Experimental and Computational

Investigations to Evaluate the

Effects of Fluid Viscosity and
Particle Size on Erosion Damage
Risa Okita1
e-mail: risa-okita@utulsa.edu

Yongli Zhang
Brenton S. McLaury
Siamack A. Shirazi
Department of Mechanical Engineering,
The University of Tulsa,
Tulsa, OK 74104-3189

Zhang et al. (2006) utilized computational fluid dynamics (CFD) to examine the validity
of erosion models that have been implemented into CFD codes to predict solid-particle
erosion in air and water for inconel 625. This work is an extension of Zhangs work and
is presented as a step toward obtaining a better understanding of the effects of fluid viscosity and sand-particle size on measured and calculated erosion ratios, where erosion
ratio is defined as the ratio of mass loss of material to mass of solid particles. The erosion
ratios of aluminum 6061-T6 were measured for direct impingement conditions of a submerged jet. Fluid viscosities of 1, 10, 25, and 50 cP and sand-particle sizes of 20, 150,
and 300 lm were tested. The average fluid speed of the jet was maintained at 10 m/s. Erosion data show that erosion ratios for the 20- and 150-lm particles are reduced as the
viscosity is increased, whereas, surprisingly, the erosion ratios for the 300-lm particles
do not seem to change much for the higher viscosities. For all viscosities considered,
larger particles produced higher erosion ratios, for the same mass of sand, than smaller
particles. Concurrently, an erosion equation has been generated based on erosion testing
of the same material in air. The new erosion model has been compared to available models and has been implemented into a commercially available CFD code to predict erosion
ratios for a variety of flow conditions, flow geometries, and particle sizes. Because
particle speed and impact angle greatly influence erosion ratios of the material,
calculated particle speeds were compared with measurements. Comparisons reveal that,
as the particles penetrate the near wall shear layer, particles in the higher
viscosity liquids tend to slow down more rapidly than particles in the lower viscosity
liquids. In addition, CFD predictions and particle-speed measurements are used to
explain why the erosion data for larger particles is less sensitive to the increased
viscosities. [DOI: 10.1115/1.4005683]
Keywords: erosion modeling, CFD, viscosity, particle size

Introduction

Oil and gas, produced from reservoirs, usually contain impurities, such as sand particles. The sand particles impinge the wall
of flow lines and remove material from the pipe wall. The removal
of material by solid particles is called erosion. The erosion damage caused by particles can be very dangerous because pipes and
fittings can rupture suddenly without prior indication of failure.
Failures caused by erosion also can result in expensive repairs and
lost production time. To save money and increase safety, many
industries are using erosion models to predict when the pipeline
and fittings are susceptible to significant erosion damage.
In the literature, there are many equations developed to predict
solid-particle erosion [14]. The erosion equations normally
depend on particle impact speed and angle, as well as particle and
material properties. However, in various flow situations fluid
properties affect particle impact speed and angle. Therefore, computational fluid dynamics (CFD) is normally used to predict particle impact speed and angle so that this information can be used in
erosion models to predict solid-particle erosion [57]. Previously,
Zhang et al. [6] conducted a series of erosion measurements and
calculations to show that erosion equations can be utilized in CFD
software to predict erosion. They conducted a series of erosion
1
Corresponding author.
Contributed by the Fluids Engineering Division of ASME for publication in the
JOURNAL OF FLUIDS ENGINEERING. Manuscript received February 24, 2011; final manuscript received December 8, 2011; published online May 29, 2012. Assoc. Editor:
Mark F. Tachie.

Journal of Fluids Engineering

tests in gas and developed a representative erosion equation. Next,

the erosion equation was implemented into a commercially available CFD code and used to predict erosion for various gas and liquid cases. However, their study was limited to only one sandparticle size and only erosion resulting from sand in either air or
water [5,6]. To extend the previous work and predict the amount
of erosion resulting for various viscous liquids and particle sizes,
several investigators have measured speeds of particles and carrier
fluid, as well as erosion ratios for the direct impingement flow geometry for a range of carrier fluid viscosities and sand-particle
sizes [8,9].
The tasks and steps in the present work are shown in Fig. 1.
CFD-based erosion calculations in this work include the following
steps and contributions: (1) Flow speed calculations and particle
tracking are performed in the CFD program. Flow calculations
provide flow information such as fluid speeds and pressure. Particle tracking calculations provide particle information, such as particle speeds and particle impact conditions. (2) Flow and particle
speeds are validated by measuring the actual fluid and particle
speeds in the direct impingement geometry for liquids with
different viscosities utilizing laser Doppler velocimetry (LDV).
(3) Erosion data were collected for 6061-T6 aluminum target material and various particle types in air for the same flow geometry.
Erosion models were generated to represent these conditions. (4)
Erosion data for the same aluminum material were collected for
the same particle types and geometry in liquids with various viscosities. (5) Erosion ratios in liquid are calculated using the CFD
velocities and applying the erosion models from step (3). Once

C 2012 by ASME
Copyright V

JUNE 2012, Vol. 134 / 061301-1

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 1

Flow chart of erosion prediction using CFD

again, at this stage, the CFD predicted erosion ratios are validated
by comparisons with actual erosion measurements from step (4).

Background

2.1 Erosion Background. Erosion by solid particles has

been of great interest in the last few decades. The mechanisms of
solid-particle erosion have been studied by numerous researchers.
There have been many ideas proposed by researchers on how erosion occurs. Finnie (1972) listed the factors which may influence
the erosion of ductile material [10]. Among these factors are particle impact angle, speed, particle size, surface properties, shape of
the surface, particle shape, target strength, concentration of particles, and nature of carrier fluids.
Particle impact velocity is probably the most important factor
for erosion. The effect of particle velocity can easily overshadow
the changes in the other factors. There are many publications
which explain the relationship of erosion and particle velocity by
an empirical power law, as shown in Eq. (1):
ER / Vn

(1)

Lindsley and Marda (1999) studied the effect of velocity on erosion rates on 7030 brass and FeC martensite. Lindsley and
Marda reported that the empirical constant n is independent of target material and erosion mechanism, but governed by test conditions, such as particle characteristics and erosion test apparatus.
According to their work, erosion resulting from both brittle cracking and plastic deformation mechanisms showed the exponent n
to be 2.9, even though the erosion mechanisms are completely different [11]. Finnie (1978) examined the effect of particle velocity
on erosion and found that the exponent n should increase with
impact angle for a given range of velocities. He also compared the
value of n from other literature and said n ranges from 2.05 to
2.44 depending on test conditions [1].
Particle impact angle is another important factor for erosion.
Finnie (1972) derived the angle function from the equation of
motion for a rigid abrasive particle striking a ductile surface. For
aluminum alloys, the model shows that the maximum erosion
occurs at 13 deg and decreases to zero at 0 deg and 90 deg. This
angle function is in good agreement with experimental data,
except for high impact angles. As a reason for the discrepancy
between measurements and the model, Finnie states that the erosion mechanism at high angles is very different from the one at
low angles. At low angles, erosion is mostly caused by the cutting
mechanism, whereas at high angles, it is caused by the surface
roughening and low cycle fatigue fracture [10].
Particle properties also greatly affect the erosion of ductile material. For example, difference in angularity of particle shape can
061301-2 / Vol. 134, JUNE 2012

cause different erosion mechanisms, which yield different erosion

ratios. Winter and Hutchings (1974) studied the effect of particle
orientation at the impact surface [12]. Erosion measurements on
mild steel and lead were performed using flat-faced angular particles. They described the different erosion mechanisms based on
the angularity of abrasives by a rake angle. The rake angle, by definition, is the angle between the perpendicular to the surface and
the leading edge of the particle. It is found that cutting is favored
when the rake angle is positive or small negative values. At large
negative rake angles, plowing rather than cutting occurs. Because
spherical particles always have negative rake angles at impact,
plowing is the only possible deformation. On the other hand, most
of the abrasives are both round and angular, therefore, the deformation can be either cutting or plowing. Particle shapes not only
influence the erosion mechanism but also influence the particle
motion in fluid. Loth (2007) conducted comprehensive studies on
the drag coefficients for nonspherical particles with both regular
and irregular shapes [13]. Gavze and Shapiro calculated hydrodynamic forces and velocities of spheroidal particles near the wall.
They found that the effect of the wall on the forces and torques
decreases as the particle shape deviates from spherical. Furthermore, the rotational motion is more retarded for nonspherical particles than for spherical particles [14].
Particle size also influences erosion of ductile material as much
as its shape. Many researchers support the idea that particle size
influences erosion rates for smaller sizes; however, after a critical
size, which is reported to be 100 lm, effect of size on erosion is
negligible [10,15].
2.2 Erosion Ratio Equations. Many investigators have studied erosion mechanisms and found that erosion is influenced by
various factors, such as particle size, particle impact speed, impact
angle, and target material properties, such as density, hardness,
and many other factors [10,1618]. Erosion can be very complex,
because erosion is influenced by so many factors. Meng and
Ludema [16] studied previous erosion models and equations in
the literature and concluded that there is no universal model that
works for all flow conditions or materials for practical uses.
Because there is no universal model which works for all materials
or flow conditions, an erosion model must be developed for each
material and condition separately.
Two erosion equations are used in this paper to predict the
erosion ratio for aluminum 6061-T6. Equation (2) is called the
E/CRC (Erosion/Corrosion Research Center) equation and originally presented by investigators of E/CRC for carbon steels [2]
but is modified in this work:
 
kg
Fs  CBH0:59  V n  Fh
(2)
ER
kg
Hv 0:1023
0:0108

(3)

1
 sin hn1  1 Hnv3 1  sin hn2
f

(4)

BH
Fh

The equation calculates the erosion ratio based on particle

impact angle and speed, as well as material Brinell hardness. ER
is the erosion ratio in kg of material loss per 1 kg of sand throughput. Fs is a sharpness factor which ranges from 0.2 to 1 and
depends on the shape of sand particles. C and n are empirical constants. In literature, n is commonly found to be between 2.0 to 2.9
[5,6,10,11]. In this work, 2.41 is used for n. BH is the Brinell hardness of the target material calculated based on Vickers hardness,
Hv of the material using Eq. (3). V is the particle impact speed
and F(h) is the function of impact angle shown in Eq. (4). Equation (4) is a modified version of the one originally proposed by
Oka et al. [3,4]. n1, n2, and n3 are empirical constants and f is the
maximum value of F(h) and is used to normalize the function:
Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Table 1

Constants for erosion equations (aluminum 6061)

Okas equation

Variables
Values

Hv (GPa)
1.12

k
65

k1
0.12

k2
2.3(Hv)0.038

k3
0.19

n1
0.71(Hv)0.14

n2
2.4(Hv)0.94

Present work (E=CRC)

Variables
150 lm
300 lm

Hv (GPa)
1.12
1.12

Fs
0.50
1.00

f
5.27
2.19

n1
0.59
0.50

n2
3.60
2.50

n3
2.50
0.50

C
1.50E07
3.28E07

 k2  k3
 
kg
V
D
k1
K  Hv 
ER
 0
 q  Fh
kg
V0
D

(5)

Fh sin hn1  1 Hv 1  sin hn2

(6)

Equation (5) is the erosion model developed by Oka et al. in 2005

[3,4]. K, k1, k2, and k3 are the constants based on particle properties and hardness of the target material. q is the density of the target material. V0 and D0 are reference values based on experiments
(104 m/s and 326 lm, respectively). F(h) is the angle function and
is shown in Eq. (6). The difference between the forms of Eqs. (4)
and (6) is that there is an additional constant, n3, as an exponent to
the Vickers. hardness in Eq. (4).
Values for the constants in each equation are shown in Table 1.
The E/CRC equation is defined for two types of sand with particle
sizes of 150 and 300 lm. Oka developed the erosion equation
based on air testing. The E/CRC equation is also developed based
on direct impingement testing in air. The details of how the
E/CRC equation was developed are shown in the next section.
The erosion equation developed through gas testing is used to predict erosion for a liquid carrier fluid. Using the particle information such as particle speed and angle for each impingement from
CFD, total erosion of the target material are calculated.

2.3 CFD Background. The first step in CFD erosion prediction is flow simulation. Time-averaged Navier-Stokes equations
applying a Reynolds stress model (RSM) are applied in this work
to calculate the flow fields presented. Equation (7) shows the modeled transport equation for the RSM.
@qu01 u0j
@t

@quk u01 u0j

@xk

DT;ij DL;ij Pij /ij eij

(7)

where
"
#
@ lt @u0l u0j
DT;ij
@xk rk @xk
"
#
@u0l u0j
@
DL;ij
l
@xk
@xk


@uj
@ui
u0j u0k
Pij q u0l u0k
@xk
@xk




e 0 0 2
2
/ij C1 q ul uj  dij k 6C2 Pij  Cij  dij P  C
k
3
3
2
eij dij qe
3

(7a)

(7b)
(7c)
(7d)
(7e)

The first term on the LHS is the local time derivative term and
the second term is the convection term. DT;ij is the turbulent diffusion term, DL;ij is the molecular diffusion, Pij is the stress production term, Uij is the pressure strain term, and eij is the dissipation
2
term. The turbulent viscosity lt is computed as lt qCl ke .
The value of Cl is 0.09. The value of rk is 0.82. C1 is 1.8 and C2
is 0.6, Pij and Cij are defined in Eqs. (2)(10) and P 1/2Pkk,
Journal of Fluids Engineering

C 1/2Ckk. The scalar dissipation rate e can be modeled as

follows:
@qe @qui e
@

@t
@xi
@xj


 
lt @e
1
e
e2
 C1e Pii   C2e q
l
2
k
re @xj
k
(8)

The values of constants in Eq. (8) are re 1:0, C1e 1:44, and
C2e 1:92. Turbulent kinetic energy k is obtained using either
Eq. (9) away from the wall or Eq. (10) near the wall:
1
k u0l u0l
2
@qk @qui k
@

@t
@xi
@xj


 
l @k
1

l t
 qe
2Pii
rk @xk

(9)

(10)

The turbulence models described above are known to be valid

only in the bulk of the flow. The existence of a wall greatly influences the nature of flow. In the near wall region, the gradients are
much larger than the core flows, and the momentum and other scalar transport occur most actively. Therefore, it is important to
accurately represent the flow in the near-wall region to have successful predictions. Fluent [19] provides numerous approaches to
model near-wall flow; however, in this work, the standard wall
function is used.
After the flow calculation is completed, particle tracking is performed. In particle tracking, CFD releases virtual particles and
computes the trajectories of these particles using a particle equation of motion. Fluent 6.3 uses an Euler-Lagrange approach, in
which the fluid phase is treated as a continuum by solving RANS,
whereas the dispersed phase is solved by tracking a large number
of particles through the calculated flow field. In this approach, the
volume fraction of volume occupied by particles is assumed to be
very low. Therefore, interactions between particles are neglected,
and the effect of the particles on the fluid motion is also neglected.
In Fluent, a particle trajectory is calculated by integrating the trajectory equations for individual particles using the instantaneous
fluid velocity u u u0 where u is the average velocity and u0 is
the velocity fluctuation. Fluent uses a model called discrete random walk (DRW) to determine the instantaneous fluid velocity. In
DRW, a particle enters eddies continuously one after another.
Each eddy is characterized by its velocity fluctuation and its lifetime. The amount of time required for a particle to pass through
an eddy is called the eddy crossing time. During particle tracking,
Fluent keeps track of the time and distance that have elapsed for a
particle in a given eddy and compares these values with the eddy
lifetime and eddy crossing time. When the smaller of the two is
reached, the particle leaves the existing eddy and enters a new
eddy [19].
When a solid particle hits the wall, there is a change in momentum after the impact. The ratio of particle velocity after the
impingement to before the impingement is called restitution coefficient. Forder et al. and Tabakoff showed in their works, that the
restitution coefficients are dependent on particle impact angle
[20,21]. Tabakoff stated in his work that the rebound dynamics of
particles can be only described in statistical manner. Grant and
Tabakoff correlated restitution coefficients and the standard
JUNE 2012, Vol. 134 / 061301-3

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 2

Schematic of the experimental facility

deviations of coefficients with particle impingement angles [22].

Chen et al. [23] evaluated the rebound models proposed by Forder
[20] and the stochastic model proposed by Grant and Tabakoff
[22]. It was found that Grant and Tabakoffs model gives more
physical results than constant restitution coefficients when the carrier fluid is low-density gas. However, generally, when the carrier
fluid is liquid, the restitution coefficients have very small influence on the final prediction of erosion because the momentum
exchange between the liquid and particles is very strong.
During the particle tracking process, when particles hit the target wall, particle-impact information, such as particle speed,
angle, location of impact, and number of impacts, is saved on
each face cell on the target wall. Erosion rates at the wall are calculated using erosion models for each face on the wall by applying user-defined functions and are stored at the cell center of the
face. An erosion rate because of a single impact is calculated
using Eq. (11).


 
kg
kg
_p 1

ER
M
ER
sm2
kg
Af

(11)

ERkg=kg on the RHS of the equation is the erosion ratio calculated directly from the erosion models based on single impact
_ p . An
information. In CFD, each particle has its mass flow rate, M
erosion ratio is multiplied by a mass flow rate of the particle
impacting and divided by an area of the face being impacted, Af.
The resulting erosion rate is in units of mass loss per time and
area (kg/sm2). The erosion rates from multiple impacts in each
cell are summed in the end and reported as an overall erosion rate
for that cell. The total erosion rates, ER(kg/sm2) is converted to
ERkg=kg by multiplying the area of entire wall and dividing by
the total mass sand flow rate.

Fig. 3

061301-4 / Vol. 134, JUNE 2012

Experimental Facility and Conditions

3.1 Experimental FacilityLiquid Testing. An experimental facility was constructed to measure erosion for carrier fluids
with different viscosities and different particle sizes. Figure 2
shows a schematic of the experimental facility. The viscous liquid
is prepared in the reservoir tank prior to conducting the experiments by mixing CMC (carboxymethyl cellulose) and water. The
density of the mixture was essentially the same as the density of
water. The viscosities of the carrier fluid tested were 1, 10, 25,
and 50 cP. The CMCwater mixture is known to be a nonNewtonian fluid; however, at the low viscosities of the current test
conditions, the CMCwater mixture behaves as a Newtonian fluid.
Sand particles are mixed with liquid in the reservoir tank. The ratio of sand particles and liquid are usually 0.1% by volume. The
slurry mixer is turned on while taking measurements to maintain
the homogeneity of the mixture. The mixture of sand and liquid
circulates in the system as follows. The mixture travels from the
reservoir tank to the hydraulic pump and from the pump to the
straight nozzle in the testing tank. The bypass valve is used to
control the flow rate in the system during the measurements. The
mixture exits the nozzle and impinges the target wall and drains
from the testing tank back to the reservoir tank. The nozzle and
target wall are completely submerged in the mixture. The nozzle
inner diameter is 8 mm, and the distance from nozzle exit to the
wall is 12.7 mm. The average fluid speed at the nozzle exit was
kept at 10 m/s.
Flat surface coupons shown in Fig. 2 are used to measure the
erosion at the target wall. The target material is Aluminum 6061.
The weight of a coupon is measured before and after each test to
determine the metal weight loss (kg). The weight loss of material
_ sand kg=s
is divided by mass flow rate of sand in the system, M
and testing time, ttest(s) to calculate the erosion mass ratio (kg/kg).
The calculation is shown in Eq. (12):
 
kg
Wbefore  Wafter

(12)
ERmass
_ sand  ttest
kg
M
The types of sand used for testing are silica flour, Oklahoma #1
sand, and California 60 mesh sand. The average particle sizes are
20, 150, and 300 lm, respectively. Microscopic pictures of these
sand particles are shown in Fig. 3. Silica flour and California 60
mesh sand have sharper particles than the Oklahoma #1 sand.
3.2 Experimental FacilityAir Testing. The erosion model
for aluminum 6061-T6 is developed based on direct impingement
test results in air. Air is used as the carrier fluid to generate erosion equations, because in such low-density and low-viscosity fluids, the particle speed and direction do not change much as
particles approach the target wall, therefore, it is easy to control
particle impact speed and angle. Figure 4 is a schematic of the
direct impingement testing in air. A compressor provides the air

Microscopic pictures of abrasive particles

Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 4
air

Schematic of facility used for erosion measurements in

Fig. 5 LDV measurement location map

flow in the nozzle. The flow rate of the gas was controlled by the
valve positioned upstream of the flow meter. Sand is injected
from the sand feeder. The pressure drop at the nozzle creates suction in the feeding tube and draws the sand particles in the nozzle.
The mixture of the sand and air flow out of the nozzle, and particles impinge the target material.
In air, it is common that there is a relative slip velocity between
a particle and the gas. The particles leaving the nozzle are traveling at a lower speed than the gas at this location. The particle
speeds, Vp, of 150- and 300-lm particles are measured by LDV
and the gas speeds were measured by a Pitot tube. The Pitot tube
was installed at the center of the nozzle approximately 3 mm
downstream of the nozzle exit. The measured gas speed at this
location by the Pitot tube was divided by 1.2 (assuming fully
developed turbulent pipe profile) to obtain an approximate average gas speed across the nozzle exit. The gas speed was measured
before sand was injected in the system, and the measured gas
speeds are not influenced by the sand particles. It should be noted
that Pitot tube is only used as a reference for calibration and the
gas velocities that are measured by the Pitot tube may not accurately represent the average gas speed at the nozzle exit and they
are not used in any calculations. The graph of the particle speed
versus gas speed is shown in Fig. 4. As shown in the graph, the
particle speed is directly proportional to the fluid speed. A linear
equation was used to fit the data points. The data show that there
is more slip between the fluid and the 300-lm particles than for
the 150-lm particles. The linear equations in Fig. 4 were used to
determine the gas speeds necessary to produce the desired particle
speeds during the erosion measurements. It should be noted that,
for these larger particles entrained in a gas, only axial velocities at
the nozzle exit were measured. For larger particles flowing in a
gas, CFD results, as well as observed erosion patterns on the
specimens, indicate that the momentum of the particles near the
target wall are not affected by the reduction in gas speed near the
wall. After the particle speeds were measured, the erosion measurements were conducted for aluminum 6061-T6 in air. From the
LDV measurements, particle speeds of 13, 24, and 42 m/s were
chosen to conduct erosion tests. At each speed, measurements
were taken for impact angles of 90 deg, 60 deg, 30 deg, and
15 deg. At each particle speed and angle, the measurements were
taken at least three times with 300 g of sand each time. Erosion ratios
(kg/kg) are calculated by mass loss of a target coupon (kg) divided
by the total mass of sand throughput (kg). Particle sizes of 150
and 300 lm are used as abrasive sands for these measurements.
3.3 Laser Doppler Velocimeter (LDV) Measurement
Conditions. To help examine erosion results for liquid cases, in a
parallel study, particle axial and radial speeds and liquid axial and
Journal of Fluids Engineering

Fig. 6

CFD mesh of the flow region and boundary conditions

radial speeds were measured utilizing LDV by Miska [8]. Aluminum particles with average sizes of 3, 120 and 550 lm were used
as seeding particles. 3-lm particles were assumed to travel with
liquid with no slip. 120- and 550-lm particles were used to measure the speeds of particles representing relatively small and large
particles. Figure 5 shows the measurement locations between the
nozzle exit and the target wall. The measurements were performed
for r 0 to r 12 mm, and z 1 to z 12.5 mm. The particle
speeds could not be measured near or on the target surface
(z 12.7 mm) because of interference of the measuring volume of
the laser beams of LDV system and the wall.
Measurements were collected for 5 different viscosities of 1,
10, 25, 50, and 100 cP. The data sampling rate is 1000 s1. The
duration of each measurement is approximately 10 to 20 s. Data
obtained utilizing the LDV is compared with CFD predictions in
Sec. 5.3.

CFD Approach and Design Input

Computational Fluid Dynamics (CFD) was utilized to predict

fluid and particle speeds and erosion of aluminum 6061-T6 for
various viscous liquids to evaluate the accuracy of the erosion
models generated from air testing. Fluent 6.3 was used for flow
simulations and particle tracking. Erosion in the form of erosion
ratios are calculated by user defined functions (UDF) created by
Zhang et al. [5].
JUNE 2012, Vol. 134 / 061301-5

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 7

Diagram of grid adaptation

Fig. 9 Erosion ratio versus viscosity for aluminum 6061 with
different particle sizes

Table 2 Erosion results for Al 6061 (coupon tests)

Erosion ration (kg=kg)
Aluminum 6061
300 lm
Viscosity (cP)
1
10
25
50

Fig. 8 Comparison of predicted particle (120 lm) speeds for

finer and original grid (1 and 50 cP)

The computational mesh and boundary conditions of the direct

impingement case is shown in Fig. 6. This geometry is modeled
as two-dimensional axi-symmetric. The diameter of the entire
flow region is 38.1 mm and its length is 25.4 mm. The nozzle diameter is 8 mm, and the distance from nozzle exit to the wall is
12.7 mm, which is the same as the experimental apparatus. In the
flow region, the grid gets finer near the target wall. The total number of cells used for this geometry is 4600 and mesh type is 100%
quadrilateral (brick mesh). For the nozzle exit plane speed, actual
LDV data are used for the inlet speed condition. The target wall is
defined as a nonslip smooth wall. The geometry is axis symmetric
and assumed to be in steady state. The Reynolds stress model
(RSM) is used to calculate the flow field, and the standard wall
function is used to simulate the flow near the target wall.
To determine the dependency of the calculated results on the
size of grid, a grid size analysis was performed. After the flow calculation was performed for the original mesh, the grid was made
finer by breaking each cell into two cells in the z direction. A diagram of how the original grid is adapted to a finer grid is shown in
Fig. 7. The last layer of cells adjacent to the wall was kept the
same so that it still fulfills the standard wall function criteria. The
mesh type is the same as the original but the total number of cells
is 18,145 for the finer mesh. The flow and particle calculations
were redone using the finer grid. Figure 8 shows particle speeds
for viscosities 1 and 50 cP at r 8 mm for the original and finer
grids. To focus on the near-wall speed, the horizontal axis of the
plots is set from 6 to 12.7 mm in these figures. From Fig. 8, it is
observed that the particle speeds are independent of grid size for
both 1 and 50 cP. Therefore, for this case, the results are not influenced by the size of mesh. The original mesh was used for all the
predictions in this paper.
061301-6 / Vol. 134, JUNE 2012

150 lm

20 lm

Value

STDV

Value

STDV

Value

3.21E06
2.56E06
2.69E06
2.83E06

4.37E07
8.12E08
1.81E07
4.04E07

1.46E06
1.07E06
7.03E07
4.29E07

2.18E07
9.21E08
6.57E08
2.39E07

9.78E08
3.83E08
2.30E08
1.28E08

CFD and Experimental Results

5.1 Measured Erosion Results in Viscous Liquids. Figure 9

presents the measured erosion mass ratio with change in viscosity
of carrier fluid for aluminum 6061-T6. Table 2 summarizes the
erosion results for the coupon tests. Measurements were repeated
more than two times (usually 4 to 6 times) for each condition for
150- and 300-lm particles and results were averaged. One measurement was taken for 20-lm particles. The standard deviation
was calculated for each test condition using the repeated measurements. The standard deviation shown in Table 2 shows the range
of the measurements and indicates repeatability. Figure 9 shows
the erosion ratio is reduced as the viscosity increases for 20- and
150-lm particles. However, for 300-lm particles, the erosion ratio
does not change as significantly as the viscosity is increased. The
erosion ratio of 20-lm sand decreases more significantly than that
of 150-lm sand. From Table 2, it should be noted that the larger
particles tend to have larger erosion ratios than smaller particles at
any viscosity.
5.2 Measured Erosion Results in Air. To develop the coefficients that are needed in the erosion equation (Eq. (2)), erosion
ratios (kg/kg) are measured for aluminum 6061-T6 in air. This
material is also used in erosion tests involving liquids. Oklahoma
#1 (150-lm, average-size) sand and California 60 mesh (300-lm,
average-size) sand are used as the erodent. A total of 900 g of
sand was injected for the erosion tests that are presented in this
section. After injecting each 300 g of sand, the weight of the coupon was measured and recorded.
Figures 10 and 11 for 150- and 300-lm sand, respectively, are
the graphs of erosion ratio for aluminum 6061 in kg/kg versus
impact angle for three different impact speeds. For both particle
sizes, higher particle impact speeds yield higher erosion ratios for
all of the impact angles. From the figures, it is also found that
Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 10 Erosion ratio versus impact angle for Al 6061 in air

(150 lm)

Fig. 11 Erosion ratio versus impact angle for Al 6061 in air

(300 lm)

erosion ratios are higher for lower impact angles for all the particle speeds. 300-lm particles give higher erosion ratios than
150 lm because of their size and shape. The microscopic pictures
of these particles are shown in Fig. 3. Note that the 300-lm particles erosion ratios have a trend that is slightly different from that
of the 150-lm particles. Erosion ratios for the 150-lm particles
increase more significantly than the 300-lm particles as the
impact angle decreases. The differences in the two types of sand
can result in not only a difference in magnitude but also in trend
of erosion ratio with impact angle (angle function).
Utilizing the information gained through erosion measurements
in air, the erosion model for aluminum was generated. At first, all
the experimental results from Figs. 10 and 11 are normalized and
plotted together in Fig. 12. The erosion ratio for each test condition is normalized by dividing it by the value of erosion ratio at
h 15 deg (corresponding to the maximum value) for each velocity (Eq. (13)). Because the results are normalized, the results for
different particle speeds are close in value to each other which
indicate that particle speed only affects the erosion equations in
magnitude but not the shape of the erosion versus impact angle
profile. The angle function shown in Eq. (3) is defined by finding
n1, n2, and n3, which fit the experimental data. Because the experimental data are normalized, the angle function must be normalized as well by dividing the function by f, which is the maximum
value of the function:
 
 
kg
kg
ERh15
(13)
ERnormalized  ER
kg
kg
After the angle functions are defined, the erosion equations are
defined by finding an empirical constant C for each sand type.
Journal of Fluids Engineering

Fig. 12 Normalized erosion ratio versus impact angle in air

150 and 300 lm

Fig. 13 Erosion ratio versus impact angle at V 5 13 m/s

(150 lm, aluminum)

These constants are listed in Table 1. Figures 13 and 14 are the

measured erosion ratio versus impact angle for 13 m/s plotted
with the newly defined E/CRC equation and Okas equation. For
150-lm particles, the new E/CRC equation tends to give higher
erosion ratios at lower angles than Okas equation. However, at
higher angles, from 25 deg to 90 deg, erosion ratios given by
Okas equation are higher than the E/CRC equation (Fig. 13). For
300-lm particles, Okas model gives much lower erosion ratios
than the new E/CRC equation (Fig. 14). This is because there is
no parameter that accounts for sharpness in Okas equation. There
is a factor, k3, which accounts for the size of particle, but it is
very small; therefore, there is not much difference in the erosion
ratios predicted by Okas model for the 150- and 300-lm particles. On the other hand, the new equation is generated based on
erosion testing in air for 300-lm sand; therefore, it is more accurate than Okas equation. The erosion equations for 150- and 300lm particles defined here are used in the CFD code to predict erosion ratios for the liquid testing.
Because there is no erosion data for 20-lm particles, the erosion equation for 150-lm particles is used to predict erosion ratios
for this sand. A sand sharpness factor of 1 is used instead of 0.5,
because 20-lm sand has much sharper edges than the 150-lm
sand (Fig. 3).
Erosion ratios in air and liquid are plotted against particle
Reynolds number in Fig. 15. Particle Reynolds number is a
dimensionless number calculated using Eq. (14). Dp is the diameter of particle, Vp is the particle velocity, and  is the kinematic
viscosity of fluid. For liquid testing, each particle has a different
impact speed and angle; therefore, average fluid nozzle exit
JUNE 2012, Vol. 134 / 061301-7

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 14 Erosion ratio versus impact angle at V 5 13 m/s

(300 lm, aluminum)

Fig. 16 Contour plots of measured particle speed for 120 lm

(1 and 100 cP)

Fig. 15

Erosion ratio versus particle Reynolds number

velocity (10 m/s) is used for Vp for all the cases. For air testing,
viscosity of room temperature air is used for all the test
conditions.
Rep

Dp Vp
v

(14)

From Fig. 15, it should be noted that in both air and liquid,
higher particle Reynolds numbers give higher erosion ratios which
is intuitive. However, closer examination reveals that air results
have a steeper profile than liquid results. Also, for liquid testing,
larger particles are not influenced by particle Reynolds number as
much as smaller particles while in air, there is no significant difference in profile for 150- and 300-lm particles. Figure 15 also
indicates that the particle Reynolds number is not the only parameter that influences erosion and other parameters such as local particle and fluid response times (perhaps local Stokes numbers) may
affect erosion ratios. This also may indicate that the erosion mechanism in liquid may be different from the one in air.
5.3 Comparison of Predicted And Measured Particle
Speeds. Although erosion is influenced by many factors, particle
impact speed is the biggest factor that affects erosion. To study
the effect of viscosity and particle size on particle speed, which
affects erosion ratio, axial and radial particle speeds of 3-, 120and 550-lm-diameter particles were measured in 1-, 50-, and
100-cP fluid viscosities and compared to the CFD predicted
speeds. Note that the results shown in this paper are speeds that
are magnitudes of the resultant velocity calculated from axial and
radial velocity components. The predicted particle impact speeds
061301-8 / Vol. 134, JUNE 2012

Fig. 17 Measured particle speed Vz at z 5 1 mm

and angles are used in CFD to calculate erosion ratios. Therefore,

it is important to compare calculated particle speeds with experimental data to make sure that calculated particle speeds match the
data.
Figure 16 shows the contour plots of 120-lm particle speeds
for 1 and 100 cP. The contours show a high speed near the nozzle
exit along the horizontal axis (centerline). The speeds in this
region (marked by the red oval) are mostly in the axial direction
(because of small radial speed component) which travels from
nozzle exit toward the target metal. The speeds near the centerline
for the 100-cP fluid are higher than the speeds in the same region
for the 1-cP fluid. For the same flow rate, the 100-cP fluids have a
laminar profile at the nozzle exit, whereas the 1-cP fluid has a turbulent profile; thus the maximum speed in this region tends to be
higher for higher viscosity liquids. However, the particles slow
down as they approach the wall because of the reduction in fluid
speed near the stagnation point. As particles move radially away
from the centerline along the target wall, there is another location
with high particle speed (shown by the black ovals). The speed of
particles in this region is high and is mostly from the radial component of speed, which travels outward from the centerline. These
high speeds are responsible for erosion that occurs in this region.
Figure 17 presents the measured particle axial speeds of 120and 550-lm particles at z 1 mm. The particle axial speeds measured at this location are used as speed inlet conditions for CFD
particle tracking simulations. The radial component of particle velocity at this location is very small. From Fig. 17, it is observed
Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 18 Measured and predicted particle speeds versus axial

distance

Fig. 19 Measured and predicted particle speeds versus axial

distance

that particle speeds in the higher viscosity liquids have higher

speeds near the centerline. This is because of the mass conservation of fluid because flow rate is constant. 550-lm particles tend
to travel with slightly lower speeds than 120-lm particles because
of slip.
Detailed comparisons of measured and predicted speeds for
120-lm particles along r 0 and 8 mm are shown in Figs. 1820.
Results for 550 lm and the other radial locations are omitted here
because the comparison of 120- and 550-lm particle speeds will
be shown in the following section.
As shown in Fig. 18, particle speeds at the centerline (r 0
mm) are higher for higher viscosity fluids than that of lower viscosity fluids near the nozzle exit (z 0 mm). However, as particles approach the wall, speeds decrease and become negligible
at the wall. The particle speeds along the centerline are axially
dominant which means most of the particles are traveling in the
horizontal direction towards the wall. For r 0 mm, CFD predictions are in a good agreement with the measured speeds.
For r 8 mm, the particle speeds are low near the nozzle exit
(z 0 mm); however, they increase near the wall and reach a
maximum speed, then decrease again very close to the wall
(Fig. 19). Although the particle speeds become smaller as they
approach the wall, they never become zero. The erosion is
caused by this nonzero particle speed at the target wall. The particle speeds in this region are mostly radial and the axial velocity components are small, and particles move nearly parallel to
the target wall. CFD predictions agree with the measurements
away from the wall, but CFD under-predicts the peaks in speed
near the wall.
Journal of Fluids Engineering

Fig. 20 Measured and predicted particle speeds versus axial

distance

Fig. 21 Measured particle speeds versus axial distance at r 5 8

mm 120 and 550 lm

To examine nearwall particle behavior, Fig. 19 is expanded

from z 10 to 12.7 mm (Fig. 20). Figure 20 shows that both
measured and predicted particle speeds reach their maximum values near the wall, and then particles slow down as they approach
the wall. The location of maximum particle speed occurs farther
away from the wall for higher viscosity liquids than for the lower
viscosity liquids. This can be explained by the fluid shear layer
formed near the wall. The layer tends to be thicker for higher viscosity fluids than lower viscosity fluids. Particle speeds are influenced by these shear layers near the wall, and they slow down
quickly as they enter these shear layers. Therefore, particles in
higher viscosity liquids tend to slow down starting at a distance
farther away from the wall than particles in lower viscosity
liquids. Another point to note here is that away from the wall, particle speeds in high viscosity fluids are higher than low viscosity
fluids. However very near the wall, particle speeds in lower viscosity fluids are higher than speeds of higher viscosity fluids. As
particles enter the shear layer near the wall, the higher viscosity
liquids exert more drag on particles than the lower viscosity fluids; as a result, particles in higher viscosity liquids slow down
more significantly than particles in lower viscosity fluids. Also,
because the shear layer as well as the distance between the maximum fluid speed and the wall are thicker for higher viscosity
liquids, it provides a larger distance for the particles to slow down
before they reach the wall. This explains why erosion ratios
reduce as liquid viscosity is increased for small particles. CFD
predicted speeds for each viscosity show the same trend.
Figure 21 presents the comparison of the measured speeds of
120- and 550-lm particles in 1- and 50-cP fluids. 100-cP results
JUNE 2012, Vol. 134 / 061301-9

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 22 Predicted particle speeds versus axial distance at r 5 8

mm 120 and 550 lm
Fig. 24 Predicted d(speed)/dz versus z distance at r 5 8
mm 120 and 550 lm (near wall)

Fig. 25 Normalized predicted erosion ratio and experimental

data for aluminum
Fig. 23 Measured d(speed)/dz versus z distance at r 5 8
mm 120 and 550 lm (near wall)

are omitted here because they have similar profiles as 50-cP

results. Figure 22 is the comparison of the predicted speeds varying viscosity. Figure 21 shows that both 120- and 550-lm particles in the 50-cP fluid reach maximum speeds at locations
further away from the wall than the particles in 1 cP. However,
the difference of locations of maximum speed for 1 and 50 cP
(distance between the two dash lines) is more obvious for 120-lm
particles than for 550-lm particles. Also, near the wall, the differences of particle speeds in 1 and 50 cP (shown by arrows) are
more significant for 120-lm particles than that of 550-lm particles. Predicted particle speeds in Fig. 22 show the same trend.
The comparison of the lengths of the arrows (which represent
the differences between the particle speeds) for the 120- and 550lm particle graphs indicates that the particle speed does not
change as much for 550-lm particles as for 120-lm particles
when the viscosity of liquid changes from 1 to 50 cP. This is an
indication that the speeds of larger particles are not affected by
the fluid viscosity as much as those of smaller particles. The bigger particles have larger inertia than smaller particles, and they
penetrate through the shear layer near the wall and do not slow
down as much as smaller particles do. This is one of the reasons
that erosion ratio does not change much with increase in viscosity
for 300-lm sand.
061301-10 / Vol. 134, JUNE 2012

To examine how much particles are slowing down as they

approach the wall, the particle acceleration with respect to
axial distance, d(speed)/dz is calculated for each test condition.
Figures 23 and 24 are the measured and predicted particle acceleration with respect to z versus axial distance z along r 8 mm near
the wall. From both figures, it can be seen that both sizes of particles slow down significantly near the wall (negative values).
Comparison of 1-cP and 50-cP profiles reveals that particles in
50 cP liquid are slowing down more rapidly than particles in 1 cP
liquid. Comparison of two particle sizes shows that 550-lm particles have less difference in deceleration between two viscosities
than those for 120-lm particles near the wall.
5.4 Predicted Erosion Results in Liquids. The erosion
equations that were developed based on gas tests (Sec. 2.2) were
implemented into Fluent CFD code and were utilized to predict
erosion for aluminum. Figures 2527 are the normalized measured
and predicted erosion ratios (volume loss/mass of sand) of aluminum versus viscosity for three particle sizes. Both measurements
and CFD results are normalized with respect to the 1-cP results to
examine the trend of results. From Fig. 25, CFD predicts that the
erosion ratio decreases with increase in viscosity for 300-lm particles, whereas measurements show the erosion ratio stays nearly
constant across the viscosity change. Predictions using both
E/CRC model and Okas model are very similar in trend. The erosion ratio trend for 150-lm particles agrees better with CFD
Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 26 Normalized predicted erosion ratio and experimental

data for aluminum

Fig. 27 Normalized predicted erosion ratio and experimental

data for aluminum

predictions (Fig. 26). Both predictions and measurements have

declining trends. For the smallest particles (20 lm), the trend of
predictions do not agree with the trend of measurements (Fig. 27).
CFD predictions show that the erosion ratio trend decreases from
1 to 10 cP, however, increases from 10 cP to 50 cP, whereas measurements show the erosion ratio is lower for higher viscosities.
The calculation from the Oka model is similar in trend to the calculation from the E/CRC model and was omitted.
5.5 Number of Impacting Particles Analysis (CFD). Besides
particle speed, the number of impacting particles is another important factor that influences erosion ratio. As viscosity changes, the
number of particles hitting the wall also changes. In the CFD simulation, 10000 particles are released at the nozzle exit for each
case, and the number of particles hitting the wall boundary was
recorded using a user-defined function (UDF). The results are
shown in Fig. 28. Figure 28 shows the number of particles impacting versus the number of impacts per particle for different particle
sizes and different viscosities. Number of impacts (horizontal
axis) is how many times each particle hits the wall. Number of
particles impacting (vertical axis) is how many particles are hitting at each number of impacts. Some particles only hit the wall
once or several times and move away from the shear layer near
the wall (low number of impacts), whereas some particles keep
hitting the wall over and over again and are stuck in the shear
layer near the wall (high number of impacts). Figure 28 shows
that 20-lm particles tend to have much higher number of impacts
than 300- and 150-lm particles. For all particle sizes, a greater
number of particles are hitting more frequently for higher viscosity liquids (50 cP) than in lower viscosity liquids (1 cP).
Ziskind et al. has studied the motion of inertial particles in
shear flow near a solid surface. The authors have solved the anaJournal of Fluids Engineering

Fig. 28 Number of particles impacting versus impact number

for 20, 150, and 300 lm

lytical equations of particle motion based on its inertia and hydrodynamic lift and drag. He states that the particle motion at a fixed
flow condition is classified either as unstable or stable. In unstable
motion, particles escape rapidly from the surface into the main
flow. The particle motion becomes unstable when the initial
velocity of particle is smaller than the local fluid velocity and wall
induced lift is larger than the inertia of the particle. When the wall
induced lift is smaller than the inertia, the particle approaches the
wall and its motion becomes stable. The instability of particle
motion depends on particle speed, inertia, and the fluid shear rate.
He has shown analytically and numerically that whether particle
motion becomes unstable and leaves the near wall region depends
on the location of particle from the wall [24,25].
Perhaps the reason for the high number of impacts for smaller
particles is because of the stable particle motion and low number
of impacts for larger particles is because of the unstable particle
motions. After particles impinge the wall and rebound, the smaller
particles do not have enough momentum to overcome the drag of
the fluid and leave the shear layer but instead get pushed back to
the wall region and repeat the impingement motion. On the other
hand, big particles have greater momentum to overcome the drag
of the fluid and leave the shear layer after a few impacts. Drag the
fluid exerts on the particle is increased for high viscous liquid;
therefore, if the particle size is the same, particles tend to impact
more frequently in high viscous liquid than the particles in low
viscous liquids.
The number of impacts for 20-lm particles is very high as compared to other particles. To demonstrate that the number of particles impacting is the major source of the difference between the
CFD predictions and measurements, the erosion ratio for aluminum was recalculated for 20-lm particles using only the first
impact and discarding the rest of the impacts for an individual particle. The results are shown in Fig. 29. The graph shows the new
JUNE 2012, Vol. 134 / 061301-11

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

Fig. 29 Measured and predicted erosion ratios for first impacts

(20 lm, Al 6061)

CFD predictions using the first impact along with experimental

data. The predictions using only the first impact agree with the experimental data much better than the original predictions using all
of the impacts.
Considering only the first impact is not a realistic scenario;
however, this new calculation is shown only to emphasize that the
number of impingement is one of the reasons for the discrepancy
between CFD and measurements. Additional theoretical and numerical analyses need to be performed in this area to better model
the particle motions near the wall in CFD.

6.

Summary and Conclusions

(1) Erosion ratios for aluminum 6061-T6 occurring from 20-,
150-, and 300-lm particles with carrier liquids of 1, 10, 25,
and 50 cP were measured. It was found that the erosion ratio decreases as viscosity increases for 20- and 150-lm
sand particles whereas 300-lm sand showed no significant
change in erosion ratio with change in viscosity. The measured and predicted particle speeds were examined and they
appear to support the observed erosion ratios.
(2) Erosion models were developed based on air testing for the
same aluminum. For aluminum, erosion models were generated for two types of sand: Oklahoma #1 and California
60. Each equation is unique with different empirical constants and angle functions. These newly developed equations are presented and were used to predict erosion ratio
using CFD.
(3) LDV measurements were examined to study how particles
behave near the wall or target material. Fluid and particle
speeds for 120- and 550-lm particles in 1-, 50-, and 100-cP
fluids were measured. The comparison of predicted and
measured particle speeds of small and large particles
showed that CFD is able to predict particle speeds comparable to the measurements. Near the wall, however, it was
found that the speeds of large particles are not affected by
viscosity as much as that of small particles. Because of
larger inertia, large particles penetrate through the shear
layer near the wall and slow down, but not as much as the
small particles, which slow down significantly as they enter
the shear layer. As a result, erosion ratios with large particles do not change as much as small particles as viscosity
changes. It was also found that higher viscosity liquids
form thicker shear layers near the wall than lower viscosity
liquids. Therefore, if particle sizes are the same, particles in
higher viscosity fluids reach a maximum speed farther
away from the wall and slow down more rapidly than particles in lower viscosity fluids near the wall. Because higher
viscosity liquids have a thicker shear layer and exert higher

061301-12 / Vol. 134, JUNE 2012

drag on particles, particles impacting the wall have lower

speeds in higher viscosity fluids than lower viscosity fluids.
This may explain why erosion ratios decrease as viscosity
increases for each particle size.
(4) CFD was also utilized to predict erosion ratio for aluminum
for particle sizes of 20, 150, and 300 lm in carrier fluids of
1, 10, 25, and 50 cP. The comparison of predicted and
measured erosion ratio trends of aluminum 6061-T6
showed that for 150-lm particles, the trend of the predicted
erosion ratio is comparable to that of the measured erosion
ratio. For the 20-lm sand, the predicted erosion ratio
decreases when viscosity increases from 1 to 10 cP, but
increases when viscosity increases from 10 to 50 cP, which
does not agree with the experimental data.
(5) The number of particles impacting the wall or target material was studied using CFD simulations. It was found that
particles in higher viscosity fluids impact the wall more frequently than particles in lower viscosity fluids. It was also
found that 20-lm particles tend to impinge the target over
and over again resulting in unrealistic predicted erosion ratio trends whereas 150- and 300-lm particles only impact a
few times and exit the region of interest near the wall. The
erosion ratios of 20-lm particles were recalculated using
information from only the first impact of each impacting
particle. The predicted erosion ratio trends using only the
first impact successfully agrees with the trend of the
observed erosion ratio of the experimental data. This
implies that the number of impacting particles is another
important factor for erosion prediction.
It should be mentioned that the erosion equations used in this
study were all generated from gas data for a range of speeds from
13 to 42 m/s. However, the magnitude and trend of erosion did
not accurately match the data for larger and very small sand in
viscous liquids. Therefore, additional investigations are needed to
determine if there is a fundamental factor causing differences
between the erosion process in liquids and gases.

Acknowledgment
The authors wish to acknowledge the support of Erosion/Corrosion Research Center (E/CRC). Comments of Dr. Edmund F.
Rybicki are also appreciated. The authors also wish to acknowledge Contributions of Edward Bower of E/CRC, Stephen Miska,
and Sepinood Torabzadehkhorasani.

References
[1] Finnie, I., and McFadden, D., 1978, On the Speed Dependence of the Erosion
of Ductile Metals by Solid Particles at Low Angles of Incidence, Wear, 48, pp.
181190.
[2] McLaury, B., 1993, A Model to Predict Solid Particle Erosion in Oilfield Geometries, M.S. thesis, The University of Tulsa, Tulsa, OK.
[3] Oka, Y. I., Okamura, K., and Yoshida, T., 2005, Practical Estimation of Erosion Damage caused by Solid Particle Impact, Part 1: Effects of Impact Parameters on a Predictive Equation, Wear, 25, pp. 95101.
[4] Oka, Y. I., and Yoshida, T., 2005, Practical Estimation of Erosion
Damage Caused by Solid Particle Impact, Part 2: Mechanical Properties
of Materials Directly Associated with Erosion Damage, Wear, 259, pp.
102109.
[5] Zhang, Y., 2006, Application and Improvement of Computational Fluid Dynamics (CFD) in Solid Particle Erosion Modeling, Ph.D. dissertation, The University of Tulsa, Tulsa, OK.
[6] Zhang, Y., Reuterfors, E. P., MClaury, B. S., Shirazi, S. A., and Rybicki, E. F.,
2007, Comparison of Computed and Measured Particle Speeds and Erosion in
Water and Air Flows, Wear, 263, pp. 330338.
[7] Zhang, Y., Okita, R., Miska, S., McLaury, B. S., Shirazi, S. A., and Rybicki,
E. F., 2009, CFD Prediction and LDV Validation of Liquid and Particle
Speeds in a Submerged Jet Impinging a Flat Surface for Different Viscosities
and Particle Sizes, Proceedings of ASME 2009, Fluids Engineering Division,
Vail, Colorado.
[8] Miska, S., 2008, Particle and Fluid Speed Measurements for Viscous Liquids
in a Direct Impingement Flow Resulting in Material Erosion, M.S. thesis, The
University of Tulsa, Tulsa, OK.

Transactions of the ASME

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

[9] Torabzadehkhorasani, S., 2009, Erosion Experiments and Calculations in Gas

and Liquid Impacts Varying Particle Size and Viscosity, M.S. thesis, The University of Tulsa, Tulsa, OK.
[10] Finnie, I., 1972, Some Observations on the Erosion of Ductile Materials,
Wear, 19, pp. 8990.
[11] Lindsley, B. A., and Marder, A. R., 1999, The Effect of Speed on the Solid
Particle Erosion Rate of Alloys, Wear, 22529, pp. 510516.
[12] Winter, R. E., and Hutchings, I. M., 1974, Solid Particle Erosion Studies Using
Single Angular Particle, Wear, 29, pp. 181194.
[13] Loth, E., 2008, Drag of Non-Spherical Solid Particles of Regular and Irregular
Shape, Powder Technol., 182, pp. 342353.
[14] Gavze, E., and Shapiro, M., 1997, Particles in Shear Flow Near a Solid Wall:
Effect of Nonsphericity on Forces and Velocities, Int. J. Multiphase Flow,
23(1), pp. 155182.
[15] Tilly, G., and Sage, W., 1970, The Interaction of Particle and Material Behavior in Erosion Process, Wear, 16, pp. 447465.
[16] Meng, H. S., Ludema, K. C., 1995, Wear Models and Predictive Equations:
Their Form and Content, Wear 181183, pp. 443457.
[17] Oka, Y. I., Matsumura, M., and Kawabata, T., 1993, Relationship between
Surface Hardness and Erosion Damage Caused by Solid Particle Impact,
Wear, 162164, pp. 688695.

Journal of Fluids Engineering

[18] Oka, Y. I., Ohnogi, H., Hosokawa, T., and Matsumura, M., 1997, The Impact
Angle Dependence of Erosion Damage Caused by Solid Particle Impact,
Wear, 203204, pp. 573579.
[19] FLUENT, 2006, FLUENT 6.3 Users Guide, FLUENT Inc., Canonsburg, PA.
[20] Forder, A., Thew, M., and Harrison, D., 1998, A Numerical Investigation of
Solid Particle Erosion Experienced within Oilfield Control Valves, Wear, 216,
pp. 184193.
[21] Tabakoff, W., 1992, High Temperature Erosion Resistance of Coatings for
Use in Gas Turbine Engines, Surf. Coat. Technol., 52, pp. 6579.
[22] Grant, G., and Tabakoff, W., 1975, Erosion Prediction in Turbomachinery Resulting from Environmental Solid Particles, J. Aircr., 12, pp.
471478.
[23] Chen, X., McLaury, B. S., and Shirazi, S. A., 2004, Application and Experimental Validation of a Computational Fluid Dynamics (CFD)-Based Erosion
Prediction Model in Elbows and Plugged Tees, Comput. Fluids, 33, pp.
12511272.
[24] Ziskind, G., Fichman, M., and Gutfinger, C., 1998, Effects of Shear on Particle
Motion near a SurfaceApplication to Resuspension, J. Aerosol Sci, 29(3),
pp. 323338.
[25] Ziskind, G., and Gutfinger, C., 2002, Shear and Gravity Effects on Particle
Motion in Turbulent Layers, Powder Technol., 125, pp. 140148.

JUNE 2012, Vol. 134 / 061301-13

Downloaded From: http://fluidsengineering.asmedigitalcollection.asme.org/ on 09/23/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use