Scientia Iranica C (2012) 19 (6), 16381652

Sharif University of Technology

Scientia Iranica
Transactions C: Chemistry and Chemical Engineering
www.sciencedirect.com

Thermodynamic optimization principle for open inverse Brayton

cycle (refrigeration/heat pump cycle)
W. Zhang a,b , L. Chen a, , F. Sun a
a
College of Naval Architecture and Power, Naval University of Engineering, Wuhan 430033, PR China
b
China Satellite Maritime Tracking and Controlling Department, Jiangyin 214431, PR China

Received 27 November 2011; revised 12 March 2012; accepted 4 July 2012

KEYWORDS Abstract A thermodynamic model for an open inverse Brayton cycle (refrigeration or heat pump cycle)
Finite time with pressure drop irreversibilities is established. There are seven flow resistances (or pressure drops)
thermodynamics; encountered by the working fluid stream for the inverse Brayton cycle. Two of these, the friction through
Pressure drop; the blades and vanes of the compressor and the expander, are related to the isentropic efficiencies.
Refrigeration/heat pump The remaining flow resistances are always present because of the changes in flow cross-section at the
cycle; compressor inlet and outlet, heat exchanger inlets and outlets and expander inlet and outlet. The analytical
Cooling/heating load; formulae about the cooling load of refrigeration cycle, the heating load of heat pump cycle and other
Coefficient of performance coefficients are derived, which indicate that the thermodynamic performance for open inverse Brayton
of refrigeration/heat pump cycle can be optimized by adjusting the mass flow rate (or the distribution of pressure losses along the
cycle; flow path). It is shown that there are optimal air mass flow rates (or the distribution of pressure losses
Optimization. along the flow path) which maximize the cooling load of refrigeration cycle, and the optimal air mass flow
rates are smaller than the one at the maximum power output of the direct Brayton cycle.
2012 Sharif University of Technology. Production and hosting by Elsevier B.V.
Open access under CC BY-NC-ND license.

1. Introduction on Thermodynamic first/second law analysis without consid-

The analyses of air (Brayton) refrigeration (heat pump)
cycles have been recently paid more attention to due to the
enormously destruction of the Ozonosphere by Freon. The
refrigeration (heat pump) cycles that were ignored at one time,
have get luciferous foreground [18] again under the condition
of forbidding the refrigeration fluid containing Fl completely
before 2030 all over the world according to the United Nation
Montreal Protocol.
Finite-Time Thermodynamics (FTT) [921] is a powerful tool
for analyzing and optimizing the performance of thermody-
namic processes and cycles while the classical ones just focus
Corresponding author. Tel.: +86 27 83615046; fax: +86 27 83638709.
E-mail addresses: lingenchen@hotmail.com, lgchenna@yahoo.com
(L. Chen).
Peer review under responsibility of Sharif University of Technology.
1026-3098 2012 Sharif University of Technology. Production and hosting by Elsevier B.V. Open access under CC BY-NC-ND license.
doi:10.1016/j.scient.2012.09.007
ering the finite constraints such as area, time, heat conduc-
tance, etc. The primary study concentrated on the direct/inverse
Carnot cycles, and then direct/inverse Brayton cycles. In the
study for the inverse Brayton cycles (refrigeration/heat pump),
much work is concerned with the FTT performance analysis
and optimization for closed, simple [2234] and regenerated
[3540] inverse Brayton cycles (refrigeration/heat pump cycles)
by taking the cooling load, cooling load density, cooling coef-
ficient of performance, heating load, heating load density and
heating coefficient of performance as optimization objectives.
In the study for the direct Brayton cycles (gas turbine cycles),
much work is concerned with the FTT performance and opti-
mization for closed, simple [4143], regenerated [44,45], inter-
cooled [46,47] and intercooled and regenerated [48,49] Brayton
cycles by taking the power, specific power, power density, effi-
ciency and ecological function as optimization objectives. But,
all those studies were concentrating on closed cycles [2249],
for which the energy losses are derived from the heat resis-
tance losses of the heat transfer process, the principle of op-
timally dividing a finite heat conductance or area inventory
of the heat exchanger between the hot and cold ends of the
plant was used by taking the power, efficiency, cooling load,
 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652 1639

model by Radcenco et al. [50]. The power and efficiency of an

Notations open-cycle regenerated gas turbine power plant was optimized
by using the similar method by Chen et al. [51]. The power and
A flow area (m2 ) efficiency of an open-cycle intercooled gas turbine power plant
a ratio of the outermost equivalent flow cross- was optimized by using the similar method by Wang et al. [52].
sections The power and efficiency of an open-cycle combined Brayton
C specific heat (kJ/(kg K))/constant and (two parallel) inverse Brayton cycles power plant was opti-
h specific enthalpy (kJ/kg) mized by using the similar method by Zhang et al. [5355]. The
K contraction pressure loss coefficient power and efficiency of an open cycle gas turbine power plant
m mass flow rate (kg/s) with a refrigeration cycle for compressor inlet air cooling was
N number of heat transfer units optimized by using the similar method by Chen et al. [56,57].
P pressure (MPa) For the open Brayton cycles, the energy losses are derived from
Q heat transfer quantity (kJ) the flow resistance losses of the flow process of working fluid,
Q heat transfer rate (kW) and the principle [5063] of optimally adjusting the flow rate
Q dimensionless heat transfer rate and the distribution of pressure losses along the flow path was
R ideal gas constant (kJ/(kg K)) used by taking the power and efficiency as optimization objec-
s specific entropy (kJ/(kg K)) tives. Up to now, there is no work about the FTT study of the
T temperature (K) open inverse Brayton cycle in open literature. The working fluid
U heat conductance (kW/K) flow process of open inverse Brayton cycle is similar to that of
u hot-side heat conductance distribution open direct Brayton cycle. Then, can we analyze its performance
V mean velocity (m/s) according to its flow process of the working fluid? Is the opti-
v volume (m3 ) mization principle of open inverse Brayton cycle different from
w specific work (kJ/kg) that of open direct Brayton cycle? Is the optimization principle
W power (kW) of open inverse Brayton cycle different from that of closed in-
W dimensionless power verse Brayton cycle? Is the difference of the optimization prin-
ciple of open and closed direct/inverse Brayton cycle different
Greek symbols from that of open and closed direct/inverse Carnot cycle? To an-
swer these questions, we will take a further step (with regard
pressure ratio
to Refs. [5057]) to establish a thermodynamic model for the
variation
open inverse Brayton cycles (refrigeration/heat pump cycles)
effectiveness of the heat exchanger
with consideration pressure drop irreversibilities along to the
ratio of specific heats
flow process of the working fluid, and the optimization princi-
efficiency
ple of the open inverse Brayton cycle will get. The cooling load
adiabatic temperature ratio
(coefficient of performance) of refrigeration cycle and the heat-
density (kg/m3 )
ing load (coefficient of performance) of heat pump cycle will
temperature ratio of the top cycle
be optimized by using the similar principle and method used
relative pressure drop
in [5063].
Subscripts
2. Physical model
a air
c compressor The proposed system is shown in Figure 1. The plant model is
ch compressor-heat exchanger expressed using pressure drop and mass flow rate distributions
e environment and temperatureentropy diagram, as shown in Figure 2.
ht heat exchanger-expander There are four components, the compressor, the expander and
in inlet two heat exchangers. The feature of the flow through each
m mean component is that there is a pressure drop which increases with
max maximum the flow rate. There are seven flow resistances encountered by
min minimum the air stream for open inverse Brayton cycle. Two of these,
opt optimal the friction through the compressor and the expander, are
out outlet related to the isentropic efficiencies c and t , respectively.
p pressure In principle, these resistances can be rendered negligible by
s entropy minimizing friction in the compressor and expander in the limit
t expander (c , t ) 1. The remaining flow resistances, however, are
0 ambient always present because of the changes in the flow cross section
1, 2, . . . , 5 state points/sequence numbers at the compressor inlet and outlet, heat exchanger inlets and
outlets and expander inlet and outlet. These resistances control
the air flow rate m.
cooling coefficient of performance, heating load and heating
coefficient of performance as optimization objectives. In practi- 2.1. Cycle assumptions
cal industry, open-cycle gas turbine power plants and refriger-
ation/heat pump plants are widely applied. The performance of Proceedings along the path followed by the working fluid,
an open-cycle simple gas turbine power plant was optimized the following assumptions are made:
by incorporating the irreversibilities due to the various pres- (1) The working fluid (air) is an ideal gas with a specific heat
sure drops distributed along the flow path into the power plant that depends on temperature and composition.
1640 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652

modeled as adiabatic and irreversible with the entropy in-

crease, 1st . This process is equivalent to the isentropic ex-
pansion 34s followed by the adiabatic throttling process
4s 4 accounting for the pressure drop, 1Pt , associated with
fluid friction through the blades and vanes of the expander.
(6) The airflow through the heat exchanger 2 (process 45) is
characterized by the overall pressure drop 1P2 . This pro-
cess is represented schematically by the throttling process
44 (entropy increase, 1s4 ) followed by the isobaric pro-
cess 4 5, which accounts for the pressure drop, 1P2 , asso-
ciated with fluid friction through the heat exchanger 2.
(7) The discharge of the air stream from the heat exchanger 2
(process 5 e) causes another pressure drop, 1Pe = P5 P0 ,
and entropy increase, 1se , at temperature Te .

2.2. Cycle analysis

Figure 1: Pressure drop and mass flow rate distributions of the open inverse The pressure drop at the compressor inlet of the air cycle is
Brayton cycle. given by:

1
1P1 = K1 0 V12 , (1)
2
where K1 is the contraction pressure loss coefficient, and V1 is
the average air velocity through the inlet flow cross section A1 ,
see Figure 1. It is assumed that the flow is highly turbulent and,
as the first approximation, K1 is a constant when the change
in the flow cross section is fixed [64]. The air mass flow rate
through the same cross-section is m = A1 0 V1 , or:
1/2
2 1/2
m = A1 0 P0 1 , (2)
K1
where 1 = 1P1 /P0 is the relative pressure drop associated
with the first flow resistance.
The modeling of the flow through the compressor continues
with the apparent compressor pressure ratio, 1 = P2 /P0 , as
an input parameter [65]. The effective pressure ratio, c =
P2 /P1 = 1 /(1 1 ), is related to the isentropic temperature
( 1)/
ratio (cs ) across the compressor, cs = T2s /T1 = c a1 a1
,
where the ratio of the air specific heats a1 = (Cp /Cv )air , the
constant pressure specific heat, Cp , is obtained by the following
equation [66]:
8.314
Cp = (3.653 1.337 103 T + 3.294 106 T 2
Figure 2: The temperatureentropy diagram and the flow resistances. 28.97
1.913 109 T 3 + 2.763 1013 T 4 ). (3)
(2) The air flows into the compressor (process 01) irreversibly
and accompanied by the pressure drop, 1P1 = P0 P1 , and The mean constant pressure specific heat, Cp , for a certain
the entropy increase, 1s1 , at the ambient temperature, T0 . process is given in Box I. And the specific heat ratio of the
(3) The air compression process 12 is adiabatic and irre- process is:
versible, leading to the entropy increase, 1sc . In Figure 2
this process is represented schematically by the isentropic Cp
= , (5)
compression 12s followed by the throttling process 2s 2, Cp R
which accounts for the pressure drop, 1Pc , associated with
fluid friction through the compressor. where R is the ideal gas constant, R = 0.287 kJ/(kg K), and a1
(4) The air flow through the heat exchanger 1 (process 23 ) is evaluated based on Eq. (5).
is characterized by the overall pressure drop 1Pch . This The specific work required by the compressor, wc = c1
process is represented schematically by the throttling RT0 (cs 1)a1 /(a1 1), can be related to the pressure drop
process 22 (entropy increase, 1s2 ) followed by the through the compressor by writing h2s = h2 , and noting that
isobaric process 2 3 at pressure P , which accounts for the c = 1Pc /P2 = (c /cs )a1 /(a1 1) 1 and c = T2 /T1 =
pressure drop, 1Pch , associated with fluid friction through 1 + (cs 1)/c . The resulting dimensionless expression for the
the heat exchanger 1. compressor power input, Wc = mwc , is:
(5) The air flows into the expander (process 3 3) irreversibly
and accompanied by the pressure drop, 1Pht , and the en- Wc a1 (cs 1) 1/2
Wc = = . (6)
tropy increase, 1s3 . The expander expansion process 34 is A1 (2/K1 )1/2 P0 (RT0 )1/2 c (a1 1) 1
 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652 1641

T +1T
T
Cp dT
Cp =
1T
T +1T T +1T

8.314
28.97 T
(3.653 1.337 103 T + 3.294 106 T 2 )dT T
(1.913 109 T 3 2.763 1013 T 4 )dT
= (4)
1T
Box I

The pressure drop associated with the flow into the heat In conclusion, the expander power output, Wt = mwt , can be
exchanger 1 is 1Pch = K2 2 V22 /2, where K2 and V2 are expressed in the dimensionless form as:
the pressure loss coefficient and the mean velocity based on
Wt
A2 of the compressor, with K2 treated as a constant. The Wt = 1/2
relative pressure drop associated with the flow out of the A1 (2/K1 ) P0 (RT0 )1/2
compressor and into the heat exchanger 1 is determined from a3 T3 t (1 1/ts ) 1/2
mass conservation, m = A1 0 V1 = A2 2 V2 . The result is: = 1 . (13)
T0 (a3 1)
1Pch c 1 The pressure drop associated with the flow into the heat
ch = = . (7)
P2 (A2 /A1 )2 (K1 /K2 )
2
1 exchanger 2 is 1P2 = K4 4 V42 /2, where K4 and V4 are
the pressure loss coefficient and the mean velocity based
The heating load of the heat pump cycle in the heat exchanger
on A4 of the expander, with K4 treated as a constant. The
1 is:
relative pressure drop 2 associated with the flow out of the
Qout = m(h2 h3 ) = mRa2 1 (T2 TH )/(a2 1) expander and into the heat exchanger 2 is determined from
mass conservation, m = A1 0 V1 = A4 4 V4 . The result is:
T2 T3
= UH T2 TH
= mRa2 (T2 T3 )/(a2 1), (8)
2 = 1P2 /P4
ln T3 TH
T3 t2 1
where a2 is evaluated based on Eq. (5), and 1 is the effective- = . (14)
ness of the hot-side heat exchanger,
2
1 t T0 (1 ch )2 (1 ht )2 (A4 /A1 )2 (K1 /K4 )
1 = 1 exp(NH ), (9) The cooling load of the refrigeration cycle in the heat exchanger
2 is
where NH is the number of heat transfer units, NH = UH /
(mCpa2 ), and UH is the heat conductance of the hot-side heat Qin = m(h5 h4 ) = mR2 a4 (TL T4 )/(a4 1)
exchanger. T5 T4
According to Eq. (8), = UL TL T4
= mRa4 (T5 T4 )/(a4 1), (15)
ln TL T5
T3 = T0 [c 1 (c TH /T0 )]. (10)
where a4 is evaluated based on Eq. (5), 2 is the effectiveness
Qout can be dimensionless and expressed as follows: of the cold-side heat exchanger,
Qout = Qout /[A1 (2/K1 )1/2 P0 (RT0 )1/2 ] 2 = 1 exp(NL ), (16)
1/2 where NL is the number of heat transfer units, NL = UL /(mCpa4 ),
= a2 1 1 (c TH /T0 )/(a2 1). (11)
and UL is the heat conductance of the cold-side heat exchanger.
The pressure drop associated with the flow into the expander is According to Eq. (15):
1Pht = K3 3 V32 /2, where K3 and V3 are the pressure loss co-
efficient and the mean velocity based on A3 of the expander, T5 = 2 TL + (1 2 )T4 . (17)
with K3 treated as a constant. The relative pressure drop, ht = Qin can be dimensionless and expressed as follows:
1Pht /P , associated with the flow out of the heat exchanger 1
and into the expander is determined from mass conservation, Qin = Qin /[A1 (2/K1 )1/2 P0 (RT0 )1/2 ]
m = A1 0 V1 = A3 3 V3 . The result is: 1/2
= a4 2 1 (TL T3 /t )/T0 /(a4 1). (18)
T3 1 The pressure drop associated with the flow out of the heat
ht (1 ht ) = . (12)
12 T0 (1 ch )2 (A3 /A1 )2 (K1 /K3 ) exchanger 2 is defined as 1Pe = K5 5 V52 /2 through the ambient
pressure, where K5 and V5 are the pressure loss coefficient
The modeling of the flow through the expander continues with
and the mean velocity based on A5 of the outlet of the heat
the apparent expander pressure ratio, 2 = P /P4 , as an in-
exchanger 2, with K5 treated as a constant. The relative pressure
put parameter [65]. The effective pressure ratio, t = P3 /P4 =
drop, 5 = 1Pe /P0 , associated with the flow out of the
2 (1 ht ), is related to the isentropic temperature ratio, ts , heat exchanger 2 is determined from mass conservation, m =
( 1)/a3
across the expander, ts = T3 /T4s = t a3 , where the ratio A1 0 V1 = A5 5 V5 , and the pressure of the outlet of the heat
of the air specific heats, a3 , in the temperature range occupied exchanger 2 is P5 = P0 (1 + 5 ). The result is:
by the expander, is correlated by the same as Eq. (5).
The specific power output of the expander is wt = t RT0 1 T5
5 = . (19)
(1 1/ts )a3 /(a3 1), which is related to the pressure drop (A5 /A1 )2 (K1 /K5 )T0
associated with the friction through the expander blades and
The expanders pressure-ratio is:
vanes. Taking note of h4s = h4 and t = T3 /T4 = 1/(1
t + t /ts ), one can get t = 1Pt /P4 = (ts /t )a3 /(a3 1) 1. t = P3 /P4 = 1 (1 ch )(1 ht )(1 2 )/(1 + 5 ). (20)
1642 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652

The rate of heat released by the exhaust from the heat ex-
changer 2s outlet, as it reaches the ambient temperature T0 , is
Q0 = m (h0 h5 ), or in dimensionless form:

Q0
Q0 =
A1 (2/K1 )1/2 P0 (RT0 )1/2
1/2
a5 1 [1 T3 /T0 /t 2 (TL T3 /t )/T0 ]
= , (21)
a5 1
where a5 is evaluated based on Eq. (5).
The overall energy balance expression for the plant is Qout =
Qin + Q0 + Win , and Win is the power input of the cycle.
The coefficient of performance for the heat pump cycle is:

Qout Qout Qout

COP 1 = = =
Win Win Qout Qin Q0
1 Figure 3: The power input, heating load, cooling load, rate of heat released by
= the exhaust and coefficient of performance for the refrigeration cycle versus the
1 Qin /Qout Q0 /Qout relative pressure drop characteristics.
a4 2 (a2 1)(TL T3 /t )

= 1
a2 1 T0 (a4 1)(c TH /T0 ) 3.1. The optimization of the refrigeration cycle
a5 (a2 1)[1 2 TL /T0 + T3 (2 1)/T0 /t ]
1
. When the inverse Brayton cycle is utilized as the refriger-
a2 1 (a5 1)(c TH /T0 ) ation cycle, Qin is the dimensionless form of the cooling load,
(22) COP 2 is the coefficient of performance for the refrigeration cy-
cle, the temperature of the cooling space is TL , and TH = T0 . The
The coefficient of performance for the refrigeration cycle is: temperature ratio of the heat reservoir is = TH /TL .
Qin Qin Qin Figure 3 illustrates the behavior of the main quantities
COP 2 = = = of interest (Win , Qout , Qin , Q0 , COP 2 ) as 1 increases. It shows
Win Win Qout Qin Q0
that Win , Qout and Q0 increase with increases in 1 , and COP 2
1 decreases with an increase in 1 ; the left quantities of interest
=
Qout /Qin Q0 /Qin 1 Qin reach its maximum values at the corresponding optimal

a2 T0 1 (a4 1)(c TH /T0 ) values of 1 .
=
a4 2 (a2 1)(TL T3 /t )
3.1.1. Optimal pressure drops at the maximum cooling load
a5 (a4 1)[T0 2 TL + T3 (2 1)/t ]
1
Figures 47 show the maximum cooling load Qin max , and the
1 . (23)
a4 2 (a5 1)(TL T3 /t ) corresponding optimal compressor inlet relative pressure drop
1opt versus the compressor pressure ratio 1 , the effectiveness
To summarize the analytical formulae of the model, one notes 1 of hot-side heat exchanger, the effectiveness 2 of cold-
the expressions for the compressors power input Eq. (6), the side heat exchanger and the temperature ratio of the heat
heating load of the heat pump cycle Eq. (11), the expanders reservoir, respectively. From Figure 4, one can see that Qin max
power output Eq. (13), the cooling load of the refrigeration increases with an increase in 1 . From Figures 57, one can see
cycle Eq. (18), and the heat rejection via the exhaust Eq. (21). that Qin max increases with an increase in 1 , and 2 decreases
1/2
Each of these quantities is proportional to 1 , which in turn with an increase in , and 1opt increases with an increase in
is proportional to the mass flow rate m. Therefore, the energy 1 , and varies smoothly as 2 increases.
interactions Wc , Qout , Wt , Qin and Q0 decrease in proportion
with the mass flow rate; for example, the net power output is 3.1.2. Pressure drops with effects of power input of the cycle and
zero when the mass flow rate is zero. size constraints
An important feature of the maximum cooling load condi-
3. Numerical examples tion identified in the proceeding section is that the coefficient
of performance, COP 2 , at 1 = (1 )opt is less than that at 1
The effects of the mass flow rate and pressure-drops on the less than (1 )opt . In fact, COP 2 is maximum when 1 = 0, i.e.,
energy interactions are examined by numerical examples. The when the entropy generated by the seven resistances is zero. In
range covered by the calculations is 0 1 0.2, 1 1 practical application, the refrigeration cycle is operating at the
10, 0.8 1 , 2 0.99, P0 = 0.1013 MPa, c = 0.9, and rated power input. The power input Win constraint can be ex-
t = 0.85. The ratio of the extreme flow cross-sections covers pressed analytically as:
the range of 0.25 a1i 4, where a1i is the dimensionless
group Win = Qout Qin Q0 = const, (25)
where Qout , Qin , and Q0 are evaluated based on Eqs. (11), (18)
1/2
A1 Ki
a1i = (i = 2, 3, 4, 5). (24) and (21), respectively.
Ai K1
The second constraint refers to the total size of the cycle
In the calculations, a13 = 1/2, a12 = a14 = a15 = 1/3 plant. The sum of the compressor inlet and expander outlet
and T0 = 300 K are set. flow areas, A1 and A4 is related to the overall size and weight.
 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652 1643

Figure 4: The maximum cooling load and optimal relative pressure drop versus Figure 7: The maximum cooling load and optimal relative pressure drop versus
compressor pressure ratio characteristics. the temperature ratio characteristics.

and seek the optimal flow area allocation ratio x defined by

1/2 1/2
A1 /K1 = xA and A4 /K4 = (1 x)A . In the calculations,
a12 = a15 = 1/3 and a13 = 1/2 which can take other
values are set, that is to say A2 , A3 and A5 vary in proportion
with A1 (or x) during the optimization process, when x varies
between 0 and 1.
The pipeline countercurrent heat exchangers are chosen; the
pipes diameter, length, number and area of the heat transfer of
the hot-side heat exchanger are set D1 , L1 , nH and F1 , and the
pipes diameter, length, number and area of the heat transfer of
the cold-side heat exchanger are set D2 , L2 , nL and F2 . The flow
areas of the hot/cold-side heat exchangers are:

A2 = nH D21 /4, A4 = nL D22 /4. (27)

The areas of the heat transfer of the hot/cold-side heat exchang-
ers are:
Figure 5: The maximum cooling load and optimal relative pressure drop versus F1 = nH D1 L1 = 4L1 A2 /D1 ,
the effectiveness of the hot-side heat exchanger characteristics. (28)
F2 = nL D2 L2 = 4L2 A4 /D2 .
The heat conductance of the hot/cold-side heat exchangers is:
UH = 1 F1 , UL = 2 F2 , (29)
where 1 and 2 are the heat transfer coefficient of the hot/cold-
side heat exchangers, and the total heat conductance inventory
can be obtained
UT = UH + UL = 1 F1 + 2 F2
= 41 L1 A2 /D1 + 42 L2 A4 /D2 . (30)
If the diameter and length of the pipes of the hot/cold-side heat
exchangers are chosen, the flow areas and the heat transfer ar-
eas can be optimized by properly allocating the pipe numbers,
as the above assumption that the flow areas of the components
of the plant vary in proportion with A1 (or x) during the opti-
mization process, A2 and A4 vary in proportion with A1 (or x)
too, then the total heat conductance inventory vary in propor-
tion with A1 (or x) too, and the following constraint referring to
Figure 6: The maximum cooling load and optimal relative pressure drop versus the total heat transfer area is set:
the effectiveness of the cold-side heat exchanger characteristics.
UT UT
U = = 1/2 P
, (31)
Instead of fixing the sum (A1 + A4 ), it is convenient to include 2xA P0 (RT0 ) 1/2 A1 (2/K1 ) 0 (RT0 )1/2
the pressure loss coefficients in the constraint, where U is treated as a constant, it can be calculated by the
A1 A4 constants 1 , 2 , D1 , L1 , D2 and L2 .
1/2
+ 1/2
= A = const, (26) The heat conductance uUT of the hot-side heat exchanger is
K1 K4 set, and the heat conductance of the cold-side heat exchanger is
1644 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652

(1u)UT , where u is the hot-side heat conductance distribution.

The effectiveness of the hot- and cold-side heat exchangers
becomes:

uU T0 (a2 1)
1 = 1 exp 1/2
,
a2 1
(32)
(1 u)U T0 (a4 1)
2 = 1 exp 1/2
.
a4 1
The combination of Eq. (25) with Eq. (26) yields:
1/2
1/2 2xa2 1 1 (c 1)
Win = Win /[A P0 (RT0 ) ]=
a2 1
1/2
2xa4 2 1 (1/ T3 /t /T0 )

(a4 1)
1/2
2xa5 1 [1 T3 /t /T0 2 (1/ T3 /t /T0 )]

a5 1
Figure 8: The coefficient of performance for the refrigeration cycle versus the
= const. (33) relative pressure drop, flow area allocation ratio and hot-side heat conductance
distribution characteristics.
Therefore:
T3 t2 x2 1
2 = . (34)

2
1 t T0 (1 ch )2 (1 ht )2 (1 x)2
Based on Eq. (26), one has:
1/2
2xa2 1 1 (c 1)
Qout = Qout /[A P0 (RT0 )1/2 ] = , (35)
a2 1
Qin = Qin /[A P0 (RT0 )1/2 ]
1/2
2xa4 2 1 (1/ T3 /t /T0 )
= , (36)
(a4 1)
Q0 = Q0 /[A P0 (RT0 )1/2 ]
1/2
2xa5 1 [1 T3 /T0 /t 2 (1/ T3 /t /T0 )]
= . (37)
a5 1
Such that the coefficient of performance of the refrigeration Figure 9: The maximum coefficient of performance and optimal relative
cycle becomes: pressure drop, flow area allocation ratio and hot-side heat conductance
distribution for the refrigeration cycle versus the compressor pressure ratio
Qin Qin 1 characteristics.
COP 2 = = =
Win Qout Qin Q0 Qout /Qin Q0 /Qin 1

a2 1 (a4 1)(c 1) Figures 912 show the maximum coefficient of performance,
= COP 2max , of the refrigeration cycle, the corresponding optimal
a4 2 (a2 1)(1/ T3 /t /T0 ) compressor inlet relative pressure drop, 1opt , the optimal area
a5 (a4 1)[1 2 / + T3 (2 1)/t /T0 ]
1
allocation ratio, xopt , and the optimal hot-side heat conductance
1 . distribution, uopt , versus the compressor pressure ratio, 1 , the
a4 2 (a5 1)(1/ T3 /t /T0 )
temperature ratio, , of the heat reservoir, the dimensionless
(38)
value, U T0 , and the dimensionless rated power input, Win .
Figure 8 shows the characteristic of the coefficient of perfor- Form Figures 912, one can see that there exists an optimal
mance, COP 2 , of the refrigeration cycle versus the compres- compressor pressure ratio, 1opt , that maximizes, COP 2max . In
sor inlet relative pressure drop, 1 , the area allocation ratio, x, addition, COP 2max decreases with increases in and Win ; 1opt
and the hot-side heat conductance distribution, u. One can see increases with a decrease in 1 , and the increases in , U T0 and
that there exist an optimal compressor inlet relative pressure Win ; xopt increases with decreases in 1 and , and an increase
drop, 1opt , or an optimal area allocation ratio, xopt , and an opti- in Win ; uopt increases with a decrease in 1 , and increases in
mal hot-side heat conductance distribution, uopt , that maximize
and Win .
the coefficient of performance, COP 2 , of the refrigeration cycle,
respectively.
The numerical maximum, COP 2max , can be obtained through 3.2. The optimization of the heat pump cycle
simultaneous solution of the equations:
When the inverse Brayton cycle is utilized as the heat pump
COP 2 /1 = 0

(39) cycle, Qout is the dimensionless form of the heating load, COP 1
COP 2 / u = 0. is the coefficient of performance for the heat pump cycle, the
 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652
Figure 10: The maximum coefficient of performance and optimal relative pres-
sure drop, flow area allocation ratio and hot-side heat conductance distribution
for the refrigeration cycle versus the temperature ratio characteristics.
Figure 11: The maximum coefficient of performance and optimal relative
pressure drop, flow area allocation ratio and hot-side heat conductance
distribution for the refrigeration cycle versus the dimensionless value, U T0 ,
characteristics.
Figure 12: The maximum coefficient of performance and optimal relative
pressure drop, flow area allocation ratio and hot-side heat conductance
distribution for the refrigeration cycle versus the dimensionless rated power
input, Win , characteristics.
temperature of the heating space is TH , and TL = T0 . The
temperature ratio of the heat reservoir is = TH /TL .
1645
Figure 13: The power input, heating load, cooling load, rate of heat released by
the exhaust and coefficient of performance for the heat pump cycle versus the
relative pressure drop characteristics.
Figure 13 illustrates the behavior of the main quantities
of interest (Win , Qout , Qin , Q0 , COP 1 ) as 1 increases. It shows
that Win and Qout increase with an increase in 1 , and COP 1
decreases with an increase in 1 .
3.2.1. The analysis of the heating load and coefficient of perfor-
mance of the heat pump cycle
According to Figure 13, one can see that Win and Qout in-
crease with an increase in 1 , and COP 1 decreases with an
increase in 1 . There is no optimal compressor inlet relative
pressure drop which maximizes the heating load and the co-
efficient of performance of the heat pump. The effects of the
parameters1 , 1 , 2 and on the main quantities of interest
(Win , Qout , Qin , Q0 , COP 1 ) are obtained in the following.
Figures 1417 illustrate the behavior of the main quantities
of interest (Win , Qout , Qin , Q0 , COP 1 ) as 1 , 1 , 2 and
increase. From Figure 14, one can see that there is an optimal
compressor pressure ratio which maximizes the coefficient of
performance of the heat pump. From Figures 1417, one can see
that COP 1 increases with an increase in 1 , and with a decrease
in . Win , Qout and Qin increase with increases in 1 and 1 ,
and with a decrease in . Qin increases with an increase in 2 .
Q0 increases with an increase in 1 , and with decreases in 2
and .
3.2.2. Pressure drops with effects of power input of the cycle and
size constraints
In practical application, the heat pump cycle is operating at
the rated power input. The power input, Win , constraint can be
expressed analytically as:
Win = Qout Qin Q0 = const, (40)
where Qout , Qin , and Q0 are evaluated based on Eqs. (11), (18)
and (21), respectively.
The second constraint refers to the total size of the cycle
plant. Related to the overall size and weight is the sum of the
compressor inlet and expander outlet flow areas, A1 and A4 .
Instead of fixing the sum (A1 + A4 ), it is convenient to include
the pressure loss coefficients in the constraint,
A1 A4
1/2
+ 1/2
= A = const, (41)
K1 K4
1646 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652

Figure 14: The power input, heating load, cooling load, rate of heat released by Figure 17: The power input, heating load, cooling load, rate of heat released by
the exhaust and coefficient of performance for the heat pump cycle versus the the exhaust and coefficient of performance for the heat pump cycle versus the
compressor pressure ratio characteristics. temperature ratio charateristics.

and A5 vary in proportion with A1 (or x) during the optimization

process, when x varies between 0 and 1.
As with the same assumptions to the refrigeration cycle, the
constraint refers to the total heat transfer area is set:
UT UT
U = = , (42)
2xA P0 (RT0 )1/2 A1 (2/K1 )1/2 P0 (RT0 )1/2
where U treated as a constant.
The heat conductance uUT of the hot-side heat exchanger is
set, and the heat conductance of the cold-side heat exchanger is
(1u)UT , where u is the hot-side heat conductance distribution.
The effectiveness of the hot- and cold-side heat exchangers
becomes:

uU T0 (a2 1)
1 = 1 exp 1/2
,
a2 1
Figure 15: The power input, heating load, cooling load, rate of heat released by
(43)
(1 u)U T0 (a4 1)
the exhaust and coefficient of performance for the heat pump cycle versus the
2 = 1 exp 1/2
.
effectiveness of the hot-side heat exchanger characteristics.
a4 1
Combination of Eqs. (40) and (41) yields:
1/2
1/2 2xa2 1 1 (c )
Win = Win /[A P0 (RT0 ) ]=
a2 1
1/2
2xa4 2 1 (T0 T3 /t )

T0 (a4 1)
1/2
2xa5 1 (1 T3 /t /T0 )(1 2 )
= const. (44)
a5 1
Therefore:

T3 t2 x2 1
2 = . (45)

2
1 t T0 (1 ch )2 (1 ht )2 (1 x)2
Based on Eq. (41), one has:

Qout = Qout /[A P0 (RT0 )1/2 ]

Figure 16: The power input, heating load, cooling load, rate of heat released by 1/2
the exhaust and coefficient of performance for the heat pump cycle versus the 2xa2 1 1 (c )
effectiveness of the cold-side heat exchanger characteristics. = , (46)
a2 1
and seek the optimal flow area allocation ratio, x, defined by Qin = Qin /[A P0 (RT0 )1/2 ]
1/2 1/2
1/2
A1 /K1 = xA and A4 /K4 = (1 x)A . In the calculations, 2xa4 2 1 (T0 T3 /t )
a12 = a15 = 1/3 and a13 = 1/2 are set, that is to say A2 , A3 = , (47)
T0 (a4 1)
 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652 1647

Figure 19: The maximum coefficient of performance and optimal relative

pressure drop, flow area allocation ratio for the heat pump cycle versus the
compressor pressure ratio characteristics.
Figure 18: The coefficient of performance for the heat pump cycle versus the
relative pressure drop, flow area allocation ratio and hot-side heat conductance
distribution characteristics.

Q0 = Q0 /[A P0 (RT0 )1/2 ]

1/2
2xa5 1 (1 T3 /T0 /t )(1 2 )
= . (48)
a5 1
So the coefficient of performance of the heat pump becomes:

Qout Qout
COP 1 = =
Win Qout Qin Q0
1
=
1 Qin /Qout Q0 /Qout
a4 2 (a2 1)(T0 T3 /t )

= 1
a2 1 T0 (a4 1)(c )
a5 (a2 1)(1 2 )(1 T3 /T0 /t )
1
. (49)
a2 1 (a5 1)(c ) Figure 20: The maximum coefficient of performance and optimal relative
pressure drop, flow area allocation ratio for the heat pump cycle versus the
Figure 18 shows the characteristic of the coefficient of perfor- temperature ratio characteristics.

mance COP 1 of the heat pump cycle versus the compressor inlet
relative pressure drop 1 , the area allocation ratio x and the hot-
side heat conductance distribution u from which, one can see
that there exists an optimal compressor inlet relative pressure
drop 1opt , or an optimal area allocation ratio xopt that maxi-
mizes the coefficient of performance COP 1 of the heat pump
cycle, respectively. When u = 1, COP 1 reaches its maximum
COP 1max , which shows that the heating load gets its maximum
without the heat exchanger 2 for the heat pump. The numer-
ical maximum COP 1max can be obtained through the equation
COP 1 /1 = 0.
Figures 1922 show the maximum coefficient of perfor-
mance, COP 1max , of the heat pump cycle, the corresponding op-
timal compressor inlet relative pressure drop, 1opt , and the
optimal area allocation ratio, xopt , versus the compressor pres-
sure ratio, 1 , the temperature ratio, , of the heat reservoir,
the dimensionless value, U T0 , and the dimensionless rated
power input, Win . Form Figure 19, one can see that there ex-
ists an optimal compressor pressure ratio, 1opt , that maximizes Figure 21: The maximum coefficient of performance and optimal relative
COP 1max . In addition, COP 1max decreases with increases in and pressure drop, flow area allocation ratio for the heat pump cycle versus the
Win , 1opt increases with an increase in Win . dimensionless value, U T0 , characteristics.
1648 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652

Figure 22: The maximum coefficient of performance and optimal relative

Figure 23: The power versus relative pressure drop characteristics for the
pressure drop, flow area allocation ratio for the heat pump cycle versus the
direct cycle and cooling load versus relative pressure drop characteristics for
dimensionless rated power input, Win , charateristics.
the inverse cycle.

4. Discussions

From the numerical examples, there exists an optimal com-

pressor inlet relative pressure drop that maximizes the cooling
load, the coefficient of performance of the refrigeration cycle or
heat pump cycle. Therefore, there are different features for the
Brayton cycles of closed or open cycle.
When the cycle is utilized as gas turbine cycle [50], the
power output reaches its maximum value at the optimal
compressor inlet relative pressure drop, the maximized power
output has additional maxima with respect to the compressor
pressure ratio. When the thermodynamic optimization is per-
formed subject to a fixed fuel flow rate, and the power plant
size is constrained, the power output and efficiency can reach
their maximums at the optimal compressor inlet relative pres-
sure drop; the maximized power output and efficiency have ad-
ditional maxima with respect to the compressor pressure ratio.
When the cycle is utilized as refrigeration cycle, the cooling
load reaches its maximum value at the optimal compressor inlet Figure 24: The efficiency versus relaive pressure ratio and flow area allocation
ratio characteristics for the direct cycle and coefficient of performance versus
relative pressure drop. When the thermodynamic optimization
relative pressure ratio and flow area allocation ratio characteristics for the
is conducted by taking the rated power input of the cycle inverse cycle.
and the plant total size as the constraints, the coefficient
of performance and the corresponding cooling load reach
at the maximum power output of the gas turbine power plant
their maximum values at the optimal compressor inlet
is different from that at the maximum cooling load of the re-
relative pressure drop and optimal hot-side heat conductance
frigeration cycle, see Figure 23. When the thermodynamic op-
distribution. Those maximum values reach their maximum
timization is conducted by taking the energy input (the energy
values at the optimal compressor pressure ratio, respectively.
released by fuel inside the combustion chamber for the direct
When the cycle is utilized as heat pump cycle, the heating
cycle, the rated power input for the inverse cycle) of the cycle
load increases with an increase in compressor inlet relative
pressure drop, there exists the optimal compressor pressure and the plant total size as the constraints, the efficiency (power
ratio which maximize the coefficient of performance of the heat output) of the direct cycle, or the coefficient of performance
pump. When the thermodynamic optimization is performed (cooling load or heating load) of the inverse cycle get their max-
subject to a fixed rated power input of the cycle, and the imum values at the optimal compressor inlet relative pressure
plant size is constrained, the coefficient of performance and drop, respectively. The optimal compressor inlet relative pres-
the corresponding heating load reach their maximum values sure drop at the maximum efficiency (power output) of the di-
at the optimal compressor inlet relative pressure drop. Those rect cycle is different from that at the maximum coefficient of
maximum values reach their maximum values at the optimal performance (cooling load or heating load) of the inverse cycle,
compressor pressure ratio, respectively. And the numerical see Figure 24. There are optimal compressor pressure ratios that
analyses also show that the coefficient of performance reaches maximize the maximum efficiency (power output) of the direct
its maximum without the heat exchanger 2 for the heat pump cycle, or the maximum coefficient of performance (cooling load
cycle. or heating load) of the inverse cycle, respectively.
As mentioned in the above analysis, the heating load in- Therefore, the optimization performance is differential be-
creases with an increase in compressor inlet relative pressure tween the direct cycle and the inverse cycle of the open Brayton
drop, and the optimal compressor inlet relative pressure drop cycle, see Table 1.
 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652 1649

Table 1: Thermodynamic optimization principle for open direct/inverse Brayton cycles.

Direct cycle Inverse cycle

Physical Open gas turbine cycle Physical Open heat pump cycle Open refrigeration cycle
parameters parameters
No constraint Fixed size No constraint Fixed size No constraint Fixed size

1 1opt 1opt 1 1 1opt 1opt 1opt

1 1opt 1opt 1 1opt 1opt 1 1opt
x xopt x xopt xopt
u u uopt

Qout No extremum Qout max

Q No extremum Q = C
Qin Qin max Qin max

W Wmax Wmax Win No extremum Win = C No extremum Win = C

COP 1 COP 1max COP 1max
No extremum max
COP 2 No extremum COP 2max

Table 2: Thermodynamic optimization principle for endoreversible direct/inverse Brayton/Carnot cycle.

Direct cycle Inverse cycle

Physical Brayton cycle Carnot cycle Physical Brayton cycle Carnot cycle
parameters parameters
Heat pump Refrigeration Heat pump Refrigeration
cycle cycle cycle cycle

Qout Qout max Qout max

W Wmax Wmax
Qin Qin max Qin max
COP 1 COP 1opt COP 1opt
opt opt
COP 2 COP 2opt COP 2opt
u 0.5 0.5 u 0.5 0.5 0.5 0.5

(Q is the total heat transfer received by the gas stream, W is the net power output, is the thermal efficiency, C is the constant in Tables 1 and 2.).

Many studies show that the optimization performance is

differential between the closed and the open direct/inverse
Brayton cycle. Figures 25 and 26 show the temperatureentropy
diagram of the closed direct/inverse constant/variable-tempe-
rature heat reservoir Brayton cycles, Figures 27 and 28 show
the temperatureentropy diagram of the closed direct/inverse
constant/variable-temperature heat reservoir Carnot cycles. For
the closed endoreversible direct/inverse Brayton/Carnot cycles,
as indicated by [2249,67,68], when the total heat conductance
inventory, UT , of the plant is a constant, the heat conductance,
UH , of the hot-side heat exchanger and the heat conductance,
UL , of the cold-side heat exchanger are equal, i.e. UH = UL =
uUT = UT /2, which maximizes the power output for the di-
rect cycle, the cooling/heating load for the inverse cycle, see
Table 2. For the irreversible closed direct/inverse Brayton/
Carnot cycle, when the total heat conductance inventory, UT ,
of the plant is a constant, the optimal heat conductance of the
hot-side heat exchanger, and the cold-side heat exchanger are
unequal, which maximize the power output for the direct cy-
cle, the cooling/heating load for the inverse cycle [18,6974].
From Tables 1 and 2, one can see that the optimization vari- Figure 25: The temperatureentropy diagram of the closed direct/inverse
Brayton cycle with constant-temperature heat reservoirs.
able (i.e. the compressor inlet pressure drop) at maximum
power output (efficiency), cooling load (the coefficient of per-
formance of refrigeration) or heating load (coefficient of perfor-
mance of heat pump) of the open direct/inverse Brayton cycles 5. Conclusions
are not identical; and the optimization variable (i.e. the heat
conductance of the heat source) at maximum power output According to the above analysis and discussion, a whole
(efficiency), cooling load (coefficient of performance of refriger- scene about the Brayton cycle study is given from direct cycles
ation) or heating load (coefficient of performance of heat pump) to inverse cycles, and simple cycles to complex cycles; the
of the closed direct/inverse Brayton cycles are identical, which contrast between Brayton cycle and other cycles are done, such
is identical to that of the closed direct/inverse Carnot cycles. as Carnot cycle. The following are obtained:
1650 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652
Figure 26: The temperatureentropy diagram of the closed direct/inverse
Brayton cycle with variable-temperature heat reservoirs.
Figure 27: The temperatureentropy diagram of the closed direct/inverse
Carnot cycle with constant-temperature heat reservoirs.
Figure 28: The temperatureentropy diagram of the closed direct/inverse
Carnot cycle with variable-temperature heat reservoirs.
1. Considering the resistances that encountered by the work-
ing fluid along the flow path of the plant, a thermodynamic
model for open inverse Brayton cycle (refrigeration or heat
pump cycle) with pressure drop irreversibilities is estab-
lished and optimized. The optimization results show that
there exists the compressor optimal inlet relative pressure
drop that maximizes the cooling load of the refrigeration cy-
cle. When the thermodynamic optimization is conducted by
taking the rated power input of the cycle and the plant total
size as the constraints, there exists an optimal compressor
inlet relative pressure drop that maximize the coefficient of
performance (cooling load) of the refrigeration cycle and the
coefficient of performance (heating load) of the heat pump
cycle, respectively. Moreover, there exists the optimal hot-
side heat conductance distribution that maximizes the co-
efficient of performance of the refrigeration cycle. There are
the optimal compressor pressure ratios that maximize the
maximum coefficient of performance (cooling load) of the
refrigeration cycle and the maximum coefficient of perfor-
mance (heating load) of the heat pump cycle, respectively.
2. For the open direct/inverse Brayton cycle, the power output
and efficiency of the direct Brayton cycle can be optimized by
adjusting the compressor inlet pressure drop and the com-
pressor ratio; and the cooling load and coefficient of per-
formance of the inverse Brayton cycle can be optimized by
adjusting the compressor inlet pressure drop and the com-
pressor ratio too. When the thermodynamic optimization is
conducted by taking the total energy input of the cycle and
the plant total size as the constraints, the efficiency (power
output), coefficient of performance (cooling load) of the re-
frigeration, coefficient of performance (heating load) of the
heat pump can be optimized by adjusting the compressor
inlet pressure drop and the compressor ratio too.
3. For the open direct/inverse Brayton cycle, the optimal com-
pressor inlet pressure drop at the maximum power output of
the direct Brayton cycle is larger than that at the maximum
cooling load of the refrigeration. When the thermodynamic
optimization is conducted by taking the total energy input of
the cycle and the plant total size as the constraints, the opti-
mal compressor inlet relative pressure drop at the maximum
efficiency of the direct Brayton cycle is closer to that at the
maximum coefficient of performance of the inverse Brayton
cycle.
4. For the closed direct/inverse Brayton cycle, when the to-
tal heat conductance inventory of the plant is a constant,
the optimal heat conductance of the hot- and cold-side heat
exchangers are equal, which maximizes the power output
for the direct cycle, the cooling/heating load for the inverse
cycle, and this conclusion is identical to that of the closed
direct/inverse Carnot cycle. But it is different from that of
the open direct/inverse Brayton cycle, for which, the opti-
mal variable (i.e. the compressor inlet pressure drop) at the
maximum power output of the direct Brayton cycle and that
at the maximum cooling load (heating load) of the inverse
Brayton cycle are different.
Acknowledgments
This paper is supported by The National Natural Science
Foundation of PR China (Project No. 10905093) and the Natural
Science Foundation of Naval University of Engineering (Grant
No. HGDYDJJ10011). The authors wish to thank the reviewers
for their careful, unbiased and constructive suggestions, which
led to this revised manuscript.
 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652
References
[1] Angelino, G. and Invernizzi, C. Prospects for real-gas reversed Brayton-
cycle heat-pumps, Int. J. Refrig., 18(4), pp. 272280 (1995).
[2] Fleming, J.S., Van der Wekken, B.J.C., Mcgovern, J.A. and Van Gerwen, R.J.M.
Air-cycle cooling and heating, part 1: a realistic appraisal and a chosen
application, Int. J. Energy Res., 22(7), pp. 639655 (1998).
[3] Fleming, J.S., Li, L. and Van der Wekken, B.J.C. Air-cycle cooling and
heating, part 2: a mathematical model for the transient behaviour of fixed
matrix regenerators, Int. J. Energy Res., 22(5), pp. 463476 (1998).
[4] Hamlin, S., Hunt, R. and Tassou, S.A. Enhancing the performance of
evaporative spray cooling in air cycle refrigeration and air conditioning
technology, Appl. Therm. Eng., 18(9), pp. 11391148 (1998).
[5] Braun, J.E., Bansal, P.K. and Groll, E.A. Energy efficiency analysis of air-
cycle heat-pump dryers, Int. J. Refrig., 25(7), pp. 954965 (2002).
[6] Spence, S.W.T., Doran, W.J. and Artt, D.W. Design, construction and testing
of an air-cycle refrigeration system for road transport, Int. J. Refrig., 27(5),
pp. 503510 (2003).
[7] Williamson, N.J. and Bansal, P.K. Feasibility of air cycle systems for low-
temperature refrigeration applications with heat recovery, Proc. Inst.
Mech. Eng. Part E, 217(E3), pp. 267273 (2003).
[8] Spence, S.W.T., Doran, W.J. and Artt, D.W. Performance analysis of a
feasible air-cycle refrigeration system for road transport, Int. J. Refrig.,
28(3), pp. 381388 (2004).
[9] Bejan, A., Entropy Generation through Heat and Fluid Flow, Wiley, New York
(1982).
[10] Andresen, B., Salamon, P. and Berry, R.S. Thermodynamics in finite time,
Phys. Today, pp. 6270 (1984).
[11] Bejan, A. Entropy generation minimization: the new thermodynamics
of finite-size devices and finite-time processes, J. Appl. Phys., 79(3),
pp. 11911218 (1996).
[12] Sieniutycz, S. and Salamon, P., Eds., Advances in Thermodynamics. Finite
Time Thermodynamics and Thermoeconomics, Taylor & Francis, New York
(1990).
[13] Berry, R.S., Kazakov, V.A., Sieniutycz, S., Szwast, Z. and Tsirlin, A.M.,
Thermodynamic Optimization of Finite Time Processes, Wiley, Chichester
(1999).
[14] Chen, L., Wu, C. and Sun, F. Finite time thermodynamic optimization
or entropy generation minimization of energy systems, J. Non-Equilib.
Thermodyn., 24(4), pp. 327359 (1999).
[15] Gordon, J.M. and Ng, K.C., Cool Thermodynamics, Cambridge Int. Science
Publishers, Cambridge, UK (2000).
[16] Sieniutycz, S. HamiltonJacobiBellman framework for optimal control
in multistage energy systems, Phys. Rep., 326(4), pp. 165285 (2000).
[17] Chen, L. and Sun, F., Advances in Finite Time Thermodynamic: Analysis and
Optimization, Nova Science Publisher, New York (2004).
[18] Chen, L., Finite-Time Thermodynamic Analysis of Irreversible Processes and
Cycles, Higher Education Press, Beijing (2005).
[19] Sieniutycz, S. and Jezowski, J., Energy Optimization in Process Systems,
Elsevier, Oxford, UK (2009).
[20] Feidt, M. Optimal thermodynamicsnew upperbounds, Entropy, 11(4),
pp. 529547 (2009).
[21] Andresen, B. Current trends in finite-time thermodynamics, Angew.
Chem. Int. Ed., 50(12), pp. 26902704 (2011).
[22] Chen, L., Sun, F., Gong, J., Chen, W. and Lai, X. Optimization of steady flow
refrigeration cycles for specified boundary conditions, J. Eng. Thermophys.,
15(3), pp. 249252 (in Chinese) (1994).
[23] Tu, Y., Chen, L., Sun, F. and Wu, C. Exergy-based ecological optimization
for an endoreversible Brayton refrigeration cycle, Int. J. Exergy, 3(2),
pp. 191201 (2006).
[24] Tu, Y., Chen, L., Sun, F. and Wu, C. Comparative performance analysis
for endoreversible simple air refrigeration cycles considering ecological,
exergetic efficiency and cooling load objectives, Int. J. Ambient Energy,
27(3), pp. 160168 (2006).
[25] Ust, Y. and Sahin, B. Performance optimization of irreversible refriger-
ators based on a new thermo-ecological criterion, Int. J. Refrig., 30(3),
pp. 527534 (2007).
[26] Ust, Y. Performance analysis and optimization of irreversible air
refrigeration cycles based on ecological coefficient of performance
criterion, Appl. Therm. Eng., 29(1), pp. 4755 (2009).
[27] Ust, Y. The effects of regeneration on the thermo-ecological performance
analysis of irreversible air refrigeration cycles, Heat Mass Transfer, 46(4),
pp. 469478 (2010).
[28] Wu, C., Chen, L. and Sun, F. Optimization of steady-flow heat-pumps,
Energy Convers. Manage., 39(56), pp. 445453 (1998).
[29] Chen, L., Ni, N., Wu, C. and Sun, F. Heating load vs. COP characteristics for
irreversible air-heat pump cycles, Int. J. Electr. Power Energy Syst., 21(2),
pp. 105111 (2001).
[30] Bi, Y., Chen, L. and Sun, F. Heating load, heating load density and COP
optimizations for an endoreversible variable-temperature heat reservoir
air heat pump, J. Energy Inst., 82(1), pp. 4347 (2009).
1651
[31] Bi, Y., Chen, L. and Sun, F. Comparative performance analysis for
endoreversible simple air heat pump cycles considering ecological,
exergetic efficiency and heating load objectives, Int. J. Exergy, 6(4),
pp. 550566 (2009).
[32] Feidt, M. Optimisation dun cycle de Brayton moteur en contact avec des
capacits thermiques finies, Rev. Gen. Therm., 35(418419), pp. 662666
(1996).
[33] Feidt, M., Costea, M. and Postelnicu, V. Comparaison entre le cycle
simple de Brayton avec apport thermique impos et avec contrainte de
temprature maximale, Oil Gas Sci. Technol.Rev. IFP., 61(2), pp. 237245
(2006).
[34] Costea, M. and Feidt, M. Optimization of gas turbine cogeneration system
for various heat exchanger configurations, Oil Gas Sci. Technol.Rev. IFP.,
67(3), pp. 517535 (2012).
[35] Chen, L., Zhou, S. and Sun, F. Performance of heat-transfer irreversible re-
generated Brayton refrigerators, J. Phys. D: Appl. Phys., 34(5), pp. 830837
(2001).
[36] Zhou, S., Chen, L., Sun, F. and Wu, C. Theoretical optimization of a
regenerated air refrigerator, J. Phys. D: Appl. Phys., 36(18), pp. 23042311
(2003).
[37] Tyagi, S.K., Chen, G.M., Wang, Q. and Kaushik, S.C. A new thermoeconomic
approach and parametric study of an irreversible regenerative Brayton
refrigeration cycle, Int. J. Refrig., 29(7), pp. 11671174 (2006).
[38] Tu, Y., Chen, L., Sun, F. and Wu, C. Optimization of cooling load and COP
for real regenerated air refrigerator, Proc. Inst. Mech. Eng. Part E, 220(E4),
pp. 207215 (2006).
[39] Chen, L., Ni, N., Sun, F. and Wu, C. Performance of real regenerated air heat
pumps, Int. J. Power Energy Syst., 19(3), pp. 231238 (1999).
[40] Sadeghy, K., Saffaripour, M. and Ramazani-Rend, R. On optimizing the
performance of an air cycle heat pump using a recuperator, Proceedings
of the International Green Energy Conference, Waterloo, Ontario, Canada,
Paper No. IGEC-1-ID07, 1216 June (2005).
[41] Chen, L., Sun, F. and Wu, C. Performance analysis of an irreversible Brayton
heat engine, J. Inst. Energy, 70(482), pp. 28 (1997).
[42] Chen, L., Zheng, J., Sun, F. and Wu, C. Optimum distribution of heat
exchanger inventory for power density optimization of an endoreversible
closed Brayton cycle, J. Phys. D: Appl. Phys., 34(3), pp. 422427 (2001).
[43] Zheng, J., Chen, L., Sun, F. and Wu, C. Power and efficiency performance
of an endoreversible Brayton cycle, Int. J. Thermal Sci., 41(2), pp. 201205
(2002).
[44] Chen, L., Zheng, J., Sun, F. and Wu, C. Power density analysis and
optimization of a regenerated closed variable-temperature heat reservoir
Brayton cycle, J. Phys. D: Appl. Phys., 34(11), pp. 17271739 (2001).
[45] Kaushik, S.C. and Tyagi, S.K. Finite time thermodynamic analysis of an
irreversible regenerative closed cycle Brayton heat engine, Int. J. Sol.
Energy, 22(34), pp. 141151 (2002).
[46] Cheng, C.Y. and Chen, C.K. Maximum power of an endo-reversible
intercooled Brayton cycle, Int. J. Energy Res., 24(6), pp. 485494 (2000).
[47] Wang, W., Chen, L., Sun, F. and Wu, C. Optimum performance comparison
of an intercooled gas turbine cycle under the power and the power density
conditions, Int. J. Ambient Energy, 27(3), pp. 115124 (2006).
[48] Tyagi, S.K., Chen, J. and Kaushik, S.C. Optimal criteria based on the
ecological function of an irreversible intercooled regenerative modified
Brayton cycle, Int. J. Exergy, 2(1), pp. 90107 (2005).
[49] Wang, W., Chen, L., Sun, F. and Wu, C. Efficiency optimization of an
irreversible closed intercooled regenerated gas-turbine cycle, Proc. Inst.
Mech. Eng. Part A, 220(6), pp. 551558 (2006).
[50] Radcenco, V., Vargas, J.V.C. and Bejan, A. Thermodynamics optimization
of a gas turbine power plant with pressure drop irreversibilities, Trans.
ASME, J. Energy Resour. Technol., 120(3), pp. 233240 (1998).
[51] Chen, L., Li, Y., Sun, F. and Wu, C. Power optimization of open-cycle
regenerator gas-turbine power-plants, Appl. Energy, 78(2), pp. 199218
(2004).
[52] Wang, W., Chen, L., Sun, F. and Wu, C. Performance optimization of
an open-cycle intercooled gas turbine power plant with pressure drop
irreversibilities, J. Energy Inst., 81(1), pp. 3137 (2008).
[53] Zhang, W., Chen, L. and Sun, F. Power and efficiency optimization
for combined Brayton and inverse Brayton cycles, Appl. Therm. Eng.,
29(1415), pp. 28852894 (2009).
[54] Chen, L., Zhang, W. and Sun, F. Power and efficiency optimization
for combined Brayton and two parallel inverse Brayton cycles, part
1: description and modeling, Proc. Inst. Mech. Eng. Part C, 222(3),
pp. 393403 (2008).
[55] Zhang, W., Chen, L. and Sun, F. Power and efficiency optimization
for combined Brayton and two parallel inverse Brayton cycles, part
2: performance optimization, Proc. Inst. Mech. Eng. Part C, 222(3),
pp. 405414 (2008).
[56] Chen, L., Zhang, W. and Sun, F. Performance optimization for an open cycle
gas turbine power plant with a refrigeration cycle for compressor inlet air
cooling. Part 1: thermodynamic modeling, Proc. Inst. Mech. Eng. Part A,
223(A5), pp. 505513 (2009).
[57] Zhang, W., Chen, L. and Sun, F. Performance optimization for an open cycle
gas turbine power plant with a refrigeration cycle for compressor inlet air
cooling. Part 2: power and efficiency optimization, Proc. Inst. Mech. Eng.
Part A, 223(A5), pp. 515522 (2009).
1652 W. Zhang et al. / Scientia Iranica, Transactions C: Chemistry and Chemical Engineering 19 (2012) 16381652

[58] Radcenco, V., Generalized Thermodynamics, Editura Tehnica, Bucharest
(1994).
[59] Bejan, A. Maximum power from fluid flow, Int. J. Heat Mass Transfer,
39(6), pp. 11751181 (1996).
[60] Bejan, A., Advanced Engineering Thermodynamics, 2nd Edn., Wiley, New
York (1997).
[61] Chen, L., Wu, C., Sun, F. and Yu, J. Performance characteristic of fluid flow
converters, J. Energy Inst., 71(489), pp. 209215 (1998).
[62] Chen, L., Bi, Y. and Wu, C. Influence of nonlinear flow resistance relation
on the power and efficiency from fluid flow, J. Phys. D: Appl. Phys., 32(12),
pp. 13461349 (1999).
[63] Hu, W. and Chen, J. General performance characteristics and optimum
criteria of an irreversible fluid flow system, J. Phys. D: Appl. Phys., 39(5),
pp. 993997 (2006).
[64] Bejan, A., Heat Transfer, Wiley, New York (1993).
[65] Gordon, C.O., Aerodynamics of Aircraft Engine Components, AIAA Press, New
York (1985).
[66] Moran, M. and Shapiro, H., Fundamentals of Engineering Thermodynamics,
Wiley, New York (2004).
[67] Chen, L., Ni, N. and Sun, F. Comparison in performances of Brayton
and Carnot thermal engine cycles. (1) Steady flow cycles with infinite
reservoirs, J. Propul. Technol., 18(4), pp. 1821 (in Chinese) (1997).
[68] Chen, L., Zhu, Z., Cao, Y. and Sun, F. Comparison in performance of
Brayton and Carnot thermal engine cycles. (2) Steady flow cycles with
finite reservoirs, J. Propul. Technol., 18(5), pp. 5761 (in Chinese) (1997).
[69] Chen, L., Sun, F. and Wu, C. Effect of heat transfer law on the performance
of a generalized irreversible Carnot engine, J. Phys. D: Appl. Phys., 32(2),
pp. 99105 (1999).
[70] Chen, L., Zhou, J., Sun, F. and Wu, C. Ecological optimization for
generalized irreversible Carnot engines, Appl. Energy, 77(3), pp. 327338
(2004).
[71] Chen, L., Sun, F. and Wu, C. Effect of heat transfer law on the
performance of a generalized irreversible Carnot refrigerator, J. Non-
Equilib. Thermodyn., 26(3), pp. 291304 (2001).
[72] Chen, L., Sun, F. and Wu, C. Optimum allocation of heat exchanger area for
refrigeration and air conditioning plants, Appl. Energy, 77(3), pp. 339354
(2004).
[73] Zhu, X., Chen, L., Sun, F. and Wu, C. Optimal performance of a generalized
irreversible Carnot heat pump with a generalized heat transfer law, Phys.
Scr., 64(6), pp. 584587 (2001).
[74] Chen, L., Zhu, X., Sun, F. and Wu, C. Exergy-based ecological optimization
for a generalized irreversible Carnot heat pump, Appl. Energy, 84(1),
pp. 7888 (2007).
Wanli Zhang received his degrees (B.S., 2004; M.S., 2006; Ph.D., 2010) in
power engineering and engineering thermophysics from the Naval University
of Engineering, PR China. His work covers topics in finite time thermodynamics
and technology support for marine machinery. He is the author or co-author of
25 peer-refereed articles (15 in English).
Lingen Chen received all his degrees (B.S., 1983; M.S., 1986; Ph.D., 1998) in
power engineering and engineering thermophysics from the Naval University
of Engineering, PR China. His work covers a diversity of topics in engineering
thermodynamics, constructal theory, turbomachinery, reliability engineering,
and technology support for propulsion plants. He has been the Director of the
Department of Nuclear Energy Science and Engineering, the Director of the
Department of Power Engineering and the Superintendent of the Postgraduate
School. Now, he is President of the College of Naval Architecture and Power,
Naval University of Engineering, PR China. Professor Chen is the author or co-
author of over 1230 peer-refereed articles (over 570 in English journals) and
nine books (two in English).
Fengrui Sun received his B.S. degree in 1958 in Power Engineering from the
Harbing University of Technology, PR China. His work covers a diversity of topics
in engineering thermodynamics, constructal theory, reliability engineering, and
marine nuclear reactor engineering. He is a Professor in the Department of
Power Engineering, Naval University of Engineering, PR China. Professor Sun is
the author or co-author of over 950 peer-refereed papers (over 440 in English)
and two books (one in English).