Vol.

122

ACTA PHYSICA POLONICA A

(2012)

No. 4

Eect of Expansion-Compression Ratio

on Performance of the Miller cycle

R. Ebrahimi

Department of Agriculture Machine Mechanics, Shahrekord University

P.O. Box 115, Shahrekord, Iran

(Received June 10, 2011)

The objective of this study is to analyze the eect of expansion-compression ratio on the performance of dual
cycle. Using nite-time thermodynamics, the relations between thermal eciency, power output and compression
ratio for an air standard Miller cycle have been derived. In the model, the nonlinear relation between the specic
heats of working uid and its temperature, the frictional loss computed according to the mean velocity of the
piston, and heat transfer loss are considered. The results show that the power output rst increases with the
increasing expansion-compression ratio and then starts to decrease. Comparisons of the power output of the
Miller, Otto, and Atkinson cycles show that if compression ratio is less than certain value, the power output
for Otto cycle is higher, while if compression ratio exceeds certain value, the power output for the Miller cycle
is higher. With further increase in compression ratio, the power output for Atkinson cycle is higher. In high
compression ratio, the power output of the Miller cycle is higher. The results obtained in the present study
provide guidance to the performance evaluation and improvement for practical internal combustion engines.
PACS: 05.70.Ln, 82.60.Fa, 88.05.b, 88.20.td

1. Introduction

A series of achievements have been made since nite-time thermodynamics was used to analyze and optimize
the performance of real thermodynamic processes, devices and cycles [15]. In 1940's, Miller [6] proposed a
dierent Otto cycle with unequal compression and expansion stroke called the Miller cycle. The Miller cycle has been put attention recently [7], and some authors have examined the nite-time thermodynamic performance of the Miller cycle. Hatamura et al. [8] report
that the Miller cycle has advantages such as higher mean
eective pressure than the Otto cycle with lower nominal
compression ratio. Fukuzawa et al. [9] described the main
technologies and performance specications for a high efciency Miller cycle gas engine as well as for the series
of engines planned in the future. Al-Sarkhi and Akash
[10] compared the performance characteristic curves of
the Atkinson cycle with those of the Miller and Joule
Brayton cycles by using numerical examples, and outlined the eect of maximizing power density on the performance of the cycle eciency. Sasaki et al. [11] report
an ecient Miller cycle with a high performance capacitor system for hybrid buses. Wu et al. [12] performed a
performance analysis and optimization of a supercharged

e-mail:

Rahim.Ebrahimi@gmail.com

Miller cycle Otto engine. Their zero-dimensional analysis

is based on the Otto cycle. They report that the Miller
cycle shows no eciency advantage and suers a penalty
in power output in the normally aspirated version.
In the supercharged Otto engine adopted for the Miller
cycle, it has again no eciency advantage but does provide increased output with a reduced propensity to engine knock problem. Ge et al. derived the performance
characteristics of the Miller cycle with heat transfer
loss [13] and with heat transfer and friction-like term
losses [14], respectively. These works were done without
considering the variable specic heats of working uid.
Ge et al. [15] also studied the eect of variable specic
heats of working uid on performance of endoreversible
Miller cycle. Chen et al. [16] built a class of generalized irreversible universal steady ow heat engine cycle
model consisting of two heating branches, two cooling
branches, and two adiabatic branches with consideration
of the losses of heat resistance, heat leakage, and internal irreversibility. The performance characteristics of the
Diesel, Otto, Brayton, Atkinson, dual and Miller cycles
were derived.
As can be seen in the relevant literature, the investigations of the eect of expansion-compression ratio on
performance of the Miller cycle do not appear to have
been published. Therefore, the objective of this study is
to examine the eect of expansion-compression ratio on
the performance of air standard Miller cycle.

(645)

R. Ebrahimi

646
2. An irreversible Miller cycle model

The pressurevolume (P V ) and the temperature entropy (T S ) diagrams of an irreversible Miller heat engine is shown in Fig. 1, where T1 , T2 , T3 , T4 , and T5 are
the temperatures of the working substance in state points
1, 2, 3, 4, and 5. Process 1 2 is an isentropic compression. The heat addition occurs in the constant volume
process 2 3. The process 3 4 is an isentropic expansion process. The heat addition occurs in two steps:
processes 4 5 and 5 1 are constant volume and
constant pressure heat additions, respectively.

The heat added per second to the working uid during

process 2 3 is
T3
(
Qin = m
t
CV dT = m
t 8.353 1012 T 3
T2

+ 5.818 108 T 2.5 2.123 107 T 2

+ 2.108 105 T 1.5 + 1.0433T + 3.024 104 T 0.5
)T 3
3.063 105 T 1 + 1.106 107 T 2
.
(3)
T2

The heat rejected per second by the working uid during

processes 4 5 and 5 1 is
[
]
T5
T4
Qout = m
t
CV dT +
Cp dT

T5

T1

=m
t 8.353 1012 T 3 + 5.818 108 T 2.5
2.123 107 T 2 + 2.108 105 T 1.5
+ 1.0433T + 3.024 104 T 0.5

)T 4
3.063 105 T 1 + 1.106 107 T 2
T5
(
12 3
8 2.5
+m
t 8.353 10 T + 5.818 10 T
2.123 107 T 2 + 2.108 105 T 1.5
Fig. 1. (a) P V diagram; (b) T S diagram for the air
standard Miller cycle.

In most cycle models, the working uid is assumed to

behave as an ideal gas with constant specic heats. But
this assumption can be valid only for small temperature
dierence. For the large temperature dierence encountered in the practical cycle, this assumption cannot be
applied. According to Refs. [17, 18], for the temperature
range of 3003500 K, the specic heat with constant volume can be written as

CV = 2.506 1011 T 2 + 1.454 107 T 1.5

7

4.245 10

T + 3.162 10

0.5

(1)

where T is the absolute temperature and the unit of CV

is J/(kg K).
According to the relation between specic heat with
constant pressure and specic heat with constant volume,
the specic heat with constant volume can be written as

= 2.506 1011 T 2 + 1.454 107 T 1.5

4.245 107 T + 3.162 105 T 0.5 + 1.3303
1.512 10 T
7

2.212 10

+ 3.063 10 T
5

,
T1

(4)

where m
t is the mass ow rate of the working uid and
the unit of m
t is kg/s.
Since Cp and CV are dependent on temperature, adiabatic exponent = Cp /CV will vary with temperature.
According to Refs. [19, 20] for small change in temperature dT , and volume dV of the working uid, the equation for reversible adiabatic process with variable can
be written as follows:
(5)

(T + dT )(V + dV )1
(
= (T + dT ) V 1 + ( 1)V 2 dV + . . . dV 2
0
)
 dV

+ . . . + . . . dV1
 0
(
)
= (T + dT ) V 1 + ( 1)V 2 dV
= T V 1 + ( 1)T V 2 dV + V 1 dT

2

dV
+ ( 1)V
dT


 0

Cp = CV + Rair

1.5

)T 5

Expanding the right hand term, one arrives to an alternative form as

1.512 104 T 1.5 + 3.063 105 T 2

3.063 105 T 1 1.106 107 T 2

(T + dT )(V + dV )1 = T V 1 .

+ 1.0433

2.212 107 T 3 ,

+ 1.3303T + 3.024 104 T 0.5

= T V 1 + ( 1)T V 2 dV + V 1 dT.

(6)

By substituting Eq. (6) into Eq. (5), we nally get

(2)

where Rair = 0.287 kJ/kg is the gas constant of the working uid.

dT
dV
= Rair
.
T
V
Integrating Eq. (7) in process i j , we have
CV

(7)

Eect of Expansion-Compression Ratio . . .

Vi
Tj
= Rair ln ,
(8)
Ti
Vj
where the temperature in the equation of CV is dened as:
/ (T )
j
T = (Tj Ti ) ln
.
(9)
Ti
The expansion-compression ratio, , the eective compression ratio, rc , and the compression ratio, rc , are dened as:
V5
T5
=
=
,
(10)
V1
T1
V1
(11)
rc =
V2
and
V5
rc =
= rc .
(12)
V2
Therefore, equations for reversible adiabatic processes
1 2 and 3 4 are as follows:
T2
CV ln
= Rair ln rc ,
(13)
T1
T3
CV ln
= Rair ln(rc ).
(14)
T4
For an ideal reciprocating heat engine cycle model, there
are no losses. However, for a real reciprocating heat engine cycle, heat transfer irreversibility between the working uid and the cylinder wall is not negligible. It can
be assumed that the heat loss through the cylinder wall
is proportional to the average temperature of both the
working uid and the cylinder wall and that the wall
temperature is constant at T0 [2123]. If the released
heat by combustion for one mass working uid is A and
the heat leakage coecient of the cylinder wall is B , the
heat added to the working uid by combustion is given
by the linear relation
Q in = m
t [A B(T2 + T4 2T0 )].
(15)
CV ln

Every time the piston moves, friction acts to retard the

motion. Considering the friction eects on the piston
in all the processes of the cycle, we assume a dissipation term represented by a friction force that is linearly
proportional to the velocity of the piston, which can be
written as follows [2426]:
dx
f = = ,
(16)
dt
where is a coecient of friction that takes into account
the global losses and x is the piston displacement. Therefore, the lost power is
dx dx
dW
=
= 2 .
(17)
P =
dt
dt dt
The mean velocity of the piston is
[
]
rc ( 1)

= 4LN = 4L 1 +
N,
(18)
rc 1
where L is total distance of the piston traveling per cycle
and L is the length of stroke compression.

647

The power output of the cycle is

Pdu = Q in Q out P
[
(
)
=m
t 8.353 1012 T13 + T33 T23 T43
(
)
+ 5.818 108 T12.5 + T32.5 T22.5 T42.5
(
)
2.123 107 T12 + T32 T22 T42
(
)
+ 2.108 105 T11.5 + T31.5 T21.5 T41.5

+ 1.0433(T3 T2 + T5 T4 ) + 1.3303(T1 T5 )
(
)
+ 3.024 104 T10.5 + T30.5 T20.5 T40.5
(
)
3.063 105 T11 + T31 T21 T41
(
)]
+ 1.106 107 T12 + T32 T22 T42
( [
] )2
rc ( 1)

16 L 1 +
(19)
N .
rc 1
The eciency of the cycle is
Pdu
du =
Q
{ in[
(
)
= m
t 8.353 1012 T13 + T33 T23 T43
(
)
+ 5.818 108 T12.5 + T32.5 T22.5 T42.5
(
)
2.123 107 T12 + T32 T22 T42
(
)
+ 2.108 105 T11.5 + T31.5 T21.5 T41.5

+ 1.0433(T3 T2 + T5 T4 ) + 1.3303(T1 T5 )
(
)
+ 3.024 104 T10.5 + T30.5 T20.5 T40.5
(
)
3.063 105 T11 + T31 T21 T41
(
)]
+ 1.106 107 T12 + T32 T22 T42
] )2 }
( [
rc ( 1)

N
16 L 1 +
rc 1
/
{ [
(
)
m
t 8.353 1012 T33 T23
(
)
+ 5.818 108 T32.5 T22.5
(
)
2.123 107 T32 T22
(
)
+ 2.108 105 T31.5 T21.5 + 1.0433(T3 T2 )
(
)
+ 3.024 104 T30.5 T20.5
(
)
3.063 105 T31 T21
(
)]}
.
(20)
+ 1.106 107 T32 T22
When = 1 or T5 = T1 , the Miller cycle becomes Otto
cycle, while T5 = T4 , the Miller cycle becomes the Atkinson cycle.
When rc , and T1 are given, T2 can be obtained from
Eq. (13); then, substituting from Eq. (3) into Eq. (15)
yields T3 ; and T4 can be found from Eq. (14); at last, T5
can be found by Eq. (10). Substituting T1 , T2 , T3 , T4 ,
and T5 into Eqs. (19) and (20) yields the power output
and thermal eciency. Therefore, the relations between

R. Ebrahimi

648

the power output, the thermal eciency and the compression ratio can be derived.
3. Results and discussion

The following constants and parameter values have

been used in this exercise: L = 72 mm, T1 = 300 K,
= 12.9 N s m1 , B = 1.4 kJ/(kg K), N = 4200 rpm,
A = 2820.5 kJ/kg, = 11.7, and M = 0.016 kg/s
[2730]. Using the above constants and range of parameters, the power output versus compression ratio
characteristic, the thermal eciency versus compression
ratio characteristic and the power output versus thermal eciency characteristic with varying the expansion-compression ratio can be plotted as in Figs. 24 (the
dashed lines in the gures denote where the cycle cannot work because T5 exceeds T4 ). One can see that the
power output versus compression ratio characteristic and
the eciency versus compression ratio characteristic are
parabolic like curves, and the power output versus eciency curve is loop shaped. They reect the performance
characteristics of a real irreversible Miller cycle. Numerical examples are shown as follows.

Fig. 2. Eect of expansion-compression ratio on the

variation of the power output with compression ratio.

Fig. 3. Eect of expansion-compression ratio on the

variation of the power output with compression ratio.

Figures 24 show the eect of expansion-compression

ratio versus compression ratio variation on the performance of cycle by considering the nonlinear relation be-

Fig. 4. Eect of expansion-compression ratio on the

variation of the power output with thermal eciency.

tween the specic heats of working uid and its temperature, the frictional loss and heat transfer loss. From these
gures, it can be found that the expansion-compression
ratio plays important roles on the performance cycle. It
is clearly seen that the eect of expansion-compression
ratio on the power output and thermal eciency is related to compression ratio. The results shows that if
compression ratio is less than certain value, the power
output decreases with increasing expansion-compression
ratio, while if compression ratio exceeds certain value,
the power output rst increases and then starts to
decrease with increasing expansion-compression ratio.
With further increase in compression ratio, the increase
of expansion-compression ratio results in increasing the
power output. The behavior of the thermal eciency versus compression ratio for various expansion-compression
ratio is qualitatively similar to that for the power output.
From these gures, we can conclude that when
expansion-compression ratio increases from 1.1 to 1.4, the
maximum power output and the maximum thermal eciency increase by 7.7% and 3.1%, respectively. With further increase in expansion-compression ratio, from 1.4 to
1.7, the maximum power output and the maximum thermal eciency decrease by 6.5% and 11.4, respectively.
The optimal compression ratio corresponding to maximum power output point, the optimal compression ratio
corresponding to maximum thermal eciency point and
the working range of the cycle increase with increasing
expansion-compression ratio. The results also show that
the optimal thermal eciency corresponding to maximum power output and the optimal power output corresponding to maximum thermal eciency rst increase
and then start to decline with increasing expansion-compression ratio. Numerical calculation shows that for
any same compression ratio, the largest power output is
for Otto when rc 4.3, is for Miller when 4.3 < rc < 27.5
and rc > 40.5 and is for Atkinson when 27.5 rc 40.5
and also the largest thermal eciency is for Otto when
rc 5.3 and is for Miller when 5.3 < rc 40.7 and
rc > 44.5 and is for Atkinson when 40.7 < rc 44.5.
According to above analysis, it can be found that the
eect of the expansion-compression ratio on the cycle
performance is obvious, and it should be considered in

Eect of Expansion-Compression Ratio . . .

practice cycle analysis in order to make the cycle model
be more close to practice.
4. Conclusions

In this study, nite-time thermodynamics analysis of

an irreversible air standard Miller cycle is established
which is closer to practice engine. In this model, eect of
expansion-compression ratio by considering the nonlinear
relation between the specic heats of working uid and
its temperature, the frictional loss and heat transfer loss
is analyzed. Numerical examples displayed the inuences
of this parameter in power output and eciency of the
cycle. The general conclusions drawn from the results of
this work are as follows:

If compression ratio is less than certain value, the

power output decreases with increasing expansion-compression ratio, while if compression ratio exceeds certain value, the power output rst increases and then starts to decrease with increasing
expansion-compression ratio. With further increase
in compression ratio, the increase of expansion-compression ratio results in increasing the power
output. The behavior of the thermal eciency
versus compression ratio for various expansion-compression ratio is qualitatively similar to that
for the power output.
The maximum power output, the maximum thermal eciency, the optimal thermal eciency corresponding to maximum power output and the optimal power output corresponding to maximum thermal eciency rst increase and then start to decline
with the increasing expansion-compression ratio.
Comparisons of the power output of the Miller,
Otto, and Atkinson cycles show that if compression ratio is less than certain value, the power output for Otto cycle is higher, while if compression
ratio exceeds certain value, the power output for
the Miller cycle is higher. With further increase in
compression ratio, the power output for the Atkinson cycle is higher. In high compression ratio, the
power output of the Miller cycle is higher. The behavior of the thermal eciency versus compression
ratio for the Miller, Otto and Atkinson cycles is
qualitatively similar to that for the power output.
The conclusions of this investigation are of importance
when considering the designs of actual Miller engines.
References

[1] B. Andresen, P. Salamon, R.S. Berry,

38, 62 (1984).

Phys. Today

649

[2] L. Chen, C. Wu, F. Sun, J. Non-Equilib. Thermodyn.

24, 327 (1999).
[3] L. Chen, F. Sun, Advances in Finite Time Thermodynamics: Analysis and Optimization, Nova Sci. Publ.,
New York 2004.
[4] R. Ebrahimi, J. Energy Instit. 84, 30 (2011).
[5] R. Ebrahimi, J. Energy Inst. 84, 38 (2011).
[6] R.H. Miller, ASME Trans. 69, 453 (1947).
[7] U. Kesgin, Int. Energy Res. 29, 189 (2005).
[8] K. Hatamura, M. Hayakawa, T. Goto, M. Hitomi,
JSAE Rev. 18, 101 (1997).
[9] Y. Fukuzawa, H. Shimoda, Y. Kakuhama, H. Endo,
K. Tanaka, Techn. Rev. 38, 146 (2001).
[10] A. Al-Sarkhi, B.A. Akash, Int. Commun. Heat Mass
Transfer 29, 1159 (2002).
[11] M. Sasaki, S. Araki, T. Miyata, T. Kawaji, JSAE Rev.
23, 451 (2002).
[12] C. Wu, P.V. Puzinauskas, J.S. Tsai, Appl. Therm.
Eng. 23, 511 (2003).
[13] Y. Ge, L. Chen, F. Sun, C. Wu, Int. Comm. Heat
Mass Transfer 32, 1045 (2005).
[14] Y. Ge, L. Chen, F. Sun, C. Wu, Appl. Energy 81, 397
(2005).
[15] Y. Ge, L. Chen, F. Sun, C. Wu, Int. J. Amb. Energy
26, 203 (2005).
[16] L. Chen, W. Zhang, F. Sun, Appl. Energy 84, 512
(2007).
[17] Y. Ge, L. Chen, F. Sun, Appl. Energy 85, 618 (2008).
[18] R. Sonntag, C. Borgnakke, G. Van Wylen, Fundamentals of Thermodynamics, 5th ed., Wiley, New York
1998.
[19] Y.-L. Ge, L. Chen, F.-R. Sun, Proc. IMechE, Part D:
J. Automob. Eng. 222, 887 (2008).
[20] Y. Ge, L. Chen, F. Sun, Appl. Energy 85, 618 (2008).
[21] R. Ebrahimi, J. Am. Sci. 5, 58 (2009).
[22] L. Chen, F. Zeng, F. Sun, C. Wu, Energy 21, 1201
(1996).
[23] S.A. Klein, Trans. ASME J. Eng. Gas Turbines
Power 113, 511 (1991).
[24] R. Ebrahimi, J. Energy Instit. 83, 1 (2010).
[25] A. Al-Sarkhi, J.O. Jaber, M. Abu-Qudais,
S.D. Probert, Appl. Energy 83, 153 (2006).
[26] L. Chen, T. Zheng, F. Sun, C. Wu, Int. J. Amb. Energy 24, 195 (2003).
[27] R. Ebrahimi, Acta Phys. Pol. A 117, 887 (2010).
[28] R. Ebrahimi, Acta Phys. Pol. A 118, 534 (2010).
[29] Y. Ge, L. Chen, F. Sun, C. Wu, J. Energy Instit. 80,
52 (2007).
[30] Y. Ust, B. Sahin, A. Kodal, J. Energy Instit. 82, 48
(2009).