31st Conference of the

European Working Group on Acoustic Emission (EWGAE) We.4.A.5

Monitoring Acoustic Emission (AE) Energy

of Abrasive Particle Impacts in a Slurry
Impingement Flow Loop
Ghazi DROUBI *, Bob REUBEN **
*School of Engineering, Robert Gordon University, Aberdeen, UK
** School of Engineering and Physical Sciences, Heriot-Watt University, Edinburgh, UK

Abstract.The estimation of energy dissipated during multiple particle impact is a

key aspect in evaluating the abrasive potential of particle-laden streams. A
systematic investigation of particle impact energy using acoustic emission (AE)
measurements was carried out to evaluate the influence of particle size, free stream
velocity, and nominal particle concentration on the amount of energy dissipated in a
carbon steel target using a slurry impingement flow loop test rig. Silica sand
particles of mean particle size (225 to 650 m) were used for impingement on the
target with nominal concentration between 1 and 5% while the free stream velocity
was changed between 4.2 and 14 ms-1.
The measured AE energy was found to be proportional to the incident kinetic
energy of the particles, although the high arrival rate involved in the slurry
impingement flow loop poses challenges in resolving individual particle impact
signatures in the AE record.The results have been reconciled with earlier work by
the authors on sparse streams where there are few particle overlaps and good control
over particle kinetic energies, by extending their model to account for different
particle carrier-fluids and to situations where arrivals cannot necessarily be resolved.

Introduction
It is well-established [e.g. 1-3]that the rate of dissipation of the kinetic energy of impacting
particles in a flow is related to the rate of material removal. Also, there is a general
agreement that the AE energy associated with particle impingement is proportional to the
incident kinetic energy 12 mv 2 [e.g. 4, 5]. Monitoring of particle impact using acoustic
emission (AE) relies upon a fraction of the incident kinetic energy of each impacting
particle dissipating as elastic waves, which propagate through the target material before
being detected by a suitably placed AE sensor. Some of the investigators in this area have
concentrated on monitoring the erosion variables [6], and others have concentrated on
monitoring the amount of erosion [7]. The current authors [5, 8] have previously developed
a model based on a probability distribution of particle impact energy, validated under
controlled conditions of impingement.The purpose of this relatively uncontrolled
impingement experiment was to assess what further adjustments need to be made in the
processing to use AE as a semi-quantitative diagnostic indicator for particle impingement in
real process flows.

1
License: http://creativecommons.org/licenses/by-nd/3.0/

Experimental method
The experimental set-up (Figure 1) used an AE system with a carbon steel target assembly
identical to those used for earlier tests using air jet [8] and slurry impingement [5]. The
flow loop consisted of a positive displacement pump (model C22BC10RMB, Mono
pumpdriven by a 1.1 kW geared motor to give an output speed of 587 rpm), standard 25
mm PVC piping, a 50 litre conical tank and choke valves. The slurry was first mixed by
recirculating it through a by-pass leg for around 20 minutes to ensure that all the solids
were suspended in the flow before diverting the flow to the bend.
Splash guard

AE sensor
Mixing tank

Pre-amplifier:
PAC-1220A
Flowmeter

Teepiece

12 bit NI, PCI6115 DAQ

Control Valve

Computer

Valves
Flow direction
Motor

Drainage
Valve

Mono pump

Figure 1: Sketch of the experimental flow loop with AE measurement system

The bend was made from 5 mm bore carbon steel inserted into the 23 mm bore PVC
pipe, a sharp (90) bend having been selected in order to localize the impingement area and
minimize the impact angle range. The pipe wall opposite to the stream was milled flat in
order to have a plane area to mount the AE sensor and the bend was machined to give an
internal bore of 5 mm with a conical transition, giving 7 mm wall thickness at the site
where the sensor was mounted. The length of the target section was 75 mm giving an
overall impingement area similar to the other studies.The AE sensor was mounted using the
magnetic clamp and coupled by means of vacuum grease to the opposite surface of the
bend directly above the impingement area and the pre-amplified data were acquired at 2.5
MS/second for a record length of 1 second. Prior to testing, the sensitivity of the sensor was
checked by performing a pencil lead break test at the bend to check the functioning of the
AE detection system and to confirm the quality of sensor coupling.
Silica sand slurry was made from 10 litres of clean water and a predetermined mass of
different particle size fractions in order to obtain the required concentration. Four different
particle size ranges were used and, for each particle size range, an impingement run was
carried out with a total of three levels of solid concentration (1, 2.5, and 5wt%) and four
different flow velocities (4.2, 6.8, 10.2, and 12.7 ms-1). The average particle launch rate was
assessed by multiplying the volumetric flow rate by the average measured concentration
and dividing by the average mass of a particle, giving values from around2.5104 to
2.5106 particles per second.

The AE energy measured was based on at least ten repeat records making a total of 120
AE records for each particle size range tested.Following each set of experiments, the rig
was drained and cleaned.
The background noise AE energy associated with particle-free water impingementand
the variability in AE energy associated with sensor removal and replacement between
experiments was assessed in three testsbetween which the sensor was demounted and
reinstalled, running clean water at each of the four flow speeds. Figure 2 shows the
recorded AE energy at each of thespeeds for each of the three experiments where each
point represents the average of ten 1-second AE energy values along with its standard
deviation. As can be seen, the variation in the energy recorded for each sensor installation
(within group variation) is small, while the variation between installations is slightly larger.

AE energy, V2sec

1.0E+05
9.0E+04

Installation 1, n=1.8

8.0E+04

Installation 2, n=1.8

7.0E+04

Installation 3, n=1.7

y = 1192.x1.7

6.0E+04
y = 740.78x1.8

5.0E+04
4.0E+04

y = 673.6x1.8

3.0E+04
2.0E+04
1.0E+04
0.0E+00
0

10

12

14

Flow speed, m/s

Figure 2: Recorded AE energy for pure water impingement in flow loop

Experimental results
For each experimental condition, the AE energy was calculated from the raw signal
(measured as an amplified voltage, V) by integrating over the entire record:

At least ten repeat 1-second records were analysed for each condition and the average value
is used in the following general analysis to establish the effects of flow speed, particle size,
and concentration, against the normal expectationthat energy will depend on the square of
the impact speed, the cube of the particle diameter (i.e. the particle mass) and be linear with
concentration expressed as mass per unit volume of water.
Figure3 shows one example of the effect of the flow speed (v) on the measured AE
energy for the largest particle size range and all concentrations. As can be seen, the
measured AE energy increases with both flow speed and concentration following
approximately the second power of flow speed. Table 2 summarises the results for all
particle size ranges and all concentrations. As can be seen, the flow speed exponent is close
to the expected value of 2 for all particle size ranges except the lowest size fraction where
the signal:noise might be expected to be low. The variation of the best fit power index for
3

all experiments along with the respective correlation coefficients are also summarised in
Table 2 which shows the weighted average exponent to be 2.
AE energy, E , V2sec

4.0E+5
C=1%

3.5E+5
3.0E+5

C=2.5%

2.5E+5

C=5%

2.0E+5
1.5E+5
1.0E+5
5.0E+4
0.0E+0
0

6
8
10
12
14
Flow speed, v, m/s
Figure 3: Effect of flow speed on the measured AE energy for the three concentrations for particle size range
600-710 m
Table 2: Exponent of flow speed dependence of measured AE energy and correlation coefficient for all
experiments (bold text data are shown in Figure 3)

Particle size
range (m)

212-250

300-425

500-600

600-710

Nominal concentration
(kg/m3)

Flow speed
exponent (n)

Curve fitting R2
value (%)

1
2.5
5
1
2.5
5
1
2.5
5
1
2.5
5

0.45
0.63
2.5
1.9
2
2
1.8
2.2
3.6
2.5
2.4

36
91
97
98
96
88
94
94
95
99
99

Table 3 lists the best fit power index for the effect of mean particle diameter on the
measured AE energy for all measurements. Generally, the energy varies with
approximately the third power of the mean particle diameter, except in the case of low
speed where there is very little particle signal (above the water noise) and where changes
are difficult to discern at all. As for the flow speed exponent the diameter exponent tends
towards the expected value of 3 at higher concentrations whereas, at the lower speeds and
concentrations, the exponent tends towards 2 (in cases where a change can be discerned),
leading to a weighted mean exponent of 2.6.
Finally, Table 4 summarises the effect of nominal solid concentration on the measured
AE energy for all particle size ranges. The resulting average values of the ten AE records at
each condition show a general increase in AE energy with concentration for all particle
sizes at all flow speeds, although there is a considerable scatter at higher flow speeds. The
nominal concentration exponent tends towards the expected value of unity except in cases
4

of larger particle sizes and flow speeds where a drop out phenomenon might play a
significant role. The weighted average exponent was 0.95.
Table 3: Exponent of particle size dependence of measured AE energy for all experiments

Nominal concentration
(kg/m3)

Flow speed (m/s)

Particle diameter
exponent ()

Curve fitting R2
value (%)

4.2
6.8
10.2
12.7
4.2
6.8
10.2
12.7
4.2
6.8
10.2
12.7

0.8
3.3
3.8
4.8
1.5
2
3.2
3.2
0.95
1.8
2.4
2.75

17
97
92
91
79
97
88
94
74
81
80
85

2.5

Table 4: Exponent of particle concentration dependence of measured AE energy for all experiments

Particle size range

(m)

212-250

300-425

500-600

600-710

Flow speed (m/s)

Solid concentration
exponent ()

Curve fitting R2
value (%)

4.2
6.8
10.2
12.7
4.2
6.8
10.2
12.7
4.2
6.8
10.2
12.7
4.2
6.8
10.2
12.7

0.76
1.5
1.6
1.6
1.4
1.1
1.3
0.9
0.25
0.37
0.46
0.3
1.4
0.7
1
0.45

80
95
98
99
99
99
99
82
84
93
99
72
99
97
98
69

Discussion
Given that the measured energy shows roughly the expected variation with speed, particle
density and particle size, it remains to be seen whether the energy measured corresponds to
what would be expected from a previously-developed log-normal distribution function[5] to
5

describe the probability distribution of particle arrival AE energy for air-propelled particles
using the same target and sensor. The mean of the log-normal distribution function was
found to be:

The expected AE energy in a population of impacts,Ecalculated , can now be obtained

using the average particle arrival rate (assumed to be the same as the launch rate) and the
mean of the energy distribution function. The measured AE energy associated with the
particles, Emeasured, was estimated by subtracting the background water impingement energy
Ewfrom the integral of the signal,
whereEw was obtained from the average of the correlation functions shown in Figure 2. The
empirical model of Turenne and Fiset[9] was used to calculate the average particle speed
for all the conditions studied.
Figures 4 and 5 show examples of the correlation between the calculated and the
measured AE energy for each of the particle sizes using the average calculated impact
speed. It is clear from these figures that the correlation slope approaches the expected value
of unity with increasing particle size. This might be explained by the fact that smaller
particle fractions (less inertia) are more vulnerable to influences of the fluid than bigger
fractions (bigger inertia), which would change the impact angle and also the proportion of
particles striking the surface, and this would also explain the lower measured values in
Figure 4where uncontrolled behaviour of particles sweeping around the bend is more
likely. Figure 6 shows the average slope of the correlation between calculated AE energy
and measured AE energy when taking all the data together. As can be seen, the slope is
close to unity, although, the calculated (expected) AE energy is slightly overestimated. This
might partly be explained by particle trajectories around the bend generally having an angle
of incidence influenced by the bulk fluid flow, resulting in a greater proportion of particles
having an angle of impact less than 90o, and thus overestimating the calculated AE energy.
Another possible reason might be that the hydraulic differences between the bend and the
slurry impingement rig result in a smaller proportion of particles actually striking the target
and contributing to AE energy due to a higher degree of particle interaction at or near the
surface, resulting in particle collisions, reduced particle impact velocities and changed
impact angles. Also, the effect of the slightly different design of the target at the bend might
provide a leakage path for AE energy reducing the amount of measured AE energy. These
factors have probably all contributed to the overestimate in the calculated AE energy and
are those which would have to be taken into account in any real application of the technique
as they are dependent on the design of the system being monitored.

Calculated AE energy, Ec, V2sec

0.025
0.02
Ec = 1.85Em
R = 0.18

0.015
0.01

0.005
0
0

0.002

0.004

0.006

0.008

Measured AE energy, Em,

0.01

0.012

V2sec

Calculated AE energy, Ec, V2sec

Figure 4: Calculated AE energy versus measured AE energy for particle size range 212-250 m

0.3
Ec= 1.31Em
R = 0.70

0.25
0.2
0.15
0.1
0.05
0
0

0.05

0.1

0.15

Measured AE energy, Em,

0.2

V2sec

Figure 5: Calculated AE energy versus measured AE energy for particle size range600-710 m

Calculated AE energy, Ec, V2sec

0.3

Ec = 1.35Em
R = 0.79

0.25
0.2
0.15
0.1
0.05
0
0

0.05

0.1

Measured AE energy, Em,

0.15

0.2

V2sec

Figure 6: Calculated AE energy versus measured AE energy for all particle size ranges investigated

Conclusion
A series of slurry impingement tests were carried out to study the effect of particle size,
flow speed, and particle concentration, on the AE energy dissipated in a carbon steel bend,
with the following broad findings:
1. The measured AE energy was found overall to be proportional to the expected
square of velocity, cube of particle size, and linear with concentration of the
incident flow over a wide range of particle sizes (125-600 m), flow speeds (4-12
ms-1), and nominal concentrations (1-5 wt%), but, again, with weaker expression for
smaller, slower particles.
2. The calculated AE energy (from the model) showed good agreement with the
measured AE energy, but with the model overestimating the energy slightly,
particularly for smaller particles. The discrepancies could be traced to details of the
design of the hydraulics and the target, and these are factors which would need to be
accounted for in any practical application.
3. In combining the fluid mechanics of particles suspended in liquid and the model,
this model of AE energy can be used as a semi-quantitative diagnostic indicator for
particle impingement in industrial equipment such as pipe bends.
References
1. Clark H M and Wong K K, Impact angle, particle energy and mass loss in erosion
by dilute slurries. Wear, 1995, 186-187(Part 2), pp. 454-464.
2. Head W J and Harr M E, The development of a model to predict the erosion of
materials by natural contaminants. Wear, 1970, 15(1), pp. 1-46.
3. Burstein G T and Sasaki K, Effect of impact angle on the slurry erosion-corrosion
of 304L stainless steel. Wear, 2000, 240(1-2), pp. 80-94.
4. Buttle D J and Scruby C B, Characterization of particle impact by quantitative
acoustic emission. Wear, 1990, 137(1), pp. 63-90.
5. Droubi M G, Reuben R L and G. White, Statistical distribution models for
monitoring acoustic emission (AE) energy of abrasive particle impacts on carbon
steel. Mechanical Systems and Signal Processing, 2012, 30, pp. 356-372.
6. Hou R, Hunt A and Williams R A, Acoustic monitoring of pipeline flows:
particulate slurries. Powder Technology, 1999, 106(1-2), pp. 30-36.
7. Ferrer F et al., On the potential of acoustic emission for the characterization and
understanding of mechanical damaging during abrasion-corrosion processes.
Wear, 1999, 231(1), pp. 108-115.
8. Droubi M G, Reuben R L and White G, Acoustic Emission (AE) monitoring of
abrasive particle impacts on carbon steel. Proceedings IMechE, Part E, Journal of
Process Mechanical Engineering, 2012,226(3), pp. 187-204.
9. Turenne S and Fiset M, Modeling of abrasive particle trajectories during erosion by
a slurry jet. Wear, 1993, 162-64 (pt. B), pp. 679-687.