AUTHOR QUERY FORM

Journal: J. Acoust. Soc. Am.

Please provide your responses and any corrections by

annotating this PDF and uploading it according to the instructions

Article Number: 042509JAS

provided in the proof notification email.

Dear Author,
Below are the queries associated with your article; please answer all of these queries before sending the proof back to AIP.
Please indicate the following:
Figures that are to appear as color online only (i.e., Figs. 1, 2, 3) _____________________________ (this is a free service).
Figures that are to appear as color online and color in print______________________ (a fee of $325 per figure will apply).
Article checklist: In order to ensure greater accuracy, please check the following and make all necessary corrections before
returning your proof.
1. Is the title of your article accurate and spelled correctly?
2. Please check affiliations including spelling, completeness, and correct linking to authors.
3. Did you remember to include acknowledgment of funding, if required, and is it accurate?
Location in

Query / Remark: click on the Q link to navigate

to the appropriate spot in the proof. There, insert your comments as a PDF annotation.

article
AQ1

Payment of page charges for papers over 12 pages is mandatory. If you will not honor the page charges then you must reduce
your paper to 12 pages.

AQ2

Please check that the author names are in the proper order and spelled correctly. Also, please ensure that each authors given
and surnames have been correctly identified (given names are highlighted in red and surnames appear in blue).

AQ3

Please define CFD in the abstract.

AQ4

Confirm change to NIST23. This is how it appears on the reference list.

AQ5

Please supply city and state of all manufacturers throughout the paper.

AQ6

Please check Refs. 23 and 30. They contain similar information, except that Ref. 30 cites authors and Ref. 23 does not.

AQ7

Please add all section numbers and letters that you are referring to here.

AQ8

Please confirm the section number and letter.

AQ9

Please confirm that  is the correct symbol here. The PDF contains a box.

AQ10

Please specify section numbers and letters.

AQ11

Please check if the  symbol is correct. The PDF showed a blank box.

AQ12

Please provide a digital object identifier (doi) for Ref(s). 36. For additional information on dois please select this link: http://
www.doi.org/. If a doi is not available, no other information is needed from you.

Thank you for your assistance.

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 1

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

AQ1

1
2

AQ2

Experimental and numerical investigations of resonant acoustic

waves in near-critical carbon dioxide

Nusair Hasan and Bakhtier Farouka)

Mechanical Engineering and Mechanics, Drexel University, Philadelphia, Pennsylvania 19014, USA

5
6

(Received 8 July 2014; revised 16 August 2015; accepted 30 August 2015; published online xx xx
xxxx)

Acoustically augmented flows represent an efficient way of enhancing mixing and mass transport
processes in supercritical fluids and has applications in the chemical separation and extraction
industry. Flow and transport induced by resonant acoustic waves in a near-critical fluid filled
cylindrical enclosure is investigated both experimentally and numerically. Supercritical carbon
dioxide (near the critical or the pseudo-critical states) in a confined resonator is subjected to
acoustic field created by an electro-mechanical acoustic transducer and the induced pressure
waves are measured by a fast response pressure field microphone. The frequency of the acoustic
transducer is chosen such that the lowest acoustic mode propagates along the enclosure. For numerical simulations, a real-fluid CFD model representing the thermo-physical and transport properties
of the supercritical fluid is considered. The simulated acoustic field in the resonator is compared
with measurements. The formation of acoustic streaming structures in the highly compressible
medium is revealed by time-averaging the numerical solutions over a given period. Due to
diverging thermo-physical properties of supercritical fluid near the critical point, large scale
oscillations are generated even for small sound field intensity. The strength of the acoustic wave
field is found to be in direct relation with the thermodynamic state of the fluid. The effects of
near-critical property variations and the operating pressure on the formation process of the
streaming structures are also investigated. Irregular streaming patterns with significantly higher
streaming velocities are observed for near-pseudo-critical states at operating pressures close to the
critical pressure. However, these structures quickly re-orients to the typical Rayleigh streaming
C 2015 Acoustical Society of America.
patterns with the increase operating pressure. V
[http://dx.doi.org/10.1121/1.4930951]

8
9
10
11
12
13
14

AQ3

15
16
17
18
19
20
21
22
23
24
25
26

[JDM]

Pages: 114

27

I. INTRODUCTION

28

Acoustic wave induced convective transport inside an

enclosure is a well-studied problem which has been investigated by several researchers.18 Hamilton et al.9 derived an
analytic solution for the secondary flow field generated by
standing waves confined by parallel plates. Acoustically
augmented convection has the application in accelerating
certain kinds of rate processes, especially the cooling process
of electronic systems under micro-gravity conditions, where
nature convections are greatly reduced or completely
eliminated. Transport mechanisms such as this can also be
employed to enhance mixing processes and to augment heat
and mass transfer through resonator walls. In recent years,
the problem has been extensively studied both from experimental7 and numerical1,8 perspectives. From a mass transport point of view, Kawahara et al.10 studied the effect of
standing wave fields on transport in an ideal gas. Convective
heat transfer from an isolated sphere in a standing sound
field was examined by Gopinath and Mills3 for large flow
Reynolds numbers. Heat transfer enhancement by acoustic
augmentation was also studied by Tajik et al.6 However,
most of these studies are limited to ideal gases at near atmospheric pressures and at present, investigations of acoustically

29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

a)

Electronic mail: bfarouk@coe.drexel.edu

J. Acoust. Soc. Am. 138 (3), September 2015

augmented transport phenomena in high pressure/supercritical fluids are absent in the literature.
Supercritical fluids are commonly used in industries
nowadays as solvents, efficient thermal storage, and transport media.11 Due to the increasing number of applications
of supercritical fluids in chemical and thermal process industries, convective transport in near-critical and supercritical
fluids has drawn a lot attention to the researchers in the
recent decade.1215 While supercritical fluids, in general, exhibit interesting physical properties,1 specific interest in CO2
is magnified by its perceived green propertiescarbon
dioxide is non-flammable, non-toxic, and relatively inert. In
addition, unlike water, the supercritical regime of CO2 is
readily accessible, given its critical temperature of only
304 K (and critical pressure of 7.38 MPa). However, the
transport dynamics of CO2 is relatively slow near the critical
point and, therefore, improvements in convective transport
(both thermal and mass) are required.16 Application of
mechanically driven acoustic waves represents a potential
efficient way of enhancing transport processes in supercritical fluid medium.17,18 The enhancement in transport processes is induced due to effects produced by compressions
and decompressions of the acoustic waves as well as the formation of the secondary flow-field (acoustic streaming).
Balachandran et al.18 studied the influence of ultrasound on
supercritical extraction based on extraction of gingerol from

0001-4966/2015/138(3)/1/14/$30.00

C 2015 Acoustical Society of America

V

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 2

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112

AQ4

113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132

a freeze-dried ginger sample. In a relatively recent experimental study, Riera et al.19,20 proposed power ultrasound
assisted supercritical fluid extraction to enhance the mass
transport in almond oil extraction. Although results from
both of these studies show that power ultrasound significantly accelerates the kinetics of the process and improves
the final extraction yield, a clear physical description of the
transport processes were not specified in either one of the
papers. Hasan and Farouk21 developed a computational fluid
dynamic model of acoustic field assisted supercritical fluid
extraction of caffeine from coffee beans in a fixed bed
extractor. However, the transport enhancement was observed
mainly due to localized convection (primary oscillatory flow
field) as no secondary flow structures were developed in the
reactor geometry.
In the present study, the generation and propagation of
mechanically driven acoustic waves in sub- and supercritical
carbon dioxide are experimentally studied and numerically
simulated. A closed, cylindrical shaped resonator filled with
carbon dioxide (at sub- and supercritical states) is considered. The primary oscillatory flow field in the enclosure is
generated by the vibration of one of the end walls of the resonator (using an electro-mechanical acoustic transducer).
The driving frequencies of the vibrating wall are chosen
such that the lowest acoustic mode propagates along the resonator and a standing wave field is produced. The amplitude
of the generated acoustic waves is measured at the pressure
anti-node (fixed end wall) by a fast response piezo-resistive
pressure field microphone. The variations in acoustic and
flow fields are also studied as a function of space and time.
Interaction of acoustic waves in highly compressible fluids
(e.g., supercritical fluids) and solid boundaries produces a
second-order steady flow field known as acoustic streaming.1,22 It is often observed in compressible media exposed
to an acoustic field. Formation of the standing wave and
acoustic streaming are numerically simulated by directly
solving the full compressible form of the Navier-Stokes
equations. The NIST23 Standard Reference Database 23 is
used to obtain the q-p-T relations for supercritical carbon
dioxide as well as the different thermo-physical and transport properties of the fluid. With the developed model, physical processes including the interaction of the wave field
with viscous effects and formation of streaming structures
are simulated. The effects of near-critical property variations
and fluid pressure on the formation process of the streaming
structures are also investigated in detail.

value, the fluid is in a supercritical state. A supercritical fluid

can thus be categorized as a single phase fluid that is dense
as a liquid, yet compressible as a gas.
Near the critical point, the thermo-physical and transport properties of fluids exhibit unusual behaviors; the specific heat (cp) and the isothermal compressibility (b) show
strong divergence, causing a vanishing thermal diffusivity
(a). The near-critical thermo-physical property variations are
also influenced by the corresponding pseudo-critical states.
A pseudo-critical state of a pure fluid can be defined as a
state in near-critical supercritical region at which the density
of the fluid is equal to its critical density (i.e., where q qc )
and the thermodynamic and transport properties have their
maximum rate of change with temperature at a constant
supercritical pressure.24 Below the temperature corresponding to the pseudo-critical state (where q=qc > 1), the fluid
has liquid-like properties while above (where q=qc < 1), it
more closely resembles a vapor (gas). Figure 1 shows the
density vs temperature diagram for supercritical carbon
dioxide with changing pressure. The horizontal line refers to
the critical density (qc 467.6 kg/m3). At atmospheric pressure, carbon dioxide behaves like an ideal gas and the
density is almost constant, while the density in the
supercritical-pressure condition widely varies across the phase
interface from the liquid or gas phase to the supercritical fluid
phase. The isothermal compressibility also shows a strong
divergence around the near-critical pseudo-critical states.
Hence, even around 10 K above the critical temperature
(304.1282 K), carbon dioxide is found to be highly compressible but the density remains high. Acoustic speed of the fluid
medium is an important factor for generation standing waves
in a resonator. The acoustic speed also exhibits divergence
around the near-critical pseudo-critical states. Due to this
wide variation of the thermo-physical and transport properties
of near-critical fluids, it is expected that the behavior of
the generated acoustic field and the formation of acoustic

II. THERMO-PHYSICAL PROPERTIES AT

SUPERCRITICAL STATE

The critical point of a fluid specifies the conditions

(temperature and pressure) at which a phase boundary ceases
to exist and above which distinct liquid and gas phases do
not exist. For pure substances, there is an inflection point in
the critical isotherm on a p-v diagram, leading to @p=@vT
@ 2 p=@v2 T 0. The critical pressure (pc) for carbon
dioxide is 7.3773 MPa, the critical temperature (Tc) is
304.1282 K and the critical density (qc) is 467.6 kg/m3.23 If
both pressure and temperature are beyond each critical
2

J. Acoust. Soc. Am. 138 (3), September 2015

FIG. 1. Variation of density, q as functions of temperature and pressure in

near-critical carbon dioxide.
Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 3

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

FIG. 2. (Color online) Schematic diagram of the problem geometry.

170

streaming structures would be significantly different than that

in ideal gases.

171

III. PROBLEM DESCRIPTION

172

186

A cylindrical resonator with a length (L) of 25.0 cm is

considered. The aspect ratio (L/D) of the resonator is 12.5
(diameter of the resonator is 2.0 cm). The schematic of the
geometry considered is shown in Fig. 2. One of the end walls
(see Fig. 2) of the resonator is vibrated at a specific frequency. The frequency of vibration is chosen such that the
lowest acoustic mode propagates through the fluid medium
(i.e., f csi / 2 L; where csi is the acoustic speed at the initial
state of the fluid). The axial symmetry of the problem
geometry is utilized and two dimensional (2D) cylindrical
coordinate (r - z) system is considered for the numerical calculations to simplify the computational efforts. Details
regarding the experimental setup used for the measurement
of the acoustic waves in sub- and supercritical carbon dioxide are discussed in Sec. IV.

187

IV. EXPERIMENTAL SETUP

188

Achieving supercritical state safely is a key point for the

success of generation and measurement of resonant acoustic
waves in supercritical carbon dioxide. The experimental
setup is designed following our previous work25 with supercritical carbon dioxide. Figure 4 shows a schematic illustration of the experimental setup.
The supercritical chamber consists of gas inlet and
outlet ports fitted with high pressure (103.4 MPa) ball valves
and the chamber is connected to a carbon dioxide tank
(p  6.0 MPa). A hand pump (HiP, 34.5 MPa, 60 mL/stroke)
is used to raise the pressure of the carbon dioxide in the
supercritical chamber from the tank pressure to a pressure

169

173
174
175
176
177
178
179
180
181
182
183
184
185

189
190
191
192
193
194
195
196
197
198
199

above the critical point of carbon dioxide. Heating tape

(Thermolyne, 0.5 in.  4 ft) wrapped around the supercritical
chamber is used to raise the temperature of the carbon dioxide in the supercritical chamber from the room temperature
(usually 295 K) to the operating temperature. The mechanically driven acoustic waves in supercritical carbon dioxide
are studied in a PTFE (Polytetrafluoroethylene) tube (inside
diameter 2.0 cm, length 25.0 cm) snugly fitted inside the
supercritical chamber. A small hole in the PTFE tubing
aligned with the gas inlet port provides flow path for the gas
to and from the supercritical chamber. The resonator is
mounted horizontally with the acoustic driver on the right
end (see Fig. 3) and a plug made of polyoxymethylene
(Delrin) is used to close the opposite end. Delrin is a hard,
high stiffness thermoplastic that prevents any absorption/
attenuation of the acoustic waves during reflection from the
end wall.
The acoustic driver is an electro-mechanical driver type
loudspeaker (CUI CMS0401KL-3X). The sinusoidal driving
signal of the acoustic driver is generated by a function generator (BK Precision 4011A) and amplified by a Crown
CE1000 type power amplifier. The signal generator is capable of providing 65.0 V sine waves up to a frequency of 5.0
MHz. A wattmeter (Powertek, ISW 8000) is connected
between the amplifier and the loudspeaker to measure the
root mean square (RMS) values of the delivered input power,
the applied voltage, the applied current and the phase angle
between the voltage and the current. An Endevco 8507C-1
series piezo-resistive pressure transducer is used to detect
and quantify the acoustic field. The transducer is installed in
the Delrin plug located at one end of the resonator. The
cross-sectional area of the microphone is approximately one
percent of the resonator area, therefore the error introduced
by the presence of the probe in the sound field is assumed

FIG. 3. (Color online) Schematic diagram of the experimental setup.

J. Acoust. Soc. Am. 138 (3), September 2015

Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233

AQ5

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 4

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

249

negligible. Lead wires from the loudspeaker and the pressure

transducer are introduced inside the supercritical chamber
through compression fittings. An Endevco 4428A signal conditioner is used to process the signal (pressure) from the
transducer and provide excited voltage.
Two type E thermocouples (Thermocouple 1 and 2,
Fig. 3) are used to measure the steady state initial temperature of the supercritical fluid. A high pressure transducer
(Omega PX309) is used to measure and monitor the steady
state pressure inside the supercritical chamber. The analog
temperature and pressure measurements are recorded, digitized and saved through a National Instrument SCB-68 terminal block and a 6052E data acquisition (DAQ) board.
High sample rate (333 kHz) of the 6052E DAQ board guarantees that the signals are recorded with high fidelity. Details
of the experimental setup are summarized in Table I.

250

V. MATHEMATICAL MODEL

251

254

The numerical model of the flow field developed in the

acoustic resonator is described below. Specific details about
the governing equations solved, the initial and boundary conditions and the numerical scheme used are also provided.

255

A. Governing equations

256

The transport processes in supercritical and near-critical

fluids can be modeled by the hydrodynamic description for
an isotropic, Newtonian, compressible, and dissipative (viscous and heat-conducting) fluid.14,2628 The governing equations corresponding to mass, momentum, and energy
balances are as follows:

234
235
236
237
238
239
240
241
242
243
244
245
246
247
248

252
253

257
258
259
260
261

@q 1 @ qrur @ quz

0;

@z
@t r @r


@ qur 1 @ qru2r
@ qur uz

@t
@z
r @r
@p 1 @ rsrr @srz shh


;

@z
r
@r r @r
 
@ quz 1 @ qrur uz @ qu2z

@z
@t
@r
r
@p 1 @ rsrz @szz


;
@z
@z r @r


@ qh0 1 @
@T
qrur h0  kf

r @r
@r
@t


@
@T
@p
quz h0  kf
U:

@z
@z
@t

(1)

(2A)

(2B)

(3)

The viscous dissipation term, U is given as

U

1@
@
rur srr uz srz  ur srz uz szz :
r @r
@z

The components of stress tensor are



1 @ rur @ uz
@ur
;

srr l0
2l
@r
@z
r @r


@ uz
@uz
0 1 @ rur

2l
;
szz l
@z
@z
r @r


 
@ uz
ur
0 1 @ rur
shh l
;

2l
r
@z
r @r


@ur @uz

srz l
:
@z
@r

262

(4)
263

(5A)
(5B)
(5C)

(5D)

Here, p is pressure and T is temperature. h0 is the total enthalpy of the fluid given by h0 h 12 u2r u2z . l is the
dynamic viscosity and l0 is the second coefficient of viscosity [l0  23 l; considering zero bulk viscosity]. ur and uz
are the velocities along the radial (r) and axial (z) coordinates, respectively, p is pressure, and T is temperature of the
fluid. The equation of state describing the q-p-T relation of
supercritical fluids (including the near-critical states) is not
well-represented by the van der Waals equation of state.29 In
this study, we use the NIST23 Standard Reference
Database 30 for the q f (p, T) relations and for evaluation
of other thermodynamic properties of supercritical and nearcritical carbon dioxide. The NIST23 (Ref. 30) equation of
state describing the q-p-T relation of carbon dioxide is based
on the equation of state proposed by Span and Wagner,31
which is mainly empirical in nature and includes special
non-analytic terms to predict the correct behavior of the fluid
to the immediate vicinity of the critical point. Two dimensional look-up tables are developed in the present study to
represent the density, q f (p, T), thermal conductivity, kf f
(p, T), dynamic viscosity, l f (p, T), specific enthalpy, h f
(p, T) and specific heat, cp f (p, T) data provided by the
NIST Standard Reference Database 23.

264

B. Initial and boundary conditions

287

Initially thermally quiescent and motion-free sub-/supercritical carbon dioxide inside the domain is considered for
the calculations. The initial state (temperature and pressure)
of the fluid is described as, Ti and pi. The wall boundaries at
the rigid end are maintained constant at the initial

288

TABLE I. Specifications of the experimental system.

No.
1
2
3
4
5
6

Part

Make and Model

Specification

Hand pump
Glands
Thermocouples
Pressure transducers
Acoustic driver
Data acquisition board

HiP 87-6-5
Conax TG-24 T(E)-A2-T
Omega EMTSS-125 (Probe Type)
Endevco 8507C-1 Omega PX309-2 KG5V
CUI CMS0401KL-3X
NI 6052E

34.5 MPa 60 mL/stroke

Pressure rating: 22 MPa
Type E3.175 mm diam.
Sensitivity: 200 mV/psi Pressure rating: 1 psi pressure rating: 13.8 MPa
Rated power: 10 W Impedance: 8 X
Sampling rate: 333 kHz

J. Acoust. Soc. Am. 138 (3), September 2015

Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286

289
290
291
292

AQ6

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 5

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

293
294
295
296
297
298

temperature of the fluid, i.e., isothermal, Tr; z L Ti .

The isothermal boundary condition is the closest approximation to an experiment and has been shown in past studies to
have the best agreement with experimental results.32,33 The
vibrating wall and the wall along the length of the resonator
are maintained adiabatic, i.e.,
@T 
@T 
 0
 :
@z z0
@r rR

299
300
301
302

For generating the acoustic waves, the vibrating wall is harmonically oscillated to model the motion of the speaker. The
moving wall (at z 0) is vibrated according to the following
equation:
zt A0 sin2pft:

303
304
305
306
307
308
309
310
311
312

(6A)

Here, A0 is the amplitude of vibration and f is the frequency

of oscillation. The frequency of oscillation of the vibrating
wall is chosen such that the lowest acoustic mode propagates
through the fluid medium (i.e., f csi / 2 L, where csi is the
acoustic speed at the initial state). For the numerical simulations, the amplitude of vibration of the moving wall is calculated from the power consumption of the electro-mechanical
speaker using the model developed by Beranek.34 Axial velocity of the fluid in the immediate vicinity of the moving
wall is given by

minimum grid size in axial direction, Dzmin 0.25 mm).

Supercritical carbon dioxide (pi 7.6 MPa; Ti 308 K) in
the cylindrical resonator with a resonant driving frequency
(f) of 391.0 Hz and vibration amplitude (A0) of 4.0 lm is
chosen as the test case. The results from the grid dependency
study are shown in Fig. 4. The maximum amplitude of the
generated acoustic wave (pmax) and the normalized axial
streaming velocity (ust/uR) are selected as the characteristic
variables indicating grid independency of the model. It is
observed that for the 250  100 grid, the characteristic variables (pmax and ust/uR) are independent of the grid size. The
time-step (Dt) for the simulations is dependent on the acoustic speed in the fluid. For the thermodynamic states of carbon
dioxide considered in the present study, the simulation timestep is around 0.51.0 ls and obeys the CFL condition with
a Courant number (C csi  Dt / Dx, where csi is the acoustic
speed for the given conditions) around 0.4.

340

VI. RESULTS AND DISCUSSIONS

357

The numerical model is validated first with results from

past studies. The validation of the numerical model as well
as the numerical and experimental results obtained from the
present study is discussed in Sec. VI A.

358

A. Model validation

362
363

317

Radial velocity of the fluid at the same location is considered

to be zero. No-slip boundary conditions are considered at the
adiabatic side wall (at r R) and at the end wall (at z L).
Symmetry boundary condition for both velocity and temperature is imposed along the resonator axis.

318

C. Numerical scheme

The numerical prediction of mechanically driven acoustic waves with the present model is compared with a previous numerical study by Aktas and Farouk.1 Aktas and
Farouk numerically investigated mechanically driven acoustic waves in atmospheric pressure nitrogen at 300 K in a two
dimensional rectangular enclosure of length, L 8.825 mm
and varying width. The frequency and amplitude of the
vibrating wall is always kept at 20 kHz and 10 lm, respectively. The length of the enclosure is chosen such that
L k/2 (k is the wavelength calculated from cs fk) and the
lowest acoustic mode propagates through the fluid medium.

319

The numerical scheme for solving the governing equations is based on the finite volume approach. The continuity,
momentum and energy equations are solved for the fluid
using the central difference scheme. The motion of the
vibrating wall is captured by a moving grid scheme near the
piston wall. The re-meshing scheme used in the simulations
is the Transfinite Interpolation scheme.35 A second order
CrankNicholson scheme (with a blending factor of 0.7) is
used for the time derivatives in the continuity, momentum
and energy equations. A convergence criterion (i.e., percentage relative error) of 104 is used for all the variables in the
iterative implicit numerical solver.
As the cylindrical resonator geometry is axisymmetric,
2D cylindrical coordinate system (r varying between 0 and
r0) is used for simulation. The problem geometry is studied
with non-uniform structured grid and fine grid is used in
vicinity of the wall boundaries to provide adequate spatial
resolution and capture the oscillating boundary layer (dv). A
grid dependency study is performed considering four different grid sizes ranging from coarse (100  30; minimum grid
size in axial direction, Dzmin 1.0 mm) to fine (400  100;

FIG. 4. (Color online) Maximum amplitude of the generated acoustic wave

(pmax) and the normalized axial streaming velocity (ust/uR) as a function of
the grid size.

@zt
2pfA0 cos 2pft:
uz t
@t
313
314
315
316

320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339

J. Acoust. Soc. Am. 138 (3), September 2015

(6B)

Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356

359
360
361

364
365
366
367
368
369
370
371
372
373

AQ7

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 6

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

374
375
376

377
378

For the case considered for validation of our present model,

the width of the rectangular enclosure is 20dv. Here, dv is the
viscous penetration depth given by
s
2l
; x 2pf :
(7)
d
qx
The streaming velocities in the enclosure are calculated
using the following formula:
ust

379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423

hqui
;
hqi

vst

hqvi
:
hqi

(8)

Here, ust and vst describes the x and y components of the

streaming velocities, respectively. The hi sign indicates the
time-averaged quantities. The time averaging is performed
during the 100th vibration cycle (between cycle #99 and
#100). The streaming velocities calculated based on the
time averaging during 80th and 100th cycles do not differ
significantly. Hence, the average mass transport velocities
are assumed to be cycle independent by this time (cycle
#100).
Spatial variation of the x component of streaming velocity at x 3 L/4 for the validation case is shown in Fig. 5(a).
In this figure, the vertical axis is the x component of the
dimensionless streaming velocity (ust/uR) and uR is given by
uR 3u02/16cs. Here, u0 is the maximum oscillatory
velocity. This reference velocity value represents the maximum streaming velocity in case of a perfect sinusoidal wave
form obtained by Rayleigh. Figure 5(b) shows the variation
of the y component of streaming velocity along the enclosure
semi-height at x L/2. The vertical axis represents the y
component of the non-dimensional streaming velocity (vstx0
/uRy0). Here x0 is the length of the enclosure and y0 is the
semi-height of the enclosure. It is observed that the dimensionless streaming velocities calculated from the present
model are in good agreement with that predicted by Aktas
and Farouk.1 Figure 5(a) indicates the existence of two
different vortical structures at x 3 L/4; one formed at the
vicinity of the side wall (inner streaming) while the other is
formed in the bulk fluid (outer streaming). The height of the
circulatory flow structures (inner streaming) observed in the
vicinity of the horizontal walls is characterized by the thickness of the acoustic boundary layer. The streaming structures
seen in the middle section of the enclosure (outer streaming)
have larger sizes. These predicted streaming structures are
also in good agreement with the results reported by earlier
studies.1,9 On an absolute scale, Aktas and Farouk1 reported
a maximum streaming velocity of approximately 0.06 m/s
with a maximum instantaneous velocity of 12 m/s in the primary oscillatory flow field in the enclosure, while the present
model slightly under predicts these velocities. The maximum
streaming velocity and the instantaneous velocity calculated
from the present model are 0.057 and 11.2 m/s, respectively.
This slight deviation is mainly due to the implicit nature of
the present numerical model (the model used by Aktas
and Farouk1 is explicit in nature and have less numerical
diffusion).
6

J. Acoust. Soc. Am. 138 (3), September 2015

FIG. 5. (Color online) Variation of the (a) x component of the streaming velocity at x 3 L/4 and (b) y component of the streaming velocity at x L/2
compared with Aktas and Farouk (Ref. 1).

B. Acoustic streaming in sub- and supercritical fluids

424

As described in Sec. I, acoustic streaming phenomena in

ideal gases are well studied. However, the same in nearcritical fluids is expected to be significantly different due to
high density and compressibility and lower viscosity of the
fluid. To thoroughly investigate the formation of streaming
structures in near-critical carbon dioxide and the transition
from the ideal gas acoustic streaming phenomena, four cases
are studied (cases 14) first, see Table II. These four cases
correspond to carbon dioxide at states ranging from ideal
gas-like to near-critical; excited at their resonant frequencies
and at various amplitudes of vibration. At this stage the initial temperature of the fluid is kept constant (Ti 308 K) and
the undisturbed pressure is varied.
The fluid in case 4 is in the supercritical state and this
case is considered as the base case for the rest of the paper.

425

Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

426
427
428
429
430
431
432
433
434
435
436
437
438
439

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 7

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

TABLE II. List of cases simulated for investigation of acoustic streaming
formation in sub- and supercritical fluids.

440
441
442
443
444
445
446
447
448
449
450
451

Case no. Ti (K) pi (MPa)

q / qc

f (Hz) xmax (lm) dv (lm)

1
2
3
4

0.0037
0.0812
0.1854
0.6224

545.3
517.6
484.2
391.0

308
308
308
308

453

346.7
17.5
8.8
4.0

72.2
15.9
11.1
8.0

0.760
0.853
1.02
3.53

In our experiments, acoustic waves are generated inside the

cylindrical shaped resonator filled with supercritical carbon
dioxide at this state (7.6 MPa and 308 K) with a speaker
power of 10.0 W. The vibrational amplitude (A0) of the moving wall corresponding to the acoustic impedance (qcs) of
the fluid and the input power at case 4 is 4.0 lm (Beranek34).
For the acoustic wave measurements in cases 13, the input
power is kept constant (10.0 W) and the vibrational amplitude is calculated using the same model.34 For the same
electro-mechanical variables of the speaker and input power,
these amplitudes can also be approximated using the amplitude for case 4 and the following equation:
A0;n

452

0.1
2.0
4.0
7.6

Pr

  2
q
fi
A0;i i
:
qn
fn

(9)

Equation (9) is based on a simple force balance on the

vibrating wall where the subscript i refers to the values

corresponding to the base case (case 4) and the subscript n

refers to the values corresponding to the nth case.
Transient variation of the pressure (gage) at the midpoint of the end wall (pressure anti-node) of the cylindrical
resonator at a pseudo-steady state for cases 14 are shown in
Fig. 6. The time-scale is normalized with the acoustic time
period (s 1=f ). Both the computed and measured pressure
transients are shown in this figure. A monotonic increase of
the maximum pressure amplitude is observed with the
increase in operating pressure. The maximum pressure
amplitude reaches approximately 5.0 kPa at supercritical
state (case 4), while it is around 0.6 kPa at atmospheric pressure (case 1). This monotonic increase is mainly due to the
increase in density of the fluid with pressure. Also at atmospheric pressure (case 1), the pressure wave form is much
sharper and shock wave-type profile is observed [Fig. 6(a)],
while at higher sub-critical pressures (cases 2 and 3), the
pressure profile is near-sinusoidal [Figs. 6(b) and 6(c)]. At
the supercritical state (case 4), the pressure profile again
becomes sharper and a shock-wave type [Fig. 6(d)]. This
behavior can be explained by the isothermal compressibility,
b 1=q@q=@pT of the fluid. At atmospheric pressure the
isothermal compressibility is high. However the compressibility quickly decreases with the increase of operating
pressure until the supercritical state is reached, where the
compressibility starts to increase again.
For all the cases discussed (cases 14), the computed
and the measured transient pressure distribution compares

FIG. 6. (Color online) Transient variation of computed and measured pressure (gage) at the end wall of the
cylindrical resonator at a pseudosteady state for (a) p 0.1 MPa, (b)
p 2.0 MPa, (c) p 4.0 MPa, and (d)
p 7.6 MPa and for four acoustic
cycles.

J. Acoust. Soc. Am. 138 (3), September 2015

Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 8

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502

well with the predictions. For the cases showing shock

wave-type profiles (cases 1 and 4), existence of a slight
variation between the measured and computed pressure
amplitude is observed at the troughs and the trenches of the
wave form. This is mainly due to the high experimental
noise in the system when the temporal pressure gradient is
high (i.e., at troughs and trenches of the wave form).
For all these cases (cases 14), the streaming flow
patterns are observed in the shape of regular structures.
Four outer steaming structures are observed in all these
cases. Variation of axial component of the non-dimensional
streaming velocity at z 3 L/4 for cases 14 is shown in
Fig. 7(a). Similar to the maximum pressure amplitude, a
monotonic increase in the streaming velocities with increasing pressure is observed in this figure. The difference in
the streaming structures are also noticed from this figure.
At atmospheric pressure (case 1), existence of the anticlockwise rotating inner streaming structure gives rise to the
negative velocity near the side wall (r/r0 1). However,
with the increase of operating pressure, the inner streaming
structures are dissolved while the outer streaming structures

expand and cover the entire domain. At a supercritical state

(case 4), the streaming structures are characterized by relatively high axial streaming velocity near the wall and a flat
plug-flow like profile in the rest of the domain as compared
to a parabolic velocity profile at the atmospheric pressure
(case 1). Variation of the radial component of the nondimensional streaming velocity at z L/2 for cases 14 is
shown in Fig. 7(b). A strong radial velocity from the bulk
fluid to the wall in the supercritical state (case 4) is observed
from this figure. This strong velocity corresponds to a jet
like flow at z L/2. This behavior is highly desired for
mixing applications, where the jet can efficiently carry
the bulk fluid to the wall boundary layer.

503

C. Effect of thermo-physical property variations

across the pseudo-critical state (q=qc 51)

516

The effect of the thermo-physical property variations

across the pseudo-critical state on the mechanically driven
acoustic waves and acoustic streaming formation is investigated in this section. Three additional cases (cases 57)
along with case 4 are considered for this study. For these
four cases, the operating pressure is supercritical and is kept
constant at p 7.6 MPa. The pseudo-critical state (i.e.,
where q=qc 1) corresponding to this pressure is at
305.43 K. Hence, the pseudo-critical state is approached by
reducing the operating temperature. The four cases correspond to fluids ranging from gas-like (i.e., q=qc < 1:0) to
liquid-like (i.e., q=qc > 1:0) properties. Table III below lists
the details of the cases studied. Here, the fluid corresponding
to cases 4 and 5 are gas-like with relatively low density, thermal conductivity, and acoustic speed, while that corresponding to case 7 is liquid-like with higher density, thermal
conductivity and acoustic speed. The fluid corresponding to
case 6 is very close to the pseudo-critical state.
Temporal variation of the computed and measured
pressures (gage) at the end wall (pressure anti-node) of the
cylindrical resonator for cases 47 are shown in Fig. 8. Two
specific phenomena are observedfirst, the pseudo-critical
state has a strong effect on the maximum pressure amplitude.
It is observed that, as the pseudo-critical state (i.e.,
T 305.43 K at p 7.6 MPa) is approached, the maximum
pressure amplitude increases. For the cases studied (cases
47), the maximum pressure amplitude is observed for case
6 (T 305.5 K) closest to the pseudo-critical state. This is
due to the high compressibility and density of the fluid as it
approaches the pseudo-critical state. The maximum pressure
amplitude decreases as the thermodynamic state of the fluid
moves away from the pseudo-critical state (case 7,
T 305.1 K). At this state, density of the fluid is very high,

518

FIG. 7. Spatial variation of the (a) axial component of the streaming velocity at z 3 L/4 and (b) radial component of the streaming velocity at z L/2
for different operating pressures (cases 14).
8

J. Acoust. Soc. Am. 138 (3), September 2015

q / qc

f (Hz) xmax (lm) dv (lm)

4
5
6
7

0.6224
0.8
0.937
1.2

391.0
352.6
325.6
373.1

308
305.8
305.5
305.1

7.6
7.6
7.6
7.6

4.0
3.81
3.83
2.26

8.0
8.1
8.3
7.9

Pr
3.53
11.26
26.60
13.66

Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

505
506
507
508
509
510
511
512
513
514
515

517

TABLE III. List of cases simulated for investigation of the effect of pseudocritical states on acoustic streaming.
Case no. Ti (K) pi (MPa)

504

519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 9

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

FIG. 8. (Color online) Transient variation of computed and measured pressure

(gage) at the end wall of the cylindrical resonator at a pseudo-steady state
for different operating temperatures at p 7.6 MPa (cases 47). Measured
pressures are shown with symbols: T 308.0 K; T 305.8 K;
T 305.5 K, and T 305.1 K.

551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570

but the compressibility is low. The second phenomenon

observed from the transient pressure is related to the pressure
wave form. The pressure wave forms corresponding to all
these cases (cases 47) show a shock wave-type profile with
a sharp rise and long tail in general. However, the pressure
wave form corresponding to the case closest to the pseudocritical state (case 6, T 305.5 K) exhibits a slightly different profile with a non-linear distortion and a shift in mean
value. This distortion of the pressure transient is mainly due
to the strong variation of acoustic speed of the near-pseudocritical fluid at this case (case 6). At near-pseudo-critical
states close to the critical point, the acoustic speed of the
fluid exhibits relatively large variations with changing pressure and temperature. Hence, in an oscillating pressure (and
hence temperature) field such as this, the acoustic speed of
the fluid also oscillates to a great extent which in turn affects
resonance and hence the standing wave field. The measured
pressure (see Fig. 8symbols) also shows similar distortion
of the wave form discussed above. High frequency disturbances are observed from the measurements in the

supercritical state (cases 47). Similar disturbances were

also reported in earlier experimental studies in supercritical
fluids.25,36 These are caused mainly due to the high compressibility of the medium and non-uniform heating of the supercritical chamber.
The quasi-steady (time-averaged) streamlines in the resonator corresponding to cases 47 are shown in Fig. 9.
Acoustic streaming structures are observed for all these
cases (cases 47). For the cases relatively far from the
pseudo-critical state (i.e., cases 4, 5, and 7), four outer
steaming structures are observed. This behavior is in accordance with that observed in Sec. VI B. However, at the nearpseudo-critical state (case 6), the streaming structures do not
exhibit the regular pattern. It is observed that the existence
of the distorted pressure field gives rise to eight outer
streaming cells in the system instead of four. Out of these
eight, four of the cells are large and prominent in shape.
While, the other four are relatively small in size and are
derived from the variation of the thermo-physical properties
(e.g., acoustic speed of the medium) due to the wave field
distortion.
The variation of the axial component of the nondimensional streaming velocity at z 3 L/4 for cases 47 is
shown in Fig. 10(a). For the cases relatively far from the
pseudo-critical state (i.e., cases 4, 5, and 7), the streaming
velocities also increase monotonically as the corresponding
pseudo-critical state is approachedsimilar to the maximum
pressure amplitude. The streaming velocity distribution at
these states shows a flat plug-like profile in the bulk fluid
with a sharp gradient near the wall. This sharp gradient indicates a very thin acoustic boundary layer (dv) in the fluid.
This is also confirmed by the analytical calculation of the
acoustic boundary layer (8.0 lm) as presented in Table III.
The plug-like profile (similar to a turbulent flow profile) is
due to the small dynamic viscosity of the supercritical medium. However, the near-pseudo-critical state (case 6,
T 305.5 K) exhibits none of these trends. The streaming
velocity gradient is relatively small near the wall and the
streaming velocity distribution in the bulk fluid exhibits a
parabolic profile (similar to the profile at atmospheric pressure, case 1). Although analytical calculation of dv provides
a similarly thin boundary layer (8.3 lm), the oscillating

FIG. 9. Cycle averaged streamlines

(acoustic streaming) in the cylindrical
resonator at a pseudo-steady state for
(a) T 308.0 K (case 4), (b)
T 305.8 K (case 5), (c) T 305.5 K
(case 6), and (d) T 305.1 K (case 7)
for an operating pressure of 7.6 MPa.

J. Acoust. Soc. Am. 138 (3), September 2015

Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612

AQ8

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 10

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

FIG. 10. (Color online) Spatial variation of the (a) axial component of the
streaming velocity at z 3 L/4 and (b) radial component of the streaming velocity at z L/2 for different operating temperatures at p 7.6 MPa (cases
47).

pressures (8.0 and 8.5 MPa) are considered for this study. To
understand the effect of operating pressure on the nearpseudo-critical acoustic streaming formation, three cases
(q=qc < 1:0, q=qc  1:0; and q=qc > 1:0) are considered
for each of these pressures. Table IV below lists the details
of the cases studied.
Temporal variation of the computed and measured pressures (gage) at the end wall (pressure anti-node) of the cylindrical resonator for cases 810 (p 8.0 MPa) are shown in
Fig. 11(a). The same features of the pressure wave form
(monotonic increase as pseudo-critical state is approached,
sharp and shock wave-type profile, etc.) are also observed in
this pressure. However, for the near-pseudo-critical state
(case 9, T 308.0 K), the non-linear variation in the pressure
wave form is not as prominent as observed for p 7.6 MPa
(case 6). The experimental measurements for this case also
confirm this observation. Temporal variation of the computed and measured pressures (gage) at the end wall
(pressure anti-node) of the cylindrical resonator for cases
1113 (p 8.5 MPa) are shown in Fig. 11(b). It is observed
that, the near-pseudo-critical effect on the pressure wave
form (i.e., the non-linear variation) is further diminished at
this pressure.
This behavior of the pressure (acoustic) wave can be
explained by the near-critical thermo-physical property variations. As the operating pressure is increased, the thermodynamic state of the fluid is moved far away from the critical
point and the thermo-physical property variations (including
the acoustic speed) at these higher pressures (e.g., p 8.0
and 8.5 MPa) are not very strong. Hence, the acoustic speed
variation of the fluid (as discussed in Sec. II) in the resonator
at the near-pseudo-critical state becomes negligible and as a
consequence of that, the pressure wave form becomes similar to that observed at the cases far from the pseudo-critical
states.
The quasi-steady (time-averaged) streamlines in the
resonator corresponding to the near-pseudo-critical cases at
two different operating pressures (cases 9 and 12) are shown
in Fig. 12. It is observed that for the cases far from the corresponding pseudo-critical states (i.e., cases 8, 10, 11, and 13)
the operating pressure has a negligible effect on the streaming structures. The regular outer streaming structures are
observed for these cases (not shown in Fig. 12). However,
for the near-pseudo-critical states (cases 9 and 12), the additional streaming cells (corresponding to the distortion of
the standing wave field) are not significant at this elevated
pressures. At p 8.0 MPa, the additional streaming cell can

623

thermo-physical properties of the near-pseudo-critical fluid

hinders resonance in a standing wave field and the formation
of a pseudo-steady boundary layer.
Variation of the radial component of the nondimensional streaming velocity at z L/2 for cases 47 is
shown in Fig. 10(b). Similar to case 4 (Sec. VI B), a strong
radial velocity from the bulk fluid to the wall in the supercritical states (cases 47) is observed. The jet like flow due
to the radial velocity strengthens as the pseudo-critical state
is approached and it becomes maximum for the near-pseudocritical state (case 6).

624

D. Effect of operating pressure

Case no.

Ti (K)

pi (MPa)

q / qc

f (Hz)

xmax (lm)

dv (lm)

Pr

625

As observed in Figs. 1 and 2, the significant variation in

thermo-physical properties near the pseudo-critical states
become milder as the pressure is increased (from the critical
pressure). The effect of this transition in thermo-physical
properties on acoustic streaming phenomena in supercritical
fluids is investigated in this section. Two different operating

8
9
10
11
12
13

307
308
308.75
309.3
311.1
312.5

8
8
8
8.5
8.5
8.5

1.2
0.933
0.8
1.2
0.936
0.8

401.1
360.9
373.0
422.8
384.1
392.5

1.97
3.13
3.42
1.78
2.76
3.10

7.6
7.9
7.8
7.4
7.7
7.7

9.58
11.99
6.95
7.13
7.57
4.95

613
614
615
616
617
618
619
620
621
622

626
627
628
629
630

10

J. Acoust. Soc. Am. 138 (3), September 2015

TABLE IV. List of cases simulated for investigation of the effect of operating pressure on acoustic streaming.

Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677

AQ9

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 11

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

FIG. 12. Cycle averaged flow-field (acoustic streaming) in the cylindrical

resonator at a pseudo-steady state for (a) p 8.0 MPa, T 308.0 K (case 9)
and (b) p 8.5 MPa, T 311.1 K (case 12).

FIG. 11. (Color online) Transient variation of computed and measured pressure (gage) at the end wall of the cylindrical resonator at a pseudo-steady
state for different operating temperatures at (a) p 8.0 MPa (cases 810).
[Measured pressures are shown with symbols: T 308.75 K;
T 308.0 K, and T 307.0 K] and (b) p 8.5 MPa (cases 1113).
[Measured pressures are shown with symbols: T 312. 5 K;
T 311.1 K, and T 309.3 K].

678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694

be noticed and is almost merged with the regular outer

streaming structure [Fig. 12(a)]. At p 8.5 MPa, the additional structure is not noticed at all.
The variation of the axial component of the nondimensional streaming velocity at z 3 L/4 corresponding to
the cases at two different operating pressures (cases 813) is
shown in Fig. 13. It is observed that the operating pressure
has a deteriorating effect on the streaming velocity. The
maximum amplitude of the non-dimensional streaming
velocity decreases with increasing pressure. Although the
additional streaming structures are not observed in the nearpseudo-critical states (cases 9 and 12), the distribution of the
axial streaming velocity remains parabolic. A slight difference between the axial streaming velocities in the bulk fluid
(along the axis of the resonator) for the cases in the gas-like
(q=qc < 1:0) and liquid-like (q=qc < 1:0) property regimes
are observed at near-critical pressure [Fig. 10(a)]. With the
J. Acoust. Soc. Am. 138 (3), September 2015

increase in operating pressure, this difference is diminished

(Fig. 13).
The observations from Figs. 10(a) and 13 are summarized and shown in Fig. 14. Figure 14 shows the variation of
the axial streaming velocity (non-dimensional) along the
axis of the resonator with reduced density for the three different operating pressures studied (p 7.6, 8.0, and
8.5 MPa). The reduced density axis represents the proximity
of the thermodynamic state to the pseudo-critical state
(q=qc 1:0). The dotted lines represent the best fitted

FIG. 13. (Color online) Spatial variation of the axial component of the
streaming velocity at z 3 L/4 for different operating temperatures at
p 8.0 MPa (cases 810) and p 8.5 MPa (cases 1113).
Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

11

695
696
697
698
699
700
701
702
703
704

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 12

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

FIG. 14. (Color online) Variation of non-dimensional axial streaming velocity in the bulk fluid (at z 3 L/4 and r 0) with reduced density q=qc at
different isobars.
705
706
707
708
709
710
711
712
713
714
715

AQ10

curves for the data points. Parameters for the curves are provided in Table V. It is observed that, the streaming velocity
in the bulk fluid (i.e., along the axis of the resonator) is maximized at the pseudo-critical point. At pressures close to the
critical pressure (e.g., p 7.6 MPa), the axial streaming
velocity is slightly higher in the gas-like property regime
(q=qc < 1:0) than that in the liquid-like property regime
(q=qc > 1:0). This variation is mainly due to the reduced
viscosity and higher compressibility of the gas-like fluid.
With the increase in pressure, the thermo-physical property
variations become weak and this trend is barely noticed.

716

E. Thermal effects due to acoustic streaming

717

The oscillatory pressure and the streaming flow-field (as

discussed in the previous sections) has a significant effect on
the temperature field in the resonator. At sub-critical pressures (cases 13), a cold zone is observed in the temperature
contours roughly at the location of the pressure node (z/L
0.5), while two hot zones are observed at the pressure
anti-nodes (z/L 0, 1). However, for the supercritical state,
the cold zone almost disappears. The phenomena observed
in the sub-critical regime has been reported in earlier studies37 and is utilized in the development of thermoacoustic
refrigerators.38 The disappearance of the cold zone in the
supercritical state is due to the high thermal conductivity (kf)
and Prandtl number (Pr) coupled with the high isothermal
compressibility of the fluid at this state.

718
719
720
721
722
723
724
725
726
727
728
729
730

The thermal effect due to acoustic streaming in nearcritical fluids is summarized in Fig. 15. The normalized maximum and minimum temperatures along the cylindrical axis
in a standing wave resonator as a function of the reduced
density of the near-critical carbon dioxide are shown in these
figures. It is observed that the cases with gas-like properties
(i.e., q=qc < 1:0) exhibits a strong temperature difference in
the resonatorwith hot zones near the pressure anti-nodes
and a cold zone near the pressure node. This temperature
difference is rather weak for the near-pseudo-critical state
(case 6) due to very high specific heat of the fluid. The
temperature difference is slightly larger in the case with
liquid-like properties (i.e., q=qc > 1:0). This is due to the
relatively high thermal conductivity, but lower specific heat
as compared to the near-critical case. At elevated pressures,
the temperature field for the cases simulated (cases 813)
also show similar trend. But this trend is relaxed as the operating pressure is increased. At the highest operating pressure
(p 8.5 MPa), the temperature difference is almost the same
for the three cases studied.

731

VII. SUMMARY AND CONCLUSIONS

751

Mechanically driven resonant acoustic waves in nearcritical supercritical carbon dioxide are investigated in this
chapter. The formation of acoustic (pressure) waves, acousticviscous boundary layer interactions, and associated flows in a
cylindrical resonator are numerically studied by solving the
unsteady, compressible NavierStokes equations in an axisymmetric x-r coordinate system. The acoustic field in the enclosure is created due to the harmonic vibration of the end
wall. The effects of the pseudo-critical state (q=qc  1:0) and
operating pressure on the acoustic field and the formed flow
structures are determined by utilizing a highly accurate numerical scheme. Acoustic waves generated by an electromechanical driver in a cylindrical resonator filled with supercritical carbon dioxide are measured using a fast-response
pressure field microphone. The experimental measurements

752

TABLE V. Parameters for best fitted curve of the data presented in Fig. 14.
just =uR j c0 c1 q=qc c2 q=qc 2 c3 q=qc 3

p 7.6 MPa
p 8.0 MPa
p 8.5 MPa

12

c0

c1

c2

c3

59.03
29. 80
23.48

225.80
64.02
50.17

281.10
32.19
25.14

111.10
-

J. Acoust. Soc. Am. 138 (3), September 2015

FIG. 15. (Color online) Variation of non-dimensional maximum (solid lines)

and minimum (dashed lines) temperatures in the bulk fluid (along the resonator axis) with reduced density q=qc at different isobars.
Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750

753
754
755
756
757
758
759
760
761
762
763
764
765
766

AQ11

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 13

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829

are compared with results from the numerical simulations

with accurately developed boundary conditions.
Several interesting phenomena are observed from the
numerical simulations and confirmed by the experimental
measurements. The observed primary oscillatory and
secondary steady flow fields in the supercritical fluid medium demonstrate significant effects of the thermodynamic
state (sub-critical/supercritical/pseudo-critical). The strength
of the acoustic wave field is found to be in direct relation
with the thermodynamic state (near-pseudo-critical/far from
the pseudo-critical state). It is observed that due to the strong
thermo-physical property variations, maximum pressure amplitude in the standing wave field increases as the thermodynamic state of the fluid approaches the corresponding
pseudo-critical state. In the near-pseudo-critical state, the
pressure wave form exhibits a non-linear and distorted
profile. In a standing wave field with near-critical fluid, the
acoustic speed of the fluid also oscillates to a great extent
which in turn affects resonance and hence the formation of
acoustic streaming structures. Far from the pseudo-critical
states, the streaming structure exhibits four counter rotating
cells with a jet like flow-field along the semi-length of the
resonator (at the pressure node)a typical Rayleigh
streaming like behavior. While near the pseudo-critical state,
irregular streaming structures consisting of eight outer
streaming cells in the resonator are observed. The evolved
flow structures are also dependent on the operating pressure.
The irregular streaming patterns are observed mainly for
near-pseudo-critical states at operating pressures close to the
critical pressure (pc 7.377 MPa). However, these structures
quickly re-orients to the regular streaming patterns (four
outer streaming cells) with the increase of operating
pressureas the singularity effect of the critical point diminishes. The acoustic streaming phenomena observed numerically and the near-critical standing wave field measurements
in this study can be utilized for mixing applications in supercritical fluid medium, especially for enhancing the transport
characteristics in supercritical fluid extraction processes.
1

M. K. Aktas and B. Farouk, Numerical simulation of acoustic streaming

generated by finite-amplitude resonant oscillations in an enclosure,
J. Acoust. Soc. Am. 116, 28222831 (2004).
2
J. Cai, X. Huai, S. Liang, and X. Li, Augmentation of natural convective
heat transfer by acoustic cavitation, Front. Energy Power Eng. China 4,
313318 (2010).
3
A. Gopinath and A. F. Mills, Convective heat transfer from a sphere due
to acoustic streaming, J. Acoust. Soc. Am. 94, 17731774 (1993).
4
M. F. Hamilton, Y. A. Ilinskii, and E. A. Zabolotskaya, Thermal effects
on acoustic streaming in standing waves, J. Acoust. Soc. Am. 114,
30923101 (2003).
5
B.-G. Loh, S. Hyun, P. I. Ro, and C. Kleinstreuer, Acoustic streaming
induced by ultrasonic flexural vibrations and associated enhancement of
convective heat transfer, J. Acoust. Soc. Am. 111, 875883 (2002).
6
B. Tajik, A. Abbassi, M. Saffar-Avval, A. Abdullah, and H. MohammadAbadi, Heat transfer enhancement by acoustic streaming in a closed cylindrical enclosure filled with water, Int. J. Heat Mass Transfer 60,
230235 (2013).
7
Y. Rafat, S. Amiri, R. Taher, and L. Mongeau, Experimental investigation of acoustic streaming flows inside a standing wave tube, J. Acoust.
Soc. Am. 135, 23452345 (2014).
8
T. Yano, Numerical study of Rayleigh-type acoustic streaming based on
a three-dimensional incompressible flow model, J. Acoust. Soc. Am. 135,
23452345 (2014).
J. Acoust. Soc. Am. 138 (3), September 2015

M. F. Hamilton, Y. A. Ilinskii, and E. A. Zabolotskaya, Acoustic streaming generated by standing waves in two-dimensional channels of arbitrary
width, J. Acoust. Soc. Am. 113, 153160 (2003).
10
N. Kawahara, A. L. Yarin, G. Brenn, O. Kastner, and F. Durst, Effect of
acoustic streaming on the mass transfer from a sublimating sphere, Phys.
Fluids 12, 912923 (2000).
11
E. J. Beckman, Supercritical and near-critical CO2 in green chemical
synthesis and processing, J. Supercrit. Fluids 28, 121191 (2004).
12
X.-R. Zhang, L. Chen, and H. Yamaguchi, Natural convective flow and
heat transfer of supercritical CO2 in a rectangular circulation loop, Int. J.
Heat Mass Transfer 53, 41124122 (2010).
13
N. Hasan and B. Farouk, Thermoacoustic transport in supercritical fluids
at near-critical and near-pseudo-critical states, J. Supercrit. Fluids 68,
1324 (2012).
14
A. Guardo, M. Coussirat, F. Recasens, M. A. Larrayoz, and X. Escaler,
CFD studies on particle-to-fluid mass and heat transfer in packed beds:
Free convection effects in supercritical fluids, Chem. Eng. Sci. 62,
55035511 (2007).
15
H. Z. Nei, S. Fatemi, M. R. Mehrnia, and A. Salimi, Mathematical
modeling and study of mass transfer parameters in supercritical fluid
extraction of fatty acids from Trout powder, Biochem. Eng. J. 40, 7278
(2008).
16
H. Sovova, Rate of the vegetable oil extraction with supercritical CO2
I. Modelling of extraction curves, Chem. Eng. Sci. 49, 409414 (1994).
17
J. A. Carcel, J. V. Garca-Perez, J. Benedito, and A. Mulet, Food process
innovation through new technologies: Use of ultrasound, J. Food Eng.
110, 200207 (2012).
18
S. Balachandran, S. E. Kentish, R. Mawson, and M. Ashokkumar,
Ultrasonic enhancement of the supercritical extraction from ginger,
Ultrason. Sonochem. 13, 471479 (2006).
19
E. Riera, Y. Golas, A. Blanco, J. A. Gallego, M. Blasco, and A. Mulet,
Mass transfer enhancement in supercritical fluids extraction by means of
power ultrasound, Ultrason. Sonochem. 11, 241244 (2004).
20
E. Riera, A. Blanco, J. Garca, J. Benedito, A. Mulet, J. A. Gallego-Juarez,
and M. Blasco, High-power ultrasonic system for the enhancement of
mass transfer in supercritical CO2 extraction processes, Phys. Procedia 3,
141146 (2010).
21
N. Hasan and B. Farouk, Mass transfer enhancement in supercritical fluid
extraction by acoustic waves, J. Supercrit. Fluids 80, 6070 (2013).
22
W. L. Nyborg, Acoustic Streaming near a Boundary, J. Acoust. Soc.
Am. 30, 329339 (1958).
23
NIST, Reference fluid thermodynamic and transport properties database,
in NIST23, vol. Version 9.0, ed., NIST, Gaithersburg, MD, 2010.
24
R. H. Sabersky and E. G. Hauptmanni, Forced convective heat transfer to
carbon dioxide near the critical point, Int. J. Heat Mass Transfer 10,
14991508 (1967).
25
N. Hasan and B. Farouk, Fast heating induced thermoacoustic waves in
supercritical fluids: Experimental and numerical studies, J. Heat Transfer
135, 081701081701 (2013).
26
B. Zappoli, S. Amiroudine, P. Carles, and J. Ouazzani, Thermoacoustic
and buoyancy-driven transport in a square side-heated cavity filled with a
near-critical fluid, J. Fluid Mechan. 316, 5372 (1996).
27
G. Accary, P. Bontoux, and B. Zappoli, Convection in a supercritical
fluid: A reduced model for geophysical flows, Phys. Fluids 19,
014104014106 (2007).
28
A. Guardo, M. Casanovas, E. Ramrez, F. Recasens, I. Maga~
na, D.
Martnez, and M. A. Larrayoz, CFD modeling on external mass transfer
and intra-particle diffusional effects on the supercritical hydrogenation of
sunflower oil, Chem. Eng. Sci. 62, 50545061 (2007).
29
K. A. Gillis, I. I. Shinder, and M. R. Moldover, Thermoacoustic boundary
layers near the liquid-vapor critical point, Phys. Rev. E 70, 021201
(2004).
30
E. Lemmon, M. Huber, and M. McLinden, NIST standard reference database 23, vol. Version 9.0, ed., NIST, Gaithersburg, MD, 2010.
31
R. Span and W. Wagner, A new equation of state for carbon dioxide
covering the fluid region from the triple-point temperature to 1100 K at
pressures up to 800 MPa, J. Phys. Chem. Ref. Data 25, 15091596
(1996).
32
A. Alexeev, A. Goldshtein, and C. Gutfinger, Heat interaction in a resonance tube, Phys. Fluids 14, 18121815 (2002).
33
A. Alexeev and C. Gutfinger, Resonance gas oscillations in closed tubes:
Numerical study and experiments, Phys. Fluids 15, 33973408 (2003).
34
L. L. Beranek, Acoustics, 2nd ed. (Acoustical Society of America, New
York, 1986).
Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

13

830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903

J_ID: JASMAN DOI: 10.1121/1.4930951 Date: 11-September-15

Stage:

Page: 14

Total Pages: 15

PROOF COPY [14-14639R1] 042509JAS

AQ12

904
905
906
907
908
909

35

C. Dyken and M. S. Floater, Transfinite mean value interpolation,

Comput. Aided Geom. Des. 26, 117134 (2009).
36
Y. Miura, S. Yoshihara, M. Ohnishi, K. Honda, M. Matsumoto, J. Kawai,
M. Ishikawa, H. Kobayashi, and A. Onuki, High-speed observation of the
piston effect near the gasliquid critical point, Phys. Rev. E74, 010101R
(2006).

37

D. S. Antao and B. Farouk, High amplitude nonlinear acoustic wave

driven flow fields in cylindrical and conical resonators, J. Acoust. Soc.
Am. 134, 917932 (2013).
38
D. S. Antao and B. Farouk, Experimental and numerical investigations of
an orifice type cryogenic pulse tube refrigerator, Appl. Therm. Eng. 50,
112123 (2013).

916

14

J. Acoust. Soc. Am. 138 (3), September 2015

Nusair Hasan and Bakhtier Farouk

ID: aipepub3b2server Time: 19:56 I Path: D:/AIP/Support/XML_Signal_Tmp/AI-JAS#150576

910
911
912
913
914
915