Renewable Energy 35 (2010) 2590e2601

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Numerical model for predicting thermodynamic cycle and thermal efficiency of

a beta-type Stirling engine with rhombic-drive mechanism
Chin-Hsiang Cheng*, Ying-Ju Yu
Department of Aeronautics and Astronautics, National Cheng Kung University, No. 1, Ta-Shieh Road, Tainan 70101, Taiwan R.O.C.

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

Article history: This study is aimed at development of a numerical model for a beta-type Stirling engine with rhombic-
Received 4 January 2009 drive mechanism. By taking into account the non-isothermal effects, the effectiveness of the regenerative
Received in revised form channel, and the thermal resistance of the heating head, the energy equations for the control volumes in
7 February 2010
the expansion chamber, the compression chamber, and the regenerative channel can be derived and
Accepted 5 April 2010
solved. Meanwhile, a fully developed flow velocity profile in the regenerative channel, in terms of the
Available online 7 May 2010
reciprocating velocity of the displacer and the instantaneous pressure difference between the expansion
and the compression chambers, is derived for calculation of the mass flow rate through the regenerative
Keywords:
Numerical model
channel. In this manner, the internal irreversibility caused by pressure difference in the two chambers
Stirling engine and the viscous shear effects due to the motion of the reciprocating displacer on the fluid flow in the
Rhombic drive regenerative channel gap are included. Periodic variation of pressures, volumes, temperatures, masses,
Beta type and heat transfers in the expansion and the compression chambers are predicted. A parametric study of
the dependence of the power output and thermal efficiency on the geometrical and physical parameters,
involving regenerative gap, distance between two gears, offset distance from the crank to the center of
gear, and the heat source temperature, has been performed.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction (3) g-type e this type of configuration has a piston and a displacer,
each in its own cylinder.
Stirling engines are power machines that operate over a closed,
regenerative thermodynamic cycle, with cyclic compression and In addition, a double-acting Stirling engine has multiple cylin-
expansion of the working fluid at different temperature levels, as ders and elongated power pistons, and can be considered as
described by Walker [1]. The working fluid in the Stirling engines a coupled a engines with thermodynamic cycle taking place
might be air, nitrogen, helium, or hydrogen. In theory, ideal Stirling between the top of one piston and the bottom of the next piston.
engines features high thermal efficiency, low emissions, and Stirling engine technology has come a long way in the past
quietness. Most importantly, a variety of heat sources can be several decades. However, new concepts and designs continue to
utilized, including solar energy, waste heat, and fossil fuels. In appear, according to the review of Ross [4]. Among all, the
particular, the solar dish power systems (for example, the SES 25- domestic-scale Stirling electric power generator [5] is particularly
kW SunCatcher [2]) typically consisting a solar dish concentrator, of great market potential.
a sun tracker, and a Stirling engine have received great attention In an ideal Stirling engine, compression and expansion cham-
from the energy-related researchers in recent years. bers were assumed to be maintained at constant volume or
In general, the Stirling engines are classified into three types of isothermal conditions during the heating and the cooling
configuration [3] as follows: processes. However, in 1859 Rankine [6] already mentioned that
neither the heating nor the cooling takes place exactly at constant
(1) a-type e this type of configuration features two pistons, each in volumes or at constant temperatures. A simple second-order
its own cylinder; analysis of the ideal Stirling engine cycle was proposed by
(2) b-type e this type of configuration has a piston and a displacer G. Schmidt in 1871 [7]. The Schmidt analysis has been used widely
in the same cylinder; and as an approximation of Stirling engine performance, and Bercho-
witz and Urieli [8] gave a detailed description of it. The Schmidt
* Corresponding author. Tel.: þ886 6 2757575x63627; fax: þ886 6 2389940. analysis takes the isothermal analysis of Stirling engines one step
E-mail address: [email protected] (C.-H. Cheng). further by assuming that the volume of the expansion and the

0960-1481/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.renene.2010.04.002
 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601 2591

Nomenclature Rt2 thermal resistance of cooling jacket ( C/kW)

t time (s)
C1, C2 coefficient of velocity equation tmax maximum computation time (s)
Cv constant volume specific heat (kJ/kg K) tp period (s)
Cp constant pressure specific heat (kJ/kg K) t0 reference time (s)
d1 diameter of displacer (m) T temperature (K)
d2 core diameter of cylinder (m) TH heat source temperature (K)
G regenerative channel gap (m) TL heat sink temperature (K)
h enthalpy (kJ/kg) U internal energy (kJ/s)
ld height of displacer (m) v velocity (m/s)
lt height of the engine (m) V volume (m3)
l1, l2, l3, l4 lengths of the linkages of the rhombic-drive W _ out power output (kJ/s)
mechanism (m) Wout network output per cycle (kJ/cycle)
L distance between the two gears (m) Y displacement (m)
Ldt length from the linkage l4 to the top surface of
displacer (m) Subscripts
Lpt length from the linkage l1 to the top surface of piston c compression chamber
(m) d displacer
m mass of working fluid (kg) e expansion chamber
m_ mass flow rate (kg/s) p piston
P pressure (kPa) r regenerative channel
Q volumetric flow rate (m3/s)
_
Q in input heat transfer rate (kJ/s) Greek symbols
Qin input heat transfer per cycle (kJ/cycle) 3 regeneration effectiveness
r, 4, z cylindrical coordinates g compression ratio
r1 radius of displacer (m) h thermal efficiency
r2 core radius of cylinder (m) m dynamic viscosity (kg/m s)
R gas constant (J/kg K) q crank angle (rad)
Rd offset distance from the crank to the center of gear (m) r fluid density (kg/m3)
Rt1 thermal resistance of heating head ( C/kW) u rotational speed (rpm)

compression space vary sinusoidally. However, it does not take into driving mechanism by means of a lever. By means of nodal ther-
account non-isothermal effects and internal irreversibility caused modynamic and kinematic analysis, the instantaneous temperature
by fluid friction and the pressure difference in different spaces of distribution of working fluid, through the heatingecooling passage,
the engine. As a result, it is not capable of dealing with the ther- conducting the cold space to hot space, was studied. Parlak, et al.
modynamic phenomena in a real engine. [13] performed a thermodynamic analysis of a g-type Stirling
Recently, the present authors built a 10-W b-type Stirling engine engine by using a quasi steady flow model. In addition, Mahkamov
with a rhombic drive (Model 10ST1), of which the photographs are [14] performed an axisymmetric computational fluid dynamics
shown in Fig 1. According to the experiences learned from the approach to the analysis of the working process of a solar Stirling
development of the engine, it is recognized that major fundamental engine. de Boer [15] tried to find the aximum attainable perfor-
problems that the designers of the Stirling engines have to face mance of Stirling engines and refrigerators. Andersen, et al. [16]
were the time consumption in the design process and the cost of presented a control-volume-based modeling in one space dimen-
fabrication. In most of the cases the determination of the geomet- sion of oscillating, compressible flow in the reciprocating machines.
rical parameters of the engines is merely based on the experience of Reinalter et al. [17] provided detailed performance analysis of
the designers or on insufficient costly experimental information. a 10 kW dish/Stirling system.
Therefore, the improvement of the engine can only be carried out in It has been recognized that these models give a certain amount
a relatively high-cost time-consuming trial-and-error process. In of information that are not easily obtained by the experiments. In
this regard, modeling the engines is expected to be one of the addition, modeling costs much lower since it can be readily used to
alternative solutions to these issues. A number of numerical models evaluate the performance of the engine design before the design is
and simulation methods for different types of the Stirling engines submitted for fabrication. The present model is referred to as the
have been proposed. The Stirling cycle machines under non- second-order model. Herein the three-dimensional CFD model is
isothermal working condition were firstly analyzed by Finkelstein not adopted because it is an expensive method. Though the three-
[9] in 1960s. His analysis was recognized as one significant devel- dimensional CFD model might provide detailed information
opment in the past several decades. His model, by means of regarding the flow pattern and temperature and pressure distri-
introducing the heat transfer coefficient, the finite heat transfer in butions, it requires much more computation efforts than the
the working space was derived. The temperature variation of the present model.
gas in the working chambers was investigated. Schulz and The rhombic drive, which is frequently used on a b-type Stirling
Schwendig [10] proposed a general simulation model for Stirling engine [18], utilizes a jointed rhomboid to convert linear motion of
cycles. A thermodynamic analysis of a Stirling engine including a reciprocating piston to a rotational of fly wheel. In the rhombic
dead volumes of hot space, cold space and regenerator was per- drive mechanism, one rigid rod is installed connecting the piston to
formed by Kongtragool and Wongwises [11]. Most recently, Kar- the top corner of the jointed rhomboid, and another rigid rod
abulut, et al. [12] presented a b-type Stirling engine with a displacer connecting the displacer to the bottom corner of the rhomboid. In
2592 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601

Fig. 1. A 10-W b-type Stirling engine with rhombic-drive (Model 10ST1) and its parts [from Power Engines and Clean Energy Laboratory (PEACE Lab.), National Cheng Kung
University].

addition, two symmetric gears of equal diameter are connected to account the non-isothermal effects, the effectiveness of the
the right and the left corners of the rhomboid fixed on the gears at regenerative channel, and the thermal resistance in the heating
an offset distance from gears center. The two gears are in contact head. Meanwhile, the internal irreversibility caused by pressure
and rotate in opposite directions. When gas pressure is applied to difference in the two chambers and the viscous shear effects due to
the piston, the top corner of the rhomboid is pushed downward; the motion of the reciprocating displacer on the fluid flow in the
the rhomboid is flattened in the direction of the piston axis; and regenerative channel gap are also included.
then it pushes on the gear wheels and causes them to rotate. At the
same time, as the gear wheels rotate, the rhomboid progresses its
change of shape and drives its bottom corner and the displacer to 2. Numerical model
move upward.
The purpose of the present study is aimed at development of A schematic diagram of the engine is illustrated in Fig. 2. The
a numerical model for predicting thermodynamic cycle and effi- working fluid contained in the engine is air. The applied model has
ciency of the b-type Stirling engine with the rhombic drive, in been divided into three control volumes: (1) expansion chamber,
which the annual gap around the displacer acts as a regenerator, so (2) regenerative channel, and (3) compression chamber. Each of the
as to evaluate the thermodynamic performance of the engine and three control volumes can be treated as an open system. It is
predict the periodic variation of pressures, specific volumes, assumed that the working fluid is an ideal gas and that the volumes
temperatures, masses, and heat transfers in the expansion and the of the chambers as well as the pressures, temperatures, and masses
compression chambers. Note that the present model takes into of the air in the expansion and the compression chambers are
 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601 2593

 
Ve ðtÞ ¼ p$r22 lt  Yp ðtÞ (3)
Gas velocity
Displacer Cylinder
profile  
Vc ðtÞ ¼ p$r22 Yd ðtÞ  ld  Yp ðtÞ (4)
vd
where r2 is the core radius of cylinder. Furthermore, differentiating
Yp(t) and Yd(t) of Eqs. (1) and (2) with respect to t yields the
Regenerative velocities of the piston and the displacer, vp(t) and vd(t), respec-
channel tively, as
Expansion
8 "
chamber TH <  2 #1=2
L l1
vp ðtÞ ¼ Rd u cos q  sin q l2 
2
  Rd cos q
d2(=2r2) : 2 2
9
 =
L l1
   Rd cos q ð5Þ
2 2 ;
Yd
8 "
d1(=2r1)
<  2 #1=2
L l4
vd ðtÞ ¼ Rd u cos q þ sin q l23    Rd cos q
: 2 2
9
 =
L l4
TL Compression    Rd cos q ð6Þ
2 2 ;
chamber

2.1. Expansion chamber

Yp
The expression of volume of the expansion chamber can be
further written by inserting Eqs. (1)e(3) as
8 "
<
Ve ¼ pr22 lt  Ldt þ Rd sinq
:
9 (7)
  2 #1=2 =
L l4
 l23    Rd cos q
2 2 ;

Using the ideal-gas equation of state, the pressure in the

expansion chamber is calculated by

me RTe
Pe ¼ (8)
Ve
Fig. 2. Schematic of b-type Stirling engine with rhombic drive. where me is the mass and Te is the temperature of the air contained
in the expansion chamber, and R is the gas constant..
Next, the air temperature in the expansion chamber (Te) is
varying with time transiently. However, all the time-varying
calculated based on the energy conservation law:
properties are uniform inside the chambers at any time instant.
Displacements of the piston and displacer connected with the   2 
dU  _ out;e  dm h þ v 
 ¼ Q_ in;e  W (9)
2 e
rhombic-drive mechanism, Yp(t) and Yd(t), are given as:
dt e dt
"  2 #1=2
L l1 where the values of h and v is chosen in accordance with the
Yp ðtÞ ¼ Lpt þ Rd sin q þ l22    Rd cos q (1) direction of the mass flow. Here, the mass flowing out of the
2 2
expansion chamber is prescribed to be positive. Therefore:
For mass leaving the expansion chamber and entering the
"  2 #1=2
L l4 regenerative channel:
Yd ðtÞ ¼ Ldt þ Rd sin q  l23    Rd cos q (2)
2 2
dm
> 0; h ¼ he ; v ¼ ve (10a)
where q is the crank angle; Rd is the offset distance from gear dt
centers to the joints between the rhomboid and the gears; l1, l2, l3, where he is the specific enthalpy of the air contained in the
and l4 represent lengths of the linkages of the rhombic-drive expansion chamber, and ve is the average velocity of the air leaving
mechanism; Ldt denotes the length from the linkage l1 to the top the expansion chamber calculated with ve ¼ jdmdt
j=½re pðr22  r12 Þ.
surface of the displacer; and Lpt is the length from the linkage l4 to For mass entering the expansion chamber from the regenerative
the top surface of the piston. channel:
The volumes of the expansion and the compression chambers,
Ve(t) and Vc(t), can be calculated, respectively, in terms of Yp(t), dm
< 0; h ¼ hj ; v ¼ vj (10b)
Yd(t), and the cross-sectional area of the cylinder: dt
2594 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601

where hj is the specific enthalpy of the air exiting the regenerative where the coefficients, C1 and C2 are determined based on the
channel, and vj is the average velocity of the air calculated with following boundary conditions:
vi ¼ jdm
dt
j=½rj pðr22  r12 Þ. Let
B:C  1 : r ¼ r1 ; vz ¼ vd ðtÞ (15a)
dU  .
j ¼ me Cv Tenþ1  Ten Dt (11)
dt e B:C  2 : r ¼ r2 ; vz ¼ 0 (15b)
where me ¼ ðmnþ1 e þ mne Þ=2 and Dt is the time step used in the where the reciprocating velocity of the displacer, vd(t), is deter-
time marching computation. In this study, Dt is typically assigned to mined by Eq. (6). Then, the velocity profile in the regenerative
be 1 106 s. channel is carried out as
With the help of Eqs. (10) and (11), Eq. (9) can be discretized into
"  2  #
the finite-difference form as 1 dP 2 r2  r12 r vd ðtÞ
2
  vZ ¼ r  r2  ln  ln r
  " nþ1 n
4m dz lnðr2 =r1 Þ r2 lnðr2 =r1 Þ
Dt Dt TH Pe Ve  Ve
Tenþ1 ¼ 1 Ten þ  vd ðtÞ
me Cv Rt1 me Cv Rt1 Dt þ ln r2 (16)
lnðr2 =r1 Þ
 #
dm v2 dm
 he þ e ; for >0 ð12aÞ In the above equation, it is noticed that the effects of pressure
dt 2 dt difference in the two chambers and the viscous shear effects due to
the motion of the reciprocating displacer on the fluid flow in the
or
regenerative channel gap are included. The volumetric flow rate Q
"   can then be calculated as
  nþ1 n
Dt Dt TH Pe Ve  Ve
Tenþ1 ¼ 1 Ten þ  Zr2  "  2 2 #
me Cv Rt1 me Cv Rt1 Dt p dP r2  r12
!# Q ¼ vZ ð2prÞ dr ¼  4 4
r2  r1 
dm vj2
dm 8m dz lnðr2 =r1 Þ
 hj þ ; for <0 ð12bÞ r1

dt 2 dt 2pvd ðtÞ 1 2  1 

 r2 ln r2  r12 ln r1  r22  r12

where TH is the heat source temperature; P e ¼ ðPenþ1 þ Pen Þ=2; and lnðr2 =r1 Þ 2 4

the specific enthalpy h is calculated by h ¼ Cp T. Note that the input vdðtÞ 1 2 1 2

þ 2p ln r2 r2  r1 ð17Þ
heat transfer rate is defined by Q_ in;e ¼ ðTH  Te Þ=Rt1 and the lnðr2 =r1 Þ 2 2
output power due to volume expansion is determined by The mass flow rate through the regenerative channel is
W_ out;e ¼ P e ðV nþ1  V n Þ=Dt. In addition, Rt1 is thermal resistance of
e e
the heating head, and in this study Rt1 is set to be 0.5  C/W.  "  2 2 #
dm rp Pe  Pc 4 4 r2  r12
However, it is important to mention that the thermal resistance is ¼  r2  r1 
dt 8m ld lnðr2 =r1 Þ
basically dependent on the heating condition, the materials, and
2pvd ðtÞr 1 2  1 

the geometry of the heating head. Hence, for a specific case, further
 r2 ln r2  r12 ln r1  r22  r12
experiments or numerical simulation should be required to deter- lnðr2 =r1 Þ 2 4

mine the value of Rt1. Eqs. (12a) and (12b) are employed to calculate vd ðtÞr 1 2 1 2
the air temperature in the expansion chamber. þ 2p ln r2 r2  r1 (18)
lnðr2 =r1 Þ 2 2
where dP/dZ has been approximated with ðPe  Pc Þ=ld . Using the
2.2. Regenerative channel
obtained value of dm/dt, the masses of the air contained in the
expansion and the compression chambers can be updated as
In the present model, the gap between the outer surface of the
displacer and the inner surface of the cylinder is referred to as the
dm n
regenerative channel. The model of regenerator used here is mnþ1
e ¼ mne  j Dt (19a)
dt
a simplified model which assumes that the gas is incompressible
throughout the interior chambers. The gas in the regeneration
dm n
channel enters at one end and exits at another, and the direction mnþ1
c ¼ mnc þ j Dt (19b)
and the mass flow rate of the gas flow in the regeneration channel dt
is dependent on the reciprocating velocity of the displacer and the where the superscripts n and n þ 1 represent two consecutive time
instantaneous pressure difference between the expansion and the steps.
compression chambers. Since the regenerative channel is usually The mass conservation principle requires that the mass flow rate
rather small, it is reasonable to approximate the fluid flow in the through the entire regenerative channel will remain constant. That
channel as a fully developed flow. It implies that vr ¼ vf ¼ 0, is,
vZ ¼ vZ ðrÞ. For a quasi-steady, fully developed, incompressible  
annular flow, the momentum equation can be simplified as follows. dm
_ in;r ¼ m
m _ ¼ 
_ out;r ¼ m  (20)
  dt 
1 dP 1 v vvz
¼ r (13) Applying energy conservation principle to the regenerative
m dz r vr vr
channel, one has
Integrating the above equation yields the velocity inside the
 2 
channel. vin v2out 
Q_ in;r ¼ mðh
_ in  hout Þjr  m
_   (21)
2 2 r
r 2 dP
vZ ¼ þ C1 ln r þ C2 (14)
4m dz Due to the upwind effects, the inlet and outlet temperatures of
the regenerative channel are different as the air flows in different
 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601 2595

Start

Te Expansion chamber Te
Tj Initial
conditions

t=t+ t

Update
Yd(t), Yp(t), Vd(t),
vd Vp(t) , vd(t), vp(t)
Stop
Calculate
Pe, Pc
Calculate Wout, No

t tmax? Calculate
Yes
dm n +1 n +1
, me , mc
Tc dt

Compression chamber Tc Calculate Ti, Tj

Ti Calculate
· · ·
Qin ,e ,Qin ,c ,Q in ,r
Calculate
Ten +1 , Tcn +1
Fig. 3. Inlet and outlet temperatures of air flowing through regenerative channel. (For
the air flowing from the expansion to the compression chambers, the inlet tempera-
ture is Te and outlet temperature is Ti. For the air from the compression to the Yes P, m, T No
expansion chambers, the inlet temperature is Tc and outlet temperature is Tj.) Converge?

Fig. 4. Flow chart of the computation process.

directions. Fig. 3 shows the inlet and outlet temperature of air
flowing through the regenerative channel. For the air flowing from to be a function of G only. Based on the theory given by Organ [20],
the expansion to the compression chambers, the channel inlet the function of e is approximated as a linear function in terms of G
temperature is Te and channel outlet temperature is Ti. For the air which is varied from 0.1 (at G ¼ 0.0008 m) to 0.8 (at G ¼ 0.0002 m).
flowing from the compression to the expansion chambers, the Hence, at G ¼ 0.0005 m, 3 is calculated to be 0.45. The regeneration
channel inlet temperature is Tc and channel outlet temperature is Tj. effectiveness is relatively low since the regenerator here is simply
Note that the values of the outlet temperatures, Ti and Tj, are an annular channel. However, note that for a more complicated
strongly dependent on the effectiveness of the regenerative regenerator, a study of the effectiveness of the regenerator is
channel. definitely necessary. In consequence, one has:
The regenerator quality is usually defined on an enthalpy basis For mass flowing from the expansion to the compression
in terms of a regeneration effectiveness as follows: chambers:

actual enthalpy change through the regenerator

3 ¼ (22)
maximum theoretical enthalpy change through the regenerator

Based on the above definition of the regeneration effectiveness,

the outlet temperatures, Ti and Tj, are determined by dm
> 0; Tin ¼ Te ; hin ¼ he ; vin ¼ ve ; Tout ¼ Ti ;
dt
dm
Ti ¼ Te þ 3ðTc  Te Þ; for >0 (23a) hout ¼ hi ; vout ¼ vi ð24aÞ
dt
and For mass flowing from the compression to the expansion
chambers,:
dm
Tj ¼ Tc þ 3ðTe  Tc Þ; for <0 (23b)
dt dm
< 0; Tin ¼ Tc ; hin ¼ hc ; vin ¼ vc ; Tout ¼ Tj ;
respectively. In general, the effectiveness of the regenerator is dt
a function of the geometry and the properties of the porous hout ¼ hj ; vout ¼ vj ð24bÞ
medium used. The design of the regenerator can be optimized to
yield a higher effectiveness [19]. As for the regenerative channel In the above equations, as already stated earlier, all the specific
considered here, the regeneration effectiveness is only dependent enthalpies are calculated by h ¼ CpT based on the associated
on the size of the gap (G) and the channel length (ld). In this study, temperatures, and all the average velocities of the air are calculated
the channel length ld is fixed at 0.07946 m. Hence, e can be assumed by v ¼ jdm
dt
j=½rpðr22  r12 Þ.
2596 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601

Eq. (21) is used to determine the heat transfer rate (Q_ in;r ) Table 2
between the air flow and the walls of the regenerative channel. As Operating variables of the base-line case (model 10ST1).

Q_ in;r > 0, it means that heat is transferred from the regenerative Engine speed, u 2000 rpm
channel walls to the air, and on the contrary, as Q_ in;r < 0, it means Compression ratio 1.7420
that heat is rejected from the air to the channel walls. Heat source temperature, TH 800 K
Heat sink temperature, TL 300 K
Initial pressure, PH, PL 101.3 kPa
Thermal resistance, Rt1 0.5  C/W
2.3. Compression chamber Thermal resistance, Rt2 0.4  C/W
Regenerative effectiveness, 3 0.45
Introducing Eqs. (1) and (2) to Eq. (4) leads to the volume of the
compression chamber:
8"
<   2 #1=2 dU  .
L l4 j ¼ mc Cv Tcnþ1  Tcn Dt
Vc ¼ pr22 Ldt þ Rd sinq     Rd cos q
l23 dt c
(29)
: 2 2
" 9 where mc ¼ ðmnþ1 þ mnc Þ=2. Note that the input heat transfer rate
  2 #1=2 = c
L l1 is defined by Q_ in;c ¼ ðTL  Tc Þ=Rt2. Since TL < Tc usually, the value of
 Lpt þ Rd sin q þ l2 
2
  Rd cos q ld
2 2 ; Q_ in;c is negative in most of the time. The output power is deter-
mined by W _ out;c ¼ ðP c ðV nþ1  V n ÞÞ=Dt. Thus, Eq. (27) reduces to
c c
(25)
"  
  Pc Vcnþ1  Vcn
From the ideal-gas equation of state, the pressure in the Dt Dt TL
Tcnþ1 ¼ 1 Tcn þ 
compression chamber is mc Cv Rt2 mc Cv Rt2 Dt
!#
mc RTc dm v2 dm
Pc ¼ (26) þ hi þ i ; for >0 ð30aÞ
Vc dt 2 dt
and, the gas temperature in the compression chamber (Tc) is
or
calculated from energy equation
"  
   
dU  2 Dt Dt TL Pc Vcnþ1  Vcn
¼ _
Q  _ out;c þ dm h þ v j
W (27) Tcnþ1 ¼ 1 Tn þ 
dt c in;c
dt 2 c mc Cv Rt2 c mc Cv Rt2 Dt
 #
where the values of h and v is chosen in accordance with the dm v2c dm
þ hc þ ; for <0 ð30bÞ
direction of the mass flow. Please be reminded that the mass dt 2 dt
flowing into the compression chamber is prescribed to be positive.
Therefore: where TL is the heat sink temperature; mc ¼ ðmnþ1 c þ mnc Þ=2 and
nþ1 n
P c ¼ ðPc þ Pc Þ=2; and Rt2 is thermal resistance of the cooling
For mass entering the compression chamber from the regener-
ative channel: jacket, which is assigned to be 0.4  C/W. Eqs. (30a) and (30b) are
used to calculate the air temperature in the compression chamber.
dm
> 0; h ¼ hi ; v ¼ vi (28a)
dt 2.4. Network output and thermal efficiency
where hi is the specific enthalpy of the air entering the compression
The network output per cycle is further determined by inte-
chamber from the regenerative channel, and vi is the average
grating P with respect to V for a cycle in accordance with the PeV
velocity calculated with vi ¼ jdm
dt
j=½rj pðr22  r12 Þ.
relations taken in the expansion and the compression chambers.
For mass leaving the expansion chamber and entering the
That is,
regenerative channel:
0 þtp
tZ 0 þtp
tZ
dm
< 0; h ¼ hc ; v ¼ vc (28b) Wout ¼ Pe dVe þ Pc dVc (31)
dt
t0 t0
where hc is the specific enthalpy of the air contained in the
compression chamber, and vc is the average velocity of the air The net heat transfer into the air per cycle is a summation of the
calculated with vc ¼ jdm
dt
j=½rc pðr22  r12 Þ. Again, the time derivative heat transfers in the expansion chamber, the compression chamber,
term on the right-hand side of equation is approximated by and the regenerative channel, which are obtained by integrating
the heat transfer rates with respect to time for a cycle:
Table 1
Geometrical variables of the base-line case (Model 10ST1).
Table 3
r2 (m) 0.0205 Computation results for the base-line case (Model 10ST1).
r1 (m) 0.02
G (m) 0.0005 Work output in expansion chamber (kJ/cycle) 1.236  103
l1(¼l2 ¼ l3 ¼ l4) (m) 0.018 Work input in compression chamber (kJ/cycle) 7.335  104
L (m) 0.042 Heat transfer in expansion chamber (kJ/cycle) 3.826  103
Rd (m) 0.0036 Heat transfer in compression chamber (kJ/cycle) 3.288  103
Lpt (m) 0.05093 Heat transfer in regenerative channel (kJ/cycle) 3.623  105
Ldt (m) 0.16374 Net heat input Qin (kJ/cycle) 5.025  104
ld (m) 0.07946 Network output Wout (kJ/cycle) 5.025  104
lh (m) 0.158 Thermal efficiency 0.131
Piston stroke (m) 0.01 Power output (W) 16.75
 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601 2597

3. Results and discussion

Table 3 shows the computation results for the base-line case. In

this table, it is found that the network output per cycle
(5.025  104 kJ/cycle) closely agrees with the net heat transfer into
the air per cycle. The numerical results completely comply with the
energy conservation law. Furthermore, it is noticed that the engine
power output is 16.75 W, and the thermal efficiency reaches merely
13.1%. Such a poor thermal efficiency is frequently encountered
with the small-scale engines similar to Model 10ST1. However, the
magnitudes of the engine power output and the thermal efficiency
may be elevated by modifying the geometrical design or by
increasing the regeneration effectiveness of the regenerative
channel.
Detailed results from the present analysis of the base-line case
are provided in Figs. 5e12. Fig. 5 illustrates the periodic displace-
ment of the displacer and the piston in the cylinder for the base-
line case. During the cycles, the displacer moves at a phase angle
ahead of the piston in order to draw the working gas traversing
back-and forth between the expansion and the compression
chambers for heating or cooling. At some instants, the displacer and
Fig. 5. Displacement of the piston and the displacer. the piston were rather close to each other. The minimum distance
between the piston and the displacer is treated to be an influential
factor affecting the dead zone volume of the engine. Less minimum
distance means smaller dead zone in the compression chamber and
0 þtp
tZ 0 þtp
tZ 0 þtp
tZ more working gas in the compression chamber being drawn into
the expansion chamber for heating provided the piston and the
Qin ¼ Q_ in;e dt þ Q_ in;c dt þ Q_ in;r dt (32)
displacer do not coincide with each other.
t0 t0 t0
Fig. 6 illustrates the volume variation of the expansion chamber
where t0 is an arbitrarily-selected reference time and tp represents and compression chamber. For the base-line case, the volume
the period of a cycle. Note that the energy conservation law variation of the compression chamber is larger than that of the
requires that during the cyclic operation the network output per expansion chamber. The volume of expansion chamber is varied
cycle should be equal to the net heat transfer into the air per cycle. from 3 to 16.1 cm3, whereas the volume of compression chamber is
This will be observed from the results later. from 2 to 20.3 cm3. In the beginning of the plot, the piston and the
The thermal efficiency of the Stirling engine can be expressed as displacer are moving downwards, and the volumes of both cham-
bers are increased. After a short while, the expansion chamber
Wout volume soon reaches its maximum and starts to decrease while the
h¼ (33)
0 þtp
tZ compression chamber volume is increased. When the volume of
Q_ in;e dt the expansion chamber reaches its minimum, the volume of the
compression chamber reaches its maximum almost at the same
t0
time.
Fig. 4 shows the flow chart of the computation process. The Fig. 7 shows the pressure variation inside the expansion and
computation starts from the initial conditions for the air inside the compression chambers. The initial pressure of the two chambers is
engine. In the beginning, the pressure in the entire engine is set to
be 101.3 kPa and the fluid temperature is 300 K. The computation
then marches to the next time step and the solutions for pressures,
masses, and temperatures in the expansion chamber, the
compression chamber, and the regenerative channel are carried out
by iteration at each time step. When the time exceeds the required
maximum time, the solution process is terminated. The state of the
system starts from the initial conditions and then achieves a stable
periodic regime eventually. The periodic variation of the properties
and heat transfers in different spaces of the engine can then be
recorded and the network output and the thermal efficiency are
further investigated.
The geometrical and the operating variables of Model 10ST1,
which is regarded as the base-line case in this study, are listed in
Tables 1 and 2, respectively. In this case, the heat source and the
heat sink temperatures are maintained at 800 and 300 K, respec-
tively. The rotational speed of the engine is set at 2000 rpm, and
the engine is operated at around atmospheric pressure. Based on
the geometrical variables given, the compression ratio is 1.742. Note
that some of the variables may be changed to investigate their
effects on the performance of the engine while other variables
are fixed. Fig. 6. Volume variation of the expansion and the compression chambers.
2598 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601

Fig. 7. Pressure variation of the expansion and the compression chambers.

Fig. 9. Temperature variation in the expansion and the compression chambers.

101.3 kPa at t ¼ 0. The pressure then is varied periodically during

cycles. The pressure difference between these two chambers is one chamber could be overcooled due to expansion of the volume. This
of the driving forces that make the fluid flow through the regen- model applies the energy equation to the control volumes. By
erative channel. For this case, no higher pressure difference is assuming the specific heat of the working gas, air, and the thermal
observed since the engine was relatively small in size. The pressure resistance in the expansion chamber and compression chamber,
difference data are shown in Fig. 8. Expansion chamber is full of the temperature variation is calculated during cycles. Since the
high temperature working gas in a cycle, so the pressure inside is temperatures of heat source and the heat sink are given, the heat
larger than that in the compression chamber usually. However, transfer in the chambers can be calculated in terms of the thermal
when the volume of the expansion chamber is further increased, resistances and the temperature difference between the fluid
the pressure inside is reduced to a certain low value. In that case, temperature and the heat source or heat sink temperature.
the pressure may be lower in the expansion chamber than in the Fig. 10 shows the magnitudes of heat transfer rates in the
compression chamber, and the pressure difference becomes expansion and the compression chambers. In this figure, the value
negative. of the heat transfer rate is positive indicating that the heat is
Fig. 9 illustrates the temperature variations in the expansion transferred from the surroundings to the chamber. On the contrary,
chamber and compression chamber. The heat source and the heat the value of the heat transfer rate is negative indicating that the
sink are maintained at 800 K and 300 K, respectively. It is inter- heat is transferred from the chamber to the surroundings. It is
esting to note that the temperature in the expansion chamber could observed that the heat transfer rate in the expansion chamber is not
be higher than 800 K at some instants, and on the contrary, the always positive and that in compression chamber is not always
temperature in the compression chamber could be lower than negative. The reason for this has already been discussed earlier.
300 K. This is because the air in the expansion chamber could be
overheated due to compression and the air in the compression

Fig. 8. Pressure difference between the expansion and the compression chambers. Fig. 10. Heat transfer rate in the expansion and the compression chambers.
 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601 2599

Fig. 13. Effects of the regenerative gap (G).

Fig. 11. Mass variation in the expansion and the compression chambers.
than 0.0003 m. This is attributed to the significant increase in the
Fig. 11 shows the variation in the masses of the working fluid in shear friction loss as the gap size is reduced. On the other hand,
the expansion and compression chambers. Total amount of mass in when the gap size is increased, the regenerative effectiveness of the
the engine remains constant; therefore, as the mass in the expan- regenerative channels would be reduced. Therefore, there should
sion chamber is decreased, the mass in the compression chamber be an optimal value for the gap size that leads to a peak value of the
will be increased. power output or the thermal efficiency. The curve of the power
Fig. 12 shows the PeV diagrams for the expansion chamber and output indeed reaches a peak value of 16.75 W at G ¼ 0.0005 m. The
compression chambers. The areas of the clockwise and counter- thermal efficiency exhibits the same tendency, and at G ¼ 0.0003 m
clockwise PeV cycles with the expansion and the compression the thermal efficiency reaches its peak value. In this figure, it is
R t0 þtp
chambers are calculated by integrating Pe dVe and found that the maximum value of the thermal efficiency of the
R t0 þtp t0
t0 Pc dVc , respectively. The network output is thus obtained engine is 16.5%.
from Eq. (31). Fig. 14 illustrates the effects of the heat source temperature (TH)
Next, effects of the size of the regenerative channel (G), heat on the engine’s power output and thermal efficiency. It is observed
source temperature (TH), the distance between the two gears (L), that the power output can be increased from 7.96 to 32.78 W as TH
and the offset distance from the crank to the center of gear (Rd) are is elevated from 600 to 1200 K. The value of TH is strictly restricted
studied and the results are shown in Figs. 13e16. by the induced thermal stress that the materials can endures. In the
Fig 13 shows the effects of the size of the regenerative gap on the figure, the power output seems to be linearly varied with TH.
power output and thermal efficiency of the engine. The regenera- The effects of the distance between the center of the gears (L) on
tive gap size is varied from 0.00021 to 0.0008 m. The power output the power output and the thermal efficiency of the engine are
and thermal efficiency increase rapidly with the gap size as G is less shown in Fig. 15. It is found that the power output is reduced from

Fig. 12. PeV diagrams for the expansion and the compression chambers.
2600 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601

Fig. 14. Effects of heat source temperature (TH). Fig. 16. Effects of the offset distance from the crank to the center of gear (Rd).

18.286 to 15.3 W, and the thermal efficiency from 0.1377 to 0.123, as 4. Concluding remarks
the value of L is increased from 0.040 to 0.044 m.
Fig. 16 shows the power output and the thermal efficiency as A theoretical model which is applied to simulation of thermo-
functions the offset distance from the crank to the center of gear dynamic cycle and performance of a beta-type Stirling engine with
(Rd). It is found that as Rd becomes larger, the power output and the rhombic-drive mechanism has been proposed in this study. Peri-
thermal efficiency are both increased. Power output can be varied odic variation of pressures, volumes, temperatures, masses, and
from 10.139 to 24.137 W with Rd changing from 0.0027 to 0.0045 m, heat transfers in the expansion and the compression chambers are
while thermal efficiency from 0.106 to 0.150. Obviously, both the predicted, and a parametric study of the dependence of the power
power output and the thermal efficiency monotonically increase output and thermal efficiency on the geometrical and physical
with Rd. parameters for a base-line case is performed.
It is interesting to mention that the distance between the Results show that by adjusting the influential parameters
centers of the gears (L) and the offset distance from the crank to the including regenerative gap, distance between two gears, offset
center of gear (Rd) exhibit opposing effects on the performance of distance from the crank to the center of gear, and the heat source
engine according to Figs. 15 and 16 These two parameters greatly temperature, the performance of the base-line case can be
affect the strokes of the piston and the displacer. In accordance with improved. The power output of the base-line case reaches a peak
the obtained results, it is noted that the combination of these two value of 16.75 W at G ¼ 0.0005 m, accompanied by a thermal effi-
parameters should be carefully designed for different requirements ciency of only 13.1%. If the thermal efficiency is of major concern,
in engine performance. the thermal efficiency can be elevated to a peak value of 16.5% at
G ¼ 0.0003 m.
It is also observed that the power output of the base-line case
can be increased from 7.96 to 32.78 W as TH is elevated from 600 to
1200 K. However, the value of TH is strictly restricted by the induced
thermal stress that the materials can endures.
Meanwhile, an increase in the distance between the center of
the gears (L) leads to a decrease in the power output and the
thermal efficiency. On the other hand, it is found that the power
output and the thermal efficiency both increase with the offset
distance from the crank to the center of gear (Rd). The two
parameters, L and Rd, exhibit opposing effects on the performance
of engine.

Acknowledgement

The authors would like to thank the Bureau of Energy, Ministry

of Economic Affairs, Taiwan, R.O.C., for their financial support under
Grant 99-D0101.

References

[1] Walker G. Stirling engines. Oxford: New York; 1980.

[2] SES SunCatcher. Available from: http://www.stirlingenergy.com.
Fig. 15. Effects of the distance between two gears (L). [3] Ross A. Making stirling engines. Ross Experimental; 1993.
 C.-H. Cheng, Y.-J. Yu / Renewable Energy 35 (2010) 2590e2601 2601

[4] Ross B. Status of the emerging technology of Stirling machines. IEEE AES [13] Parlak N, et al. Thermodynamic analysis of a gamma type Stirling engine in
System Magazine 1995;June. non-ideal adiabatic conditions. Renewable Energy 2009;34(1):266e73.
[5] Berchowitz DM. The development of a 1 kW electrical output free piston [14] Mahkamov K. An axisymmetric computational fluid dynamics approach
Stirling engine alternator unit. In: The 18th intersociety energy conversion to the analysis of the working process of a solar Stirling engine. Journal
engineering conference, Orlando, Florida; 1983. p. 897e901. of Solar Energy Engineering e Transactions of the ASME 2006;128
[6] Rankine WJM. Thermodynamics. London: Transactions of Royal Society; 1859. (1):45e53.
[7] Schmidt G. Theorie der Lehmannschen calorischen maschine. Zeit Des Ver- [15] de Boer PCT. Maximum attainable performance of stirling engines and
eines deutsch Ing 1871;15:1e12. refrigerators. Journal of Heat Transfer e Transactions of the ASME 2003;125
[8] Urieli I, Berchowitz D. Stirling cycle engine analysis, Alternative sources of (5):911e5.
energy. Bristol: Adam Hilger; 1984. p. 71. [16] Andersen SK, Carlsen H, Thomsen PG. Control volume based modelling in one
[9] Finkelstin T. Thermodynamic analysis of Stirling engines. Journal of Spacecraft space dimension of oscillating, compressible flow in reciprocating machines.
and Rockets 1967;4(9):1184. Simulation Modelling Practice and Theory 2006;14(8):1073e86.
[10] Schulz S, Schwendig F. A general simulation model for stirling cycles. Journal [17] Reinalter W, et al. Detailed performance analysis of a 10 kW dish/Stirling
of Engineering for Gas Turbines and Power e Transactions of the ASME system. Journal of Solar Energy Engineering e Transactions of the ASME
1996;118(1):1e7. 2008;130(1).
[11] Kongtragool B, Wongwises S. Thermodynamic analysis of a Stirling engine [18] Hargreaves CM. The Philips Stirling engines. New York: Elsevier; 1991.
including dead volumes of hot space, cold space and regenerator. Renewable [19] Jones JD. Optimization of Stirling engine regenerator design. In: Proceedings
Energy 2006;31(3):345e59. of the 25th Intersociety Energy Conversion Engineering Conference (IECEC-
[12] Karabulut H, Aksoy F, Ozturk E. Thermodynamic analysis of a beta type Stir- 90); August 12e17, 1990. p. 359e65.
ling engine with a displacer driving mechanism by means of a lever. [20] Organ AJ. The regenerator and the Stirling engine. London: Mechanical
Renewable Energy 2009;34(1):202e8. Engineering Publications; 1997.