See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/292345265

Combining COMSOL Modeling with Acoustic

Pressure Maps to Design Sono-reactors

Article in Ultrasonics Sonochemistry · January 2016

DOI: 10.1016/j.ultsonch.2016.01.036

CITATIONS READS

8 580

2 authors:

Zongsu Wei Linda K Weavers

University of California, Los Angeles The Ohio State University
34 PUBLICATIONS 223 CITATIONS 79 PUBLICATIONS 1,944 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

New AOP investigation View project

luo shuang View project

All content following this page was uploaded by Zongsu Wei on 16 February 2016.

The user has requested enhancement of the downloaded file.

 Ultrasonics Sonochemistry 31 (2016) 490–498

Contents lists available at ScienceDirect

Ultrasonics Sonochemistry
journal homepage: www.elsevier.com/locate/ultson

Combining COMSOL modeling with acoustic pressure maps to design

sono-reactors
Zongsu Wei, Linda K. Weavers ⇑
Department of Civil Environmental and Geodetic Engineering, The Ohio State University, Columbus, OH 43210, USA

a r t i c l e i n f o a b s t r a c t

Article history: Scaled-up and economically viable sonochemical systems are critical for increased use of ultrasound in
Received 10 September 2015 environmental and chemical processing applications. In this study, computational simulations and acous-
Received in revised form 23 January 2016 tic pressure maps were used to design a larger-scale sono-reactor containing a multi-stepped ultrasonic
Accepted 29 January 2016
horn. Simulations in COMSOL Multiphysics showed ultrasonic waves emitted from the horn neck and tip,
Available online 29 January 2016
generating multiple regions of high acoustic pressure. The volume of these regions surrounding the horn
neck were larger compared with those below the horn tip. The simulated acoustic field was verified by
Keywords:
acoustic pressure contour maps generated from hydrophone measurements in a plexiglass box filled with
Ultrasound
COMSOL Multiphysics
water. These acoustic pressure contour maps revealed an asymmetric and discrete distribution of acous-
Hydrophone tic pressure due to acoustic cavitation, wave interaction, and water movement by ultrasonic irradiation.
Acoustic field The acoustic pressure contour maps were consistent with simulation results in terms of the effective
Cavitation scale of cavitation zones (
10 cm and <5 cm above and below horn tip, respectively). With the mapped
acoustic field and identified cavitation location, a cylindrically-shaped sono-reactor with a conical bot-
tom was designed to evaluate the treatment capacity (
5 L) for the multi-stepped horn using COMSOL
simulations. In this study, verification of simulation results with experiments demonstrates that coupling
of COMSOL simulations with hydrophone measurements is a simple, effective and reliable scientific
method to evaluate reactor designs of ultrasonic systems.
Ó 2016 Elsevier B.V. All rights reserved.

1. Introduction and operational parameters (e.g., frequency) impact optimizing

performance of ultrasonic systems [15–20]. Of the available com-
Many laboratory studies have reported the chemical processing putational tools, COMSOL Multiphysics applies a finite element
of materials, water contaminants, and waste streams using ultra- method to solve different physics and engineering problems (e.g.,
sound [1–3]. However, few studies report methods to scale up acoustic propagation and heat transfer) governed by partial differ-
these bench-scale studies to larger systems. The most commonly ential equations (PDEs). The numerous modules and corresponding
used bench-scale device (e.g., horn type probe) for sonication has analytical solutions in the software allow it to combine different
low energy efficiency, localized cavitation, and a non-uniform phenomena into one model, which is required to simulate ultra-
acoustic field in the reactor [4–6]. In our previous work, a scaled- sonic systems that feature electromechanical and elastic mechani-
up multi-stepped horn was designed and characterized showing cal effects [21,22]. Therefore, COMSOL Multiphysics has been
higher energy efficiency, multiple cavitational zones, and more applied to simulate acoustic fields and sonochemistry in reactors
widely distributed acoustic pressure as compared to typical horns and has provided results consistent with laboratory measurements
[7]. To date, there are still limited strategies that have been inves- [15,16,18].
tigated to design new ultrasonic devices [7,8], improve reactor A hydrophone is a piezoelectric device that detects sound pres-
performance [9–12], and scale up sonolytic processes [13,14]. sure underwater and converts the pressure signals to electrical sig-
In the design process, computational simulations are used to nals. Hydrophone measurements are used to determine an acoustic
investigate how different reactor geometries, horn configurations, pressure distribution in solution and through frequency spectral
analysis, locate cavitation regions [23–25]. Bubble oscillations in
an acoustic field, together with shock waves/micro-jets that follow
⇑ Corresponding author at: The Ohio State University, Department of Civil
bubble collapse, introduce many subharmonic/harmonic frequen-
Environmental & Geodetic Engineering, 470 Hitchcock Hall, 2070 Neil Avenue,
Columbus, OH 43210, USA. cies and a broad range of frequencies (i.e., background noise)
E-mail address: weavers.1@osu.edu (L.K. Weavers). [26–29]. This emitted broadband signal is indicative of transient

http://dx.doi.org/10.1016/j.ultsonch.2016.01.036
1350-4177/Ó 2016 Elsevier B.V. All rights reserved.
 Z. Wei, L.K. Weavers / Ultrasonics Sonochemistry 31 (2016) 490–498 491

Nomenclature

an normal acceleration of solid horn (m s2) R0 bubble radius at equilibrium (m)

Ai amplitude in radius change for ith harmonic sE elastic compliance (6  6 matrix; m2 N1) in a constant
c speed of ultrasound propagation in the water (m s1) electric field
cE elastic coefficients (6  6 matrix; Pa) at constant electric S strain vector (6  1 matrix; m m1)
field strength T stress vector (6  1 matrix; Pa)
d piezoelectric strain constant (3  6 matrix; m V1) t time (s)
dt transposed piezoelectric strain constant matrix (6  3; u particle displacement (m)
m V1) x defined power series
D electric flux density vector (3  1 matrix; C m2)
e dielectric permittivity (3  6 matrix; C m2) Greek letters
eiu alternating current (AC) a characteristic exponent
et transposed dielectric permittivity matrix (6  3; C m2) b ratio of driving frequency to bubble oscillation fre-
E electric field intensity vector (3  1 matrix; V m1) quency
f frequency of ultrasound (Hz) c ratio of specific heats
fh bubble oscillation frequency (Hz) eS dielectric permittivity matrix (3  3; F m1) at constant
fR resonance frequency of bubble oscillation (Hz) mechanical strain
FV force per volume (N m3) eT dielectric permittivity matrix (3  3; F m1) at constant
m integral number mechanical stress
n integral number l fluid viscosity (Pa s)
n unit vector q water density (kg m3)
P acoustic pressure (Pa) qm material density (kg m3)
PA maximum acoustic pressure (Pa) qs density of horn rod (kg m3)
Pstat hydrostatic pressure (Pa) r surface tension (N m1)
Pvapor vapor pressure (Pa) u phase difference (rad)
q dipole source (m s2) ui phase difference for ith harmonic (rad)
R bubble radius at time t (m) x angular frequency (rad s1)

cavitation [30]. Hydrophone measurements of acoustic emissions 2.1.1. Applied physical modules
have been used to characterize acoustic fields and sonochemical A piezoelectric effect is a phenomenon in which an applied
reactivity in many ultrasonic systems [23,30,31]. stress on a piezoelectric material induces electric polarization or
The coupling of computational simulation with mapping the an applied electric field induces a dimensional change in the piezo-
acoustic field using hydrophone measurements provides a method electric material [34]. In an ultrasonic transducer, the piezoelectric
for designing ultrasonic reactors. This work presents a protocol for material, often a lead zirconate titanate (PZT) ceramic, generates a
a sono-reactor design using this coupled method. First, acoustic mechanical strain under an applied electrical field (i.e., alternating
field surrounding the newly designed multi-stepped horn was sim- current or AC). Thus, these electromechanical behaviors of the iso-
ulated in COMSOL Multiphysics to evaluate ultrasound propaga- tropic PZT are expressed by linearized constitutive equations as
tion and the resulting cavitation zone in water. The simulation follows [34,35]:
results were then verified using acoustic pressure maps from (
T ¼ cE S  et E
hydrophone measurements in a plexiglass box, followed by spec- ð1aÞ
tral analysis of ultrasound signals to determine the cavitation D ¼ eS þ eS E
region and scope. Finally, the configuration of an approximately
(
sized sono-reactor was proposed and modeled. We propose this t
S ¼ sE T þ d E
method for reactor design as a rational way to design and charac- ð1bÞ
D ¼ dT þ eT E
terize sono-reactors.
where T is the stress vector (6  1 matrix; Pa), S is the strain vector
(6  1 matrix; m m1), E is the electric field intensity vector (3  1
2. Methodology
matrix; V m1), D is the electric flux density vector (3  1 matrix;
C m2), cE is the elastic coefficient (6  6 matrix; Pa) at constant
2.1. COMSOL simulation
electric field strength, et is the transposed dielectric permittivity
matrix (6  3; C m2), e is the dielectric permittivity (3  6 matrix;
An ultrasonic system, composed of a transducer and a horn,
C m2), eS is the dielectric permittivity matrix (3  3; F m1) at con-
involves different physical phenomena [6,21,22]. The piezoelectric
stant mechanical strain, sE is the elastic compliance (6  6 matrix;
material in the transducer converts electricity into mechanical
m2 N1) in a constant electric field, d is the transposed piezoelectric
t
vibrations which pass through the ultrasonic horn rod and are
amplified at the end of the horn [22]. These amplified mechanical strain constant matrix (6  3; m V1), d is the piezoelectric strain
waves (i.e., ultrasonic waves) are emitted and propagate through a constant (3  6 matrix; m V1), and eT is the dielectric permittivity
medium, such as water. Thus, three different modules were matrix (3  3; F m1) at constant mechanical stress.
selected to simulate these physical effects in the COMSOL Multi- The vibration generated in the piezoelectric transducer is then
physics software (version 4.2): (1) a piezoelectric material module transmitted to the horn rod. Assuming both the stainless steel
for the transducer; (2) a linear elastic material module for the horn structure of the horn rod and PZT are isotropic and elastic, their lin-
rod; and (3) a pressure acoustics module for water [32,33]. Each ear elastic behavior is governed by Newton’s Second Law [32,33]:
module is governed by its own equations that describe the specific
qm x2 u  r  T ¼ FV ei/ ð2Þ
physics as discussed in the following section.
492 Z. Wei, L.K. Weavers / Ultrasonics Sonochemistry 31 (2016) 490–498

where qm is the material density (kg m3), x is the angular fre- the horn tip was defined to be Z = 0 and horizontal planes were
quency (rad s1), u is the particle displacement (m), FV is the force defined as X–Y planes. A manual positioning system with a resolu-
per volume (N m3), and ei/ indicates the AC. tion of 2 cm was used to position the hydrophone accurately dur-
The pressure acoustics module has been used to simulate ultra- ing acoustic field mapping. The origin of the hydrophone was just
sound propagation in water. The acoustic wave equation is given as below the horn tip (X, Y, Z = 0, 0, 0). With the manual positioning
follows [33,35,36]: system, the hydrophone was then moved in the X–Y plane at
  2 cm intervals, followed by movements in the Z-direction (vertical)
1 x2 P
r   rP þ q þ ¼0 ð3Þ to map another X–Y plane. A full-scan of an X–Y plane was accom-
q qc2 plished through line scans in the x- or y-axis. X–Y planes below the
where q is the density of water (kg m3), c is the speed of ultra- horn tip (Z = 4 cm), at the horn tip (Z = 0 cm), and above the horn
sound propagation in water (m s1), P ¼ P A cosðxtÞ is the acoustic tip (Z = +4 cm) were scanned to generate acoustic field maps for
pressure (Pa; P A is the maximum acoustic pressure and t is time, the multi-stepped horn in the water tank. Hydrophone readings
s), and the dipole source q (m s2) is optional. For our setup, there in these scans were acquired as root mean square values by the
is no polarization (q = 0) for the longitudinal ultrasonic waves [36]. oscilloscope. Operational conditions such as power input and
water volume were constant for all measurements. The tempera-
ture of water in the tank varied from 18 °C to 22 °C depending
2.1.2. Assigned boundary conditions and initial inputs
on the length of sonication. Such temperature change was not
The boundary conditions set to couple the three modules are
found to alter hydrophone readings.
based on COMSOL Modeling Guides [32,33] and previous simula-
tion studies [16,18]. A structure-acoustic boundary was set to the
interface between the ultrasonic horn and water [33,37]. Specifi- 2.3. Acoustic emission
cally, the movement of the horn and surrounding solution was
coupled at the interface: The acoustic emission method was used to determine the cavi-
  tation region in the hydrophone-mapped acoustic field. Frequency
1
n   rP þ q ¼ an ð4Þ is a critical factor to determine the shape of a sound signal. At low
qs power intensity, a sinusoidal shape for a sound signal converted
where n is the normal unit vector, qs is the density of horn (kg m3), from AC indicates one dominant frequency (i.e., 20 kHz in our sys-
and an is the normal acceleration of the solution (m s2). Likewise, tem) and a linear vibration for bubbles. When a high intensity dis-
the stress exerted from the surrounding solution on the horn is sub- torts the linear system, multiples of the driving frequency (i.e.,
jected to the acoustic pressure changes in the solution as follows: ultraharmonics) are generated [40]. Beyond a threshold value, sub-
harmonics appear [40]. The numerical analysis of bubble oscilla-
T n¼Pn ð5Þ tion at subharmonic and ultraharmonic frequencies is explained
Displacements at the interface between the water and the wall of in the SI. Collapse of cavitation bubbles induces shock waves and
the tank were set to zero (u = 0 or P = 0), assuming the tank material micro-jets forming a noisy background. These bubble oscillations
with a large acoustic impedance sufficiently absorbed incident and collapses generate a broadband signal (i.e., an elevated base-
ultrasonic waves. Boundary conditions for surfaces contacting air line), which is indicative of transient cavitation [41]. Both the ele-
were also set to P = 0 [33]. The displacement at the joint between vated baseline and sharp peaks at the driving, subharmonic, and
the piezoelectric material and the stainless steel horn was set to ultraharmonic frequencies are characteristics of an observed
the same value [8,38,39]. The default temperature was 293.15 K. hydrophone spectrum from high power ultrasound. The frequency
The liquid, horn, and transducer domains were assigned to linear spectral analysis was carried out with PeakFit software (version
water media, piezoelectric material (PZT-5H), and stainless steel 4.12) which uses a Fast Fourier Transform (FFT) algorithm. We
material (AISI 4340), respectively. The input information of these assume that all sound signals and bubble dynamics are harmonic
materials is summarized in Table S1 of supporting information (SI). by using FFT. A wavelet transform algorithm, which analyzes
sound signals in both time and frequency domains, needs to be
used when bubble motions are not in steady state [42,43].
2.2. Experimental verification

2.2.1. Ultrasonic system 3. Results and discussion

As shown in Fig. S1, a Branson BCA 900 series power supplier
(1000 W at maximum) was used to transmit electrical power to a 3.1. Acoustic field modeling
Branson 902R Model ultrasonic transducer (20 kHz) which was
connected to a multi-stepped horn. The ultrasonic horn was placed COMSOL simulations were first conducted to estimate ultra-
at the center of a water tank (61 cm  61 cm  45 cm, 167.5 L) sound penetration distance and cavitation locations for the
made of plexiglass. A Reson TC4013 type hydrophone (Reson A/S, multi-stepped horn. The modeling result is a valuable reference
Denmark) was used to measure acoustic pressure in the water for the subsequent experimental design. In the simulation, several
tank. The hydrophone was connected to a TDS 5000 Tektronix assumptions were made: (1) there is no energy loss due to piezo-
oscilloscope (Tektronix Inc., USA) which recorded and displayed electric effects or transmission of mechanical energy from the
the sound signals at a sampling frequency of 125 kHz. Another typ- transducer to the horn rod; thus, simulation results may overesti-
ical horn-type ultrasonic system (Sonic Dismembrator 550, Fisher mate particle displacements for both the piezoelectric material and
Scientific) was used to determine the cavitation threshold follow- stainless steel horn rod; (2) the acoustic pressure distribution in
ing the method of Ashokkumar et al. [30]. the tank is symmetric and damping of the ultrasonic waves is
neglected; (3) there are no cavitation bubbles generated in the
2.2.2. Experimental procedure tank; and (4) water movement in the tank is negligible.
Approximately 150 L of water was filled to a depth of 40 cm in In the construction of 2D half geometry model (Fig. S2), the
the plexiglass tank and was left overnight allowing for air satura- ultrasonic horn irradiates water in a cylindrical volume with a
tion. The multi-stepped horn was submerged to the depth of diameter of 31 cm and a height of 36 cm. Fig. 1 shows the simu-
16 cm (from horn tip to water surface). The depth right below lated acoustic pressure distribution in an X–Z plane (vertical)
 Z. Wei, L.K. Weavers / Ultrasonics Sonochemistry 31 (2016) 490–498 493

Fig. 1. Simulation of acoustic pressure distribution in X–Z plane (units for color labels and axes are Pa and mm, respectively). (For interpretation of the references to color in
this figure legend, the reader is referred to the web version of this article.)

Fig. 2. Simulation of acoustic pressure distribution in X–Y planes at different depths (1 — X–Y plane near water surface; 2 — X–Y plane in the middle of horn neck; 3 — X–Y
plane at Z = 0 cm; 4–6 — X–Y planes at Z < 0 cm; units for color labels and axes are Pa and mm, respectively). (For interpretation of the references to color in this figure legend,
the reader is referred to the web version of this article.)

where red or blue indicates a high absolute acoustic pressure. Due water forming ripples. The acoustic pressure decreases from the
to the propagation of ultrasonic waves, the red and blue colors center to the edges due to the wave interactions at the boundaries
oscillate temporally in those regions. Therefore, the term ‘‘high where displacement of tank material was set to zero. Thus, color
acoustic pressure region” indicates both red and blue areas unless changes from red to yellow and blue to cyan in the acoustic pres-
noted otherwise. As shown in Fig. 1, high acoustic pressure regions sure modeling simulations reflect the effect of constructive/
surrounding the horn neck and below its tip were observed. At destructive interferences that are induced by the wave interaction
regions further from the probe, ultrasonic waves propagate in the between the multi stepped horn and the geometrical characteris-
494 Z. Wei, L.K. Weavers / Ultrasonics Sonochemistry 31 (2016) 490–498

50% 75% 100% tics of the vessel. Fig. 2 compares acoustic fields in X–Y planes at
15 15 15
different depths, where plane 3 is at Z = 0 cm. Ultrasonic waves
10 10 10 emitting from the horn neck generate a large high acoustic pres-
sure region in plane 2. In plane 2, the distance of the dark-
5 5 5 colored region extends to approximately 10 cm as opposed to
65 cm in other X–Y planes. The simulated acoustic pressure maps
Z, cm

0 0 0
indicate that areas surrounding the ultrasonic horn neck (Z > 0 cm)
-5 -5 -5 are more likely to generate multiple cavitation zones and increase
cavitation volumes.
-10 -10 -10
2cm
-15 -15 -15 5cm
10cm 3.2. Acoustic field mapping
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
Vrms, V Vrms, V Vrms, V The acoustic pressure distribution surrounding the multi-
stepped horn was verified using hydrophone measurements in
Fig. 3. Acoustic pressure distribution in Z-direction at different distances (2 cm, the plexiglass tank. Hydrophone readings were recorded as root
5 cm and 10 cm) and power levels (50%, 75% and 100%; red dotted line is the mean square values. Thus, values are reported as positive values
cavitation threshold value of 0.63 V).
as opposed to alternating values shown in simulations. First, the

(a)
2.0
V

20 1.80
Hydrophone reading,

1.60
1.5 10 1.40
1.20
Y, cm

0
1.0 1.00
0.80
-10
0.63
0.5
20 -20 0.40
10
-30 0 0.20
-20
cm

-10 -10
0 -30 -20 -10 0 10 20 30
X, cm 10 -20
Y,

20
30 X, cm
(b)
2.0
V

20 1.80
Hydrophone reading,

1.60
1.5
10 1.40
1.20
Y, cm

0
1.0 1.00
0.80
-10
0.63
0.5
20 -20 0.40
10
-30 0 0.20
-20
cm

-10 -10
0 -30 -20 -10 0 10 20 30
X, cm 10 -20
Y,

20
30 X, cm
(c)
2.0
V

20 1.80
Hydrophone reading,

1.60
1.5
10 1.40
1.20
Y, cm

1.0 0
1.00
0.80
-10
0.5 0.63
20 0.40
10 -20
-30 0 0.20
-20
cm

-10 -10
0 -30 -20 -10 0 10 20 30
X, cm 10 -20
Y,

20
30 X, cm

Fig. 4. 3D (left) and contour (right) mapping of hydrophone measurements in plexiglass tank (This scan was carried out at room temperature with 50% power input from
power supply to transducer; a — X–Y plane at Z = +4 cm; b — X–Y plane at Z = 0 cm; c — X–Y plane at Z = 4 cm).
 Z. Wei, L.K. Weavers / Ultrasonics Sonochemistry 31 (2016) 490–498 495

0.2
-2 -2
< 0.04W cm < 0.04W cm
100

Amplitude, V
0.1

Magnitude
0.0 10

-0.1
1
-0.2
1.0
-2 1000 -2
0.04W cm 0.04W cm
Amplitude, V

0.5

Magnitude
100
0.0

-0.5 10

-1.0
1.0
-2 1000 -2
0.31W cm 0.31W cm
Amplitude, V

0.5

Magnitude
100
0.0

-0.5 10

-1.0

-2 -2
0.74W cm 1000 0.74W cm
2
Amplitude, V

Magnitude

0 100

-2 10

-2 -2
1.16W cm 1000 1.16W cm
2
Amplitude, V

Magnitude

0 100

-2 10

-0.10 -0.05 0.00 0.05 0.10 0 10 20 30 40 50 60

Time, ms Frequency, kHz

Fig. 5. Ultrasonic waveforms (left) and frequency spectra (right) observed in water at different power intensities (convex feature in the waveform results from the sum of
waveforms in different frequencies to the original waveform; units for magnitude of frequency spectra are arbitrary).

acoustic pressure from Z = 15 cm to Z = +15 cm was measured at

different distances to the horn neck (i.e., 2 cm, 5 cm, and 10 cm), as
Area above cavitation threshold, %

100 Z = +4cm shown in Fig. 3. Apparently, the radial region of the horn neck
Z = 0cm
(0 6 Z 6 +15 cm) exhibited higher acoustic pressure compared to
Z = - 4cm
the regions below the horn tip. The fluctuating pressure magni-
80
tudes along the multi-stepped horn suggest that the constructive
interference of ultrasonic waves resulted in a high acoustic pres-
60 sure region while destructive interference resulted in a low pres-
sure region [44–46]. In addition, the acoustic pressure decayed
with distance (2 cm > 5 cm > 10 cm) consistent with the simulated
40
ultrasound propagation in Fig. 2. However, the higher power input
does not intensify the acoustic pressure. This unexpected observa-
20 tion probably reflects the scattering of sound by a large amount of
cavitation bubbles thereby reducing sound propagation [23]. Such
a nonlinear relationship between acoustic pressure and power
0
input has also been observed in previous studies [23,47]. They
0 5 10 15 20 25 30
attributed the nonlinearity to the acoustic energy dissipated into
Horizontal distance from horn neck, cm frequencies beyond the hydrophone detection limit and the shield-
ing effect of cavitation bubbles that limits the propagation of ultra-
Fig. 6. Percentage of cavitation zones in different X–Y planes (% of cavitation
zones = Measurements not less than 0.63 V in a X–Y plane/total measurements in
sound in water-filled vessels.
the X–Y plane  100%; 0.63 V is measured cavitation threshold using acoustic In addition to vertical mapping, horizontal propagation of ultra-
emission method). sonic waves in water is also depicted in 3D and contour plotting
496 Z. Wei, L.K. Weavers / Ultrasonics Sonochemistry 31 (2016) 490–498

(Fig. 4). As shown in Fig. 4, a decreasing intensity from the tank increasing power input, bubble oscillations depart from this linear
center to its edges was observed. Particularly at Z = 0 cm, it was nature producing convex waveforms (Fig. 5). The addition of shock
obvious that the center area below the horn tip exhibited the high- waves, micro-jets, and micro-streaming after collapse of cavitation
est acoustic pressure levels; at Z = +4 cm, the horn neck emitted bubbles further increases the degree of irregularity of acoustic
ultrasonic waves and created a large high acoustic pressure region; waveforms.
at Z = 4 cm, the acoustic pressure distribution was more dis- Frequency spectra in Fig. 5 are consistent with the waveforms.
persed without obvious spots of higher intensity. The observation At low power intensities, the driving frequency (f) and ultrahar-
of a larger scale of high pressure region at Z > 0 cm and a more dis- monic frequency (2f) were observed. As power intensity was
crete distribution of acoustic pressures at Z 6 0 cm were consistent increased (P0.04 W cm2), subharmonic frequencies were also
with the simulation results in Fig. 2. However, a standing wave present. The number of subharmonic and ultraharmonic frequen-
pattern of propagation was not observed due to the following cies increased significantly at a power intensity of 0.31 W cm2.
acoustic effects [23,48]: (1) cavitation shielding due to the pres- At 0.74 W cm2, the baseline was elevated to a magnitude of
ence of cavitation bubbles interferes with ultrasound propagation approximately 10, showing the feature of a broadband signal that
(e.g., sound intensity attenuation and sound velocity reduction is an indicator of transient cavitation [30]. Therefore, transient cav-
resulting from scattering at the bubble-water interface); (2) colli- itation is present at power intensities of 0.74 W cm2 and higher,
sions between emitted ultrasonic waves from the horn neck and which is similar to 0.70 W cm2 observed by Ashokkumar et al.
reflected waves from the tank wall disrupt the applied acoustic [30]. While the threshold is somewhere between 0.31 W cm2
pressure; and (3) agitation of water by acoustic streaming drifts and 0.74 W cm2, we defined the acoustic intensity of 0.74 W cm2
vibrating molecules off their original positions resulting in the dis- as the ‘‘threshold” for transient cavitation, which corresponds to a
crete and asymmetric distribution of acoustic pressure. Hodnett hydrophone reading of 0.63 Volt. Even though this defined thresh-
et al. [23] also show an asymmetric but reproducible distribution old may underestimate the power intensity of transient cavitation,
of the acoustic field in the characterization of a reference ultrasonic setting a threshold slightly high ensures necessary properties for a
cavitation vessel. Mhetre and Gogate [49] in recent work present a sufficient design.
non-uniform cavitation activity distribution in traditional dosime- Using the cavitation threshold defined, a cavitation zone was
try tests using potassium iodide (KI) in a large-scale sonochemical identified. As shown in Fig. 3, the threshold for transient cavitation
reactor (72 L in volume). It seems hydrophone measurements are was plotted as a red dotted line. The measurements higher than
capable of generating acoustic field maps consistent with tradi- the cavitation threshold were generally located between Z = 0 cm
tional chemical methods. and Z = +15 cm which is along the neck of the horn. The cavitation
region along the neck (Z > 0 cm) extended up to 10 cm from the
3.3. Cavitation threshold and reactive region horn axis while the cavitation zone below the horn tip (Z 6 0 cm)
extended up to 5 cm laterally from the axis (75% and 100% power
After mapping the acoustic field in the large water tank, the inputs). At 50% power input, up to 10 cm of cavitation region was
next step was to evaluate the effective range of the cavitation also observed right below the horn tip (5 cm < Z 6 0 cm) recon-
zones based on the threshold value of cavitation which was deter- firming the shielding effect of cavitation bubbles. In Fig. 4, regions
mined using acoustic emissions. Fig. 5 shows acoustic waveforms higher than the cavitation threshold are cyan and warmer colors.
acquired on the oscilloscope and the corresponding spectra. The At Z = +4 cm, there was a large area with cyan to red colors sur-
waveforms are sinusoidal to irregular in shape and become more rounding the horn neck. In contrast, the cavitation zones were
irregular with increasing acoustic intensity (<0.04–1.16 W cm2). much smaller at Z = 0 cm and Z = 4 cm. In order to quantitatively
In an ideal system, the sinusoidal AC input is converted via a trans- describe the cavitation regions in an X–Y plane, the ratio of col-
ducer into a sinusoidal vibration that is propagated through the lected data points higher than the threshold value to the total
ultrasonic horn to aqueous solution. Without dissipation, water number of scanned points was calculated (Fig. 6). At Z = +4 cm,
movement and cavitation bubbles, the hydrophone captures a the cavitation zone covered >85.0% of a 10 cm  10 cm area. The
sinusoidal sound signal that is displayed on the oscilloscope. With percentage of the zone above the transient cavitation threshold

Fig. 7. Simulation of acoustic pressure propagation in proposed reactor configuration (red — up to +1.59  105 Pa; blue color — down to 1.59  105 Pa). (For interpretation
of the references to color in this figure legend, the reader is referred to the web version of this article.)
 Z. Wei, L.K. Weavers / Ultrasonics Sonochemistry 31 (2016) 490–498 497

dropped to 73.2% and 49.5% when distances were extended to cant impact in the reactor design because absorption and reflection
12 cm and 20 cm from the horn axis, respectively. If >85% is of incident waves on the reactor wall will change the acoustic pres-
selected as a reasonable percentage of a zone undergoing transient sure distribution in the reactor. Therefore, incorporating necessary
cavitation in the reactor, a cylindrically-shaped reactor with a properties such as water viscosity, heat production, cavitation bub-
10 cm radius would be designed to fit the multi-stepped horn. In bles, cavitation shielding, and reactor materials into simulations is
the X–Y planes that did not cross the horn neck, the percentage necessary to improve simulation results.
of cavitation zones dropped dramatically. For example, at 10 cm
from the horn axis, the percentage of cavitation zones was 47.9% Acknowledgments
and 26.4% for Z = 0 cm and Z = 4 cm, respectively. Even though a
relatively low percentage was observed at Z 6 0 cm, both X–Y This work was supported by the Office of Naval Research (ONR
planes featured a high acoustic pressure center below the horn Contract No. N00014-04-C-0430) and Ohio Sea Grant College
tip. Thus, a shrinking shaped bottom, such as a conical shape, could Program.
be introduced to the reactor design to increase the percentage of
total cavitation volume. In addition, a cone-shaped bottom is ben-
eficial for solution circulation and mass transfer inside the reactor, Appendix A. Supplementary data
as verified in our previous studies [50,51].
With those conditions considered, we propose a cylindrically- Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ultsonch.2016.01.
shaped reactor with a 10 cm diameter and a conical bottom with
5 cm in depth (21 cm in total depth). As shown in frame 8 of 036.
Fig. 7, the treatment volume for this design was approximately
5.0 L, which is nearly 100-fold greater than the reactor volume References
for a typical ultrasonic horn [50,51]. We further verified the design
[1] T.J. Mason, J.P. Lorimer, Applied Sonochemistry: The Use of Power Ultrasound
using COMSOL software. Using COMSOL we simulated the acoustic in Chemistry and Processing, Wiley-VCH, Verlag GmbH, Weinheim, 2002.
pressure distribution and ultrasound propagation (frame 1–7 in [2] K.S. Suslick, Ultrasound: Its Chemical, Physical, and Biological Effects, VCH
Fig. 7) in the reactor. As shown in Fig. 7, the majority of the reactor Publishers, New York, 1988.
[3] R.Y. Xiao, Z.S. Wei, D. Chen, L.K. Weavers, Kinetics and mechanism of
was covered by high acoustic pressure regions in red (up to sonochemical degradation of pharmaceuticals in municipal wastewater,
+1.59  105 Pa) and blue (down to 1.59  105 Pa) colors suggest- Environ. Sci. Technol. 48 (2014) 9675–9683.
ing that multiple reactive zones exist and a large cavitation volume [4] D. Chen, L.K. Weavers, H.W. Walker, Ultrasonic control of ceramic membrane
fouling by particles: effect of ultrasonic factors, Ultrason. Sonochem. 13 (2006)
can be generated in the reactor if applying high intensity ultra-
379–387.
sound. The animation of pressure propagation starts in frame 1 [5] H.N. McMurray, B.P. Wilson, Mechanistic and spatial study of ultrasonically
and ends at frame 7. Frames 1 and 7 are identical indicating a com- induced luminol chemiluminescence, J. Phys. Chem. A. 103 (1999) 3955–3962.
plete cycle of propagation. The cyclic propagation of ultrasound [6] J.M. Mason, A. Tiehm, Advances in Sonochemistry, Jai Press, Connecticut, 2001.
[7] Z. Wei, J.A. Kosterman, R. Xiao, G.Y. Pee, M. Cai, L.K. Weavers, Designing and
suggests a reproducible acoustic pressure distribution which is a characterizing a multi-stepped ultrasonic horn for enhanced sonochemical
key design factor for sono-reactors. performance, Ultrason. Sonochem. 27 (2015) 325–333.
[8] S.L. Peshkovsky, A.S. Peshkovsky, Matching a transducer to water at cavitation:
acoustic horn design principles, Ultrason. Sonochem. 14 (2007) 313–322.
[9] S.I. Nikitenko, C. Le Naour, P. Moisy, Comparative study of sonochemical
4. Applications and limitations reactors with different geometry using thermal and chemical probes, Ultrason.
Sonochem. 14 (2007) 330–336.
[10] J. Rooze, E.V. Rebrov, J.C. Schouten, J.T.F. Keurentjes, Effect of resonance
This study describes a method of using COMSOL simulations frequency, power input, and saturation gas type on the oxidation efficiency of
and acoustic pressure mapping from hydrophone measurements an ultrasound horn, Ultrason. Sonochem. 18 (2011) 209–215.
for sono-reactor design. The COMSOL simulations showed regions [11] V.S. Sutkar, P.R. Gogate, Design aspects of sonochemical reactors: Techniques
for understanding cavitational activity distribution and effect of operating
of high acoustic pressure similar to the acoustic pressure maps cre- parameters, Chem. Eng. J. 155 (2009) 26–36.
ated in the plexiglass tank, suggesting this coupling method may [12] R. Xiao, D. Diaz-Rivera, L.K. Weavers, Factors influencing pharmaceutical and
be used as a tool in the design and characterization of an ultrasonic personal care product degradation in aqueous solution using pulsed wave
ultrasound, Ind. Eng. Chem. Res. 52 (2013) 2824–2831.
system. In addition, the multi-stepped horn with a 10 cm radiation
[13] P.R. Gogate, V.S. Sutkar, A.B. Pandit, Sonochemical reactors: important design
radius and 5.0 L treatment capacity with a high expected amount and scale up considerations with a special emphasis on heterogeneous
of cavitation in the reactor shows great potential for large-scale systems, Chem. Eng. J. 166 (2011) 1066–1082.
applications through an array of these horns. The next step is to [14] T.J. Mason, A. Collings, A. Sumel, Sonic and ultrasonic removal of chemical
contaminants from soil in the laboratory and on a large scale, Ultrason.
build the sono-reactor with this proposed configuration and quan- Sonochem. 11 (2004) 205–210.
titatively evaluate its performance through traditional calorimetry, [15] L. Csoka, S.N. Katekhaye, P.R. Gogate, Comparison of cavitational activity in
dosimetry, and sonochemical processing of a model compound. different configurations of sonochemical reactors using model reaction
supported with theoretical simulations, Chem. Eng. J. 178 (2011) 384–390.
Although the overall trends of simulated acoustic pressure [16] J. Klima, A. Frias-Ferrer, J. Gonzalez-Garcia, J. Ludvik, V. Saez, J. Iniesta,
maps are consistent with experimental measurements, accurate Optimisation of 20 kHz sonoreactor geometry on the basis of numerical
modeling of these systems needs further development. First of simulation of local ultrasonic intensity and qualitative comparison with
experimental results, Ultrason. Sonochem. 14 (2007) 19–28.
all, an immediate challenge for future COMSOL simulations is to [17] F.J. Trujillo, K. Knoerzer, A computational modeling approach of the jet-like
couple bubble dynamics with the acoustic wave equation. Kumar acoustic streaming and heat generation induced by low frequency high power
et al. utilized a continuum mixture model and a diffusion limited ultrasonic horn reactors, Ultrason. Sonochem. 18 (2011) 1263–1273.
[18] Y.C. Wang, M.C. Yao, Realization of cavitation fields based on the acoustic
model to explore the behavior of cavitation bubbles in a flow, sug- resonance modes in an immersion-type sonochemical reactor, Ultrason.
gesting a possible methodology to couple bubble dynamics with Sonochem. 20 (2013) 565–570.
the acoustic wave equation in a sono-reactor [52,53]. Second, a [19] R. Jamshidi, B. Pohl, U. Peuker, G. Brenner, Numerical investigation of
sonochemical reactors considering the effect of inhomogeneous bubble
simplified transducer was created in current simulations
clouds on ultrasonic wave propagation, Chem. Eng. J. 189–190 (2012) 364–
(Fig. S2). Modeling of the transducer to reflect its inside structure 375.
is beyond the scope of this paper, but is an aspect to be developed [20] R.Y. Xiao, M. Noerpel, H.L. Luk, Z.S. Wei, R. Spinney, Thermodynamic and
in future studies. In addition, ideal conditions were applied for all kinetic study of ibuprofen with hydroxyl radical: a density functional theory
approach, Int. J. Quantum. Chem. 114 (2014) 74–83.
physical modules used and reactor material was not considered in [21] N. Feng, Ultrasonics Handbook, first ed., Nanjing University Press, Nanjing,
the simulations. The material and shape used may have a signifi- 1999.
498 Z. Wei, L.K. Weavers / Ultrasonics Sonochemistry 31 (2016) 490–498

[22] K.F. Graff, Wave Motion in Elastic Solids, Dover Publications Inc, New York, [38] Z.Q. Fu, X.J. Xian, S.Y. Lin, C.H. Wang, W.X. Hu, G.Z. Li, Investigations of the
1975. barbell ultrasonic transducer operated in the full-wave vibrational mode,
[23] M. Hodnett, M.J. Choi, B. Zeqiri, Towards a reference ultrasonic cavitation Ultrasonics 52 (2012) 578–586.
vessel. Part 1: preliminary investigation of the acoustic field distribution in a [39] Z. Lin, Theory and Desing of Ultrasonic Horn, Science Press, Beijing, 1987.
25 kHz cylindrical cell, Ultrason. Sonochem. 14 (2007) 29–40. [40] A. Eller, H.G. Flynn, Generation of subharmonics of order one-half by bubbles
[24] M. Hodnett, B. Zeqiri, A strategy for the development and standardisation of in a sound field, J. Acoust. Soc. Am. 46 (1969) 722–727.
measurement methods for high power cavitating ultrasonic fields: review of [41] E.A. Neppiras, Subharmonic and other low-frequency emission from bubbles
high power field measurement techniques, Ultrason. Sonochem. 4 (1997) 273– in sound-irradiated liquids, J. Acoust. Soc. Am. 46 (1969) 587–601.
288. [42] V.S. Moholkar, M. Huitema, S. Rekveld, M.M.C.G. Warmoeskerken,
[25] M. Juschke, C. Koch, Model processes and cavitation indicators for a Characterization of an ultrasonic system using wavelet transforms, Chem.
quantitative description of an ultrasonic cleaning vessel: Part I: Eng. Sci. 57 (2002) 617–629.
experimental results, Ultrason. Sonochem. 19 (2012) 787–795. [43] V.S. Moholkar, S.P. Sable, A.B. Pandit, Mapping the cavitation intensity in an
[26] J. Frohly, S. Labouret, C. Bruneel, I. Looten-Baquet, R. Torguet, Ultrasonic ultrasonic bath using the acoustic emission, AIChE J. 46 (2000) 684–694.
cavitation monitoring by acoustic noise power measurement, J. Acoust. Soc. [44] S. Dahnke, F.J. Keil, Modeling of three-dimensional linear pressure fields in
Am. 108 (2000) 2012–2020. sonochemical reactors with homogeneous and inhomogeneous density
[27] D.L. Miller, Ultrasonic detection of resonant cavitation bubbles in a flow tube distributions of cavitation bubbles, Ind. Eng. Chem. Res. 37 (1998) 848–864.
by their 2nd-harmonic emissions, Ultrasonics 19 (1981) 217–224. [45] S. Dahnke, K.M. Swamy, F.J. Keil, A comparative study on the modeling of
[28] Y. Son, M. Lim, J. Khim, M. Ashokkumar, Acoustic emission spectra and sound pressure field distributions in a sonoreactor with experimental
sonochemical activity in a 36 kHz sonoreactor, Ultrason. Sonochem. 19 (2012) investigation, Ultrason. Sonochem. 6 (1999) 221–226.
16–21. [46] S.W. Dahnke, F.J. Keil, Modeling of linear pressure fields in sonochemical
[29] E. Cramer, W. Lauterborn, Acoustic cavitation noise spectra, Appl. Sci. Res. 38 reactors considering an inhomogeneous density distribution of cavitation
(1982) 209–214. bubbles, Chem. Eng. Sci. 54 (1999) 2865–2872.
[30] M. Ashokkumar, M. Hodnett, B. Zeqiri, F. Grieser, G.J. Price, Acoustic emission [47] M. Hodnett, R. Chow, B. Zeqiri, High-frequency acoustic emissions generated
spectra from 515 kHz cavitation in aqueous solutions containing surface- by a 20 kHz sonochemical horn processor detected using a novel broadband
active solutes, J. Am. Chem. Soc. 129 (2007) 2250–2258. acoustic sensor: a preliminary study, Ultrason. Sonochem. 11 (2004) 441–454.
[31] G.J. Price, M. Ashokkumar, M. Hodnett, B. Zequiri, F. Grieser, Acoustic emission [48] K.W. Commander, A. Prosperetti, Linear pressure waves in bubbly liquids –
from cavitating solutions: Implications for the mechanisms of sonochemical comparison between theory and experiments, J. Acoust. Soc. Am. 85 (1989)
reactions, J. Phys. Chem. B. 109 (2005) 17799–17801. 732–746.
[32] COMSOL, Acoustics Module User’s Guide, version 4.2, 2012. [49] A.S. Mhetre, P.R. Gogate, New design and mapping of sonochemical reactor
[33] COMSOL, COMSOL Multiphysics Modeling Guide, version 4.2, 2012. operating at capacity of 72 L, Chem. Eng. J. 258 (2014) 69–76.
[34] S.O.R. Moheimani, A.J. Fleming, Piezoelectric Transducers for Vibration Control [50] Z.Q. He, S.J. Traina, J.M. Bigham, L.K. Weavers, Sonolytic desorption of mercury
and Damping, Oxford University Press, Oxford, UK, 2006. from aluminum oxide, Environ. Sci. Technol. 39 (2005) 1037–1044.
[35] M.W. Nygren, Finite element modeling of piezoelectric ultrasonic [51] G.Y. Pee, S. Na, Z. Wei, L.K. Weavers, Increasing the bioaccessibility of
transducers (Master thesis), Norwegian University of Science and polycyclic aromatic hydrocarbons in sediment using ultrasound,
Technology, 2011. Chemosphere 122 (2015) 265–272.
[36] L.E. Kinsler, A.R. Frey, A.B. Coppens, J.V. Sanders, Fundamentals of Acoustics, [52] P. Kumar, S. Khanna, V.S. Moholkar, Flow regime maps and optimization
fourth ed., John Wiley & Sons, New York, NY, 2000. thereby of hydrodynamic cavitation reactors, AIChE J. 58 (2012) 3858–3866.
[37] M. Yao, Analysis and experiment of resonant sonochemical cell (Master [53] V.S. Moholkar, A.B. Pandit, Modeling of hydrodynamic cavitation reactors: a
thesis), National Cheng Kung University, 2009. unified approach, Chem. Eng. Sci. 56 (2001) 6295–6302.

View publication stats