You are on page 1of 9

Applied Energy 139 (2015) 272280

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Determination of adequate regenerator for a Gamma-type Stirling


engine
Ramla Gheith a, Fethi Aloui b,, Sassi Ben Nasrallah a
a
b

Universit de Monastir, cole Nationale dIngnieurs de Monastir, Laboratoire LESTE, Avenue Ibn El Jazzar 5019, Monastir, Tunisia
University of Lille North of France, UVHC, LAMIH CNRS UMR 8201, Campus Le Mont Houy, F-59313 Valenciennes Cedex 9, France

h i g h l i g h t s
 A Gamma-Stirling engine is investigated to optimize its operation.
 A stainless steel of a porosity of 85% was used as material for the regenerator.
 Asymmetry of heat transfer inside regenerators consumes a part of produced energy.
 Central composite rotatable design was adopted to minimize this phenomenon.
 The heating temperature is a most signicant factor in the study.

a r t i c l e

i n f o

Article history:
Received 24 July 2013
Received in revised form 20 October 2014
Accepted 11 November 2014
Available online 11 December 2014
Keywords:
Stirling engine
Regenerator
Constituting material
Porosity
Optimization

a b s t r a c t
This paper deals with an optimization of the Stirling engine regenerators. Firstly, different materials are
experimented (Stainless Steel, Copper, aluminum and Monel 400). The engine performances and the state
of each material after 15 h of use are considered. The Stainless steel was the material that best satises
these two conditions. Five regenerators in stainless steel with different porosities were manufactured and
experimented (95%, 90%, 85%, 80% and 75%). Porosity that gives the best trade-off between maximizing
the engine brake power, maximizing the heat transfer and minimizing the pressure drops, was retained.
Thus, the regenerator in stainless steel with porosity of 85% was considered as the most suitable matrix
maximizing the Stirling engine performances and minimizing heat and friction losses.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
Stirling engine regenerators are very complex to modeling and
to designing. It will require a large number of equations to describe
their thermodynamic operation. Since it is a balancing act of several factors, some researches have proposed the best regenerator
qualities [14] allowing optimal Stirling engine performances. It
must have:
High thermal capacity to minimize temperature variation [5],
high thermal conductivity to minimize temperature gradients
[6],
large surface area to minimize temperature differences between
the regenerator and the working uid [7],
Corresponding author at: University of Lille North of France, UVHC, LAMIH
CNRS UMR 8201, Campus Le Mont Houy, F-59313 Valenciennes Cedex 9, France.
E-mail address: Fethi.Aloui@univ-valenciennes.fr (F. Aloui).
http://dx.doi.org/10.1016/j.apenergy.2014.11.011
0306-2619/ 2014 Elsevier Ltd. All rights reserved.

small dead volume with dense matrix to maximize pressure


variation [7,8a,b],
a highly porous matrix with minimum resistance to ow [9,10].
Evidently, it is impossible to satisfy all these properties at one
time; thus any regenerator must be a compromise [3].
Chen et al. [11] has demonstrated that the insertion of a metallic matrix helps the system to become more stable to possible disturbances in the mass ow rate supply of hot uid. Timoumi et al.
[12] studied the performances of a Stirling engine regenerator for
porosity ranging from 0.6112 to 0.9122 by varying its corresponding wire mesh from 0.0090 to 0.0035. The decrease in mesh porosity leads to the highest friction factor and pressure drop. In spite of
the higher pressure drop we have better power and thermal efciency because we have better heat transfer. Chang and Yang
[13] studied a Ross Yoke drive engine for solar applications. They
proposed a detailed analysis of the inuence of the regenerator
parameters on the engine performances. An increase in matrices

R. Gheith et al. / Applied Energy 139 (2015) 272280

273

Nomenclature
E
r
T
P
Vol
V
N

engine efciency, 3.227


compression degree
temperature, C
pressure, bar
volume, m3
engine speed, RPM
Stirling engine speed, rpm

Subscript
C
Cold

Carnot
cold side

porosity leads to the friction factor increases and the pressure drop
loss decrease. A selection of a small wire diameter meshes may
reduce the pressure drop for high porosity matrix. The regenerator
effectiveness can be manipulated by varying wire diameter and
wire length, which in turn changes the wetted surface area.
Bangert [14] note that the regenerator effectiveness has been
recognized as an important factor to Stirling engines performances.
The regenerator effectiveness depends on porosity, permeability
and material of the porous matrix. Temperature distribution in
the regenerator can be measured to determine the regenerator
effectiveness. Furthermore, the magnitude of the regenerator effectiveness is also dependent on the operating speed. Abdulrahman
et al. [15] studied experimentally the inuence of the foam structure (including material, porosity and permeability) on the overall
performance of the Stirling engine regenerator. He proved that
when the regenerator porosity increases beyond a critical point
the engine performance decreases, due to increased external conduction and lack of thermal transfer with the working uid. Rebeiro
et al. [16] experimented different regenerator material: cellular
ceramic substrate with regular square channels, steel scourers
and stainless steel wood. They tested the performances of these
regenerators function of mean pressure between 0 and 10 bar. They
showed that the cellular ceramics may offer an alternative to traditional regenerator materials to reduce the overall system costs. They
demonstrated that the pressure drop increase with the porosity
decrease and the ow resistance depends on porosity, regularity
of the porous material and the nature of packing of the solid material. Gheith al. [17] studied different regenerator materials and demonstrated that the Stirling engine regenerator is very sensitive to its
material characteristics. Stirling engine efciency can be calculated
in different way. The must commonly used method is the Carnot one
[1820]. An ideal Stirling cycle, consisting of two isotherms and two
isochors, has thermal efciency determined by the temperature
interval between the heater and the cooler and the compression
degree r = Vheater/VCooler.

T1  T2
T 1 T 1 T 2 =c  1lnr

In order to propose the most suitable regenerator (materials


and porosity), different matrixes were manufactured. Firstly, four
regenerators in different materials (stainless steel, copper, aluminum and Monel 400) were experimented. The most suitable material was determined. From the predetermined material, ve
matrixes with different porosities (95%, 90%, 85%, 80% and 75%),
were manufactured and then experimented. The efciency and
the output brake power of a c-type Stirling engine were investigated for different porosities. The most suitable regenerator (material and porosity) for the c-type Stirling engine will be proposed.
Our choice will be based on different criteria already established.

Hot
hot side
w-input
water input circuit
w-output
water output circuit
i
initial lling pressure
com
compression space
Exp
expansion space
R1, R2, R3, R4 position of thermocouples on regenerator side 1
R5, R6, R7, R8 position of thermocouples on regenerator side 2

2. Experimental facility
2.1. Stirling engine set-up
The double-cylinder, c-type Stirling engine is the ST05G of Viebach society. This engine can provide 500 W of brake power and
can reach a maximum rotation speed of 600 rpm. Its main compartments and metrology are presented in Fig. 1. The engine is
composed of two pistons with a draft shaft of 90, each piston slide
in one separate space. The compression space is cooled by a circuit
of water and the expansion space is heated by an electrical resistance delivering a maximal power of 3.5 KW. Both pistons are
linked by a classical crank-rode system. The regenerator is located
between the cooler and the heater. It is constituted of a porous
medium with a xed porosity. It is an annular spaces crossed twice
by the working uid (Fig. 2).
2.2. Regenerator metrology
Type-K thermocouples with diameter of 0.5 mm, 0.25 mm and
0.0254 mm, were used. These diameters are enough small that
the thermocouples thermal inertia do not mislead measurements
of instantaneous temperatures. Two thermocouples were
implanted respectively upstream the expansion space (Thot) and
in the compression space (Tcold). Other two thermocouples were
used to measure the temperatures of the inlet (Tw-inp) and the outlet cooling water (Tw-out). Eight thermocouples were skinned symmetrically (Fig. 3) up to 1 mm inside the regenerator matrix
without touching the material. They allow measurement of the
working uid temperature passing through the regenerator. Different matrixes with different material were experimented (Fig. 3).
Two pressure transducers were installed in the compression space
and after the expansion space. The st one is located at the end of
the compression cylinder. It is a Druck sensor comprising a membrane. It can measure a pressure up to 20 bars and gives an answer
between 0 volt and 5 V. The second is placed upstream the expansion space. This sensor cannot support temperature higher than
80 C. So it was placed after a cooling tower to decrease the temperature of the working uid. The variation of the Stirling engine
rpm is made through an electrical dissipation system. This system
is constituted of 10 electrical resistances of 55 W, each one placed
in parallel, in order to obtain the variation of electrical dissipation,
and then the Stirling engine rpm.
3. Determination of adequate regenerator material
The Stirling engine was experimented with different regenerators material (Table 1) at the following experimental conditions
(Table 2). The rst matrix is made of Stainless steel, which is the
most commonly used material as Stirling engine regenerators.

274

R. Gheith et al. / Applied Energy 139 (2015) 272280

Water output
Tcold
Pcomp.

Cooler

Tw-inp.

Tw-out.
Regenerator
side 2 :
Series of 4
thermocouples

Water input
TR5
TR6

TR1

TR2

TR7

TR3

TR8

TR4

Regenerator
side 1 :
Series of 4
thermocouples

Pexp.
Thot
Heating
system

Heater

Fig. 1. Design of the c-type Stirling engine.

De

Di

A
Coupe AA
Fig. 2. Cut of the regenerator: external diameter: De = 98  103 m, internal
diameter:
Di = 133.5  103 m,
height
H = 50  103 m,
total
volume
V = 3.227  104 m3.

The second matrix is made of copper, which has an important thermal conductivity. The third matrix is constituted of aluminum,
which has a very low density, does not oxidize like copper and/
or like steel, and can be efciently and ecologically recycled. The
last matrix is formed of Monel 400, which is a Nickelcopper alloy
with high strength and excellent oxidation resistance. All experimental measurements are made for a steady operation of the
engine.
Viscous dissipation is the heat internal generation when the
working gas is forced to ow through the porous medium (regenerator). This dissipation generates a pressure drop across the

engine. The pressure drop caused by singularities inside the motor


are considered constant for all materials and for the same experimental conditions. The engine pressure drop increases with the
initial lling pressure and the heating temperature (Table 3). All
studied regenerators cause approximately the same loss for the
same experimental conditions. The minor recorded differences
are due to the use of different wire diameters for each material.
To limit the pressure drop in a Stirling engine regenerator, a compromise must be adopted: reducing the engine frequency, reducing
the mass of the working uid, reducing the temperature variation
in the regenerator, etc.
For all experimented regenerators, the engine brake power is
recorded for the variations of initial lling pressure and heating
temperature (Fig. 4). The engine brake power is enhanced by both
parameters. The increases of these parameters raise respectively
the mass of the working gas participating to the processes of heat
exchange in the regenerator and the thermal energy which can be
given to the working gas. The use of the Stainless Steel regenerator
produce the best brake power (about 308 W for Pi = 8 bar and
Th = 400 C). However, the Aluminum regenerator generates the
worst brake power (about 203 W for Pi = 8 bar and Th = 400 C).
The used material present different thermo-physical proprieties
so for the same experimental conditions different brake power
range are recorded (depending on used material thermal conductivity and thermal capacity).
The photos of regenerator matrices made from copper are presented in Fig. 3 before being experimented and presented in Fig. 5
after 15 h of experimentation. It is clear in Fig. 5, that the matrices
in copper oxidized after few hours of use. The copper matrix

275

R. Gheith et al. / Applied Energy 139 (2015) 272280

a
4 skinned
thermocouples

4 skinned
thermocouples

Aluminum

Stainless Steel

Copper

Monel 400

Annular space : regenerator

Fig. 3. Regenerator instrumentation and constituting material.

Table 3
Pressures drop recorded for studied regenerators for different initial lling pressure
and heating temperatures.

Table 1
Characteristics of the tested porous structures.
Materials with porosity of 90%
Proprieties

Stainless Steel
304L

Copper

Aluminum

Monel
400

Density (kg m3)


Thermal capacity
(J kg1 K1)
Thermal conductivity
(W m1 K1)

7.850
477

8.920
385

2.700
902

8.800
430

26

390

237

22

Th = 400 C
Stainless Steel
Copper
Aluminum
Monel 400

Pi = 5 bar
Stainless Steel
Copper
Aluminum
Monel 400

Table 2
Experimental conditions.
Parameters

Values

Heating temperature
Initial lling pressure
Cooling water ow rate

300 C, 400 C and 500 C


3 bar, 5 bar and 8 bar
8.1 l/min

cannot be recommended to be used as a Stirling engine regenerator. In fact, heated at 500 C, the copper oxidizes quickly (Fig. 5)
because of the working uid (air) contains about 21% of oxygen.
This material oxidation changes quickly the physical characteristics of the copper, and then leads to heat exchanges deterioration.
The aluminum matrix is formed by small diameter laments. These
later heated up to 500 C melt to small masses causing inhomogeneous porosity. The whole matrix is narrowed (Fig. 6).
The use of regenerator in Aluminum is limited by its melting
temperature but it can be recommended for Stirling engine functioning at low heating temperature (>400 C). After 15 h of experiments with a temperature below 500 C the matrices of stainless
steel and Monel 400 undergo no change in the geometry of the
manufactured matrix. The stainless steel and the Monel 400 regenerators, have good thermal efciencies, respectively of 31% and
25.5%. These two materials do not present a problem of oxidation.

Pi = 3 bar

Pi = 5 bar

Pi = 8 bar

0.1082
0.1093
0.1088
0.1040

0.1234
0.1245
0.1236
0.1236

0.132
0.136
0.1305
0.1326

TH = 300 C

TH = 400 C

TH = 500 C

0.1101
0. 1097
0.1151
0.1110

0.1234
0.1245
0.1236
0.1236

0.1423
0.1466
0.1407
0.1447

Both materials present a good thermal capacity which allows to


the regenerator to absorbs/desorbs the maximal thermal energy
from/to the working gas.
We proved the results found in Timoumis papers which specify
that a regenerator must have a high heat capacity and a low thermal conductivity. Decreasing the material thermal conductivity
reduces conduction losses. Increasing material heat capacity
increases regenerator effectiveness.
Synthesis
Based on the results obtained in this section, the stainless
steel can be considered as the best material used as Stirling engine
regenerator. Although the thermo-physical properties of stainless
steel accentuate the inhomogeneity of the heat transfer in each
section of the regenerator, this material can still get the best performance for the c-type Stirling engine, and presents no risk of oxidation. Thus, the following test will be based on Stainless Steel
material with different porosities. To improve heat exchange of
the regenerator and to reduce internal losses by conductivity, a
material with high heat capacity and low thermal conductivity
must be chosen. The stainless steel and the Monel 400 verify both
statements.

276

R. Gheith et al. / Applied Energy 139 (2015) 272280

Fig. 4. Stirling engine brake power for all experimented regenerator materials; For Pi (3 bar, 5 bar, 8 bar) and Th xed at 400 C and for Th (300 C, 400 C, 500 C) for Pi xed at
5 bar.

4. Determination of adequate regenerator porosity

Fig. 5. Photo of the regenerators of copper with different porosity after about 15 h
of test.

Fig. 6. Photo of the regenerators of Monel 400 (a) and of aluminum (b) after 15 h of
test with a heating temperature below 500 C.

The used Stainless steel matrices characteristics are depicted in


Table 4 and the experimental conditions presented in Table 5. Generally, the pressure drop is function of the uid velocity, the physical properties of the uid (viscosity and density), the average
porosity [11], the specic area and the wire diameter [6].
Figs. 7 and 8 show that the pressure drop in the experimented
engine increases with the initial lling pressure and the engine
velocity and decrease with the regenerator porosity.
Generally, the pressure drop in a Stirling engine is function of
the uid velocity, the physical properties of the uid (viscosity
and density), the average porosity [11], the specic area and the
material wire diameter [6]. The resistance to the uid ow gives
rise to a pressure drop in the working gas. According to Darcys
law the pressure drops will rise linearly with volume ow rate or
uid velocity. The increases of velocity raise the turbulence in
the porous media and consequently the pressure drop throws it.
The pressure drop increase with the initial lling pressure
(Fig. 7) and with the engine speed (Fig. 8). For a xed lling pressure, an increase of porosity lead to a fall of the friction coefcient
and the pressure drop. The rise of the pressure drop and the ow
resistance depends on the porosity, the regularity of the porous
material and the manner of disposition of the solid.
The regenerator performances can be performed by establishing
the minimum temperature difference between material matrix and
working uid it is necessary to expose the maximum surface area
of matrix, therefore matrix should be nely divided. In order to
choice the adequate regenerator porosity a compromise must be
recognized. For minimum ow losses a small highly porous matrix
is required where as for a minimum dead space in the regenerator
a dense matrix is required. A nely divided matrix is required for
maximum heat transfer between solid matrix and working gas.
For the same heating temperature (400 C), the working uid
temperature decreases with the regenerator porosity. For a

277

R. Gheith et al. / Applied Energy 139 (2015) 272280


Table 4
Five stainless steel matrixes with different porosities are manufactured.
Material: Stainless Steel
Porosity (%)

Mass (kg)

Wire diameter (m)

Hydraulic diameter (m)

Thermal conductivity (W m1 K1)

95
90
85
80
75

0.125
0.253
0.380
0.507
0.633

2.5e4
2.5e4
2.5e4
2.5e4
2.5e4

0.95
0.90
0.85
0.80
0.75

1.32
2.62
3.92
5.22
6.52

0.13

Pi = 3 bar
Pi = 5 bar

0.125

Pressure drop [bar]

Pi = 8 bar
0.12
0.115
0.11
0.105
0.1
0.095
75

80

85

90

95

Matrices porosities [%]


Fig. 7. The pressure drop vs. the matrices porosity for different charge pressure.

0.12

Pressure drop [bar]

0.118
0.116
0.114
0.112
0.11

V = 370 rpm
V = 340 rpm
V = 320 rpm
V = 300 rpm

3 bar, 5 bar and 8 bar. The pressure has no signicant inuence


on the temperature variation in a Stirling engine regenerator
(Fig. 9). Indeed, increasing the initial lling pressure led to phenomena with contradictory effects on the performance of the Stirling engine. It increases the mass of the working uid passing
through the regenerator and increases the engine rpm. The rst
phenomenon enhance the quantity of heat exchanged between
the regenerator and the working uid and the second phenomenon
decrease the exchanging duration between the working uid and
the regenerator material. The heating temperature increase causes
the increase of the working uid temperature (Fig. 10).
The Stirling engine regenerator is evaluated from its efciency
which is dened as the ratio of the amount of heat effectively
transferred during an engine cycle by the maximum amount of
heat transferred during the same cycle. The regenerator is considered ideal if its efciency is about 1. This parameter depends
mainly on the regenerator porosity, the regenerator temperature
variation and the regenerator volume. The regenerator efciency
decreases with the porosity and increase with the charge pressure
and with the heating temperature (Figs. 11 and 12).
Increasing the porosity of the regenerator causes an increase of
the pore volume and the decrease in contact area between the
material and the inner walls of the Stirling engine. This effect
reduces the residence time of the uid working in the regenerator,
and therefore losses due to the internal and external conduction.
For the same heating temperature (400 C), we nd that the temperature of the working uid decreases as the porosity of the
regenerator.
The pressure variations have a slight inuence on the change of
working gas temperature crossing the regenerator. The increase of
initial charge pressure causes simultaneously: the increase of the
mass of the working uid passing through the regenerator and
the ow velocity. The rise in the speed of the Stirling engine causes

0.108

280
Pi = 3 bar

80

85

90

Matrix porosities [%]


Fig. 8. The pressure drop vs. the matrices porosities for different engine speed (V).

Table 5
Experimental conditions.
Parameters

Values/designation

Heating temperature
Inlet cooling temperature
Initial lling pressure
Cooling water ow rate
Material constituting material
Regenerator porosities

300500 C
12 C
38 bar
8.16 l/min
Stainless steel
95%, 90%, 85%, 80%, 75%

Pi = 5 bar

260

95

Temperature TR4 [C]

0.106
75

Pi = 8 bar

240

220

200

180

160

75

80

85

90

95

Matrix porosities [%]


porosity of 75% the regenerator temperature is about 268 C,
270 C and 278 C for an initial lling pressure of respectively

Fig. 9. Working uid temperature crossing the regenerator for different stainless
steel porosities (temperature recorded by thermocouples TR1) for Pi = 3, 5 or 8 bar
and Th = 400 C).

278

R. Gheith et al. / Applied Energy 139 (2015) 272280

350

320

Th= 300C
Th= 400C
Th= 500C

Pi = 5 bar

280

Brake power [W]

300

Temperature TR4 [C]

Pi = 3 bar

300

250

200

Pi = 8 bar

260
240
220
200
180
160

150

140
100

120
75

80

85

90

95

75

Fig. 10. Working uid temperature crossing the regenerator for different stainless
steel porosities (temperature recorded by thermocouples TR1) for Th = 300, 400 or
500 C.

Pi = 3 bar

55

Pi = 5 bar
50

Regenerator efficacity [%]

80

85

90

95

Matrix porosities [%]

Matrix porosities [%]

Pi = 8 bar

45
40
35
30

Fig. 13. Brake power function of the porosity for different initial lling pressure.

the reduction of the residence time of the working uid in the


regenerator (reduced heat exchange time between the porous
matrix and the working uid).
The working uid temperature is inversely proportional to the
regenerator porosity. The amount of heat stored by the material
decreases with the porosity due to the decrease of heat capacity
and heat transfer coefcient by convection. Increasing the porosity
decreases the thermal capacity of the regenerator; causing the
decrease of the stored heat.
The brake power output of the c-type Stirling engine for different stainless steel porosities obtained for different initial lling
pressure and heating temperature are respectively presented in
Figs. 13 and 14. Increasing the initial lling pressure or the heating
temperature, the brake power delivered by the engine increases for
all matrices tested. The matrix porosity has contradictory effects
on Stirling engine performances:

25
20

75

80

85

90

95

Matrix porosities [%]


Fig. 11. Regenerator efciency evolution vs. stainless steel matrix porosities for
different initial charge pressure.

Th= 300C
Th= 400C
Th= 400C

300

60

T = 300C
T = 400C
T = 500C

55

250

50

Brake power [W]

Regenerator efficacity [%]

Increase of the thermal losses by external conduction through


engine walls: negative effect.
Increase of the amount of heat exchanged between the working
uid and the regenerator: positive effect.
Increase of the heat quantity stored in the regenerator: positive
effect.
Increase of the friction losses: negative effect.

45
40
35
30

200

150

25
20
100

15
75

80

85

90

95

Matrix porosities [%]


Fig. 12. Regenerator efciency evolution vs. stainless steel matrix porosities for
different heating temperature.

75

80

85

90

95

Matrix porosities [%]


Fig. 14. Stirling engine brake power function of the porosity for different heating
temperature.

279

R. Gheith et al. / Applied Energy 139 (2015) 272280

26.1
Th=500C

E C [%]

240
26

25.9
74

220

76

78

80

82

84

86

88

90

92

94

96

Matrix porosities [%]


210

24

200

95%
90%
85%
80%
75%

190
180

300

310

320

330

340

350

360

E C [%]

Brake power [W]

230

Th=400C

23.95
23.9
23.85

74

76

78

80

20.45

370

82

84

86

88

90

92

94

96

Matrix porosities [%]

rpm [rpm]

For porosity lower then a critical value depending on functioning conditions, the increase in porosity generates a positive balance
of different observed effects, which implies an increase in engine
power Stirling Gamma. Beyond the porosity critical value, the negative effects overcome the positive effects, which cause the
decrease of engine power.
The Stirling engine brake (Fig. 15) power increase with the
engine speed (rpm) until an optimal value.
Different contradictory effects interfere when the engine rpm
increase.
Decrease of the thermal losses by external conduction through
the engine walls: positive effect.
Decrease of the exchanging duration between the working uid
and the exchanging surface: negative effect.
Decrease of the heat quantity stored in the regenerator: negative effect.
Increase of the friction losses: negative effect.
Initially, the increase of the engine speed generates a positive
balance of different observed effects, which implies an increase
in Stirling engine power. Reaching a limit speed value, the negative
effects outweigh the positive ones causing the decrease of engine

23.95

E C [%]

Th=300C
Fig. 15. Brake power evolution vs. regenerator porosity and engine RPM.

20.4
20.35
20.3
74

76

78

80

82

84

86

88

90

92

94

96

Matrix porosities [%]


Fig. 17. Carnot efciency evolution vs. regenerator porosity for different heating
temperature (Pi = 5 bar).

power. These changes are similar to those observed in internal


explosion engines (petrol or diesel).
Figs. 16 and 17 present the engine Carnot efciency for different
regenerator porosity and for different initial lling pressure and
heating temperature. The variations of initial lling pressure do
not have an important inuence on the Carnot efciency. Only an
increase of about 0.5% of efciency is recorded for an increase of
5 bar. The efciency is increase of about 6% of an increase of the
heating temperature of about 200 C. For all studied cases the
engine Carnot efciency is maximal for regenerator porosity about
85%.
Synthesis
With reference to the obtained results the optimal porosity is
these of 85%. It presents the best compromise between minimizing
the mechanical and thermal losses and maximizing the amount of
exchanged heat. The maximum brake power and efciency are
obtained of maximum heating temperature (Figs. 14 and 16) and
the maximum initial lling temperature (Figs. 13 and 17). These
results can be generalized to any type of Stirling engine working
with the same range of operation parameters.

23.9

EC [%]

Pi = 3 bar

5. Conclusion

Pi = 5 bar

23.85

Pi = 8 bar

23.8

23.75

75

80

85

90

95

Matrix porosities [%]


Fig. 16. Carnot efciency evolution vs. regenerator porosity for different initial
lling pressure (Th = 400 C).

Four matrices made of different materials have been experimented as regenerator. These materials are: Stainless Steel, Copper, Aluminum and Monel 400. The experimental results show
that the regenerator Monel 400 stainless steel and Copper has
the best thermal efciency and engine brake power but the presence of oxygen in the working uid has a great handicap leading
to the rapid oxidation of the material, and after the modication
of the thermo-physical properties of the matrix. This degrades
extremely heat exchange, and therefore the mechanical power of
the Stirling engine. The regenerator Aluminum has a thermal
efciency acceptable, and does not oxidize. However, its use is
somewhat limited by its melting point. The engine power is very
sensitive to the nature of the material constituting the regenerator

280

R. Gheith et al. / Applied Energy 139 (2015) 272280

and the Stainless Steel is the best material that can be used as
Stirling engine regenerator working in this rage of heating
temperature. Based on various experiments, it can be concluded
that the matrix Stainless Steel is best placed to constitute the
Gamma Stirling engine regenerator. The experimentation of different stainless porosities, shows that a porosity of 85% presents the
maximal brake power and engine efciency.
A maximal heating temperature (500 C) and maximal initial
lling pressure (8 bar) applied to a stainless steel regenerator of
a porosity of 85% offer the optimal Stirling engine performances
(brake power of 320 W and a Carnot efciency over 26%).
Acknowledgements
This work was supported by the laboratories LAMIH CNRS
UMR 8201 (University of Valenciennes et du Hainaut-Cambresis,
France) and LESTE (ENIM, University of Monastir, Tunisia). These
supports are gratefully acknowledged.
References
[1] Andersen SK, Carlsen H, Thomsen PG. Numerical study on optimal Stirling
engine regenerator matrix design taking into account the effects of matrix
temperature oscillations. Energy Convers Manage 2006;47:894908.
[2] Organ AJ. Thermodynamics and gas dynamics of the Stirling cycle
machine. Cambridge University Press; 1992.
[3] Bangert K. Production of optimised metal foams for Stirling engine
regenerators. Mini project Report. The University of Shefeld; 2010.
[4] Frigerio S, Mehl M, Ranzi E, Schweiger D, Schedler J. Improve efciency of
thermal regenerators and VOCs abatement systems: an experimental and
modeling study. Exp Therm Fluid Sci 2007;31:40341.
[5] Waele ATAM. Finite heat-capacity effects in regenerators. Cryogenics
2012;52:17.

[6] Chang WS. Porosity and effective thermal conductivity of wire screens. ASME, J
Heat Transfer 1990;112:59.
[7] Park SH, Lee YS. An approximate thermal analysis of Stirling engine
regenerators. KSME J 1993;7(2):13343.
[8] (a) Tlili I, Timoumi Y, Ben Nasrallah S. Thermodynamic analysis of the Stirling
heat engine with regenerative losses and internal irreversibilities. Int J Eng Res
2008;9:45. http://dx.doi.org/10.1243/14680874JER01707;
(b) Tlili (b) I, Timoumi Y, Ben Nasrallah S. Analyse and design consideration of
mean temperature differential Stirling engine for solar application. Renew
Energy 2008;33:191121.
[9] Dallaire J, Gosselin L, da Silva AK. Conceptual optimization of a rotary heat
exchanger with a porous core. Int J Therm Sci 2010;49:45462.
[10] Martins LS, Ordonez JC, Vargas JVC, Parise JAR. Thermodynamic optimization
of a regenerator heat exchanger. Appl Therm Eng 2012;4546:4251.
[11] Chen WL, Wong KL, Po LW. A numerical analysis on the performance of a
pressurized twin power piston gamma-type Stirling engine. Energy Convers
Manage 2012;62:8492.
[12] Timoumi Y, Tlili I, Ben Nasrallah S. Analysis and design consideration of mean
temperature differential Stirling engine for solar application. Renew Energy
2008;33:191121.
[13] Cheng CH, Yang HS. Optimization of geometrical parameters for Stirling
engines based on theoretical analysis. Appl Energy 2012;92:395405.
[14] Bangert K. Production of optimised metal foams for Stirling engine
regenerators. Mini project Report. University of Shefeld; 2010.
[15] Abdulrahman SA, Zhibin Y, Jaworski AJ. Selection and experimental evaluation
of low-cost porous materials for regenerator applications in thermoacoustic
engines. Mater Des 2011;32:21728.
[16] Ribeiro AM, Neto P, Pinho C. Mean porosity and pressure drop measurement in
packed beds of monosized spheres: side wall effects. Int Rev Chem Eng
2010;2(1):406.
[17] Gheith R, Aloui F, Ben Nasrallah S. Study of the regenerator constituting
material inuence on a gamma type Stirling engine. J Mech Sci Technol
2012;26(4):12515.
[18] Rider GT, Huper C. Stirling engines. LondonNew York: E. and F.N. Spon; 1983.
[19] Urielli I, Berchowitz DM. Stirling cycle engine analysis. Bristol: Taylor & Francis
Group; 1984.
[20] Organ A. The regenerator and the Stirling engine. New York: John Wiley &
Sons; 1997.

You might also like