Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Rui Kang
Department of Refrigeration and
Cryogenic Engineering,
School of Energy and Power Engineering,
Xi’an Jiaotong University,
Xi’an 710049, China
e-mail: kangrui996996@stu.xjtu.edu.cn
in the plain fin channels is not less than 200 mm. The thickness of 2.4 Data Reduction and Model Validation. To examine the
the covered plate, which is denoted by symbol b, is 1 mm for all accuracy of the model, the numerical f and j factors are calculated
the studied fin channels. and then are compared with available experimental data. The pro-
In the computational domain shown in Fig. 2, the no-slip cedure of the data reduction is as follows:
boundary condition is adopted at the interface between the fluid The Reynolds number and the Prandtl number are defined as
and solid. The mass-flow-inlet condition, pressure-outlet condi-
tion, and constant temperature boundary condition are used on the GDh
Re ¼ (6)
inlet, outlet, and covered-plate walls, respectively. The periodic l
boundary condition for plain fin and OSF fin are applied on the
two sides of the fluid and solid area. lcp
Pr ¼ (7)
k
2.3 Numerical Procedure. The CFD preprocessor GAMBIT
2.4.6 and solver FLUENT 15.0 are employed for grid generation and The Fanning f factor is determined by
simulation, respectively. Structured meshes are generated with
hexahedral elements in the computational domain, as shown in Dp qDh
f ¼ (8)
Fig. 3. The SIMPLEC algorithm is used for solving the G2 2L
pressure–velocity coupling equations, and the second-order
upwind scheme is selected to discretize the differential equations. where Dp is the pressure drop between the inlet and outlet of the
The grid independence tests are carried out to obtain a more accu- plate-fin channel.
rate solution. The solution is considered to have converged when The j factor is calculated by
the residuals for each field variable are less than 106.
j ¼ StPr2=3 (9)
hc;app
St ¼ (10)
Gcp
1 1
hc;app ¼ (11)
go 1 b A
U ks 2Aw;cp
Q
U¼ (12)
ADTLMTD
Q sec
Table 2 Calculation of the required areas for plain fin and OSF gf ;def ¼ (17)
fin Q sec ;fict
Area Plain fin OSF fin where the fictitious heat transfer rate is calculated by
ð
Aw;cp 2sL 2sL
Q sec ;fict ¼ hc ðAÞðTbase ðAÞ Tbulk ðAÞÞdA (18)
L s
A 4Lðh þ s 2tÞ 4Lðh þ s 2tÞ þ 4t 1 h þ 2t A sec
l 2
L s The local heat transfer coefficient is determined as
A sec 4Lðh tÞ 4Lðh tÞ þ 4t 1 h þ 2t
l 2
qð AÞ
hc ð AÞ ¼ (19)
Tw A Tbulk ð AÞ
ð Þ
A sec
ef ¼ g (22)
Afb f ;def
Fig. 7 Behavior of the ideal fin efficiency in terms of Re num- Fig. 9 Local dimensionless temperature difference on the sec-
ber for the plate fins ondary fin surface
Fig. 11 Behavior of conventional fin effectiveness in terms of Fig. 13 Local dimensionless temperature difference over the
Re number for the plate fins heat transfer surfaces of plain fin
Q ¼ Qpri þ Q sec (23) From Eq. (28), the actual fin effectiveness is affected not only
by the temperature differences hsec and hpri but also by hc,sec and
Q ¼ qpri Apri þ q sec A sec (24) hc,pri, which represent the mean coefficient of convection heat
transfer prevailing at the secondary surface and that over the pri-
Q ¼ qA (25) mary surface, respectively. It can also be found that when suppos-
ing hpri ¼ hbase and hc,sec ¼ hc,pri, the actual fin effectiveness just
Here, the primary surfaces include the primary fin surfaces and represents the fin efficiency. Here, it is necessary to discuss the
the surfaces of the covered plates in a plate-fin channel. validity of the two questionable assumptions carefully before
Equations (23) and (25) yield investigating the fin performance by means of the actual fin
effectiveness.
Apri A sec Figures 12 and 13 present the local dimensionless temperature
q ¼ qpri þ ef ;act (26)
A A difference over the primary surfaces in both the plate-fin channels.
The positions of the studied surfaces are given in Fig. 8. It is seen
where ef,act is called the actual fin effectiveness and is defined by from Figs. 12 and 13 that the hpri ( ¼ hpri/hbase) values for both
plate-fin channels are close to unity, which means that there is
q sec only a small extent of temperature loss over the primary surfaces.
ef ;act ¼ (27)
qpri The assumption hpri ¼ hbase is barely acceptable in the analysis of
the fin performance. Figures 14 and 15 show the local heat trans-
The actual fin effectiveness is physically the ratio of the heat fer coefficient over the heat transfer surfaces in the OSF fin chan-
flux over the secondary surfaces and that over the primary nel and the plain fin channel, respectively. It can be seen from
Fig. 14 that for the OSF fin, the heat transfer coefficient over the
secondary surface is more than five times as high as that prevail-
ing at the primary surface. Even for the plain fin, the hc,sec is about
four times higher than hc,pri, as shown in Fig. 15. Evidently, the
idealization of hc,sec ¼ hc,pri, which runs through the conventional
fin efficiency and fin effectiveness concepts, strongly restricts the
analysis on the fin performance of plate-fin surfaces.
Figure 16 shows the actual fin effectiveness of an OSF fin and a
plain fin with the same cross section in terms of Re number with
the same inlet temperature difference. It is seen that the ef,act val- Fig. 18 ef,act versus c with Affe/t2 as a parameter for OSF fins
ues for the OSF fin are more than 3.4, and for the plain fin are
more than 3.2, in the concerned Re range. These results reflect the
extremely high effectiveness of the plate-fin surfaces used in determine the cross-sectional shape of the fin channel, have compre-
plate-fin heat exchangers. It is also clear from Fig. 16 that the hensive effects on the heat transfer performance. The effect of fin
actual fin effectiveness of the OSF fin is higher than that of the thickness-to-length ratio d is investigated with univariate analysis.
plain fin in the range of Re < 1000, which illustrates that OSF fin Following Ref. [37], the OSF fins with different cross sections
performs better than plain fin in low Re region. Besides, the ef,act can be distinguished by the dimensionless parameter Affe/t2 ( ¼ 1/
values for the two types of fin surfaces increase initially and then a þ 1/(2c) 2). The Affe is the front-fin-end area in a unit cell of
decrease as Re increases. The maximum value of ef,act for the OSF the fin, as illustrated in Fig. 17. For a given Affe/t2, the ratios a and
fin appears when the Re number reaches about 550, and the opti- c are mutually related. Thus, the analysis of fin performance can
mum Re number for the plain fin is in the vicinity of 1200. be conducted efficiently. In addition, it is necessary to keep the
same mass flux for varied fin-channel shapes in the analysis.
The actual fin effectiveness of OSF fin versus fin density with
4.3 Geometrical Effects of OSF Fin on Actual Fin G ¼ 3 kgm2s1 and with Affe/t2 as a parameter is presented in
Effectiveness. The geometrical attributes of OSF fin have been Fig. 18. When Affe/t2 100, ef,act increases with the increasing of
introduced in the foregoing. In this section, the effects of fin fin density. In the region of Affe/t2 > 100, however, ef,act initially
thickness-to-height ratio a and fin density c are investigated with the increases and then decreases with an increasing of fin density,
method given in Ref. [37] because these two parameters, which which indicates that higher fin density (or larger fin height) cannot
Acknowledgment
The authors acknowledge the financial support provided by
Funds of National Key Technology R&D Program of China
(2012BAA08B03), National Natural Science Foundation of China
(51376142), and the Specialized Research Fund for the Doctoral
Program of University by MOE China (20130201110069).
Fig. 20 Actual fin effectiveness of OSF fin in terms of fin
thickness-to-length ratio
Nomenclature
A¼ area (m2)/total heat transfer surface area (m2)
always contribute to higher fin performance. This can be Ac ¼ free-flow area (m2)
explained by the fact that for the groups of OSF fin channels with Afb ¼ the area of fin base (m2)
larger Affe/t2, the decrease of mean heat flux on the secondary Affe ¼ the front-fin-end area of OSF fin
surfaces is more considerable than that on the primary surfaces in Aw;cp ¼ the wall area of the covered plate (m2)
fin channels with an increase of fin density according to the inter- b¼ thickness of the covered plate
mediate results. It is also seen from Fig. 18 that the OSF fins with cp ¼ specific heat at a constant pressure (J kg1 K1)
Affe/t2 ¼ 120 perform better than other groups when fin density c Dh ¼ hydraulic diameter of fin channel (m)
is more than 0.07. The actual fin effectiveness of OSF fin versus f¼ Fanning friction factor
fin density with Affe/t2 ¼ 140 and with mass flux G as a parameter G¼ mass flux (kg m2 s1)
is shown in Fig. 19. With G ¼ constant, ef,act passes through a h¼ height of the OSF fin (m)
peak value as the c increases. The fin density that contributes to hc ¼ heat transfer coefficient (W m2 K1)
peak ef,act increases with increasing values of mass flux. hc;app ¼ apparent heat transfer coefficient (W m2 K1)
Figure 20 presents the actual fin effectiveness versus fin j¼ Colburn factor
thickness-to-length ratio for OSF fin. The Re number is employed l¼ fin length of the OSF fin (m)
as a parameter since the fins have the same cross section. The L¼ total length of fin channel (m)
studied d range is from 0.013 to 0.05. It can be seen that the ef,act p¼ pressure (Pa)
increases with an increase of d when the Re number is less than Pr ¼ Prandtl number
500. In the range of Re 500, the actual fin effectiveness passes q¼ heat flux (W m2)
through peak values as the ratio d increases. The d value that cor- Q¼ heat transfer rate (W)
responds to the peak ef,act decreases with the increase of Re num- Re ¼ Reynolds number
ber. These results indicate that the OSF fins with relatively large d s¼ fin spacing (m)
perform better in low Re region, while those with small d are St ¼ Stanton number
more appropriate to operate under the condition of relatively high t¼ fin thickness (m)
Re number. As shown in Fig. 20, the envelope curve is plotted T¼ temperature (K)
approximately in dashed line, which helps to select the OSF fins Tbulk ¼ bulk temperature of fluid (K)
conveniently in the design of a plate-fin heat exchanger. u¼ velocity (m s1)