Professional Documents
Culture Documents
Optimizing The Performance of A Standing Wave Loudspeaker Driven Thermoacoustic Heat Pump
Optimizing The Performance of A Standing Wave Loudspeaker Driven Thermoacoustic Heat Pump
ISSN 2229-5518
Abstract— This paper investigates the design and optimization steps of a thermoacoustic refrigerator. Matlab code will be used for
optimizing the stack length and its position. DeltaEC version 6.3b11 will be used to do the code used for simulating the refrigerator to
identify the optimized operating conditions such as the mean pressure and the oscillating pressure. Behavior of changing the operating
conditions effect on the performance is discussed.
—————————— ——————————
NOMENCLATURE
A Resonator area, [m^2] Greek Letters
a Sound velocity, [m/s] β Thermal expansion coefficient, [K^-1]
B Porosity, … γ Specific heats ratio, …
D Driver ratio, … δk Thermal penetration depth, [mm]
d Resonator diameter, [mm] δv Viscous penetration depth, [mm]
e Specific internal energy, [kJ/kg] λ Wave length, [m]
f Frequency, [Hz] µ Dynamic viscosity, [Pa. s]
IJSER
h Specific Enthalpy, [kJ/kg] ρ Density, [kg/m^3]
K Thermal conductivity, [W/(m.K)] θ Phase angle, [degree]
L Length, [m] ω Angular frequency, [rad/s]
P Pressure, [Pa] ∆T Temperature Difference, [K]
Pr Prandtl number, … Subscripts
Q Thermal power, [W] 0 Spacing
T Temperature, [K] 1 Amplitude
U Volume flow rate, [m^3/s] k Thermal
v Velocity, [m/s] m Mean
W Acoustic power, [W] n Dimensionless
X Stack position, [mm] s Stack
y The Half stack spacing, [mm] v Viscous
1 INTRODUCTION
IJSER
C.O.P
AC No
Amplifier
Yes
2
δ kn D sin (2Xsn )[ ΔTmn tan (Xsn ) 1 Pr Pr
Q
End
cn
8 A (1 Pr ) Lsn B (1 γ) 1 Pr
(1)
(1 Pr Pr δ 2kn )]
Fig. (2) Design steps for thermoacoustic refrigerator
2
W δk nD Lsn ( -1 )Bco s(X )2 [ ΔTmn tan (Xsn ) 1]
n
4
sn
Lsn A B (1 Pr ) Equation (1) and (2) are used through a matlab code to op-
δ k nD2 sin (Xsn )2 Pr timize the stack parameters. Fig. (3) shows the effect of dimen-
(2) sionless stack length and dimensionless stack position on the
4 A B (1 Pr )
stack performance. From this figure; the value of the normal-
The coefficient of performance C.O.P is the ratio between ized stack center position is set to Xsn = 0.4. This is because a
the gained cooling power and the comsumed work: greater value will lead to difficulties in fabricating, so when
Q cn
Xsn = 0.4 is chosen, the dimensionless stack length, Lsn should
C .O . P (3) be about 0.13 to give stack performance of 3.5702.
W n
IJSER © 2016
http://www.ijser.org
462
International Journal of Scientific & Engineering Research Volume 7, Issue 9, Septemper-2016
ISSN 2229-5518
12 5 RESULTS
5.1 Mean Pressure Effect
xn = 0.1
10 xn = 0.15 Increasing the mean pressure will decrease the tempera-
xn = 0.2 ture difference as shown in fig. (4).
xn = 0.25
8 xn = 0.3 The driver ratio (D) is the ratio between the amplitude
xn = 0.35
pressure and the mean pressure. The acoustic energy is direct-
C.O.P)s
xn = 0.4
6 xn = 0.45 ly proportional to drive ratio as in (2), so increasing mean
pressure will decrease the driver ratio and this leads to de-
4 creasing acoustic power and thus increasing C.O.P as shown
in fig. (5). The larger temperature difference is desired, but the
coefficient of performance should be optimized.
2
IJSER
the help of the free simulation software DeltaEC version
6.3b11 and using the default DeltaEC convergence tolerance of
10^-8 to do the code used for simulating the design. For Del-
taEC program: the pressure equation is concluded from the
momentum equation of fluid mechanics and the flow rate
equation is concluded from the continuity equation of fluid
mechanics. The equations are integrated numerically along the
x coordinate to form solutions P1(x) and U1(x). DeltaEC has
more complicated momentum and continuity equations that
include additional effects as the vicous and thermal losses re-
sulting from the dissipation of acoustic power along the sides
of ducts. It uses different equations for the different segments
to fit local conditions.
The study considers a constant cooling load, and calculate
the incoming acoustic power that varies with changing the
operating conditions. The boundary conditions in our theoret- Fig. (4) Mean Pressure effect on the temperature difference
ical model are:
θ (P) = θ (U) = 0 at the beginning segment to enforce
the resonance frequency to be obtained [2] [3].
Pressure amplitude, P1 = 2 kPa.
-0.01
Re ( U )
IJSER
Im ( U ) / 10
-0.015
2000
total power only changes with x in heat exchanger segments
1000 as shown in fig. (11). and it is constant for the regions at the
0 regions where there are no heat exchangers.
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
-1000
5.7 Temperature distribution through the resonator
-2000
Resonator Length (m) As shown in fig. (12); the resonator can be classified into
-3000
three regions. The ambient duct is the first region (the duct
-4000 before the hot heat exchanger); the cold duct (the duct after the
-5000 cold heat exchanger) is the second region and the third region
is the stack where the temperature gradient takes place.
Fig. (8) Amplitude pressure distribution through the resonator
IJSER © 2016
http://www.ijser.org
464
International Journal of Scientific & Engineering Research Volume 7, Issue 9, Septemper-2016
ISSN 2229-5518
Temperature (K)
refrigerator. Mean pressure of 2 bar, amplitude pressure of
2500 Pa and the four guesses and targets in sec. 4 are used to 290
for helium than air. This demonstrates the high temperature 265
differences for helium as the Prandtl number is proportional 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
Resonator Length (m)
directly with the viscous penetration depth and reversely to
the thermal penetration depth. The gas particle velocity in case
of helium is higher than air particles velocity and this draws Fig. (12) Temperature distribution through the resonator
more input acoustic power for helium which will have lower
performance. Helium is preferred in low temperatures but this Lsn = 0.22, X sn = 0.3, B = 0.75
60
will lead to high power consumption. A stack spacing to
55
thermal penetration value of 4 is optimum for both the tem-
50
perature difference and the temperature difference for air and
IJSER
40 Ai r
35 Hel ium
6 Lsn = 0.13, Xsn = 0.4, B = 0.75 at the resonance frequency
5.5 30
5 25
4.5 20
4
Acoustic Power (W)
15
3.5
10
3
5
2.5
2
0
0 0.6 1.2 1.8 2.4 3 3.6 4.2 4.8 5.4 6
1.5
2y0/ δ k
1
0.5
0 Fig. (13) Temperature difference for air and helium versus
0 0.1 0.2 0.3 0.4 0.5 0.6
Resonator Length (m)
0.7 0.8 0.9 1 1.1 stack spacing to thermal penetration depth
1.05
2
C.O.P
0.9 Helium
1
0.75
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
0.6
-1
0.45
Resonator Length (m)
-2 0.3
-3 0.15
-4 0
0 0.6 1.2 1.8 2.4 3 3.6 4.2 4.8 5.4 6
-5
2y0/ δk
-6
IJSER
30 [8] Namdar, A., Kianifar, A., & Roohi, E. (2015). Numerical investigation
of thermoacoustic refrigerator at weak and large amplitudes consid-
20 ering cooling effect. Cryogenics, 67, 36-44.
[9] Piccolo, A., & Pistone, G. (2006). Estimation of heat transfer coeffi-
10 cients in oscillating flows: The thermoacoustic case. International
Journal of Heat and Mass Transfer, 49(9), 1631-1642.
0
0 1 2 3 4 5
Cool i ng load (W)
6 CONCLUSION
A thermoacoustic refrigerator is designed and then it is
simulated through DeltaEC to get optimized operating condi-
tion. The following conclusions can be drawn:
The temperature difference should be put into considera-
tion when optimizing the performance.
Increasing the mean pressure will increase the perfor-
mance for the thermoacoustic refrigerator.
Increasing the amplitude pressure will decrease the per-
formance for the thermoacoustic refrigerator.
The stack is the place where the temperature gradient
takes place and the power is consumed through it in
adition to the thermal and viscous losses at the resonator
walls.
Air gives higher performance coefficients than helium
with lower temperature differences across the stack.
IJSER © 2016
http://www.ijser.org