Applied Thermal Engineering 29 (2009) 210–215

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Performance of a prototype Stirling domestic refrigerator

Sun Le’an *, Zhao Yuanyang, Li Liansheng, Shu Pengcheng
National Engineering Research Center of Fluid Machinery and Compressors, Xi’an Jiaotong University, Xi’an 710049, PR China

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

Article history: The V-type integral Stirling refrigerator (VISR) consisting of an expansion cylinder, a compression cylin-
Received 22 June 2007 der and a heat exchanger in-between was developed and tested in this research. The expansion and com-
Accepted 16 February 2008 pression pistons are V-shaped and driven by a crank-shift mechanism. A cold head connected to the
Available online 18 March 2008
expansion piston provides the cooling capacity. The refrigerator has a semi-hermetic configuration.
The parameters such as the power consumption and the coefficient of performance (COP) are investigated
Keywords: under various rotating speeds and charged pressures. The results would help us for the optimal design
Stirling cycle
and operation of the V-type integral Stirling refrigerator, when it is applied to the domestic refrigeration
Simulation
Performance
system. The characteristics curves of the prototype VISR could be applied to other usages.
V-type integral Stirling refrigerator Ó 2008 Published by Elsevier Ltd.

1. Introduction case of the STFSR, the displacer is driven by the fluid force or gas-
eous pressure which is indirectly induced by the piston motion and
In the past decades the Stirling cycle cryocoolers have been the stroke is thus variable and depends upon the operating
developed rapidly because of its high efficiency, fast cool-down, conditions.
small size, light weight, low power consumption, high reliability, So far, some analysis had been done on the V-type Stirling refrig-
and so on. And now the Stirling cycle cryocooler is widely used erator. Ataer et al. [13] analyzed the V-type Stirling cycle refrigera-
as a cooler for infrared sensors and high temperature superconduc- tor using the control volume method. Huang and Chen [14] derived
ting devices and also applied to the domestic and commercial a transfer-function for the integral-type Stirling refrigerator (ITSR).
refrigerators due to its environmental friendliness and applicabil- Walker and co-worker [15,16] simulated the performance of the
ity to the wide temperature range. Stirling cryocoolers with two-phase, two-component working fluid.
Domestic refrigeration system used Stirling cycle was first stud- They applied the V-type Stirling cycle refrigerator to the cryogenic
ied by Finkelstein and Polonski [1], and the optimum performance temperature range. The V-type Stirling cycle applicable to the
parameters for the simple Stirling cycle refrigerator and heat domestic refrigeration systems are also investigated about the ef-
pumps have also been presented as a function of operating condi- fect of the working fluid on its performance but only helium not
tions [2–7]. A free-piston Stirling cooler (FPSC) with Rankine com- nitrogen. The purpose of this work is to investigate the thermal per-
pressor has been investigated [8–10]. And a portable cooler box formance of the V-type Stirling cycle refrigerator (VISR) used he-
was discussed [11]. lium or nitrogen as a working fluid. The difference between us
The Stirling cycle is fundamentally different from the Rankine and previous researches is that our temperature range is over
cycle used in conventional refrigerators. No phase change occurs 20 to 60 °C, and the effect of nitrogen is also investigated.
in Stirling cycle. The food is kept below 30 °C for long time and In this prototype, the cold head is installed between the expan-
the conventional vapor-compression machines cannot cool-down sion port and the regenerator and the electric heater covers the
up to 20 °C. The cost for installation of the conventional one is ris- cold head. The compression and expansion pistons are driven by
ing because of necessity of the double-stage compressor for attain- a crank-shaft mechanism functioning with a fixed phase between
ing 30 °C [12]. compression and expansion processes. Using helium or nitrogen
A V-type integral Stirling refrigerator (VISR) was developed for as a working fluid, the thermodynamic process was optimized by
pursuing low cost and easy manufacture. The VISR is different from the theoretical analysis and computer simulation. And the simula-
the split-type free-displacer Stirling refrigerator (STFSR) regarding tion results were compared with experimental ones.
the driving mechanism of the displacer. The displacer of the VISR is
directly driven by a motor through a crank-shaft mechanism. In
2. Description of the prototype refrigerator

* Corresponding author. Tel.: +86 29 82663795; fax: +86 29 82668724. It is well known that the working fluid of the Stirling cycle is gas
E-mail addresses: [email protected], [email protected] (S. Le’an). and no phase-transition occurs during the cycle. The ideal Stirling

1359-4311/$ - see front matter Ó 2008 Published by Elsevier Ltd.

doi:10.1016/j.applthermaleng.2008.02.036
 S. Le’an et al. / Applied Thermal Engineering 29 (2009) 210–215 211

cycle consists of two isothermal and two isochoric heat transfer Table 1
processes where the isochoric heat transfer takes place in a regen- Design specifications of the V-type Stirling cycle refrigerator

erator. The gas absorbs thermal energy from the environment dur- Compression piston diameter 65 mm
ing the isothermal expansion and releases heat to the environment Compression piston stroke 50 mm
during the isothermal compression process. Therefore, the perfor- Expansion piston diameter 40 mm
Expansion piston stroke 50 mm
mance of a Stirling cycle cooler would be determined by the tem- Expansion piston length 110 mm
peratures at the warm and cold heads. Wire screen material of regenerator Stainless steel
The VISR consists of the expansion room, the cold head, the Wire screen mesh no. of regenerator 200
regenerator, the cooler, the compression room, the expansion Wire diameter of regenerator 0.1 mm
Cold-end temperature 238 K
and compression pistons, and a slider-crank mechanism which is
shown in Figs. 1 and 2.
The design specifications of the prototype Stirling refrigerator
are listed in Table 1 and the diameter and length of the regenerator
are described in Table 2. The charged pressures are varied to inves- Table 2
The specifications for three configurations of the regenerator

Configurations 1 2 3
Regenerator length (mm) 80 60 50
Regenerator diameter (mm) 55 50 40

tigate the effects of the configuration difference on the

performance.

3. Development of the model

3.1. Description of basic model

The isothermal model of the Stirling refrigerator developed by

Schmidt is appropriate to add the effects of the additional compo-
nents, because the Schmidt model is well known. The model would
be subsequently improved for taking account of a pressure drop,
shuttle heat loss, etc. They are calculated separately and inserted
into the results described by Smith to fit the analyzed result to
the data [17].
The system design analysis of the VISR can be carried out
assuming the followings:
Fig. 1. Schematics of the Stirling cycle refrigerator.
1. The gas temperatures in the compression room and gas passage
are uniform and thus the regenerator inlet temperature Tri is
assumed to be equal approximately to the gas temperature in
the compression room Tc and in the passage Tt, i.e. Tri  Tt  Tc.
2. The gas temperature Te in the expansion room is approximately
equal to the outside wall temperature TL of the cold head, i.e.
Te  TL.
3. The wall temperatures at both sides of the regenerator, Tri and
Tro, are approximately constant. Thus, Tw  Tri  Tt  Tc;
Tro  Te  TL.
4. The working fluid behaves as an ideal gas, PV = RT.
5. The regenerator is an energy-storage element made from
packed wire screens. The flow is 1D flow and there is no axial
heat transfer.

Taking into pressure drop through the system due to flow resis-
tance, the average pressure of starting operation pav is supposed to
be equal to the charged pressure pch. But pressures at the respec-
tive components are depending on the pressure drop. The nomen-
clatures are used along the reference [17]. The expansion volume
Ve is assumed to be changed sinusoidally.
VE
V e ð/Þ ¼  ½1 þ cosð/Þ ð1Þ
2
where VE is the swept volume of the expansion room and / is the
crank angle,
2  p  t
/ ð2Þ
Fig. 2. The photo of the Stirling cycle unit. s
212 S. Le’an et al. / Applied Thermal Engineering 29 (2009) 210–215

The volume in the compression room Vc is given as follows: where k are a frictional coefficient, l a pipe length, and di its inner
diameter. The k is given as follows:
w  VE
V c ð/Þ ¼  ½1 þ cosð/  aÞ ð3Þ 1
2 k¼ 2
1:14 þ 2 lg dDi
where w is the ratio of the swept volume in the compression room
to that in the expansion room and a is the phase angle between Ve
where D is the roughness value of a pipe wall, and D = 0.1–0.5,
and Vc. In addition, there is some dead volume Vd in the regenerator.
depending on materials.
S is the dimensionless dead volume, s the temperature ratio in the
And cooling capacity QE was calculated by following formula:
compression and expansion room, h minimal pressure angle, d Z 2p
parameter, which are given as follows n d sin h
QE ¼ f pe ð/ÞdV e ¼ p VE pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð14Þ
60 eav 1 þ 1  d2
T C X V dj 0
S¼ ; ð4Þ Q C ¼ sQ E ð15Þ
VE T dj
TC W ¼ Q E  Q C ¼ ðs  1ÞQ E ð16Þ
s¼ ; ð5Þ
TE
  where f is the rotational frequency of the crank, and n the rotational
w  sinðaÞ
h ¼ tan1 ; ð6Þ speed.
s þ w  cosðaÞ The net cooling capacity Qnet of the VISR can be evaluated by
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
s2 þ 2  s  w  cosðaÞ þ w2 subtracting the heat losses from the available cooling capacity.
d¼ : ð7Þ
sþwþ2  S The heat losses are given by the heat conduction loss Qcond of the
The pressure ratio is expressed in the following: regenerator, the regenerator’s enthalpy flow loss Qenth, the shuttle
heat loss Qshutt of the expansion piston and the system flow resis-
pmax =pmin ¼ ð1 þ dÞ=ð1  dÞ ð8Þ tance loss Qres.
The average pressure in the system pav is presented Q net ¼ Q E  Q enth  Q cond  Q shutt  Q res ð17Þ
rffiffiffiffiffiffiffiffiffiffiffiffi
1þd The losses Qenth, Qcond and Qshutt can been calculated according to
pav ¼ pmin  ð9Þ
1d the reference [12]. The loss of Qres is calculated by the following
Therefore, the instantaneous pressure in the system p(/) is also gi- formula:
ven by the average pressure pav V En
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi Q res ¼ Dp ð18Þ
30
1  d2
pð/Þ ¼ pav  ð10Þ where n is the rotational speed.
1 þ d  cosð/  hÞ
The actual input power Nac is described as follows:
W
3.2. Development of the model N ac ¼ ð19Þ
gT gm gel
The total pressure drop across the system is given by the sum- where gT is the isothermal efficiency of the machine, gm the
mation of the frictional flow resistance and bends as follows [18]: mechanical efficiency of the machine and gel the motor efficiency
X of the machine.
Dp ¼ Dpm þ Dpl ð11Þ
The COP of the VISR can be evaluated by the following equation:
The flow resistance of the respective components is given as Q net
follows: COP ¼ ð20Þ
N ac
t2p
Dpm ¼ n q ð12Þ
2 4. Experimental research
where q is a gas density, n a flow resistance coefficient, t2p a mean
gas flow velocity in the components. 4.1. Test system
At the respective outlets of the compression room, the regener-
ator, the cold head, the cooler, and the expansion room, the gas is The prototype VISR was constructed for investigating its perfor-
compressed. And thus n is given as follows: mance and power consumption. The temperatures at the compres-
sion and expansion rooms, at the inlet- and outlet-ports of
 2
A2 regenerator and their inter-space were measured with the cali-
n ¼ 0:5 1 
A1 brated PT100 thermocouples. The pressures were monitored with
Piezo-electric dynamic pressure sensors installed at the outlet-
At the respective inlets of the expansion room, the regenerator, the ports of the compression and expansion rooms. The discharged
cold head, the cooler, and the expansion room, the gas is expanded. pressure at the crankcase was measured by the pressure gauge
So n is given as follows: with 0.001 MPa accuracy which was calibrated by a dead weight
 2 tester. The rotational speed and shaft power were measured with
A2
n¼ 1 a torque sensor. The input voltage, current and the power con-
A1
sumptions of the cooler were recorded with accuracies of ±0.1%,
where A1 describes a sectional area of the compression room, the ±0.2% and ±1%, respectively. The wattmeter and uninterrupted
regenerator, the cold head, the cooler, and the expansion room, power supply were used and the frequency was changed by the
and A2 expresses a sectional area of each outlet and inlet pipes. transducer. The real time data were recorded by the National
The flow resistance of pipe (Dpl) is given as follows: Instruments M Series data acquisition (DAQ) system, i.e. NI–DAQ
8.1. In order to minimize the heat flow from the ambient to the
t2p l cold head, the heater was insulated with polyurethane foam.
Dpl ¼ k q ð13Þ
2 di Fig. 3 shows the schematic diagram of the experimental system.
 S. Le’an et al. / Applied Thermal Engineering 29 (2009) 210–215 213

Pressure Sensor 1
T

PT100 1
T
Cooler
PT100 2
P
T

Pressure Gauge Compressor PT100 3

Frequency Electromotor
T
Transducer
PT100 4
Torque sensor T

PT100 5
T
Shaft Power
Speed Indicator Expander PT100 6
Indicator Heater F

Diagram Wattmeter
P

Pressure sensor 2 Computer

Fig. 3. Schematic diagram for experiments of the VISR.

4.2. Experiment procedure Fig. 5 shows the simulation and experiment results of the cool-
ing capacity and COP as a function of the rotational speed under
Using helium and nitrogen as working fluids, the rotational fre- the Conf. 1 in which the pressure of nitrogen or helium is changed
quency and charged pressure are changed in the experiments un- at 1.0 MPa. It is observed that in case of helium the cooling capacity
der the three kinds of configurations given in Table 2. The increases with rotational speed but the COP has a maximum. In
charged pressures of helium and nitrogen in the refrigerator were case of nitrogen, both of the cooling capacity and the COP have
changed from 0.25 MPa to 1.4 MPa and the rotational frequency an optimum. And in cases of nitrogen, the optimum rotational
was 10–40 Hz. speeds for the cooling capacity and the COP are in the range of
500–700 rpm. But for helium, the optimum rotational speeds for
5. Results and discussions COP are at 800–1000 rpm. The cooling capacity of helium is greater
than that of nitrogen, because the specific heat capacity of helium
Fig. 4 shows the variation of the compression and expansion is larger than that of nitrogen, and the flow resistance of helium is
pressures with the crank angle when helium or nitrogen is charged much smaller than nitrogen. The difference in optimum points be-
at 1.0 MPa. And the rotational speed is 600 rpm in case of Conf. 1 tween nitrogen and helium is caused by the difference of the total
(shown in Table 2). pressure drop across the system from the compression room to the
In Fig. 4, the pressure difference between the compression and expansion room and its total pressure drop of nitrogen is larger
the expansion means the pressure loss. And the pressure loss for than that of helium. This pressure drop is in proportion to the
nitrogen is larger than that of the helium. square of working fluid velocity and a mass flow. The mass flow

Compression Pressure 14 Compression Pressure

14
Expansion Pressure Expansion Pressure
13 13

12 12
PRESSURE (MPa)

PRESSURE (MPa)

11 11

10 10

9 9

8 8

7 7

6 6
0 60 120 180 240 300 360 0 60 120 180 240 300 360
CRANK ANGLE (degree) CRANK ANGLE (degree)
a. Helium b.Nitrogen

Fig. 4. Pressure variation in the compression and expansion rooms with the crank angle: (a) Helium and (b) Nitrogen.
214 S. Le’an et al. / Applied Thermal Engineering 29 (2009) 210–215

COP in case of nitrogen of nitrogen is larger than that of helium. We investigated the influ-
120 COP in case of helium ence of the charged pressure of both helium and nitrogen on the
1.0
110 performance, i.e. at 0.4, 0.5, 0.6, 0.7, 0.9 MPa. The trend of cooling
100
0.9 capacity and COP against charged pressures was similar to the case
0.8 of 1.0 MPa, although their relative values are different at the differ-
90
Cooling Capacity(W)

ent charged pressure.

80 0.7
Fig. 6 shows the dependence of the cooling capacity on the
70 0.6 charged pressure under the Conf. 1 at 800 rpm rotational speeds.

COP
60 0.5 It is shown that the cooling capacity increases with the charged
50 pressure up to 1.0 MPa in case of nitrogen and up to 1.3 MPa in
0.4
40 case of helium. Fig. 7 shows the cooling capacities in case of
0.3 1.0 MPa helium, under the Conf. 1, Conf. 2 and Conf. 3. It is ob-
30 Cal cooling capacity of nitrogen
Cal cooling capacity of helium 0.2 served from Fig. 7, that the Conf. 2 is the optimal configuration
20 Test cooling capacity of nitrogen
Test cooling capacity of helium 0.1
even in simulation.
10
We can see from Figs. 6 and 7 that the experimental cooling
0 0.0
capacity is about 80% of that of the calculation. The manufacturing
100 200 300 400 500 600 700 800 900 1000 1100
Rotational Speed (rpm)
defects in the machine would cause the deterioration of the perfor-
mance compared with the predicted ones. And the inner leakage of
Fig. 5. Variations of cooling capacity and COP with rotational speed. the working fluid from the working space to the crankcase causes
also the decrease of cooling capacity.

6. Conclusions
COP in case of nitrogen
110 1.2 Based on the simulation and experiments the following could
COP in case of helium
100 be concluded:
90 1.0
80 1. To obtain the optimal configuration for each VISR components
in the refrigeration system, each component must be matched
Cooling Capacity

70 0.8
each other to satisfy the required specifications.
60
2. It is shown that the cooling capacity increases with the charged
COP

50 0.6
pressure up to 1.0 MPa in case of nitrogen and up to 1.3 MPa in
40 case of helium. The Conf. 2 is shown to be the optimal configu-
30
Cal cooling capacity of nitrogen
0.4 ration for the cooling capacity. The result shows that helium is
20 Cal cooling capacity of helium better than nitrogen.
test cooling capacity of nitrogen 3. The optimum rotational speed for the cooling capacity is differ-
10 0.2
test cooling capacityof helium
0 ent between nitrogen and helium. The COP has a peak value
around 600 rpm for nitrogen, and for helium around 900 rpm.
-10 0.0
0.2 0.4 0.6 0.8 1.0 1.2 1.4 4. These results show that the VISR is applicable to the domestic
Charged Pressure (MPa) refrigeration.

Fig. 6. Variations of cooling capacity and COP with charged pressure, and Cal in the
figure caption means calculated and Conf. the configuration in the Table 2. Acknowledgements

The authors are grateful to Prof. He Yaling, Prof. Takeshi KAWAI

and Assistant Prof. Liu Yingwen of Xi’an Jiaotong University for
their valuable comments.
Cal cooling capacity of Conf. 3
Cal cooling capacity of Conf. 2
References
Cal cooling capacity of Conf. 1
Test cooling capacity of Conf. 3
Test cooling capacityof Conf. 2
[1] T. Finkelstein, C. Polonski, Development and testing of a stirling cycle machine
160 with characteristics suitable for domestic refrigerators. Report W/M (3A).u.5,
Test cooling capacity of Conf. 1 1.0
English Electric Company Ltd., Whetstone, 1959.
140 [2] C. Wu, Maximum obtainable specific cooling load of a refrigerator, Energy
Conver. Manage. 36 (1) (1995) 7–10.
Cooling Capacity (W)

0.8
120 [3] L. Chen, C. Wu, F. Sun, Cooling load versus COP characteristics for an
irreversible air refrigeration, Energy Conver. Manage. 39 (1–2) (1998) 117–
100 125.
0.6 [4] J. Chen, Minimum power input of irreversible Stirling refrigerator for given
COP

cooling load, Energy Conver. Manage. 39 (12) (1998) 1255–1263.

80 [5] N. Ni, L. Chen, C. Wu, F. Sun, Performance analysis for endoreversible closed
0.4 regenerated Brayton heat pump cycles, Energy Conver. Manage. 40 (1999)
60 393–406.
COP of Conf. 3 [6] S.C. Kaushik, A state of art study report on finite time thermodynamics, Indian
40 COP of Conf. 2 0.2 Institute of Technology, Delhi, India, Centre for Energy Studies, 1999.
[7] S.C. Kaushik, S.K. Tyagi, S.K. Bose, M.K. Singhal, Performance evaluation of
20
COP of Conf. 1 rereversible Ericsson and Stirling heat pump cycles, Int. J. Thermal Sci. 41
(2002) 193–200.
0.0
[8] E. Oguz, F. Ozkadi, Experimental investigation of a Stirling cycle cooled
200 400 600 800 1000
domestic refrigerator, Ninth International Refrigeration and Air Condition
Rotational Speed (rpm) Conference at Purdue, USA, 2002, R19-3.
[9] B.J. Huang, C.W. Lu, Split-type free-displacer Stirling refrigerator design using
Fig. 7. Cooling capacity and COP as function of rotational speed. linear network analysis, Cryogenics 36 (1996) 1005–1017.
 S. Le’an et al. / Applied Thermal Engineering 29 (2009) 210–215 215

[10] S.K. Tyagi, S.C. Kaushik, M.K. Singhal, Parametric study of irreversible Stirling [14] B.J. Huang, H.Y. Chen, Modeling of integral-type Stirling refrigerator using
and Ericsson cryogenic refrigeration cycles, Energy Conver. Manage. 43 (2002) system dynamics approach, Int. J. Refrig. 23 (2002) 632–641.
2297–2309. [15] G. Walker, Stirling cycle cooling engine with two-phase, two-component
[11] D.M. Berchowitz, J. McEntee, S. Welty, Design and testing of a 40 W free-piston working fluid, Cryogenics 14 (1974) 459–462.
Stirling cycle cooling unit. 20th IIR Conference at Sydney, Australia, [16] G. Walker et al., Microcomputer simulation of Stirling cryocoolers, Cryogenics
1999. 29 (1989) 846–849.
[12] Lange, Th. J. de, Vries, M.J. de, The application of the Stirling cycle to [17] G. Walker, O.R. Fauvel, G. Reader, E.R. Bingham, The Stirling alternative: power
refrigeration machines: A design methodological analysis, in: Proceedings of systems, refrigerants, and heat pumps, Gordon and Breach Science Publishers,
sixth ISEC, 1993, pp. 285–290. Switzerland, 1994.
[13] Ö Ercan Ataer, H. Karabulut, Thermodynamic analysis of the V-type Stirling- [18] Yongjiang Yu, The piston compressor, Mechanical Industry Press, Beijing, 1982
cycle refrigerator, Int. J. Refrig. 28 (2005) 183–189. (Chap. 14, in Chinese).