanes with SiO2 as erodent for (a) Carbon Steel at different concentrations,
(b) Erosion contour for carbon steel, SS316 and SS304 at Cw =25 %.

3.4 Erosion wear at impeller shrouds

-8
7.0x10
Erosion Rate (kg-s-1m-2)

-8
Carbon steel_SiC
6.0x10
SS316_SiC
The erosion wear variation rate at the front shroud 5.0x10
-8
SS304_SiC
and back shroud of the pump has the same -8
4.0x10
patterns as that for the pump casing i.e., erosion
wear increases as the concentration increases, SiO2
-8
3.0x10

creates more wear than SiC (Figs. 13c and 14c). 2.0x10
-8

Also, Carbon Steel, SS316, and 304 observe 1.0x10

-8

erosion wear due to erodent materials in an 0.0

increasing fashion (Fig. 13a and Fig. 14b 0 5 10 15 20 25 30
respectively). As far as the overall average wear Concentration (%)
at the shrouds is concerned, the observations a)
-6
1.2x10
are no different than those which were observed
Erosion Rate (kg-s-1m-2)

Carbon steel_SiO2
in case of the casing and the vanes. But, the 1.0x10
-6
SS316_SiO2
erosion rate at the front shroud is found to be
8.0x10
-7 SS304_SiO2
maximum at the outer edge of the shroud
regardless of the pump material or erodent 6.0x10
-7

material, as can be seen from Fig. 15 to Fig. 18.

-7
The reason for this is the same as was for vanes. 4.0x10

Due to increased velocity at the outer edges, the 2.0x10

-7

erosion is more whereas the back shrouds

0.0
receive very fewer solids particles than the front 0 5 10 15 20 25 30
shroud. This causes less wear at the back shroud Concentration (%)
than a front shroud. b)

393
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

-6 -6
Erosion Rate (kg-s-1m-2) 1.6x10 1.4x10

Erosion Rate (kg-s-1m-2)

Carbon steel_ SiC
-6 Carbon steel_SiO2
1.4x10 Carbon steel_ SiO2 1.2x10
-6

-6 SS316_SiO2
1.2x10 SS316_SiC -6
1.0x10
SS304_SiO2
1.0x10
-6 SS316_SiO2 -7
8.0x10
8.0x10
-7 SS304_SiC
-7
-7
SS304_SiO2 6.0x10
6.0x10
-7
-7 4.0x10
4.0x10
-7
-7 2.0x10
2.0x10
0.0 0.0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Concentration (%) Concentration (%)
c) b)
-6
1.6x10
Fig. 13. Erosion rate in front shroud of carbon steel,

Erosion Rate (kg-s-1m-2)

-6
Carbon steel_ SiC
SS316 and SS304 (a) For erodent particle SiC, (b) For 1.4x10 Carbon steel_ SiO2
erodent particle SiO2, (c) Comparison of erosion rate 1.2x10
-6
SS316_SiC
for erodent particle SiC and SiO2. 1.0x10
-6
SS316_SiO2
8.0x10
-7 SS304_SiC
-7
1.0x10 SS304_SiO2
-7
Carbon steel_SiC 6.0x10
Erosion Rate (kg-s-1m-2)

8.0x10
-8 SS316_SiC 4.0x10
-7

SS304_SiC -7
-8 2.0x10
6.0x10
0.0
-8 0 5 10 15 20 25 30
4.0x10
Concentration (%)
2.0x10
-8 c)
Fig. 14. Erosion rate in back shroud of carbon steel,
0.0
0 5 10 15 20 25 30 SS316 and SS304 (a) For erodent particle SiC, (b) For
Concentration (%) erodent particle SiO2, (c) Comparison of erosion rate
a) for erodent particle SiC and SiO2.

(a) (i) Cw = 5 % (ii) Cw = 15 % (iii) Cw = 25 %

(b) (i) carbon steel (ii) SS316 (iii) SS304

Fig. 15. Erosion rate at front shroud with SiC as erodent for (a) Carbon Steel at different concentrations, (b)
Erosion contour for carbon steel SS316 and SS304 at Cw = 25 %.

394
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

(a) (i) Cw = 5% (ii) Cw = 15% (iii) Cw = 25%

(b) (i) carbon steel (ii) SS316 (iii) SS304

Fig. 16. Erosion rate at front shroud with SiO2 as erodent for (a) Carbon Steel at different concentrations, (b)
Erosion contour for carbon steel SS316 and SS304 at Cw = 25 %.

(a) (i) Cw = 5% (ii) Cw = 15% (iii) Cw = 25%

(b) (i) carbon steel (ii) SS316 (iii) SS304

Fig. 17. Erosion rate at Back Shroud with SiC as erodent (a) For carbon steel at different concentrations, (b)
Erosion contour for carbon steel, SS316 and SS304 at Cw =25 %.

(a) (i) Cw = 5 % (ii) Cw = 15 % (iii) Cw = 25 %

(b) (i) carbon steel (ii) SS316 (iii) SS304

Fig. 18. Erosion rate at Back Shroud with SiO2 as erodent (a) For carbon steel at different concentrations, (b)
Erosion contour for carbon steel, SS316 and SS304 at Cw =25 %.

395
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

3.5 Comparison of erosion wear at different

-7
3.5x10
Spiral casing_SiC
components of the pump

Erosion Rate (kg-s-1m-2)

-7
3.0x10 Front shroud_SiC
-7 Back shroud_SiC
2.5x10
Apart from the individual analysis of various Single vane_SiC
components of the pump, a cumulative
-7
2.0x10
comparison of erosion wear incurred at various 1.5x10
-7

components of the pump has also been carried -7

1.0x10
out and graphs between erosion wear, and
concentrations have been drawn in Figs. 19-21
-8
5.0x10
for every component for all pump material and 0.0
erodent particles considered in the study. The 0 5 10 15 20 25 30
Concentration(%)
figures show that the pump casing receives
a)
maximum wear than the rest of the components. -6
5.0x10
At the highest concentration of 25 %, the casing Spiral casing_SiO2

Erosion Rate (kg-s-1m-2)

receives around 7-8 times higher wear than 4.0x10
-6 Front shroud_SiO2
vanes and shrouds, which is due to the relatively Back shroud_SiO2
more interaction of the dispersed phase with the 3.0x10
-6
Single vane_SiO2
casing while the pumping action is taking place.
Whereas, the vanes receive relatively higher 2.0x10
-6

erosion wear than the shrouds as shrouds

interact with the solid’s particles least. But in 1.0x10
-6

comparison to the casing, both of the

components can be considered to receive almost 0.0
0 5 10 15 20 25 30
the same amount of wear for all cases of target Concentration(%)
materials of pump and erodent particles. b)
Fig. 20. Erosion rate of different parts of pump for SS316
4.0x10
-7 (a) SiC erodent particle, (b) SiO2 erodent particle.
-7 Spiral casing_SiC -7
3.5x10 1.2x10
Erosion Rate (kg-s-1m-2)

Front shroud_SiC
Spiral casing_SiC
Erosion Rate (kg-s-1m-2)

-7
3.0x10 Back shroud_SiC -7
1.0x10 Front shroud_SiC
-7 Single vane_SiC Back shroud_SiC
2.5x10 -8
-7
8.0x10 Single vane_SiC
2.0x10
-8
1.5x10
-7 6.0x10
-7 -8
1.0x10 4.0x10
-8
5.0x10 -8
2.0x10
0.0
0 5 10 15 20 25 30 0.0
Concentration(%) 0 5 10 15 20 25 30
Concentration(%)
a)
-6 a)
6.0x10 -6
4.0x10
Spiral casing_SiO2
Sprial casing_SiO2
Erosion Rate (kg-s-1m-2)

-6
Erosion Rate (kg-s-1m-2)

5.0x10 Front shroud_SiO2 Front shroud_SiO2

-6
-6 Back shroud_SiO2 3.0x10
4.0x10 Back shroud_SiO2
Single vane_SiO2 Single vane_SiO2
-6
3.0x10 2.0x10
-6

-6
2.0x10
-6
-6
1.0x10
1.0x10

0.0 0.0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Concentration(%) Concentration(%)
b) b)
Fig. 19. Erosion rate of different parts of pump for carbon Fig. 21. Erosion rate of different parts of pump for SS304
steel (a) SiC erodent particle, (b) SiO2 erodent particle. (a) SiC erodent particle, (b) SiO2 erodent particle.

396
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

3.6 Effect of particle size on erosion wear higher than at any other component of
the pump. Thus, the design of the casing
Further to understand the effect of particle size is more critical than the other
of erosion wear, a limited study is conducted on components as far as erosion wear to
different target materials of Carbon steel, SS316 save overall cost is concerned.
and SS304 for the different particle sizes of SiO2
 The vanes of the pumps observe more
(50-250 µm) at fixed Cw = 10%. The results so
variation of wear from the tip to the top
obtained have been drawn in Fig. 22.
for all the three steel. Thus, these
1.5x10
-6 components are to be designed more
Carbon steel_SiO2 carefully by the designers of such pumps.
Erosion Rate (kg-s-1m-2)

 The study also depicts that the edges of

-6
1.2x10 SS 316_SiO2

SS 304_SiO2
the front shrouds are to be designed
9.0x10
-7
carefully for erosion wear as they receive
more wear.
-7
6.0x10
 All pump parts showed that SS304 is the
-7
most suitable material for pumps handling
3.0x10 sand slurries to receive minimum wear in
hydraulic power plants.
0.0
50 100 150 200 250 300  The steels investigated in the current
Particle size (microns) work also showed more resistance to
Fig. 22. Effect of solid particle size on single vane for erosion due to SiC when compared to SiO2
SiO2 at Cw = 10 % and V = 10 m/s. at all pump locations.

It is clearly seen that as the particle size

increases the amount of erosion wear also Acknowledgement
increases for all target materials. The results so
obtained are consistent with Gandhi et al. [9] The authors would like to acknowledge the
and Zhong and Minemura [11]. Whereas, the TEQIP-III for funding the project. Authors also
effect of particle size is more pronounced for like to thank Kirloskar Brothers Limited, Pune,
Carbon steel than the rest of the two target India, who not only supplied the experimental
materials on account of its lesser toughness. data for the current work but also pointed out
the problems faced by their pumps. This gave
the authors a motivation to work towards
4. CONCLUSIONS minimization of the associated wear. In
particular, the contribution of Dr. S. N. Shukla in
In this work, a comparative erosion wear study for terms of technology transfer of the knowledge is
different pump materials has been carried out much appreciated.
through comprehensive numerical hydrodynamic
analyses. The results are found to be consistent
with many researches available in literature [2,6- REFERENCES
8]. The following are the major conclusions that
can be made from the present work.
[1] B. Kishor, G.P. Chaudhari, S.K. Nath, Slurry
erosion of thermo-mechanically processed
 The numerical investigation developed in
13Cr4Ni stainless steel, Tribology International,
the current work predicts that the erosion vol. 93, pp. 50–57, 2016, doi:
rate at the locations in the pump parts 10.1016/j.triboint.2015.08.048
where the flow suddenly changes
direction is significantly higher than that [2] S. Gupta, A. Khandelwal, A.K. Ghose, I.
Chakrabarty, Slurry Erosion Behavior of
at nearby locations.
Destabilized and Deep Cryogenically Treated Cr-
 It is found that for all the three common Mn-Cu White Cast Irons, Tribology in Industry,
steel, the erosion at the spiral casing is vol. 38, no. 4, 486-495, 2016.

397
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

[3] S.R. More, D.V. Bhatt, J.V. Menghani, Study of the [14] D.H. Mesa, A. Toro, A. Sinatora, A.P. Tschiptschin,
Parametric Performance of Solid Particle Erosion The effect of testing temperature on corrosion–
Wear under the Slurry Pot Test Rig, Tribology in erosion resistance of martensitic stainless steels,
Industry, vol. 39, no. 4, 471-481, 2017, doi: Wear, vol. 255, iss. 1-6, pp. 139–145, 2003, doi:
10.24874/ti.2017.39.04.06 10.1016/S0043- 1648(03)00096-6
[4] S. Aribo, I. Adedapo, C. Nwogwugwu, O. [15] S. Kumar, Evalution of erosion wear of cetrifugal
Olaniran, A. Olaseinde, O. Ige, P.A. Olubambi, pump using cfd, International Journal of Advance
Erosion-corrosion Behaviour of Dual Phase Research in Science and Engineering, vol. 6, iss.
Medium Carbon Steel using a Designed Slurry Pot, 12, pp. 1374-1381, 2017.
Tribology in Industry, vol. 40, no. 2, 239-246,
[16] Y. Xiao, B. Guo, S.-H. Ahn, Y. Luo, Z. Wang, G. Shi,
2018, doi: 10.24874/ti.2018.40.02.07
Y. Li, Slurry Flow and Erosion Prediction in a
[5] A.A. Noon, M.-H. Kim, Erosion wear on Centrifugal Pump after Long-Term Operation,
centrifugal pump casing due to slurry flow, Wear, Energies, vol. 12, iss. 8, pp. 1-17, 2019, doi:
vol. 364-365, pp. 103-111, 2016, doi: 10.3390/en12081523
10.1016/j.wear.2016.07.005
[17] J. Kumar, G. Tiwari, A. Rawat, V.K. Patel, Effect of
[6] R. Tarodiya, B.K. Gandhi, Numerical simulation of Swirl Vanes Angle on Erosion Behavior of AISI 316
a centrifugal slurry pump handling solid-liquid Pipe Bend, Materials Today: Proceedings, In
mixture: Effect of solids on flow field and Press, 2020, doi: 10.1016/j.matpr.2020.01.026
performance, Advanced Powder Technology, vol.
[18] Y.K. Baghel, J. Kumar, B. Kishor, A. Rawat, V.K.
30, iss. 10, pp. 2225-2239, 2019, doi:
Patel, Effect of Hot Forging on the Slurry Erosion
10.1016/j.apt.2019.07.003
Wear of AISI 316 and AISI 410 Steel, Materials
[7] V. Gautam, A. Kumar, L. Prasad, V.K. Patel, An Today: Proceedings, In Press, 2020, doi:
Experimental Investigation on Slurry Erosion 10.1016/j.matpr.2020.02.366
Wear Characteristics of Brass Alloy, Materials
[19] A. Tudor, R. Nehriu, I. Radu, V. Dumitru, A Wear
Today: Proceedings, vol. 4, iss. 9, pp. 9879–
study case of ceramic ball seat valve, Tribology in
9882, 2017, doi: 10.1016/j.matpr.2017.06.286
Industry, vol. 25, no. 3&4, pp. 83-88, 2003.
[8] W.K. Chan, Detection of cavitation erosion in
[20] D. Kekes, P. Psyllaki, M. Vardavoulias, Wear Micro-
centrifugal pumps, International Journal of Heat
Mechanisms of Composite WC-Co/Cr-NiCrFeBSiC
and Fluid Flow, vol. 9, iss. 1, pp. 74-77, 1988,
Coatings. Part I: Dry Sliding, Tribology in
doi: 10.1016/0142-727X(88)90033-1
Industry, vol. 36, no. 4, pp. 361-374, 2014.
[9] B.K. Gandhi, S.N. Singh, V. Seshadri, Study of the
[21] D. Kekes, P. Psyllaki, M. Vardavoulias, G. Vekinis,
parametric dependence of erosion wear for the
Wear Micro-Mechanisms of Composite WC-Co/Cr-
parallel flow of solid–liquid mixtures, Tribology
NiCrFeBSiC Coatings. Part II: Cavitation
International, vol. 32, iss. 5, pp. 275–282, 1999,
Erosion, Tribology in Industry, vol. 36, no. 4, pp.
doi: 10.1016/s0301-679x(99)00047-x
375-383, 2014.
[10] A. Rawat, S.N. Singh, V. Seshadri, Erosion wear
[22] G. Tiwari, J. Kumar, V. Prasad, V.K. Patel, Derivation
studies on high concentration fly ash slurries,
of cavitation characteristics of a 3mw prototype
Wear, vol. 378–379, pp. 114-125, 2017, doi:
francis turbine through numerical hydrodynamic
10.1016/j.wear.2017.02.039
analysis, Materials Today: Proceedings, In Press,
[11] Y. Zhong, K. Minemura, Measurement of erosion 2020, doi: 10.1016/j.matpr.2020.02.297
due to particle impingement and numerical
[23] A.R. Al-Obaidi, Investigation of effect of pump
prediction of wear in pump casing, Wear, vol.
rotational speed on performance and detection of
199, iss. 1, pp. 36-44, 1996, doi: 10.1016/0043-
cavitation within a centrifugal pump using
1648(96)06974-8
vibration analysis, Heliyon, vol. 5, iss. 6, pp. 1-19,
[12] K.V. Pagalthivarthi, P.K. Gupta, V. Tyagi, M.R. Ravi, 2019, doi: 10.1016/j.heliyon.2019.e01910
CFD Predictions of Dense Slurry Flow in
[24] A.R. Al-Obaidi, Experimental Investigation of the
Centrifugal Pump Casings, International Journal of
Effect of Suction Valve Opening on the Performance
Mechanical and Mechatronics Engineering, vol. 5,
and Detection of cavitation in the Centrifugal Pump
no. 3, pp. 538-550, 2011.
Based on Acoustic Analysis Technique, Archives of
[13] D.R. Kaushal, T. Thinglas, Y. Tomita, S. Kuchii, H. Acoustics, vol. 44, no. 1, pp. 59–69, 2019,
Tsukamoto, CFD modeling for pipeline flow of fine doi: 10.24425/aoa.2019.126352
particles at high concentration, International Journal
[25] A.R. Al-Obaidi, Monitoring the Performance of
of Multiphase Flow, vol. 43, pp. 85–100, 2012, doi:
Centrifugal Pump under Single-Phase and
10.1016/j.ijmultiphaseflow.2012.03.005
Cavitation Condition: A CFD Analysis of the

398
 J. Kumar et al., Tribology in Industry Vol. 42, No. 3 (2020) 382-399

Number of Impeller Blades, Journal of Applied [32] V. Batalovic, Erosive Wear Model of Slurry Pump
Fluid Mechanics, vol. 12, no. 2, pp. 445-459, Impeller, Journal of Tribology, vol. 132, iss. 2, p.
2019, doi: 10.29252/jafm.12.02.29303 5, 2010, doi: 10.1115/1.4001167
[26] A.R. Al-Obaidi, Effects of Different Turbulence [33] Y.I. Oka, T. Yoshida, Practical estimation of
Models on Three-Dimensional Unsteady erosion damage caused by solid particle impact
Cavitating Flows in the Centrifugal Pump and Part 2: Mechanical properties of materials
Performance Prediction, International Journal directly associated with erosion damage, Wear,
of Nonlinear Sciences and Numerical vol. 259, iss. 1-6, pp. 102–109, 2005a, doi:
Simulation, vol. 20, iss. 3-4, pp. 487–509, 2019, 10.1016/j.wear.2005.01.040
doi: 10.1515/ijnsns-2018-0336
[34] Y.I. Oka, K. Okamura, T. Yoshida, Practical
[27] M.A. Rayan, M. Shawky, Evaluation of wear in a estimation of erosion damage caused by solid
centrifugal slurry pump, Proceedings of the particle impact Part 1: Effects of impact
Institution of Mechanical Engineers, Part A: Journal parameters on a predictive equation, Wear, vol.
of Power Engineering, vol. 203, iss. 1, pp. 19-23, 259, iss. 1-6, pp. 95–101, 2005, doi:
1989, doi: 10.1243/PIME_PROC_1989_203_003_02 10.1016/j.wear.2005.01.039
[28] Z. Shen, W. Chu, X. Li, W. Dong, Sediment erosion in [35] R. Gupta, S.N. Singh, V. Sehadri, Prediction of
the impeller of a double-suction centrifugal pump – uneven wear in a slurry pipeline on the basis of
A case study of the Jingtai Yellow River Irrigation measurements in a pot tester, Wear, vol. 184, iss.
Project, China, Wear, vol. 422–423, pp. 269–279, 2, pp. 169-178, 1995, doi: 10.1016/0043-
2019, doi: 10.1016/j.wear.2019.01.088 1648(94)06566-7
[29] Z. Wang, Z. Qian, Effects of concentration and size
of silt particles on the performance of a double-
suction centrifugal pump, Energy, vol. 123, pp. Nomenclature
36-46, 2017, doi: 10.1016/j.energy.2017.01.142
Cw Concentration by weight
[30] K.A. Kazim, B. Maiti, P. Chand, A correlation to
predict the performance characteristics of Ε(𝛼) Erosion rate (kg/m2-s-1)
centrifugal pumps handling slurries, Proceedings F 1& F 2 Blending function
of the Institution of Mechanical Engineers, Part Hv Hardness of solid particle
A: Journal of Power Engineering, vol. 211 iss. 2, 𝑆𝑚 Mass addition from a source term
pp. 147-157, 1997, doi: CS Carbon steel
10.1243/0957650971537060 SS Stainless steel
[31] J.F. Gulich, Effect of Reynolds Number and α Particle impact angle,
Surface Roughness on the Efficiency of v Impact velocity (m/s),
Centrifugal Pumps, Journal of Fluids 𝜇𝑡 Turbulent viscosity,
Engineering, vol. 125, iss. 4, pp. 670-679, 2003, 𝛤𝑘 , 𝛤𝜔 Effective diffusivities,
doi: 10.1115/1.1593711
τr Particle relaxation time

399