Sie sind auf Seite 1von 14

Energy Conversion and Management 49 (2008) 35853598

Contents lists available at ScienceDirect

Energy Conversion and Management


journal homepage: www.elsevier.com/locate/enconman

Design and optimization of thermoacoustic devices


Hadi Babaei, Kamran Siddiqui *
Department of Mechanical and Industrial Engineering, Concordia University, Montreal, Canada

a r t i c l e

i n f o

Article history:
Received 6 December 2007
Accepted 21 July 2008
Available online 11 September 2008
Keywords:
Thermoacoustics
Sustainable refrigerator
Design procedure

a b s t r a c t
Thermoacoustics deals with the conversion of heat energy into sound energy and vice versa. It is a new
and emerging technology which has a strong potential towards the development of sustainable and
renewable energy systems by utilizing waste heat or solar energy. Although simple to fabricate, the
designing of thermoacoustic devices is very challenging. In the present study, a comprehensive design
and optimization algorithm is developed for designing thermoacoustic devices. The unique feature of
the present algorithm is its ability to design thermoacoustically-driven thermoacoustic refrigerators that
can serve as sustainable refrigeration systems. In addition, new features based on the energy balance are
also included to design individual thermoacoustic engines and acoustically-driven thermoacoustic refrigerators. As a case study, a thermoacoustically-driven thermoacoustic refrigerator has been designed and
optimized based on the developed algorithm. The results from the algorithm are in good agreement with
that obtained from the computer code DeltaE.
2008 Elsevier Ltd. All rights reserved.

1. Introduction
Thermoacoustic is a branch of science dealing with the conversion of heat energy into sound energy and vice versa. Device that
converts heat energy in sound or acoustic work is called thermoacoustic heat engine or prime mover and the device that transfers
heat from a low temperature reservoir to a high temperature reservoir by utilizing sound or acoustic work is called thermoacoustic
refrigerator. Although the thermoacoustic phenomenon was discovered more than a century ago, the rapid advancement in this
eld occurred during the past three decades when the theoretical
understanding of the phenomenon was developed along with the
prototype devices based on this technology [1,2]. The thermoacoustic technology has not reached the technical maturity yet, as a result, the performance of thermoacoustic devices is still lower than
their convectional counterparts. Thus, signicant efforts are
needed to bring this technology to maturity and develop competitive thermoacoustic devices. There are several advantages of heat
engines and refrigerators based on thermoacoustic technology as
compared to the conventional ones. These devices have fewer components with at most one moving component with no sliding seals
and no harmful refrigerants or chemicals are required. Air or any
inert gas can be used as working uids which are environmentally
friendly. Furthermore, the fabrication and maintenance costs are
low due to inherent simplicity of the thermoacoustic devices.
The main components of a typical thermoacoustic engine or
refrigerator are a resonator, a stack of parallel plates and two heat
* Corresponding author. Tel.: +1 514 848 2424x7940; fax: +1 514 848 3175.
E-mail address: siddiqui@encs.concordia.ca (K. Siddiqui).
0196-8904/$ - see front matter 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2008.07.002

exchangers. A half wavelength (or a quarter wavelength) acoustic


standing wave is generated in the resonator. The thermoacoustic
phenomenon takes place in the stack when a nonzero temperature
gradient imposed along the stack plates (i.e. parallel to the direction of the sound wave propagation) interacts with the sound wave
oscillations. The heat exchangers are responsible of transferring
heat in and out of a thermoacoustic device at their desired temperatures, thus maintaining a given temperature gradient along the
stack.
Thermoacoustic refrigerators can be classied based on the
source of the acoustic energy input. If the acoustic energy is provided by a thermoacoustic engine, the refrigerator is called thermoacoustically-driven thermoacoustic refrigerator (TADTAR).
Whereas, if the acoustic energy is provided by an acoustic driver
e.g. a loudspeaker, it is termed as acoustically-driven thermoacoustic refrigerator. During the past decades, several acoustically-driven thermoacoustic refrigerators have been developed [35].
Although the form of energy consumed in these refrigerators is
acoustic, the energy source for the acoustic driver is typically electrical from conventional energy resources. During recent years,
there is an increased interest in the development of thermoacoustically-driven thermoacoustic refrigerators. These devices are built
by coupling a thermoacoustic refrigerator to a thermoacoustic engine. Thermoacoustic engines are capable of producing acoustic
energy from any source of heat energy. Thus, the primary energy
source to drive the refrigerator could be conventional or unconventional that includes industrial waste heat, solar energy and fossil
fuels. If the heat source for the thermoacoustic engine is the industrial waste heat or solar energy then this device has two major
advantages. Firstly, it does not require any addition conventional

3586

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

Nomenclature
A
a
BR
COP
COPR
cp
csolid
DR
E_ 2
DE_ 2
f
H_ 2
HX
K
k
L
l
Pm
PA
p1
Pr
Q
rh
S
S_
T
DT
rT
U1
u1
x1
xc

cross-sectional area (m2)


speed of sound (m/s)
blockage ratio
coefcient of performance
coefcient of performance relative to Carnot
isobaric heat capacity of the working gas (J kg1 K1)
heat capacity of the stack plates (J kg1 K1)
drive ratio
work ux (W)
produced or consumed work ux (W)
resonant frequency (Hz)
total energy ux (W)
heat exchanger
thermal conductivity of the working gas (W m1 K1)
wave number (m1)
length (m)
half of the plate thickness (m)
mean pressure (Pa)
antinode pressure amplitude (Pa)
pressure amplitude (Pa)
Prandtl number
heat ux (W)
hydraulic radius (m)
surface area (m2)
entropy ux (W/K)
temperature (K)
temperature difference (K)
temperature gradient (K/m)
volume ow rate (m3/s)
velocity amplitude (m/s)
gas displacement amplitude (m)
stack center position (m)

energy resource and secondly, by utilizing the waste heat, the


amount of total waste heat rejected to the thermal energy sink will
be reduced which will increase the overall performance of the entire system. Thus, a complete thermoacoustic refrigeration system
in which the heat engine (which operates on waste heat) drives a
refrigerator and the entire system has no harmful affects on the
environment can be termed as a sustainable refrigeration system. In contrast to the acoustically-driven thermoacoustic refrigerator which has one moving component i.e. the acoustic driver,
thermoacoustically-driven thermoacoustic refrigerator has no
moving parts thus; chances of mechanical failure are extremely
low.
Recently, some efforts have been made to develop heat engines
that operate on waste heat. Symko et al. [6] designed and developed a thermoacoustic heat engine that utilizes heat from a microcircuit to produce sound. Hatazawa et al. [7] proposed a heat
engine that utilizes waste heat from a four-stroke automobile gasoline engine. Adeff and Hoer [8] developed a prototype thermoacoustic refrigeration system that operates on the solar energy.
Babaei et al. [9] have proposed a thermoacoustic refrigeration system for a gas turbine trigeneration system that operates on the
waste heat from the gas turbine. It has been demonstrated that
the thermoacoustic refrigeration system has the ability to enhance
the overall efciency of a trigeneration system by 5%.
Some recent theoretical studies have demonstrated the strong
potential of thermoacoustic devices in energy conservation and
reduction of harmful emissions. A study shows that if all the industrial waste heat above 140 C in Netherlands can be used in thermoacoustic devices, this would save 16 PJ per year which
corresponds to the saving of more than 5 billion m3 of natural

y0

a
b
dk
dv

es
c
C

gth
k

l
P

q
x

half of the plate spacing (m)


thermal diffusivity (m2 s1)
thermal expansion coefcient (K1)
thermal penetration depth (m)
viscous penetration depth (m)
plate heat capacity ratio
ratio, isobaric to isochoric specic heats
normalized temperature gradient
thermal efciency
wavelength (m)
dynamic viscosity (kg m1 s1)
perimeter (m)
density (kg m3)
angular frequency (rad s1)

Subscripts, superscripts
a
ambient
c
cold
con
consumed
crit
critical
d
duct
eng
engine
gen
generated
h
hot
m
mean
n
normalized
pro
produced
r
resonator
ref
refrigerator
s
stack
t
total

gas [10]. It is estimated that over 32 billion liters of fuel is consumed annually for the operation of vehicle air-conditioners in
the US alone. Modern vehicle refrigeration systems use R-134a,
with a global warming potential still 1300 times that of carbon
dioxide [11]. Zoontjens et al. [12] theoretically investigated the
feasibility of using thermoacoustic devices as the air conditioning
system of an automotive by utilizing the automotive waste heat.
They concluded that the thermoacoustic refrigerator has a strong
potential to replace the existing automotive air conditioning
systems.
Although thermoacoustic devices are easy to build and maintain, designing of these devices involves signicant technical challenges. These challenges become more substantial when designing
a thermoacoustically-driven thermoacoustic refrigerator. This is
attributed to the complicated thermoacoustic theory which is not
directly applicable for design purposes. Thus, a systematic approach is necessary to design and optimize thermoacoustic
devices.
Wetzel and Herman [13] developed a design algorithm for
acoustically-driven thermoacoustic refrigerators. They developed
the design algorithm by using the simplied linear thermoacoustic
model, and normalizing the position and length of the refrigerator
stack and the equations of the total power ow and consumed
acoustic power in the stack. By applying the algorithm, the designer can decide the stack position and length at the given temperatures of heat exchangers to have the maximum performance
of the stack. The geometrical parameters such as stack plate thickness and spacing as well as the cross-sectional area of the resonator can also be estimated. In this study, however, it is not described
how the desired cooling power of the refrigerator, the stack

3587

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

consumed acoustic power and the total power ow in the stack are
correlated, and under which conguration of the refrigerator stack
this correlation is valid.
Tijani et al. [14] also described the design algorithm for acoustically-driven thermoacoustic refrigerators by considering a correlation between the desired cooling power of the refrigerator, the
stack consumed acoustic power and total power in the stack,
which is different from that of Wetzel and Herman [13]. It is however, not well described how this correlation is derived and at
which refrigerator conguration it may be applied.
The above design algorithms are applicable only to design
acoustically-driven thermoacoustic refrigerators. These algorithms
cannot be used in designing a thermoacoustically-driven thermoacoustic refrigerator (TADTAR), as designing of TADTAR involves
more parameters and it is more challenging than the acoustically-driven thermoacoustic refrigerators. Therefore, to design
and develop efcient sustainable thermoacoustic refrigeration systems, a detailed design and optimization procedure is necessary.
To the best of authors knowledge no such design and optimization
procedure or algorithm is available.
In this paper, a comprehensive systematic procedure has been
developed for the design and optimization of thermoacoustic devices by applying the simplied linear thermoacoustic model. This
procedure which is mainly intended to design and optimize a thermoacoustically-driven thermoacoustic refrigerator (TADTAR) can
also be used to design and optimize individual thermoacoustic engines and acoustically-driven thermoacoustic refrigerators. It
should be noted that the procedure presented in this study provides a more comprehensive discussion on the design and optimization of acoustically-driven thermoacoustic refrigerators than
previous studies. The design procedure which is based on the energy and entropy balances applied on different components of
the device is a simple and effective tool to design and optimize a
thermoacoustic device to meet its requirements. The goal of
designing a thermoacoustically-driven thermoacoustic refrigerator
is to meet the required cooling power at the desired cooling temperature and at the given heat input temperature while rejecting
some heat to the environment.
The developed algorithm not only provides a step by step procedure to design and optimize a thermoacoustic device but also enables to evaluate the inuence of different parameters on the
behavior and performance of the device.
Finally, a thermoacoustically-driven thermoacoustic refrigerator is designed and optimized based on the developed procedure
and simulated by the computer code DeltaE to compare and verify
the design parameters.
It is worth mentioning that using DeltaE to design thermoacoustic devices from scratch needs tremendous amount of effort especially in the case of thermoacoustically-driven thermoacoustic
refrigerator. The presented procedure signicantly reduces the
technical challenges associated with the designing of thermoacoustic devices.

the plate. The length of the plate is assumed equal to the peak to
peak displacement of gas parcels (2 * jx1j) oscillating along the
plate (see Fig. 1a). The gure shows a magnied view of a single
plate stack and a gas parcel oscillating next to it in a half wavelength thermoacoustic refrigerator. The gas parcel oscillates under
the inuence of standing wave generated by the acoustic power input. The heat energy is transferred in and out of the device by the
cold and ambient heat exchangers located at the edges of the stack
plate, respectively. The temperatures of the heat exchangers impose a temperature gradient along the stack plate (rT). The variation of the pressure and velocity magnitudes of the acoustic wave
along the resonator is shown in Fig. 1b.
In a real thermoacoustic device, the oscillations are sinusoidal;
but for simplicity, the square wave motion is considered to explain
the basic thermodynamic cycle that a gas parcel undergoes. The
gas parcel experiences two adiabatic processes while moving along
the solid plate and two irreversible constant pressure processes
while exchanging heat with the solid plate [2]. Two temperatures
are important to the parcel. The temperature of the gas parcel after
adiabatic compression and expansion (imposed by the sound wave
and related to the sound wave pressure oscillation) and the local
temperature of the solid plate (imposed by the heat exchangers)
adjacent to the gas parcel after adiabatic compression and expansion and displacement of the gas parcel. Note that the acoustic
wave is responsible for both adiabatic compression and expansion,
and the displacement of the gas parcel along the solid plate. If the
temperature of the gas is higher than that of the plate, heat ows
from the gas to the plate. If the temperature of the gas is lower
than that of the plate, heat ows from the plate to the gas. Thus,
it is the imposed temperature gradient rT along the plate that
makes a thermoacoustic device to operate as an engine or a refrigerator. A zero or low temperature gradient is the condition for a
refrigerator and a high temperature gradient is the condition for
an engine. If rT along the plate be selected in such a way that
the temperature change along the plate (2rTjx1j) as seen by the
parcel just matches the parcels temperature change due to adia

batic compression and expansion 2 Tqm bpcp1 , no heat would ow
m

between the parcel and the solid plate. This temperature gradient
is called the critical temperature gradient and is dened as [2],

2. Thermoacoustic principle
Phasing plays an important role in the operation of thermoacoustic devices. To attain a proper phasing in a thermoacoustic device, a rather poor thermal contact is essential between the gas
parcel and its adjacent solid plate. This imperfect thermal contact
causes the heat ow between the gas and the plate, not to produce
instantaneous changes in the gas temperature. Instead, the heat
ow creates a time phasing between temperature, pressure and
displacement needed to drive the gas parcels through a thermodynamic cycle [2].
Consider a solid plate aligned in the direction of the acoustic
wave propagation with an imposed temperature gradient rT along

Fig. 1. (a) Schematic of a thermoacoustic refrigerator with a single plate stack, (b)
variation of pressure and velocity amplitudes along the resonator tube, solid line:
pressure, dashed line: velocity.

3588

rT crit

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

T m bxp1
qm cp u1

Usually the length of the plate is larger than the displacement of a


given gas parcel. Thus, the heat transfer from one end of the plate to
the other end is occurred by a series of gas parcels along the plate as
depicted in Fig. 2. In Fig. 2, gas parcel A absorbs heat from the plate
at location a and transfers it to the plate at location b. Half a cycle
later, the adjacent gas parcel B picks this heat from the location b
(at this instant, parcel A is at the location a). This heat is transferred
to location c by the parcel B from where the gas parcel C picks it and
delivers to the location d. Thus, the heat is transferred from one end
to the other by gas parcels as bucket brigade. It should be noted that
the plate is used only for the temporary storage of heat [2].

The simple linear expressions for total power ow H_ 2 (i.e. total


energy ux) through the stack, and the acoustic power DE_ 2 (work
ux) produced in (or consumed by) the stack are expressed by
the following equations [2,15]:

"
#
p

p dv 
1 Pr Pr Pres
_H2 A dk T m b j p1 kU 1 j
p
C
 1 Pr 
4 rh A1 es 1 PrK
y0
1 Pr
 AK Asolid K solid rT
3
2
3
0
1
2
2
A Ls 4c  1 j p1 j dk x @
C
qm j U 1 j dv x5

DE_ 2
p  1A 
cpm 1 es
4 rh
A2 K
1 Pr K
4
where C is the normalized temperature gradient dened as

rT
rT crit ,

2.1. Simplied linear model of thermoacoustic devices

K 1  drhv 2rdv2 , A and Asolid are the uid and solid cross-sectional
h

Consider a stack of parallel plates with x axis along the direction


of the acoustic wave propagation and y axis perpendicular to the
plane of the stack. The plate thickness is equal to 2l and the plate
spacing is equal to 2y0. The simplied thermoacoustic model is
developed by linearizing momentum, continuity and heat ow
equations, and considering the following three assumptions [2].
First, it is assumed that y0dj, y0 dv where dk is the thermal
penetration depth dened as the thickness of the layer around
the stack plate through which the heat can diffuse in the uid,
whereas, dm is the thickness of the layer around the stack plate
where the viscous effects are signicant. This assumption is called
the boundary layer approximation and typically in thermoacoustic
devices, dj 6 y0 6 2dj [2]. Second, the length of the stack is considered to be signicantly less than the wavelength of the standing
acoustic wave (i.e. Ls  k) such that it does not perturb the acoustic
standing wave (short stack approximation). With this approximation and assuming standing wave phasing between pressure and
velocity, the velocity and pressure can be expressed as [2],

p1 PA coskx;




l
PA
sinkx
u1 1
y0
qm a

Finally, it is assumed that the stack is short enough that p1 and u1


could be regarded as independent of x within the stack, and the
temperature difference along the stack is less than the stack mean
temperature (i.e. DTTm). So the thermophysical properties of the
gas are assumed to be independent of x within the stack. Thus, p1,
u1 and thermophysical properties are evaluated at the stack mid
point, i.e. the stack mid temperature [2].

Fig. 2. Mechanism of heat transfer by the gas parcels along the stack plate of a
thermoacoustic device.

areas in the stack, respectively, and es is the plate heat capacity ratio
q
q c K
dened as, es q cm p K .
solid solid

solid

By assuming that all dimensions of the resonator are much larger than the penetration depths and the temperature gradient
along the axis of the resonator is zero, the acoustic power loss
per unit surface area of the resonator can be estimated as [15],

 2
dE_ 2
1 U 1 
1 j p1 j2
c  1dj x
 qm   dv x 
4
4 cpm
dS
A

The rst term on the right hand side of Eq. (5) represents the energy
dissipated due to the viscous shear and the second term on the right
hand side represents the energy dissipated due to the thermal
relaxation.
It should be noted that the stack plates are assumed ideal so
that the plate heat capacity ratio es is zero and the last term in
the right hand side of Eq. (3) which represents the axial conduction
in the stack plates and the working gas is neglected. These two
terms have a negligible effect on the calculations [13]. By considering only ideal gases close to their critical point as the working gas,
the parameter Tmb in Eq. (3) can be set equal to unity [2].

3. Design and optimization procedure


Besides the available features from previous studies, following
new features are applied in the present study to develop the comprehensive design and optimization procedure for thermoacoustic
devices.
The simplied linear thermoacoustic model is used to evaluate
the engine part of the device. All the dimensions in the direction of
the acoustic wave propagation including the length and position of
the stacks and heat exchangers are normalized. The normalized
acoustic power equation is applied to estimate the dissipated
acoustic power in the heat exchangers. The equation estimating
the acoustic power losses in the resonators wall surface area is
normalized (see Eq. (9)). A comprehensive discussion is presented
to correlate the desired cooling power and the required heat input
to the total power ow in the stack and the acoustic power ow, by
applying the energy balance on the cold and hot heat exchangers
(see Eqs. (11), (15), (16), (18), and (20)). The normalized engine
stack position and length are selected by applying the energy balance on the whole device and these selections are then modied to
have the engine stack performs at the maximum efciency at the
given temperatures of the heat exchangers. This behavior is also
shown by applying the entropy balance and energy balance on a
device, simultaneously. It is shown that the engine stack position
and length could be selected to have the minimum entropy generation within the system while producing the required acoustic
power to run the system at the desired temperatures of the heat

3589

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

exchangers (Eq. (33)). The designer could also estimate the required heat input at the desired temperature to run the engine section of the device (Eq. (35)).
Heat exchangers are the least understood components of thermoacoustic devices and their proper designing is a critical task,
as the literature contains very little experimental or analytical
guidance. Swift [2] originally argued that the optimum length of
a heat exchanger should be equal to the peak to peak gas displacement amplitude at the heat exchanger location. The ambient heat
exchanger is always closer to the velocity node, so the peak to peak
gas displacement at its location is smaller than that of the cold heat
exchanger in a thermoacoustic refrigerator. On the other hand, the
ambient heat exchanger transfers more heat compared to the cold
heat exchanger in a thermoacoustic refrigerator as it must transfer
both the transferred heat by the cold heat exchanger and portions
of dissipated acoustic power in the device to the outside environment. So it is safe to say that by assuming the same heat transfer
coefcient and temperature difference between the solid plate
and the working gas, the ambient heat exchanger requires more
heat transfer area compared to the cold heat exchanger. The same
argument can be raised for thermoacoustic engines. The heat
transferred by the ambient heat exchanger in a thermoacoustic engine is smaller than that transferred by the hot heat exchanger
while the hot heat exchanger is always closer to the velocity node.
For the design procedure, following assumptions are made for
heat exchangers. All heat exchangers are assumed parallel plate.
The length of the cold heat exchanger and ambient heat exchanger
in a thermoacoustic refrigerator are assumed equal to the peak-topeak gas displacement amplitude at the cold heat exchanger location. The length of the hot heat exchanger and ambient heat
exchanger in a thermoacoustic engine are assumed equal to the
peak-to-peak gas displacement amplitude at the ambient heat exchanger location. The blockage ratio of heat exchangers is assumed
equal to that of their respective stack.
In the following subsections, normalization of thermoacoustic
parameters and equations are described rst followed by the
description of the energy balance (rst law of thermodynamics)
and entropy balance (second law of thermodynamics) on the selected control volumes of the device.
3.1. Normalization
The energy ux equations (Eqs. (3)(5)) indicate that there are
different sets of independent parameters that play important roles
in evaluating the performance of a thermoacoustic system. These
parameters can be categorized in three main categories. Geometrical variables, which are stack plate thickness and spacing, position
and length of the stack, and stack cross-sectional area. Material related variables that include thermophysical properties of the working gas and the stack. Design related variables which are resonance
frequency, mean pressure and pressure amplitude of the working
gas, mean temperature and temperature difference along the stack
and the desired cooling power of the system [13]. Due to a large
number of design parameters, the number of independent param-

eters can be reduced through normalization. Table 1 shows the


independent variables, normalizing parameters and the normalized form [16]. In the present study, the maximum value for the
normalized stack position (measured from pressure antinode)
and the normalized stack length are assumed equal to 0.5 to avoid
large viscous dissipation which decreases the overall performance
of the device.
As recommended in the literature, the normalized plate spacing
(i.e. blockage ratio) was set equal to 0.8 [2,4,17].
The normalized thermal penetration depth and the normalized
viscous penetration depth can be assumed in the range 0.51 and
0.5 Pr2 to Pr2, respectively [2].
As the thermoacoustic model is based on the linear wave theory, to avoid nonlinearities, it is recommended that the drive ratio
(DR = p1/pm) should be smaller than 3% so that the acoustic Mach
number and acoustic Reynolds number would be smaller than
0.1 and 500, respectively [18,19].
The normalized temperature difference along the refrigerator
stack and engine stack are assumed in the range 00.17 and
0.350.95, respectively. The mean temperature along the refrigerator stack and engine stack are assumed in the range 288303 K,
and 390600 K, respectively. It should be noted that to satisfy
the third assumption described in the previous section, thermophysical properties of the gas inside the refrigerator stack (and resonator) are calculated based on the refrigerator stack mean
temperature and the thermophysical properties of the gas inside
the engine stack are calculated based on the engine stack mean
temperature.


DT=Ls
rrT T can be
The normalized temperature gradient, C r
T
crit

crit

expressed as the function of other normalized parameters as [2],

DT n
BRc  1Lsn cotxcn

The above equation shows that for a stack with specied length and
position, there is a range of normalized temperature differences at
which the stack operates as a refrigerator (C < 1) and there is a
range of normalized temperature differences at which it operates
as an engine (C > 1).
By dividing the total power and acoustic power equations (Eqs.
(3) and (4)) by the product APma, and assuming a parallel plate
stack (rh = y0), the following normalized equations for the total
power ow through the stack H_ 2n and the acoustic power produced in (or consumed by) the stack DE_ 2n are expressed as [16],

"
#
p

p
Pr 
_H2n  1 dkn DR2 sin2xcn  C 1 Pr
p  1 Pr  dvn
8c
1 PrK
1 Pr
2

DE_ 2n 

1
C
dkn DR2 Lsn 4BRc  1Cos2 xcn @
p  1A
4c
1 Pr K
p#
2
Sin xcn Pr

BR  K

Table 1
Normalized parameters
Independent parameters

Normalizing parameters

Normalized parameters

Length and position


Plate spacing
Penetration depths
Pressure amplitude
Temperature difference along the stack
Temperature gradient along the stack
Power

k
2p

Normalized length and position


Blockage ratio
Normalized penetration depths
Drive ratio
Normalized temperature difference
Normalized temperature gradient
Normalized power

The sum of plate spacing and thickness


Half of the stack plate spacing
Mean pressure
Mean temperature of the stack
Critical temperature gradient
The product of mean pressure, sound velocity and gas cross-sectional area in the stack

3590

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

In this study, by using the same normalizing parameters as for Eqs.


(7) and (8), the normalized acoustic power dissipated in a half
wavelength resonator is estimated as,

DE_ 2n;r  

p p 2 dk
PrDR
8c
rhr




 
c  1p
dk
DR2
8c
r hr

The energy dissipated in the resonator is proportional to the wall


surface area of the resonator. So to determine the normalized
acoustic power dissipated in a quarter wavelength resonator, the
above equation should be divided by two.
3.2. Energy balance
One purpose of the analyses presented in this section is to correlate some important thermoacoustic parameters. The required
heat input to the device is correlated with the total energy ux
through the engine stack and acoustic energy ux, by applying
the energy balance on the hot heat exchanger. The desired cooling
power of the device is correlated with the total energy ux through
the refrigerator stack and the acoustic energy ux by applying energy balance on the cold heat exchanger. The analyses cover all
possible characteristics that could be assumed for thermoacoustic
engines and refrigerators. These relationships would also provide a
better estimation of the engine and refrigerator performance.
Consider a thermoacoustic engine with its stack located near
the left pressure antinode of the resonator as illustrated in Fig. 3.
The energy balance is applied on the control volume outlined with
the dashed line which encloses the hot heat exchanger. Thus,

Q h H_ 2;s H_ 2;hd

10

where H_ 2;hd is the acoustic power leaving the control volume into
the hot duct, which is dissipated in the hot duct. If it is assumed that
the heat generated by the dissipation of the acoustic power in the
hot duct DE_ 2;hd is rejected to the environment through the hot
duct walls then, H_ 2;hd DE_ 2;hd and we have Q h H_ 2;s DE_ 2;hd . As
the hot duct is always a small portion of the resonator, the magnitude of DE_ 2;hd is very small and can be neglected, thus,

Q h H_ 2;s

11

This implies that the total energy ux into the engine stack is
approximately equal to the heat input by the hot heat exchanger.
It should be noted that Eq. (11) will also be applicable when the engine stack is placed near the right-end pressure antinode or when
the resonator walls are insulated.
The ratio of the acoustic work produced by the engine stack to
the energy ux delivered to the system by HXh is dened as the
thermal efciency of the engine stack, gth,s, expressed as,

gth;s

DE_ 2;s;eng DE_ 2n;s;eng

Qh
Q hn

12

where Qh(Qhn) is dened in Eq. (11).


Fig. 4a and b shows two possible congurations of a thermoacoustic refrigerator, i.e. the refrigerator stack located near the left
pressure antinode or right pressure antinode of the resonator,
respectively. Note that the acoustic power to the refrigerators
(either by an engine or a loud speaker) is provided from the left
side of the resonator in both cases. In other word, the engine stack
(or loud speaker) is located on the left side of the refrigerator stack.
A thermoacoustic refrigerator with its stack located near the left
pressure antinode is illustrated in Fig. 4a. The energy balance is applied on the control volume outlined with the dashed line in Fig. 4a
which encloses the cold heat exchanger. Thus,

Q c H_ 2;s H_ 2;cd

13

where Qc is the desired cooling power, H_ 2;s is the total energy ow


towards the stack which is equal to the sum of heat extracted from
the cold heat exchanger and the heat produced by the acoustic
power dissipation in the cold heat exchanger, minus the acoustic
power enter the control volume. H_ 2;cd is the acoustic power leaving
the control volume into the cold duct. This acoustic power is dissipated in the cold duct. If it is assumed that the heat generated by
the dissipation of the acoustic power in the cold duct DE_ 2;cd is rejected to the environment through the cold duct walls then,

H_ 2;cd DE_ 2;cd

14

and,

Q c  DE_ 2;cd H_ 2;s

15

Since the actual net cooling power is the amount of energy ux removed from the cold heat exchanger and pumped uphill by the
stack, the actual cooling power of the device is Q c  DE_ 2;cd in this
case.
If the cold duct is insulated then it could be assumed that DE_ 2;cd
is not rejected to the environment and it appears as a load on the
cold heat exchanger. This heat leaves the cold duct and enters the
control volume and ows into the stack.
Thus, H_ 2;cd 0 and,

Q c H_ 2;s
Fig. 3. Schematic of a thermoacoustic engine. The control volume is outlined with
dashed lines which encloses the hot heat exchanger (HXh).

16

In other conguration of the thermoacoustic refrigerator, the stack


can be located near the right pressure antinode of the device as

Fig. 4. Schematic of a thermoacoustic refrigerator with two possible congurations: (a) Refrigerator stack located near the left pressure antinode, (b) refrigerator stack
located near the right pressure anitnode. The control volume is outlined with dashed lines which encloses the cold heat exchanger (HXc).

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

shown in Fig. 4b. Consider the control volume outlined with the
dashed line which encloses the cold heat exchanger. If it is assumed
that the heat generated by the dissipation of the acoustic power in
the cold portion of the resonator is rejected to the environment,
then from the energy balance, we have,

Q c H_ 2;s  H_ 2;sum

17

where H_ 2;sum is the acoustic power entering the control volume


which is the sum of the acoustic power to be dissipated by the cold
heat exchanger, stack, ambient heat exchanger and ambient duct.
Neglecting the acoustic power dissipated in the ambient duct, and
dening DE_ 2;t as the acoustic power dissipated/consumed in the
cold heat exchanger, stack, ambient heat exchanger, we have,

Q c H_ 2;s  DE_ 2;t

18

This equation implies that the total energy ux of the stack is the
cooling power of the system and the acoustic power consumed by
the stack and the heat exchangers. If the cold duct is insulated,
DE_ 2;cd is not rejected to the environment through the cold duct
walls, and it appears as a load on the cold heat exchanger. Thus,

H_ 2;sum DE_ 2;t DE_ 2;cd

19

and,

Q c H_ 2;s  DE_ 2;t DE_ 2;cd

20

The ratio of the cooling power of a thermoacoustic refrigerator to


the consumed acoustic power by the stack is dened as the coefcient of performance of the refrigerator stack.

COPs

Cooling power Normalized cooling power

DE_ 2;s;ref
DE_ 2n;s;ref

21

where cooling power is dened in Eqs. (15), (16), (18) and (20).
Considering a thermoacoustically-driven thermoacoustic refrigerator, the acoustic power produced by the thermoacoustic engine
must be consumed by the thermoacoustic refrigerator and the resonator. That is,

DE_ 2;s;eng  DE_ 2;HXh DE_ 2;HXa;eng


DE_ 2;s;ref DE_ 2;HX DE_ 2;HX
c

a;ref

DE_ 2;r

22

The left hand side is the net acoustic power output of the thermoacoustic engine. That is, the total acoustic power available for the
refrigeration purpose. The rst three terms on the right hand side
are the total acoustic power consumed by the refrigerator stack
and its exchangers and the last term is the dissipated acoustic
power in the resonator. The above equation could be summarized
as,

DE_ 2;t;eng DE_ 2;t;ref DE_ 2;r

23

Thus,

DE_ 2;pro DE_ 2;con

24

3591

where DE_ 2;pro is the acoustic power produced in the engine and
DE_ 2;con is the acoustic power consumed in the refrigerator and resonator. In the normalized form, the above equation can be expressed as,

DE_ 2n;pro  APm aeng DE_ 2n;con  APm aref

25

Parameters A and Pm are the same for both engine and refrigerator
whereas, the speed of sound is different in both stacks due to the
difference in the mean temperatures of the stacks. Eq. (25) can further be simplied as,

DE_ 2n;pro


aref
DE_ 2n;con
aeng

26

3.3. Entropy balance


To determine the entropy generation within a thermoacoustic
device, the entropy balance and energy balance are applied on
two control volumes. The rst control volume (system I) is the
acoustic power producing system which consists of the hot heat
exchanger, engine stack and the engine ambient heat exchanger,
as outlined in Fig. 5a with the dashed line. The second control volume (system II) is the acoustic power consuming system which
consists of the resonator, cold heat exchanger, refrigerator stack
and ambient heat exchanger, as outlined in Fig. 5b by the dashed
line. Since the entropy change of a steady state control volume is
zero, the second law of thermodynamics indicates that the entropy
leaving the control volume must equal the sum of the entropy
entering the control volume and the entropy generation within
the control volume. It is useful to mention that there is no entropy
associated with energy transfer as work [20].
Following equations show the energy balance on systems I and
II, respectively.

Q h Q a;eng DE_ 2;pro


Q c DE_ 2;con Q a;ref Q r

27
28

where Qr represents the amount of dissipated acoustic power in the


resonator leaving through the resonators wall in the form of heat
energy at the ambient temperature.
Following equations show the entropy balance on systems I and
II, respectively.

Q a;eng Q h _

Sgen;eng
Ta
Th
Q a;ref Q r Q c _

Sgen;r;ref
Ta
Ta
Tc

29
30

By substituting the values of Qa,eng and Qa,ref from Eqs. (27) and (28)
into Eqs. (29) and (30), respectively, we get,



Ta
T a S_ gen;eng Q h 1 
 DE_ 2;pro
Th


Ta
DE_ 2;con
T a S_ gen;r;ref Q c 1 
Tc

31
32

Fig. 5. Two thermodynamic systems outlined by dashed lines, (a) system I, acoustic power producing system (engine), (b) system II, acoustic power consuming system
(resonator and refrigerator).

3592

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

Considering a thermoacoustically-driven thermoacoustic refrigerator (TADTAR), the total entropy generation within the TADTAR
could be estimated by adding the entropy generation in the two
systems (since the two systems form a TADTAR). Thus,

T a  S_ gen;t T a  S_ gen;eng T a  S_ gen;r;ref


 



Ta
Ta
Qh 1 
Qc 1 
Th
Tc
DE_ 2;con  DE_ 2;pro 

By plotting COPs and COPRs as functions of xcn,ref, Lsn,ref, DTn,ref,


the normalized refrigerator stack length and position at the desired
normalized temperature difference is selected. The resonator
cross-sectional area (and then the resonator hydraulic radius)
can be determined by,

Ar
33

To satisfy Eq. (24) which implies that the produced acoustic power
by the engine stack must be equal to the consumed acoustic power
by the other components of the device, the engine characteristics
must be selected to make the last term on the right hand side of
Eq. (33) equal to zero. Thus, the rst term on the right hand side
of Eq. (33) shows the total entropy generation within the TADTAR.
To have the maximum efciency, the engine characteristics must be
selected to make the entropy generation minimum. For an acoustically-driven thermoacoustic refrigerator (ADTAR), the total entropy
generation could be determined by applying Eq. (32).
In the following subsections, the developed design and optimization algorithm is explained in detail and, an example is presented as a case study to demonstrate the functioning of the
algorithm. Some important issues related to the designing are also
discussed.
3.4. Design and optimization algorithm
The algorithm proposed by Wetzel and Herman [13] to design
acoustically-driven thermoacoustic refrigerators was used as the
base model to develop a new systematic comprehensive algorithm
in this study. The developed algorithm can be used to design and
optimize not only thermoacoustically-driven thermoacoustic
refrigerator but also individual thermoacoustic engines and
acoustically-driven thermoacoustic refrigerators. Furthermore,
the present algorithm includes new features for designing an
acoustically-driven thermoacoustic refrigerator by incorporating
correlations between different design parameters based on the
energy balance, that were not available in the previous studies.
The complete design algorithm is shown in Fig. 6. As mentioned
earlier, meeting the required cooling power at the desired cooling
temperature and at the given hot heat temperature while rejecting
some heat to the environment can be dened as the goals of designing a thermoacoustically-driven thermoacoustic refrigerator.
The design procedure starts with the refrigerator section of the
device followed by the resonator and the heat engine. As a rst
step, the designer must pick the working gas, blockage ratio, thermal penetration depth and drive ratio of the device (similar to
previous studies). Using the values of the given cooling temperature (i.e. the temperature of HXc), heat input temperature (i.e.
the temperature of HXh) and, the surrounding ambient temperature (i.e. the temperature of HXa), the mean temperatures of the
refrigerator and engine stacks can be calculated. The thermophysical properties of the working gas are then computed in the refrigerator section of the device (and the resonator) and engine section
of the device based on the refrigerator and engine stack mean temperatures, respectively.
As mentioned above, the designing process starts with the
refrigerator. In the energy balance section, two congurations are
presented for a thermoacoustic refrigerator (Fig. 4a and b). For each
conguration, two conditions are presented, i.e. cold duct insulated
and uninsulated. The relationship between cooling power, total energy ux to the stack and the acoustic power for all cases are also
presented (see Eqs. (15), (16), (18) and (20)). The designer must
select the desired conguration and apply the appropriate energy
balance on HXc.

Qc
Q cn  Pm  a  BR

34

In the next step, the normalized acoustic power dissipated in the


refrigerator stack DE_ 2n;s;ref is determined by applying Eq. (8). This
equation is also used to determine the normalized acoustic power
dissipated in heat exchangers DE_ 2n;HXc DE_ 2n;HXa;ref by substituting
the normalized heat exchanger length and position in Eq. (8) and
assuming no temperature gradient along the exchangers (i.e. along
the direction of acoustic wave propagation). The value DE_ 2n;t;ref
which is the sum of the above mentioned values can then be
determined.
To design the resonator, select the resonance frequency of the
thermoacoustic device. Compute the thermophysical properties
of the working gas in the resonator based on stack mean temperature of the refrigerator. Calculate the thermal penetration depth at
the resonators wall. The length of the resonator can be set equal to
half wavelength or quarter wavelength of the resonant standing
wave in the device. Determine the normalized acoustic power dissipated in the resonator DE_ 2n;r by using Eq. (9).
The resonance frequency of the acoustic standing wave is an
important design parameter. Although it has been selected at this
stage, it could be modied afterwards if necessary. Higher resonance frequency results in lower penetration depth i.e. small plate
spacing in the stack which increases the manufacturing challenge,
however, higher resonance frequency increases the power density
and reduces the length of the resonator. In the next step, the total
consumed acoustic power DE_ 2n;con can be computed by summing
the dissipated acoustic power in the refrigerator stack, its heat
exchangers DE_ 2n;t;ref and the resonator DE_ 2n;r .
The purpose of the heat engine is to produce the acoustic power
that is consumed by the device. Once the total consumed acoustic
power is computed DE_ 2n;con , the acoustic power to be produced by
the engine DE_ 2n;pro can be estimated by using Eq. (26). This
parameter serves as a basis to design the heat engine which can
meet the given requirements.
To design the engine stack of the device, estimate the stack efciency based on the energy balance applied on the hot heat exchanger using Eq. (12). The results from this energy balance
would also be used in estimating the heat input to the engine. In
the next step, the length and position of the engine stack must
be selected that could produce the estimated acoustic power at
the desired heat exchangers temperature while having the maximum possible efciency. The appropriate length and position for
the engine stack can be selected by plotting the engine thermal
efciency (gth), the engine normalized thermal efciency (gthn)
and DE_ 2n;pro as functions of xcn,eng, Lsn,eng, D Tn,eng. At this stage,
the entropy balance described in the previous section is applied
on the device to evaluate the overall entropy generation within
the thermoacoustic device (Eq. (33)). This analysis is useful to
examine the variation of the generated entropy as a function of
xcn,eng, Lsn,eng, so the engine stack position and length are selected
to minimize the entropy generation while providing the required
acoustic power at the given temperatures of the heat exchangers.
The point of the minimum entropy generation is the same as the
point of the maximum efciency of the device.
After selecting the optimized engine stack length and position,
estimate the amount of heat input to the hot heat exchanger to
produce the estimated acoustic power at the desired temperatures.
The heat input can be computed by,

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

3593

Fig. 6. Schematic of developed design and optimization algorithm for thermoacoustically-driven thermoacoustic refrigerators.

Q h Q hn  Ar  BR  Pm  aeng

35

In the last step, the plate thickness and plate spacing of engine and
refrigerator stacks are estimated.
As mentioned earlier, the present algorithm can also be used to
design an acoustically-driven thermoacoustic refrigerator. In this
case, the same procedure and steps are used to calculate the total
acoustic power consumed by the refrigerator stack, heat exchangers and the resonator DE_ 2n;con . A loud speaker is then selected
based on this total acoustic power to run the apparatus.
It is useful to mention that this study is more comprehensive
compared to previous studies for designing and optimizing thermoacoustic refrigerators; since, it uses the cooling power and energy

ux relations for different congurations of the thermoacoustic


refrigerator.
3.5. Case study
A case study is presented to demonstrate the working of the
developed procedure. The case study comprised of designing a
TADTAR with 30 W of cooling power at the desired cooling temperature of 277 K and the desired hot heat exchanger temperature of
623 K. The ambient heat exchangers are assumed to operate at
300 K. Helium at a mean pressure of 700 kPa is selected as the
working gas, which is one of the recommended gases for thermoa-

3594

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

coustic devices [21]. The other parameters required at the beginning of the designing procedure are set as follows; BR = 0.8,
dkn = 0.66 and DR = 0.03, which are consistent with the previous
studies [17,22]. Based on the given temperatures for hot, cold
and ambient heat exchangers, the mean temperature of the refrigerator stack is 288.5 K (DTn,ref = 0.08) and the mean temperature of
the engine stack is 461.5 K (DTn,eng = 0.7).
As mentioned in the previous section, the design procedure
starts with the designing of the refrigerator section. The second
conguration of the refrigerator is selected with no insulation on
the cold duct. That is, the refrigerator stack is located near the right
pressure antinode of the resonator (see Fig. 4b), and the dissipated
acoustic power in the resonator is rejected to the environment
through the resonators wall.
The next step is the selection of the normalized refrigerator
stack length and position by plotting COPs and COPRs as functions
of xcn,ref, Lsn,ref at the desired normalized temperature difference.
Fig. 7a and b show the variation of COPs and COPRs as the function
of Lsn,ref at different xcn,ref and DTn,ref = 0.08. The plots show that by
shifting the stack center away from the pressure antinode (i.e.
increasing xcn,ref), the stack length must be increased to have the
performance peak. However, the peak magnitude decreases with
increasing xcn,ref. By selecting the length of the stack to have the
performance peak, two problems arise. First, as the gures show,
the COPs values are very sensitive to the stack length near the peak.
A slightly smaller stack length causes a sharp decrease in the performance of the refrigerator. Second, the apparatus does not perform as a refrigerator at higher values of the normalized
temperature differences (discussed in a later section). Taking into
considerations these issues, it is decided to assume xcn,ref = 0.11
and Lsn,ref = 0.035 (COPs = 4.47, COPRs = 0.37). The cross-sectional
area of the resonator is computed by using Eq. (34), which for
the present case is equal to 0.0123 m2.
The variation of the normalized consumed acoustic power
DE_ 2n;con is plotted versus the resonance frequency at the selected
specications of the refrigerator stack in Fig. 8. The values of
DE_ 2n;con are computed using Eqs. (8) and (9), as described in the
previous section. The plot shows that at a given refrigerator stack
temperature difference, the consumed acoustic power decreases
by increasing the resonance frequency. As the resonance frequency
increases, the viscous and thermal penetration depths decreases
causing the acoustic power dissipated in the resonator to decrease.
Thus, it is desirable to have a thermoacoustic device operating at

Fig. 8. Normalized consumed acoustic power DE_ 2n;con versus the resonance
frequency (f) at xcn,ref = 0.11, Lsn,ref = 0.035 and DTn,ref = 0.08.

higher resonance frequency. However, higher frequency results


in lower peak-to-peak displacement of the gas particles which
could affect the performance of the heat exchangers [2].
In the present study, the resonance frequency of 400 Hz is selected. For this frequency, the length of the resonator based on
the half wavelength of the acoustic standing wave is estimated
to be 1.25 m. The normalized consumed acoustic power and normalized acoustic power to be produced by the engine are estimated to be DE_ 2n;con 3:2  106 and DE_ 2n;pro 2:55  106 ,
respectively.
The engine stack is considered to be located near the left pressure antinode. In the next step, the length and position of the engine stack is selected for the given normalized temperature
difference. The selection is done from the graphs of gth,s, gthn,s
and DE_ 2n;pro versus the normalized engine stack length (Lsn,eng) at
different values of xcn,eng for the given DTn,eng, that are obtained
by applying the energy balance on the hot heat exchanger. Fig. 9
shows the variations of gth,s, gthn,s and DE_ 2n;pro versus Lsn,eng at

Fig. 7. Normalized length of the refrigerator stack (Lsn,ref) versus (a) coefcient of performance of refrigerator stack (COPs), (b) coefcient of performance relative to Carnot
cycle (COPRs), at different normalized stack center positions (xcn,ref) at DTn,ref = 0.08.

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

3595

short stack close to the pressure antinode to long stack away from
the pressure antinode. Based on the combination of Lsn,eng and
xcn,eng that can produce the required acoustic power, the corresponding values of gth,s and gthn,s can be estimated from Fig. 9b
and c, respectively, to evaluate the stack performance. Fig. 9 also
shows that for a specied stack center position there is not more
than one stack length that could produce the required acoustic
power at the desired normalized temperature difference. The results show that at the combination xcn,eng = 0.14 and Lsn,eng = 0.09,
the stack performance is the best i.e., gth,s = 18% and gthn,s = 0.347.
Fig. 9a shows that certain combinations of Lsn,eng and xcn,eng cannot
produce the required acoustic power. It also shows that below a
certain value of xcn,eng, there is no length of stack that could produce the required acoustic power, which in the present case is
xcn,eng 6 0.06. Thus, for a specied thermoacoustic refrigerator,
there could be an engine stack with specied position and length
that produces the required acoustic power with the maximum possible efciency. Thus, the device must generate the least entropy at
these specications.
To check if the total entropy generation S_ gen;t in TADTAR is
minimum, S_ gen;t is computed using Eq. (33) at the selected combinations of xcn,eng and Lsn,eng at which the required DE_ 2n;pro is obtained. The S_ gen;t is plotted as a function of xcn,eng in Fig. 10. For
xcn,eng 6 0.06, there is no engine stack length that could produce
the required acoustic power of the system. Fig. 10 shows that the
device would generate minimum entropy by placing the engine
stack at xcn,eng = 0.14 with the corresponding value of Lsn,eng = 0.09,
which conrms the above mentioned discussion. Thus, it can be
concluded that while selecting the position and length for the engine stack that produces the required acoustic power, the designer
should conrm that the entropy generation of the device is minimum at the selected specications.
In the nal step, the amount of heat input to the hot heat exchanger (Qh) to produce the required acoustic power at the desired
temperatures is estimated using Eq. (35). For the present TADTAR,
Qh = 162.2 W. Thus, for the thermoacoustically-driven thermoacoustic refrigerator to produce the cooling power of 30 W,
162.2 W of heat input is required.
The plate thickness and plate spacing of the refrigerator stack
are estimated to be approximately 0.10 mm and 0.42 mm, respectively. The plate thickness and plate spacing of the engine stack are
estimated to be about 0.16 and 0.63 mm, respectively.

Fig. 9. Normalized length of the engine stack (Lsn,eng) versus (a) normalized acoustic
power produced DE_ 2n;pro , (b) thermal efciency of engine stack (gth,s), (c)
normalized thermal efciency of engine stack (gthn,s), at different normalized stack
center positions (xcn,eng) at DTn,eng = 0.7.

different xcn,eng and at DTn,eng = 0.7. Fig. 9a shows that at a given


stack position, the acoustic power produced by the engine decreases with an increase in the stack length, whereas, as a given
stack length, the acoustic power produced by the engine increases
as the stack moves away from the pressure antinode. For the estimated value DE_ 2n;pro 2:55  106 in the present study, several
combinations of the stack length and position are available from

Fig. 10. Total entropy generation in the device S_ gen;t versus normalized stack
center positions (xcn,eng).

3596

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

3.6. Effects of stack temperature difference (DTn) on the design of a


thermoacoustic device
As described in the design algorithm, the designer must select
the temperatures of all heat exchangers (i.e. stack temperature difference) at the beginning of the design procedure. The equations
show that the stack temperature difference (DTn) is an important
design parameter and has a signicant inuence on the performance of the respective engine or refrigerator. Therefore, it is
important for a designer to have a good understanding of the inuence of DTn on the performance of the device. Since the design procedure is based on selected values of DTn for engine and
refrigerator, the effect of DTn on their performance cannot be evaluated in the previous sections. In this section, the effects of DTn on
the overall design of the device are discussed in detail.
The impact of the refrigerator stack temperature difference
(DTn,ref) on the performance of the refrigerator stack is illustrated
in Fig. 11a and b, where the variations of COPs and COPRs are plotted as a function of Lsn,ref at different DTn,ref and at a given value of
xcn,ref. The gure shows that the peak performance of the refrigerator stack reduces by increasing the stack temperature difference.
The trends in the given gure also indicate that for a given stack
temperature difference, the COP drops to zero if the length of the
stack is lower than a certain value. That is, for a stack to operate
as a refrigerator, the length of the stack should be higher than a
cutoff value (also see Eq. (6)). The gure shows that the cutoff value of the stack length increases with decreasing the stack temperature difference. In the case study, at DTn,ref = 0.08, the best
performance of the refrigerator stack is at Lsn,ref = 0.025. However,
if the temperature difference is increased, the stack may not perform as a refrigerator. This could happen when developing the actual device as the actual stack temperature difference may vary
from its designed value. Therefore, it is safer to select the stack
length slightly larger than that correspond to the peak performance. Therefore, in the case study the length of the refrigerator
stack was selected as Lsn,ref = 0.035.
The inuence of stack temperature difference on the resonance
frequency is shown in Fig. 12, where the variation of the normalized consumed acoustic power DE_ 2n;con is plotted versus the normalized refrigerator stack temperature difference (DTn,ref) at
different resonance frequencies at xcn,ref = 0.11 and Lsn,ref = 0.035.
At a given resonance frequency, the consumed acoustic power decreases with an increase in the stack temperature difference. As the

Fig. 12. Normalized consumed acoustic power DE_ 2n;con versus normalized refrigerator stack temperature difference (DTn,ref) at different resonance frequencies, at
xcn,ref = 0.11, Lsn,ref = 0.035.

temperature difference along the refrigerator stack increases, the


thermal penetration depth decreases, causing a reduction in the
acoustic power consumed in the stack.
The inuence of the engine stack temperature difference on the
performance of the engine stack is shown in Fig. 13. In this gure,
the variations of gth,s, gthn,s and DE_ 2n;pro are plotted versus the normalized engine stack length (Lsn,eng) at different values of D Tn,eng.
This gure shows that the peak efciency of the engine stack increases by increasing the normalized temperature difference along
the engine stack. The trends in Fig. 13 also indicate that for a given
stack temperature difference, the stack efciency drops to zero if
the length of the stack is greater than a certain value. That is, for
a stack to operate as an engine, the length of the stack should be
lower than a cutoff value (also see Eq. (6)). The gure shows that
the cutoff value of the stack length decreases with decreasing the
stack temperature difference.
The inuence of stack temperature difference for engine and
refrigerator on the heat input to the device is shown in Fig. 14.
The heat input is plotted as a function of normalized engine stack

Fig. 11. Normalized length of the refrigerator stack (Lsn,ref) versus (a) coefcient of performance of refrigerator stack (COPs), (b) coefcient of performance relative to Carnot
cycle (COPRs), at different values of normalized refrigerator stack temperature difference (DTn,ref) at xcn,ref = 0.11.

3597

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

Fig. 14. Heat input (Qh) to the device versus normalized engine stack temperature
difference (DTn,eng) for different values of normalized refrigerator stack temperature
difference (DTn,ref).

4. DeltaE
The computer code DeltaE can be used to simulate the devices
designed and optimized by the procedure presented in this study.
DeltaE solves the one-dimensional wave equation in gas or liquid,
based on the low amplitude acoustic approximation in user dened geometries [23]. The desired parameters initially selected
and the parameters computed by using the design and optimization algorithm developed in this study are summarized in Table
2. Also presented in the table are the values obtained from DeltaE
for comparison. A good agreement is observed between the developed procedure and the computer code DeltaE. Although the
parameters calculated by the computer code DeltaE are slightly
different from those of estimated by the developed procedure, it
is reasonable to say that the developed procedure can serve as a
great tool to design and optimize thermoacoustic devices since
designing a TADTAR by using the computer code DeltaE to meet
the designers requirements requires tremendous numbers of trials
and errors making this job tedious. The small differences between
the two approaches are mainly due to the assumptions that were
made to linearize and simplify the governing equations to develop
the design and optimization procedure. The inaccurate expression
used to estimate the temperature difference between the metal
and working gas in the heat exchangers in the computer code
DeltaE could be another reason for the deviations [23]. One or

Table 2
Comparison between results from present algorithm and DeltaE simulations

Fig. 13. Normalized length of the engine stack (Lsn,eng) versus (a) thermal efciency
of engine stack (gth,s), (b) normalized thermal efciency of engine stack (gthn,s), (c)
normalized acoustic power produced DE_ 2n;pro , at different normalized engine stack
temperature difference (DTn,eng) at xcn,eng = 0.09.

temperature difference at different normalized refrigerator stack


temperature difference at the specied engine and refrigerator
stack positions and lengths. The gure shows that by increasing
DTn,eng or DTn,ref or both, the required heat input to the device
increases.

f
T HXh
T HXa;eng
DTn,eng
T HXc
T HXa;ref
DTn,ref
Qh
Qc
DE_ 2;s;eng
DE_
2;s;ref

gth,s (%)
COPs
Overall efciency (%)

Present algorithm

DetlaE

400
623
300
0.7
277
300
0.08
164.7
30
29.7
6.72
18
4.47
18.5

402.7
630
300.5
0.708
277
303.3
0.09
166.1
30
32.9
6.9
19.8
4.3
18.1

3598

H. Babaei, K. Siddiqui / Energy Conversion and Management 49 (2008) 35853598

more parameters such as stack center position, stack length or


resonator cross-sectional area can be adjusted to meet the desired
values in DeltaE.
5. Conclusion
Thermoacoustic devices operate by the energy conversion
between heat and sound, and have no harmful effects on the environment. The designing of thermoacoustic devices involves significant technical challenges. In the present study, a comprehensive
design and optimization algorithm is developed for designing thermoacoustic devices. The unique feature of the present algorithm is
its ability to design thermoacoustically-driven thermoacoustic
refrigerators that can serve as sustainable refrigeration systems.
In addition, new features based on the energy balance are also included to design individual thermoacoustic engines and acoustically-driven thermoacoustic refrigerators. The algorithm is based
on the simplied linear thermoacoustic model. It includes different
correlations based on the energy balance for different device congurations. Another important feature of the algorithm is the
implementation of the entropy balance on the device to rene
the optimization process. A step-by-step design and optimization
procedure is described which is followed by a case study in
which a thermoacoustically-driven thermoacoustic refrigerator is
designed and optimized to demonstrate the working of the algorithm. The results from the algorithm are in good agreement with
that obtained from the computer code DeltaE.
Acknowledgement
This research is funded by a grant from the Concordia University to Kamran Siddiqui.
References
[1] Rott N. Thermoacoustics. Adv Appl Mech 1980;20:13574.
[2] Swift GW. Thermoacoustic engines. J Acoust Soc Am 1988;84:114579.

[3] Garret SL, Adeff JA, Hoer TJ. Thermoacoustic refrigerator for space application.
J Thermophys Heat Transfer 1993;7:595.
[4] Hoer TJ. Thermoacoustic refrigerator design and performance. Dissertation
University of California, San Diego; 1986.
[5] Tijani MEH, Zeegers JCH, De Waele ATAM. Construction and performance of a
thermoacoustic refrigerator. Cryogenics 2002;42:5966.
[6] Symko OG, Abdel-Rahman E, Kwon YS, Emmi M, Behunin R. Design and
development of high-frequency thermoacoustic engines for thermal
management of microelectronics. Microelectron J 2004;35:18591.
[7] Hatazawa M, Sugita H, Ogawa T, Seo Y. Performance of a thermoacoustic sound
wave generator driven with waste heat of automobile gasoline engine. Trans
Jpn Soc Mech Eng 2004;70:2929.
[8] Adeff JA, Hoer TJ. Design and construction of a solar powered
thermoacoustically driven thermoacoustic refrigerator. J Acoust Soc Am
2000;107:L3742.
[9] Babaei H, Siddiqui K, Chishty WA. Sustainable thermoacoustic refrigeration
system for gas turbine power plants. In: 17th symposium of industrial
application of gas turbines (IGAT); Banff, Canada; 2007.
[10] http://www.ecn.nl/leadmin/ecn/units/eei/Onderzoeksclusters/
Restwarmtebenutting/b-07-007.pdf.
[11] Johnson VH. Heat generated cooling opportunities in vehicles. Society of
Automotive Engineers; 2002.
[12] Zoontjens L, Howard C, Zander A, Cazzolate B. Feasibility study of an
automotive thermoacoustic refrigerator. In: Proceedings of acoustics.
Busselton, Australia, November 911; 2005.
[13] Wetzel M, Herman C. Design optimization of thermoacoustic refrigerators. Int
J Refrig 1997;20:321.
[14] Tijani MEH, Zeegers JCH, De Waele ATAM. Design of thermoacoustic
refrigerators. Cryogenics 2002;42:4957.
[15] Swift GW. Thermoacoustics: a unifying perspective for some engines and
refrigerators. The Acoustical Society of America, NY: Melville; 2002.
[16] Olsen JR, Swift GW. Similitude in thermoacoustics. J Acoust Soc Am
1994;95:140512.
[17] Worlikar A, Knio O. Numerical simulation of a thermoacoustic refrigerator. J
Comput Phys 1996;127:42451.
[18] Sergeev SI. Fluid oscillations in pipes at moderate Reynolds number. Fluid Dyn
1966;1:1212.
[19] Merkli P, Thomann H. Transition to turbulence in oscillating pipe ow. J Fluid
Mech 1975;68:56776.
[20] Cengel YA, Boles MA. Thermodynamics: an engineering approach. Fourth
edition in SI units. McGraw Hill; 2002.
[21] Belcher JR, Slaton WV, Raspet R, Bass HE, Lightfoot J. Working gases in
thermoacoustic engines. J Acoust Soc Am 1999;105:267784.
[22] Tijani MEH, Zeegers JCH, De Waele ATAM. The optimal stack spacing for
thermoacoustic refrigeration. J Acoust Soc Am 2002;112:12833.
[23] Ward WC, Swift GW. Design environment for low-amplitude thermoacoustic
engine. J Acoust Soc Am 1994;95:36714.

Das könnte Ihnen auch gefallen