OPTIMIZATION OF DIAMETER RATIO FOR ALPHA-TYPE STIRLING ENGINES

VLAD MARIO HOMUTESCU*, DAN-TEODOR BLNESCU*

* Gheorghe Asachi Technical University of Iassy
Department of of Thermotechnics, Thermal Engines and Refrigeration
e-mail: mariohomutescu@gmail.com

Abstract: The problem of the constructive dimensioning and optimization of the Stirling machines is still
insufficiently studied. The paper analyzes the alpha-type Stirling engine from the point of view of the constructive
optimization, by choosing the optimum ratio between the diameters of the two cylinders in the machine composition.
This theoretical analysis showed that, by analyzing the functioning of the engine with either an isothermal model or
the adiabatic model without or with losses taken into consideration, an optimum value of the diameter ratio exists, for
which the work produced by the engine takes a maximum value. The optimum diameter ratio is calculated for a
numerical example.
Key words: alpha-type Stirling engine, diameter ratio optimization, adiabatic model
1. Introduction
The thermodynamic Stirling cycle [6] is composed of two isothermal processes (that take place at the
minimum and maximum temperatures of the cycle, T
m
and T
M
) linked by two isochoric processes (that took
place at the minimum and maximum volumes occupied by the gas inside the machine, V
m
and V
M
), as on
Fig. 1.
Irrespective of the architecture and the motive drive type, all Stirling engines include in their
functional unit construction (equivalent to an internal combustion monocylinder engine) the cylinder and the
displacer, the cylinder and the power piston, the heater and the cooler heat exchangers and the heat
regenerator, key components connected as shown on Fig. 2 [4].
In the case of using a crank mechanism, the indicator diagram of the engine looks like the one
presented also in Fig. 1.
The optimization of the Stirling engine functioning is a problem of major interest. Several papers
studying various aspects of the Stirling engine optimization (e.g. the optimization of the heat exchangers -
cooler, heater and regenerator - for irreversible cycles) were published [1], [2], [3].

1
0
1
0.5
1 2/4 0 1/4
p/p
max
V/V
M
4
3
2
T
M
= ct.
T
m
= ct.
0.6 V
[10
-3
m
3
]
0.5 0.7
p,
[MPa]
8
6
10


Fig. 1 - Direct Stirling thermodynamic cycle, for chosen compression ratio V
M
/ V
m
= 2, and an indicator diagram,
calculated for a numerical example
7
8
6
3
9
4
2
10
5
1


Fig. 2 - Alpha-type Stirling engine:
1 - crankshaft; 2 - displacer rod; 3 - displacer; 4 - expansion chamber; 5 - heater;
6 - regenerator; 7 - cooler; 8 - compression chamber; 9 - power piston; 10 - rod
2. Stirling Engine Performances Calculated with an Adiabatic Model
The adiabatic physico-mathematical model used for thermal performances calculation [7] is based on
the following hypotheses: the working fluid is the ideal gas, all processes occur ideally, the temperature
inside the heater is constant, the temperature inside the cooler is constant, the heat regenerator temperature is
constant and equal to the logarithmic mean of cooler and heater temperatures, the momentary pressure is the
same inside all engine chambers and the pistons move accordingly to known laws. The expansion and
compression spaces do not exchange heat with the neighboring spaces, this particularity also giving the name
of the model.

T
T
reg
T
h
T
k
V
c
, m
c
V
reg
V
h
V
e
, m
e
V
k
T
c
T
e
c - k k - reg reg - h
h - e


Fig. 3 - Schematic diagram of the alpha-type Stirling engine

The hypotheses concerning the temperatures inside the engine are shown as well in graphical manner
on Fig. 3. On Fig. 3 letters V, T and m were used to mark the volumes, temperatures and masses while the
subscripts e, c, k, reg, h refer to the expansion and the compression spaces, cooler, regenerator and heater
(they all correspond to English initials, excepting k for cooler).
The model uses the differential equation of mass conservation for the working agent, the equation of
state applied to the gas inside heat exchangers and the differential law of conservation of energy written for
the adiabatic chambers.
According to the adopted hypotheses, inside the compression and expansion chambers the gas
exchanges work with the surrounding environment (through piston movement) and enthalpy with the
neighboring heat exchangers (cooler or heater). The internal energy of the gas inside the adiabatic chambers
changes, as a consequence of mass and temperature variations. The heat exchanged by each of these two
chambers is zero, conforming to the adiabatic hypothesis.


The adiabatic model produces five differential equations:
(1)
|
.
|

\
|
+ =

c
c
k c
c
dV p dp
k
V
T R
dm
1
,

(2)
|
.
|

\
|
+ =

e
e
e h
e
dV p dp
k
V
T R
dm
1
,

(3)
|
|
.
|

\
|
+ + + +
|
|
.
|

\
|
+
=


h
h
reg
reg
k
k
e h
e
k c
c
e h
e
k c
c
T
V
T
V
T
V
k
T
V
T
V
T
dV
T
dV
p k
dp ,

(4)
|
|
.
|

\
|
+ =
c
c
c
c c
m
dm
V
dV
p
dp
T dT
c
,

(5)
|
|
.
|

\
|
+ =
e
e
e
e e
m
dm
V
dV
p
dp
T dT
e
.

The composed subscripts c-k and h-e refer to the dimensions describing the separating sections
between the compression and expansion chambers and their adjacent heat exchangers.
Eqs. (1), (2), (3), (4) and (5) form the system of differential equations of the adiabatic physico-
mathematical model of the Stirling engine. The unknown functions are the pressure p, the masses m
c
and m
e

inside the compression and expansion chambers and the temperatures T
c
and T
e
in the same chambers. The
system is nonlinear, because there are several terms in the differential equations that have an order higher
than one. The system has variable coefficients and the conditional temperatures T
c-k
and T
h-e
of the agent that
passes through the surfaces c-k and h-e depend on the sense of the gas flow. The conditional temperatures
take the expressions:

c k c
T T =

if 0 >
k c
m (or 0 <
c
dm );
(6)
k k c
T T =

if 0 <
k c
m (or 0 >
c
dm );
h e h
T T =

if 0 >
e h
m (or 0 >
e
dm );
e e h
T T =

if 0 <
e h
m (or 0 <
e
dm ).

The system can be solved only by numerical integration. If the values of the unknown functions are
adopted for a certain point in time, the problem is an initial value one and the numerical solution can be
found with a Runge-Kutta method. The solution is obtained after several iterations, each of them using the
previous one's results as initial values and thus getting closer to the result as the analysis goes on.
The useful effect of the Stirling engine is represented by the work L yielded per cycle, calculated
with:

(7)
, d )) ( V (
d
d
) ( p d )) ( V (
d
d
) ( p
)) ( V ( d ) ( p L
e c
t
o o
o
o o o
o
o
o o
t t
} }
}
(

+
(

=
= =
2
0
2
0


where
- V
t
is the total volume occupied by the working agent inside the machine and
- is the crankshaft position angle.
The adiabatic efficiency of the Stirling engine can be calculated with the following general relation:

(8)
e
c
e
c
h
k
h
k
h
ad t
L
L
L
L
Q
Q
Q
Q
Q
L
= + = = + = = 1 1 1 1 q ,

where Q
k
= L
c
and Q
h
= L
e
are the heats cyclically exchanged inside the cooler and the heater and the works
exchanged inside the compression and expansion chambers and L is the work cyclically performed by the
engine.
A closer picture of the real functioning of the engine can be obtained by taking into consideration
various kinds of losses. Losses can be calculated through decoupled models. These models suppose that each
loss is produced by an independent cause.
Results provided by the adiabatic model in this paper were corrected by taking into account two
kinds of losses, determined with models presented by Urieli [7]. The decreasing of the work yielded inside
the expanded chamber as consequence of pumping losses inside heat exchangers, as well as the decreasing of
the yielded work caused by the heat exchange inside non-ideal cooler and heater were considered.

3. Optimization of the Diameter Ratio
The chosen optimization criteria were the maximum work yielded and the maximum efficiency of
the alpha-type Stirling engine.
The paper analyzes a Stirling engine for which we know:
- the working agent (constant R);
- the temperatures T
k
and T
h
of the heat sources;
- the dimensions of the motion mechanism (identical for both displacer and power piston, that as a
consequence share the same stroke);
- the mass of the working agent inside the machine (chosen).
In order to optimize the diameter ratio the maximum of the total volume occupied by the gas inside
the machine was maintained constant. The volumes of the heat exchangers are constant and were calculated
as ratios of the maximum volume of the expansion chamber. The maximum volume of the expansion
chamber used for calculating the heat exchangers volumes was determined for the case when both pistons
share the same diameter.
We use the diameter ratio:

(9)
p
d
d
d
d
= c ,

where d
d
is the diameter of the displacer piston and d
p
is the diameter of the power piston.
At the optimization analysis the total volume occupied by the working agent inside the machine was
considered constant. This condition means that, for engines working with the same mass of agent, the
pressure inside the machine at ambient temperature is the same.
The numerical modeling of the adiabatic machine was used. All machine performances must be
represented as functions of
d
[5]. The maximum values of these functions were considered as optimums.
Because the performances depend on the heat exchanger volumes also, the optimum values depend on these
volumes too.
The isothermal physico-mathematical model of the Stirling engine was also used for calculation. The
same hypotheses were used, except that the temperatures inside compression and expansion chambers were
considered to be constant.

4. Numerical Example
An alpha-type Stirling engine (Fig. 1) featuring the following dimensions was chosen: d
p
= 0.073 m;
d
d
= 0.073 m; r = 0.0365 m; l
1
= l
2
=
0.15 m; f
TDCp
= f
BDCd
= 0.001 m; V
h
= V
k
= 0.1 V
e max
; V
reg
= 1.2 V
e max
, where
V
e max
= maximum volume of the expansion chamber, d = cylinder diameter,
r = crankshaft radius, l = rod length, f = dead space length. V
e max
determined for the chosen Stirling engine
geometry is a constant value for the optimization calculations. TDC and BDC stand for top and bottom dead
center.
L,
[J/cycle]
600
0
800
3 2

d

=

1
.
1
5
7
1
1
d 0
adiabatic
4
isothermal
400
200
with losses

d

=

1
.
0
8
3
2

Fig. 4 Variation of the work on
d
ratio

,
[-]
0.4
0
0.6
3 2

d

=

1
.
7
6
5
8
1
d
0
adiabatic
4
isothermal
0.2
with losses

d

=

1
.
0
9
0
0

Fig. 5 Variation of the adiabatic thermal efficiency on
d
ratio

d
= 1.766

d
= 0.566

d
= 1.1571
0.1 [10
-3
m
3
]
p,
[MPa]
4
0
2
6
0.3 0.2
8
10
V
c
V
e

d
= 1.766

d
= 0.566

d
= 1.1571
0.1 [10
-3
m
3
]
p,
[MPa]
4
0
2
6
0.3 0.2
8
10

Fig. 6 Indicator diagrams inside compression and expansion chambers for different diameter ration
d


The machine works with a total mass of hydrogen m = 0.0025 kg (R
H2
= 4121 J/(kg K) ) between
temperatures T
h
= 773 K and T
k
= 310 K. The main dimensions of the engine were chosen from a project that
aimed to build a Stirling engine with similar characteristics with the internal combustion engine of the
Romanian-made Dacia 1300 vehicle.

0.5 [10
-3
m
3
] 0.8 0.6 0.7
V
t
0.9

d
= 1.766

d
= 0.566

d
= 1.1571
p,
[MPa]
4
2
6
8
10


Fig. 7 Indicator diagram inside the whole engine for different
d
values

The following dimensions were adopted in order to calculate the work losses: cooler length 0.08 m,
regenerator length 0.07 m, heater length 0.08 m, inside diameter of the cooler and heater pipes 0.002 m. The
wire mesh of the regenerator have 400 wires of 0.0254 10
-3
m in diameter per inch. The pressure drop for
regenerator matrix is calculated based on experimental data taken from [8]. The revolution speed was
considered 1000 rpm.
The most important numerical results obtained by optimization of the Stirling machine are presented
in Fig. 4 and Fig. 5. The variations of the terms L and as functions of
d
ratio show that both functions
reach their maximum values.
The influence of
d
ratio over the functioning of the Stirling engine was stressed using the indicator
diagrams p(V). In Fig. 6 and Fig. 7 the indicator diagrams inside compression and expansion chambers (c
and e) and inside the whole engine (t) for the optimum (adiabatic) value
d
and for another two values of
diameter ratio were presented. Values 0.566 and 1.766 were obtained by imposing d
p
= 0.05 m, respectively
d
d
= 0.05 m.

5. Conclusions
The work cyclically performed by the alpha-type Stirling engine and the adiabatic thermal efficiency
reach their maximum values for diameter ratios above 1; so, the diameter of the displacer piston must be
greater than the diameter of the power piston.
The maximum value of the work depends on the physico-mathematical model used for numerical
simulation (e.g. the isothermal model and the adiabatic model without or with losses taken into
consideration).
Values of diameter ratios for which the work performed and the thermal efficiency are null can be
calculated.
The thermal efficiency of the alpha-type Stirling engine is close to the optimum value for a large
range of diameter ratios.
The difference between the maximum and minimum cycle pressure diminishes for greater displacer
piston diameters.

R E F E R E N C E S
1. C o s t e a M., F e i d t M., P e t r e s c u S., Synthesis on Stirling Engine Optimization. Vol. Thermodynamic
Optimization of Complex Energy Systems, edited by Bejan A., Mamut E., NATO Science Series, 3. High
Technologies,
vol. 69, Kluwer Academic Publishers, Dordrecht / Boston / London, 1999,
p 403...410.
2. F e i d t M., L e S a o s K., C o s t e a M., P e t r e s c u M., Optimal Allocation of Heat Exchanger Inventory
Associated with Fixed Power Output or Fixed Heat Transfer Rate Input. Int. J. Applied Thermodynamics, Vol.
5, (No. 1), pp. 25-36, 2002, ISSN 1301-9724.
3. F i n k e l s t e i n Th., O r g a n A., Air Engines. ASME Press, New York, 2001.
4. H o m u t e s c u C. A., S a v i t e s c u Gh., J u g u r e a n u E., H o m u t e s c u V.M., Introducere n maini
Stirling. Ed. CERMI, Iai, 2003.
5. H o m u t e s c u V. M., Optimization of Diameter Ratio for the Vuilleumier Machine Based on the Isothermal
Functioning. The 3rd International Conference "Advanced Concepts in Mechanical Engineering", 5-6 iunie
2008, in Bul. I.P.I., Tom LIV (LVIII), Fasc. 2, p. 475 ... 482, ISSN 1011-2855.
6. P o p e s c u Gh., Maini Stirling. Ed. Bren, Bucureti, 2001.
7. U r i e l i I., Stirling Cycle Machine Analysis. http://www.ent.ohiou.edu/ ~urieli/index.html and
http://www.sesusa.org/DrIz/index.html, excerpts of Urieli I., Berchowitz D.M., Stirling Cycle Machine
Analysis. Athens, Ohio, 1984, retrieved November 2009.
8. W a l k e r G., Stirling Engines, Oxford University Press, Oxford, 1980.