A thermoacoustic oscillator powered by vaporized water and ethanol

Daisuke Noda and Yuki Ueda

Citation: Am. J. Phys. 81, 124 (2013); doi: 10.1119/1.4766940

View online: http://dx.doi.org/10.1119/1.4766940
View Table of Contents: http://ajp.aapt.org/resource/1/AJPIAS/v81/i2
Published by the American Association of Physics Teachers

Related Articles
Vector acoustic intensity around a tuning fork
Am. J. Phys. 81, 99 (2013)
A low cost remote sensing system using PC and stereo equipment
Am. J. Phys. 79, 1240 (2011)
Demonstrating superposition of waves and Fourier analysis with tuning forks and MacScope II
Am. J. Phys. 79, 552 (2011)
Calibrating the frequency of tuning forks by means of Lissajous figures
Am. J. Phys. 79, 517 (2011)
Acoustic scattering from a spherical lens irradiated by a finite transducer: Focusing effect and refraction
Am. J. Phys. 79, 401 (2011)

Additional information on Am. J. Phys.

Journal Homepage: http://ajp.aapt.org/
Journal Information: http://ajp.aapt.org/about/about_the_journal
Top downloads: http://ajp.aapt.org/most_downloaded
Information for Authors: http://ajp.dickinson.edu/Contributors/contGenInfo.html

Downloaded 13 Mar 2013 to 150.108.161.71. Redistribution subject to AAPT license or copyright; see http://ajp.aapt.org/authors/copyright_permission
A thermoacoustic oscillator powered by vaporized water and ethanol
Daisuke Noda and Yuki Uedaa)
Graduate school of Bio-Applications and Systems Engineering, Tokyo University of Agriculture
and Technology, 2-24-16 Nakacho, Koganei, Tokyo 184-8588, Japan
(Received 21 December 2011; accepted 26 October 2012)
We measure the temperature difference required to drive a thermoacoustic oscillator containing air,
water vapor, and liquid water as the working fluids. The oscillator is composed of a large tube
containing an array of narrow tubes connected at one end to a tank of liquid water. When the water
is heated, the temperature difference across the tube array increases until thermoacoustic
oscillations occur. The temperature difference at the onset of oscillation is measured to be 56  C,
significantly smaller (by 200  C) than the temperature measured when the tank is filled with dry
air instead of water. The temperature difference can be further reduced to 47  C by using ethanol
instead of water. VC 2013 American Association of Physics Teachers.
[http://dx.doi.org/10.1119/1.4766940]

I. INTRODUCTION and is described in a story published in 1776.8 The instru-

ment is composed of an iron bowl and a barrel and is sche-
The propagation of an acoustic wave in a narrow tube can matically illustrated in Fig. 1(b). The barrel is mounted on
lead to a variety of thermoacoustic phenomena. As an exam- the iron bowl and both its ends are open. A mesh screen is
ple, Fig. 1(a) shows a device that is used to demonstrate ther- located in the barrel and covered with rice grains, and liquid
moacoustic oscillations.1–3 It is composed of a large tube water is placed in the iron bowl. When the water is externally
containing an array of narrow tubes, called a stack. When a heated, the gas in the Kibitsunokama begins to oscillate
heat source (such as a flame) is used to establish a tempera- spontaneously and emits a sound similar to the lowing of cat-
ture difference across the stack, a critical point is reached tle; this sound is used for fortune telling. In Japan, high
when the gas inside the tube spontaneously begins to oscil- school and undergraduate students have replicated this spon-
late, resulting in the emission of an acoustic wave from the taneous oscillation.
tube. The structure of this device is quite simple and it can The basic components of the Kibitsunokama are similar to
be used to explain the physical basis1 of thermoacoustic phe- those of the thermoacoustic oscillator—each has a large tube
nomena to students interested in investigating thermoacous- containing narrow flow channels—however, the temperature
tic engines and heat pumps.4–6 differences required to drive them are quite different. For the
The temperature difference across the stack at the onset of Kibitsunokama, the temperature difference necessary to
thermoacoustic oscillation DTo , depends on various aspects induce oscillations is only around 100  C, significantly
of the oscillator, such as the length and position of the stack, smaller than that required to drive the thermoacoustic oscil-
and is typically between 200  C and 500  C. To achieve such lator (200–500  C). The reason for such a large difference
a high temperature difference, combinations such as liquid in DTo is not obvious, but the fact that the Kibitsunokama
nitrogen and ambient air, or ambient air and burning gas are contains water, whereas the thermoacoustic oscillator does
used as the heat sink and heat source, respectively.1–3 In this not indicates that the addition of water is a key factor in the
study, we demonstrate a method of reducing DTo to below reduction of DTo .
60  C, which facilitates the demonstration of thermoacoustic In this study, we construct a thermoacoustic oscillator and
oscillations. The basic idea for this method is taken from a experimentally determine the value of DTo using air, water,
Japanese instrument called “Kibitsunokama.” and diluted ethanol as working fluids. We find that DTo
The Kibitsunokama is an instrument used in historical Jap- decreases by more than 200  C when using water and air
anese shrine rituals. This instrument has a long history; it is compared to using air alone. Moreover, the use of diluted
mentioned in the diary of a Buddhist monk written in 15687 ethanol further reduces DTo by an additional 10  C.

II. EXPERIMENTAL SETUP

A schematic diagram of our setup is shown in Fig. 2. We
construct a Sondhauss-tube-type thermoacoustic oscillator9–11
composed of a tube of length 100 mm and radius 24 mm con-
nected to a tank of volume 380 cm3 (a tank is used because it
allows for easy storage of liquid water). The stack used for the
present experiment is made of ceramic and is composed of
many flow channels. The cross section of the channels is a
square of side length 0.96 mm; the length of the stack is
20 mm. The stack is located at the end of the tube nearest the
tank, and a cold-temperature heat exchanger is attached to its
cold side. The water is heated using an electric heater that has
Fig. 1. Schematic illustrations of: (a) a conventional thermoacoustic oscilla- a power E_ el supplied to it. Thermocouples are used to measure
tor and (b) a Kibitsunokama. the temperatures at both ends of the stack (TC and TH ) and a

124 Am. J. Phys. 81 (2), February 2013 http://aapt.org/ajp C 2013 American Association of Physics Teachers
V 124

Downloaded 13 Mar 2013 to 150.108.161.71. Redistribution subject to AAPT license or copyright; see http://ajp.aapt.org/authors/copyright_permission
 Fig. 2. Schematic illustration of the experimental setup.

Fig. 3. Inverse of Q-value versus temperature difference DT ¼ TH  TC in

pressure sensor (JTEKT Co. PMS-5M) is mounted to measure the thermoacoustic oscillator filled with dry air.
the (oscillatory) pressure inside the tube.

pressure as functions of time t while maintaining the input

III. EXPERIMENT electrical power at E_ el ¼ 40 W. The measured temperatures
are shown in Fig. 4(a). Up until t ¼ 710 s (dotted line on
To begin, we used dry air as a working fluid to obtain the graph) there is no pressure oscillation; the temperature of TC
temperature difference DTo for the onset of thermoacoustic is relatively stable at 26  C, while TH steadily increases. At
oscillation under conventional conditions. Although water is t ¼ 710 s, when TH reaches 82  C, the working fluids inside
illustrated in Fig. 2, the tank had no water during this experi- the setup spontaneously begin oscillating with a frequency of
ment. Because our setup has no high-temperature heat 181 Hz. Thus, when using water in the oscillator we find
exchanger, it is difficult to increase temperature TH beyond DTo ¼ 56  C, more than 200  C less than our estimate of DTo
110  C and we observed no oscillation with TH ¼ 110  C. when using only air in the oscillator. It is worth noting that
Therefore, we estimated DTo by focusing on the relation the value of TH at the onset of the oscillation (TH;o ) is lower
between the Q-factor and DT ¼ TH  TC . than the boiling point of water.
Atchley et al.12 and Biwa et al.13 have demonstrated that As can be seen in Fig. 4(a), TH remains relatively stable at
the Q-factor in a thermoacoustic oscillator can be expressed approximately 82  C after t ¼ 710 s. Conversely, TC steadily
as increases from 26  C at t ¼ 710 s to 44  C at t ¼ 1 200 s. After
Es t ¼ 1200 s, both temperatures remain fairly stable and the
Q ¼ 2pf ; (1) amplitude of oscillatory pressure was found to have a con-
_
E d  E_ g stant value. In other words, the smallest temperature differ-
ence DT  with which thermoacoustic oscillations are
where f is the resonant frequency of the gas column in the maintained in the present experiment was 42  C (when water
setup, Es is its stored acoustical energy, E_d is the rate of dis- is used as one of the working fluids).
sipated acoustic energy in the column, and E_g is the rate of
acoustic energy generation in the stack due to thermoacous-
tic energy conversion.3 Because E_g increases with increasing
DT, there is a critical value of DT when E_d ¼ E_g , at which
point the value of Q approaches infinity. This critical value
can be regarded as DTo .12,13
To obtain the Q-factor for our setup, a loudspeaker is
placed outside the oscillator and the acoustic waves were fed
into our setup. The loudspeaker is driven with a sweep signal
and the pressure in the tube is measured with the pressure
sensor. Using a fast Fourier transform analyzer (Onosokki
Co. Ltd. DS-2000), we obtain a resonance curve from which
we can calculate the Q-factor. Repeating this experiment for
different DT values allows us to extrapolate to find the (ap-
proximate) temperature difference when Q ! 1.
Figure 3 shows our experimental results plotted as 1/Q ver-
sus DT. As seen in this figure, the value of 1/Q decreases with
increasing DT. Assuming a linear relation and extrapolating to
1=Q ! 0, we find DTo  290  C for our thermoacoustic os-
cillator filled with dry air. We note that the frequency at the
peak of the resonance curve when DT ¼ 88  C was 166 Hz,
indicating that the resonance mode in the oscillator was of the
Helmholtz type.14
Next, we place 20 cm3 of liquid water in the tank. We Fig. 4. Temperatures TH and TC shown as functions of time t in the thermoa-
measure the temperatures TH and TC and the oscillatory coustic oscillator using working fluids of water (a) and diluted ethanol (b).

125 Am. J. Phys., Vol. 81, No. 2, February 2013 D. Noda and Y. Ueda 125

Downloaded 13 Mar 2013 to 150.108.161.71. Redistribution subject to AAPT license or copyright; see http://ajp.aapt.org/authors/copyright_permission
 As mentioned above, the frequency of oscillation at the From a thermodynamic point of view, thermoacoustic
onset was 181 Hz. However, this oscillation frequency varied oscillators can be regarded as heat engines1 because they
slightly with time, first increasing (up to 185 Hz) and then convert thermal energy (heat) into acoustic waves that can in
decreasing and finally stabilizing at 171 Hz. Such a frequency turn do useful work, such as moving a speaker cone to pro-
change can be attributed to a change in the speed of sound in duce electrical energy. Recently, thermoacoustic oscillators
the experimental setup, which depends on temperature and have been designed to produce energy from waste heat,16,17
molar fraction of vaporized water in the working gas. and hence reductions in DTo and DT  have been pursued by
The total weight of the setup, including water, remained thermoacousticians.13,16 The present results indicate that
approximately the same before and after the experiment. How- using water or ethanol as one of the working fluids can
ever, we noticed that the stack was wet after the experiment, reduce these temperature differences while preserving the
whereas it was dry at the beginning of the experiment. These simplicity of the thermoacoustic energy converter.
facts indicate that air with a relative humidity of 100% became
the working fluid, and that condensation and evaporation ACKNOWLEDGMENTS
occurred continuously in the stack and tank. In addition,
Raspet et al.15 give a theoretical argument that condensation The authors would like to express their cordial thanks to
and evaporation contribute to the enhancement of the thermoa- the anonymous referees for helpful comments that improved
coustic effect. Hence, the large reduction of DTo can likely be this manuscript.
attributed to the phase change of water. In the future, a quanti-
tative comparison between the experimentally obtained and a)
Electronic mail: [email protected]
the theoretically calculated values of DTo may be presented. 1
J. Wheatley, T. Hofler, G. W. Swift, and A. Migliori, “Understanding
As a final experiment, we used 20 cm3 of diluted ethanol some simple phenomena in thermoacoustics with applications to acoustical
instead of water as the working fluid to see if this might fur- heat engines,” Am. J. Phys. 53, 147–162 (1985).
2
ther reduce DTo ; the diluted ethanol had a density of A. Jeromen, “A simplified thermoacoustic engine demonstration,” Am. J.
0:83 g=cm3 before the start of the experiment. The experi- Phys. 71, 496–499 (2003).
3
G. W. Swift, “Thermoacoustic engines,” J. Acoust. Soc. Am. 84, 1145–
ment was repeated while again holding E_ el ¼ 40 W and the 1180 (1988).
results are shown in Fig. 4(b). Once again we see that ini- 4
J. Wheatley, T. Hofler, G. W. Swift, and A. Migliori, “An intrinsically irre-
tially TH increases steadily while TC remains relatively con- versible thermoacoustic heat engine,” J. Acoust. Soc. Am. 74, 153–170
stant at around 25  C. When t ¼ 410 s (dotted line on graph), (1983).
5
thermoacoustic oscillation is detected at a frequency of 159 G. W. Swift, Thermoacoustics: A Unifying Perspective for Some Engines
Hz. The onset occurs for DTo ¼ 47  C, almost 10  C smaller 6
and Refrigerators (Acoustical Society of America, Pennsylvania) (2002).
S. L. Garrett, “Resource Letter TA-1: Thermoacoustic engines and refrig-
than the experiment with water as the working fluid. After
erators,” Am. J. Phys. 72, 11–17 (2004).
the onset of oscillation, TH stabilizes at around 72  C while 7
Unspecified Buddhist monks, Tamoinnikki (Nara) (1478–1618, re-edited
TC begins increasing until about t ¼ 650 s, when both temper- in 1935).
atures level off and increase much more slowly. At 8
A. Ueda, Ugetsu Monogatari (Kyoto, 1776)
9
t ¼ 1 500 s, the temperatures remain almost constant and the J. W. S. Rayleigh, Theory of Sound, 2nd ed. (Dover Publications, 1945),
amplitude of the oscillation also takes on a constant value. Vol. II, pp. 230–231.
10
At t ¼ 1800 s we have TH ¼ 77  C and TC ¼ 56  C, giving a N. Rott and G. Zouzoulas, “Thermally driven acoustic-oscillations, part 4:
Tubes with variable cross-section,” Z. Angew Math. Phys. 27, 197–224
smallest temperature difference of DT  ¼ 21  C. (1976).
11
N. Sugimoto and R. Takeuchi, “Marginal conditions for thermoacoustic
IV. SUMMARY oscillations in resonators,” Proc. R. Soc. London, Ser. A 465, 3531–3552
(2009).
We estimate and measure the temperature difference DTo 12
A. Atchley, H. Bass, T. Hofler, and H. Lin, “Study of a thermoacoustic
for the onset of thermoacoustic oscillation under three differ- prime mover below onset of self]oscillation,” J. Acoust. Soc. Am. 91,
ent conditions. Under the first condition, dry air is used as 734–743 (1992).
13
T. Biwa, D. Hasegawa, and T. Yazaki, “Low temperature differential ther-
the working fluid; under the second condition, air and water
moacoustic Stirling engine,” Appl. Phys. Lett. 97, 034102–1–3 (2010).
are used; and under the third condition, air and diluted etha- 14
G. Vandegrift, “Experimental study of the Helmholtz resonance of a vio-
nol are used. The experimental results show that the values lin,” Am. J. Phys. 61, 415–421 (1993).
of DTo for these experiments are 290  C, 56  C, and 47  C, 15
R. Raspet, W. V. Slaton, C. J. Hickey, and R. A. Hiller, “Theory of inert
respectively. Furthermore, the smallest measured tempera- gas-condensing vapor thermoacoustics: Propagation equation,” J. Acoust.
ture difference while maintaining thermoacoustic oscillations Soc. Am. 112, 1414–1422 (2002).
16
(DT  ) was 42  C and 21 C under the second and third condi- K. de Blok, “Novel 4-stage traveling wave thermoacoustic power gener-
ator,” in Proceedings of FEDSM2010, 73–79 (ASME, Montreal, Canada,
tions, respectively. Hence, we can conclude that the intro- 2010).
duction of water or ethanol into a thermoacoustic oscillator 17
K. Nakamura and Y. Ueda, “Design and construction of a standing-wave
has a large effect on the reduction of the temperature differ- thermoacoustic engine with heat sources having a given temperature
ence required to drive the oscillator. ratio,” J. Therm. Sci. Technol. 6, 416–423 (2011).

126 Am. J. Phys., Vol. 81, No. 2, February 2013 D. Noda and Y. Ueda 126

Downloaded 13 Mar 2013 to 150.108.161.71. Redistribution subject to AAPT license or copyright; see http://ajp.aapt.org/authors/copyright_permission