Techniques of heat transfer enhancement and their application chapter 6

1
1/72
Techniques of Heat Transfer Enhancement and their Application
Chapter 6
Externally Finned Tubes
Prof. Min Zeng
2/72
1. Introduction
2. The Geometric Parameters and the Reynolds Number
3. Plain Plate-Fins on Round Tubes
4. Plain Individually Finned Tubes
Outline
5. Enhanced Plate Fin Geometries with Round Tubes
6. Enhanced Circular Fin Geometries
7. Oval and Flat Tube Geometries
8. Row Effects-Staggered and Inline Layouts
10. Performance Comparison of Different Geometries
11. Conclusions
9. Heat Transfer Coefficient Distribution

2
3/72
1. Introduction
Finned-tube heat exchangers have been used for
heat exchange between gases and liquids (single or
two phase) for many years.
Figure 6.1 Finned tube geometries used with circular tubes: (a) plate fin-and-
tube used for gases, (b) individually finned tube having high fins, used for gases.
(From Webb [1987]) (c) Low, integral-fin tube.
4/72
A plain surface geometry will increase the air-
side hA value by increasing the area (A). Use of
enhanced fin surface geometries will provide
higher heat transfer coefficients than a plain surface.
To maintain reasonable friction power with low-
density gases, the gas velocity is usually less than 5
m/s.
Important basic enhancement geometries include
wavy and interrupted fins. Variants of the
interrupted strip fin are also used with finned-tube
heat exchangers for heat exchange to a tube-side
fluid.

3
5/72
Figure 6.2 Air-side geometries used in finned tube heat exchangers: (a) spine-fin, (b)
slit type OSF, (c) wavy fins, (d) convex louver fin, (e) Louver fins brazed to extruded
aluminum tube, (f) interrupted skive fin integral to extruded aluminum tube.
6/72
Because the gas-side heat transfer coefficient may
be 5 to 20% that of the tube-side fluid, the use of
closely spaced, high fins is desirable.
High fin efficiency can be obtained, if the fin
material has high thermal conductivity, e.g.,
aluminum or copper. If steel fins are required, fin
efficiency considerations will dictate shorter or
thicker fins.
Operational constraints, such as gas-side fouling
may limit the fin density. Air-conditioning
applications use 500 to 800 fins/m while process air
coolers are usually limited to 400 fins/m.

4
7/72
Dirty, soot-laden gases may limit the fin density to
200 fins/m. Different correlations are required for
the Figure 6.1a and b geometries.
The fin material used also depends on the operating
temperature and the corrosion potential. Listed
below are the fin and tube materials used in a
variety of applications:
(1) Residential air conditioning: Aluminum fins
and copper or aluminum tubes
(2) Automotive air conditioning: Aluminum fins
and aluminum tubes
8/72
(3) Automotive radiators: Aluminum fins brazed
to aluminum tubes, or copper fins soldered to brass
tubes.
(4) Process industry heat exchangers: Air-cooled
condensers that may use aluminum fins on copper
or steel tubes
(5) Boiler economizers (锅炉省煤器) and heat-
recovery exchangers: Steel fins on steel tubes
required by the higher operating temperature

5
9/72
In addition to describing the various fin geometries
and their performance characteristics, this chapter
compares the performance of alternative heat
exchanger and fin configurations. Then heat
exchanger and enhanced fin geometries that will
yield the highest performance per unit heat
exchanger core weight are identified. Finally,
possible improvements in the air-side surface
geometry are considered.
10/72
2. The geometric parameters and the
Reynolds number
2.1 Dimensionless Variables
The flow pattern in finned-tube heat exchangers is
very complex, due to its three-dimensional nature
and flow separations. The use of enhanced fin
geometries introduces further complications.
The geometric and flow variables that affect the
heat transfer coefficient and friction factor are the
following.

6
11/72
(1) Flow variables: Air velocity (u), viscosity (μ),
density (ρ), thermal conductivity (k), and specific
heat (cp).
(2) Tube bank variables: Tube root
diameter (do), transverse tube pitch (St), row
pitch (Sl), tube layout (staggered or inline), and the
number of rows (N).
(3) Fin geometry variables: For a plain fin, these
are the fin pitch (pf), fin height (e), fin thickness (t).
If, for example, an enhanced wavy fin geometry is
used the added variables are the wave
height (ew), the wave pitch(pw), and the wave shape.
12/72
Thus, there are seven geometry variables for a plain
fin (excluding the tube layout) and five flow
variables. Two additional variables are introduced
to account for the wavy fin geometry.
Dimensional analysis specifies that the number of
possibly important dimensionless groups is the
number of variables minus the number of
dimensions.
The dimensionless flow variables typically used in
correlations are the Reynolds number and the
Prandtl number. For heat transfer, one has the
Nusselt number or the Stanton number. For
pressure drop, one uses the friction factor.

7
13/72
2.2 Definition of Reynolds Number
The basic definition of the Reynolds number
is LcG/, where Lc is a characteristic dimension
and G is usually defined as the mass velocity in the
minimum flow area.
As previously stated, there are seven dimensions
associated with a plain fin geometry, and nine for
the wavy fin geometry. Hence, there is no unique
characteristic dimension. Hence, there are eight
possible values of Lc for definition of the Reynolds
number of the plain fin geometry.
14/72
Two different characteristic dimensions have been
used to define Lc in the Reynolds number. They are
the tube diameter (do) or the hydraulic
diameter (Dh).
Kays and London [1984] choose to use the
hydraulic diameter for the characteristic dimension
for all situations, including bare and finned-tube
banks. There is no evidence to suggest that
hydraulic diameter is a better choice.
In fact, there is evidence that the tube diameter may
be a better choice for finned-tube banks.

8
15/72
For fully developed flow in tubes, one defines
laminar and turbulent regimes. Do such regimes
also exist for finned-tube banks?
To evaluate this, consider the case of the Figure
6.1a geometry.
Figure 6.1 (a) plate fin-and-tube used for gases
16/72
Assume that the geometry uses do=19 mm with an
equilateral triangular pitch of St=44.45 mm, and
472 fins/m with 0.2 mm thickness. Assume air
enters the exchanger at 3 m/s and 20°C. The mass
velocity in the minimum flow area (G) is 7.34
kg/m2s and the hydraulic diameter is 3.68 mm.
The Reynolds numbers based on do and Dh are
8710 and 1640, respectively. Is the flow laminar or
turbulent? Based on ReDh, one would say it is
laminar. But, based on tube diameter (Red), one
would say it is turbulent. In reality, it exhibits some
of both characteristics.

9
17/72
If the tubes were not present, the flow geometry
would be a parallel plate channel, for
which Dh=3.82 mm. The Reynolds number is 1588,
which is clearly laminar. However, the tubes shed
eddies, which wash over the fin surface and provide
mixing of the flow.
If the Reynolds number based on hydraulic
diameter were dominant over the Reynolds number
based on tube diameter, one would expect that the
Nu and/data for different fin pitches would tend to
fall on one line. It is shown below that this is not
the case.
18/72
2.3 Definition of the Friction Factor
This course strives to use only the Fanning friction
factor (f), defined as follow
However, other friction factor definitions are
frequently used for tube banks (bare and finned). A
common definition for tube banks is given the
symbol ftb. It is related to the Fanning friction factor
by the equation
(6.1)
(6.2)
where N is the number of tube rows in the flow direction and L is the
flow depth. For bare or finned tubes, L=Sl (N−1)+de, where de is
diameter over the fins. For a bare tube bank, de=do.

10
19/72
3. Plain Plate-Fins on Round Tubes
Figure 6.1a shows the finned-tube geometry with
continuous, plain plate fins in a staggered tube
layout. An inline tube geometry is seldom used
because it provides substantially lower performance
than the staggered tube geometry.
3.1 Effect of Fin Spacing
Rich [1973] measured heat transfer and friction
data for the Figure 6.1a geometry having plain fins,
four rows deep, on 12.7-mm-diameter tubes
equilaterally spaced on 32-mm centers. The tubes
and fins were made of copper, and the fins were
solder-bonded to min.
20/72
Figure 6.3 Heat transfer and friction characteristics of a four-
row plain plate fin heat exchanger for different fin spacings.

11
21/72
Rich proposed that the friction drag force is the
sum of the drag force on a bare tube bank (Δpt) and
the drag caused by the fins (Δpf). The difference
between the total drag force and the drag force
associated with the corresponding bare tube bank is
the drag force on the fins. Thus, the friction
component resulting from the fins is given by
(6.3)
The term Δpt is that measured for a bare tube bank
of the same geometry, without fins. Both Δp drop
contributions are evaluated at the same minimum
area mass velocity.
22/72
Figure 6.4 Plot of the j factor and the fin friction vs. ReSI

12
23/72
Figure 6.4 shows that the j factor is a function of
velocity in the minimum flow area (Gc), and is
essentially independent of fin spacing. At the same
mass velocity (Gc), the bare tube bank heat transfer
coefficient is 40% larger than that of the finned-tube
bank.
Normally, the j/f ratio will increase as the fin
spacing is reduced, because the fractional parasitic
drag associated with the tube is reduced. Use of the
Reynolds number based on Sl has no real
significance, as all geometries tested had the same Sl.
24/72
Figure 6.4 may be regarded as evidence that the
Reynolds number based on hydraulic diameter will
not correlate the effect of fin pitch.
In a later study, Rich [1975] used the same heat
exchanger geometry with 551 fins/m to determine
the effect of the number of tube rows on the j factor.
Figure 6.5 shows the average j factor (smoothed
data fit) for each exchanger as a function of ReSI.
The row effect is greatest at low Reynolds numbers
and becomes negligible at ReSI>15,000.

13
25/72
Figure 6.5 Average heat transfer coefficients for plain plate-
finned tubes (571 fins/m) having one to six rows. Same
geometry dimensions as Figure 6.4.
26/72
3.2 Correlations for Staggered Tube Geometries
Correlations to predict the j and f factors vs.
Reynolds number for plain fins on staggered tube
arrangements were developed by McQuiston [1978],
Gray and Webb [1986], Kim et al. [1999], and Wang
et al. [2000a].
The McQuiston and the Gray and Webb heat
transfer correlations are comparable in accuracy.
However, the Gray and Webb friction factor
correlation is much more accurate than that of
McQuiston.

14
27/72
The Gray and Webb [1986] heat transfer correlation
(for four or more tube rows of a staggered tube
geometry) is
0.328 0.502 0.031
4 0.14Re ( ) ( )
t
d
l o
S s
j
S d
 
 (6.3)
Equation 6.3 assumes that the heat transfer
coefficient is stabilized by the fourth tube row,
hence the j factor for more than four tube rows is the
same as that for a four-row exchanger. The
correction for rows less than four is based on
correlation of the Figure 6.5 data, and is given by
28/72
0.092 0.031 0.607(4 )
4
0.991[2.24Re ( ) ]
4
N
N
d
j N
j
  
 (6.4)
The Gray and Webb [1986] friction correlation
assumes that the pressure drop is composed of two
terms. The first term accounts for the drag force on
the fins, and the second term accounts for the drag
force on the tubes. The friction factor of the heat
exchanger is given by
(1 )(1 )
f f
f t
f
A A t
f f f
A A p
    (6.5)
where
0.521 1.318
0.508Re ( )
t
f d
o
S
f
d

 (6.6)

15
29/72
3.3 Correlations for Inline Tube Geometries
Schmidt [1963] reports data and a correlation for the
inline geometry. However, little use exists for an
inline tube arrangement. This is because tube bypass
effects substantially degrade the performance of an
inline tube arrangement.
3.4 Plain Individually Finned Tubes
3.4.1 Plain Individually Finned Tubes
Extruded fins or helically wrapped fins on circular
tubes, as shown by Figure 6.1b, are frequently used
in the process industries and in combustion heat-
recovery equipment.
30/72
Figure 6.1 (b) individually finned tube having high fins, used for gases
Both plain and enhanced fin geometries are used. A
staggered tube layout is used, especially for high
fins (e/do>0.2).
The recommended correlations for a staggered tube
layout are made by Briggs and Young [1963] for
heat transfer and Robinson and Briggs [1966] for
pressure drop.

16
31/72
0.319 0.2 0.11
0.134Re ( ) ( )
d
s s
j
e t

 (6.7)
Equation 6.7 is based on airflow over 14 equilateral
triangular tube banks and covers the following
ranges: 1100≤Red≤18,000, 0.13≤s/e≤0.63,
1.0≤s/t≤6.6, 0.09≤e/do≤0.69, 0.01≤t/do≤0.15,
1.5≤St/do≤8.2. The standard deviation was 5.1 %.
0.316 0.927 0.515
9.47Re ( ) ( )
t t
tb d
o d
S S
f
d S
 
 (6.8)
The isothermal friction correlation of Robinson and
Briggs [1966], which we have rewritten in terms of
the tube bank friction factor, is
32/72
3.4.2 Low Integral-Fin Tubes
Rabas et al. [1981] developed more
accurate j and f correlations for low fin heights and
small fin spacings. The correlations are given below
with the exponents rounded off to two significant
digits.
0.234 0.25 0.76 0.73 0.71 0.38
3.805( ) ( ) ( ) ( ) ( ) ( )
o c o o t
e e t l
d G d d S
s e
f
d s d S S


 (6.10)
1.12 0.26 0.67 0.47 0.77
0.292( )( ) ( ) ( ) ( ) ( )
o c e e
o o
d G d d
s s t
j
d e s d t

 (6.9)
Where .The friction correlation is
0.415 0.0346( / )
e
n d s
  
The equations are valid for staggered tubes with N≥6,
5000≤Red≤25,000, 1.3≤s/e≤1.5, 0.01≤s/t≤0.06, e/do≤0.10,
0.01≤t/do≤0.02, and 1.3≤St/do≤ 1.5.

17
33/72
3.5 Enhanced Plate Fin Geometries with Round
Tubes
The wavy (or herringbone) fin and the offset strip
fin (also referred to as parallel louver) geometries
are the major enhanced surface geometries used on
circular tubes.
The combination of tubes plus a special surface
geometry establishes a very complex flow geometry.
The heat transfer coefficient of the wavy fin is
typically 50 to 70% greater than that of a plain (flat)
fin.
34/72
3.5.1 Wavy Fin
There are two basic variants of the wavy fin
geometry as illustrated in Figure 6.6. They are the
smooth wave and the herringbone configurations.
Figure 6.6 Two basic geometries of the wavy fin;
(a) herringbone wave, (b) smooth wave.

18
35/72
Much work has been done on the herringbone wave
geometry. However, very limited work has been
done on the smooth wave geometry.
Goldstein and Sparrow [1977] used a mass transfer
technique to measure local and average mass
transfer coefficients on a model having herringbone
configuration. At Re=1000 (based on fin spacing),
the wave configuration yielded a 45% higher mass
transfer coefficient compared with the plain fin
counterpart.
They proposed that the enhancement results from
Goetler vortices that form on concave wave surfaces.
36/72
3.5.2 Offset Strip Fins
The OSF concept (also known as “slit fins”) has
been applied to finned-tube heat exchangers with
plain fins for dry cooling towers and for refrigerant
condensers.
Figure 6.7 Comparison of the heat transfer coefficient for the OSF
and plain fin geometries for 9.5-mm-diameter tubes, 525 fins/m,
and 0.2-mm fin thickness as reported by Nakayama and Xu [1983].

19
37/72
At 3 m/s air velocity, the OSF provides a 78%
higher heat transfer coefficient than the plain fin.
For the same louver geometry, the OSF will provide
a higher heat transfer coefficient, when used in the
plate-and-fin-type heat exchanger.
The OSF shown in Figure 5.4 provides 150% higher
heat transfer coefficient than the plain fin at the
same velocity.
Generalized empirical correlations for j and f vs. Re
have not been developed for OSF geometry on
round tubes.
38/72
3.5.3 Louvered Fin
The louver geometry discussed in Section 5.3 has
been applied to finned-tube heat exchangers. Louver
patterns are formed on the fin area between tubes.
Care must be exercised in louvering the fin surface,
because the louvers can cut the conduction path
from the tube.

20
39/72
40/72
3.5.4 Vortex Generators
For circular finned tubes, a low heat transfer
coefficient exists in the wake region behind the
tubes, especially at low Reynolds numbers.
Vortex generators on the fin surface help to reduce
the width of the wake zone, and thus improve heat
transfer in the wake region.
Studies of vortex generators on circular finned tubes
have shown that the performance improvement is
not as great (relative to a plain fin surface) as for
flow in channels without any tubes.

21
41/72
An explanation for this is that in the circular finned
tube geometry, horseshoe vortices generated in front
of the tube cause longitudinal horseshoe vortices,
which increase the heat transfer over the fins along
the path. These horseshoe vortices provide
significant enhancement on the fins, relative to that
provided by vortex generators.
Figure 6.8 Fin configurations tested by Lozza and Melo [2001]: (a)
louver fin A, (b) louver fin B, (c) louver fin with vortex generator.
42/72
3.6 Enhanced Circular Fin Geometries
Figure 6.9 Enhanced circular fin geometries: (a) plain circular fin;
(b) slotted fin; (c) punched and bent triangular projections; (d)
segmented fin; (e) wire loop extended surface.
Figure 6.10 Segmented or spine fin geometries used in air
conditioning applications, (a) (From LaPorte et al. [1979].) (b)
Described by Abbott et al. [1980] and tested by Eckels and Rabas
[1985]

22
43/72
Curve Geometry St/do de/do Fins/m
1 Plain fin 1.0 3.47 787
2 Figure 6.10a 0.91 3.47 787
3 Figure 6.10b 0.95 2.45 787
Figure 6.11 j and f vs. Red characteristics of the Figure 6.10
enhanced fin geometries compared with a plain, circular fin
geometry as reported by Eckels and Rabas [1985]
44/72
Figure 6.12 Comparison of segmented fins (staggered and inline
tube layouts) with plain, staggered fin tube geometry as reported by
Weierman et al. [1978]. St/do=2.25, e/do=0.51, s/e=0.12, ws/e=0.17.

23
45/72
3.7 Oval and Flat Tube Geometries
Oval and flat cross-sectional tube shapes are also
applied to individually finned tubes.
Circular Oval
Tube dia. (d) 29 199.9/35.2
Fin. ht. (e) 9.8 10/9.3
Fin. htk. (t) 0.4 0.4
Face pitch (St/d) 1.03 1.05
Row pitch (Sl/d) 1.15 1.04
Fins/m 312 312
Tube Bank Dimensions (mm)
Figure 6.13 Heat transfer and friction characteristics of circular
and oval finned tubes in a staggered tube layout as reported by
Brauer [1964]
3.7.1 Oval vs. Circular Individually Finned Tubes
46/72
The oval tubes gave 15% higher heat transfer
coefficient and 25% less pressure drop than the
circular tubes.
The performance advantage of the oval tubes results
from lower form drag on the tubes and the smaller
wake region on the fin behind the tube.
The use of oval tubes may not be practical unless
the tube-side design pressure is sufficiently low.

24
47/72
3.7.2 Flat Extruded Aluminum Tubes with Internal
Membranes
Higher design pressures are possible using flattened
aluminum tubes made by an extrusion
process. Figure 6.2f shows a patented finned tube
concept made from an aluminum extrusion. Such
tubes can be made with internal membranes, which
strengthen the tube and allow for a high tube-side
design pressure.
48/72
The Figure 6.2e flat tube geometry offers significant
advantages over the Figure 6.2a or Figure 6.2b strip
fin geometries on round tubes.
(1) The airflow is normal to all of the narrow strips
on the Figure 6.2e geometry, which is not the case
for the Figure 6.2a spine fin design. Further, the
wake dissipation length decreases in the direction of
the fin base in the Figure 6.2a geometry.

25
49/72
(2) A low-velocity wake region does not occur
behind the tubes of the Figure 6.2e flat tube
geometry. The low wake velocity of the Figure 6.2a
and b geometries causes a substantial reduction of
the heat transfer coefficient, as documented by
Webb [1980].
50/72
(4) Further, the low projected area (投影面积) of the
Figure 6.2e flat tube will result in lower profile drag.
(3) The fraction of the Figure 6.2e surface that is
louvered is substantially greater than in the Figure
6.2b geometry. If a greater area distribution of
louvers were provided in the Figure 6.2b geometry,
the fin efficiency would substantially decrease. This
is because the slits would cut the heat conduction
path from the base tube.

26
51/72
3.8 Row Effects-Staggered and Inline Layouts
The published tube bank correlations are generally
for deep tube banks and do not account for row
effects. The heat transfer coefficient will decrease
with rows in an inline bank due to the bypass effects.
However, the coefficient increases with number of
tube rows in a staggered bank. This is because the
turbulent eddies shed from the tubes cause good
mixing in the downstream fin region.
As an approximate rule, one may assume that the
heat transfer coefficient for a staggered tube bank of
the Figure 6.2a and b finned tubes has attained its
asymptotic value at the fourth tube row.
52/72
Inline tube banks generally have a smaller heat
transfer coefficient than staggered tube banks. At
low Red (Red<1000) with deep tube banks (N≥8),
Raba and Huber [1989] show that the heat transfer
coefficient may be as small as 60% of the staggered
tube value.
The heat transfer coefficient of the inline bank
increases as the Red is increased; at Red=50,000
with N≥8, the inline to staggered ratio may approach
0.80.

27
53/72
There is a basic difference in the flow phenomena in
staggered and inline finned-tube banks.
Figure 6.13 NuPr−1/3 vs. Red for inline and staggered banks of
circular finned tubes with plain fins, as reported by Brauer [1964]
30%
100%
54/72
Figure 6.14 Flow patterns observed by Brauer [1964] for (a)
staggered and (b) inline finned tube banks.
Dead Spaces
Bypass stream

28
55/72
Figure 6.15 Row effect of inline tube banks for the Figure
6.20 segmented fin tubes (St/do=St/do=2.25).
The bypass effect reduces the performance the inline geometry
as the number of rows increases!!!
56/72
3.9 Heat Transfer Coefficient Distribution
Although the standard fin efficiency calculation
assumes the heat transfer coefficient is constant over
the fin surface area, local measurements have shown
this is not the case. The flow accelerates around the
tube and forms a wake region behind the tube. This
causes local variations of the heat transfer
coefficient.
Figure 6.16 Two-row finned tube geometry simulated by Saboya and Sparrow [1974]

29
57/72
Figure 6.17 Local Sherwood number distribution measured by
Saboya and Sparrow [1974] for a one-row plate fin geometry
The tube is located at 0.27≤x/L≤0.73.
58/72
Figure 6.18 Distribution of mass transfer coefficients (m3/m2-hr)
on a two-row plate finned tube measured by Krückels and
Kottke [1970]. (a) Red=1160, (b) Red=5800

30
59/72
Figure 6.19 Distribution of mass transfer coefficients (m3/m2-hr)
on a single circular finned tube (212 fins/m) measured by
Krückels and Kottke [1970]. (a) Red=1940, (b) Red=9700
60/72
3.10 Performance Comparison of Different
Geometries
Spine Wavy Slit CLF OSF Louver
Figure 6.2a 6.2c 6.2b 6.2d 6.20b 6.20c
do (mm) 9.52 9.52 9.52 9.52
b (mm) 3.46 3.46
a (mm) 7.87 8.38
t (mm) 0.76 0.76 0.76 0.76 0.76 0.76
St (mm) 25.4 23.62 25.4 25.4 21.84 21.84
Sl (mm) 25.4 20.60 21.60 21.60
Lp (mm) 1.98 1.59 1.59
nL 4 5 5
 (degrees) 20 20
a: Major diameter for rectangular tube cross section, m; b: Minor diameter for rectangular tube
cross section, m; Lp: Strip flow length of OSF or louver pitch of louver fin, m; nL: Number of
louvers in airflow depth, dimensionless.
Table 6.1 Heat Exchanger Geometries Compared by Webb and Gupte [1990]
3.10.1 Geometries Compared

31
61/72
Figure 6.20 Brazed aluminum heat exchanger having (a) 4.0-mm-
minor-diameter extruded aluminum tube (courtesy Modine
Manufacturing Co.) used with either the (b) OSF or (c) louver fins.
62/72
3.10.2 Analysis Method
Following are the sources of the various correlations,
or data, used to predict the air-side heat transfer and
friction characteristics:
(1). Figure 6.2a spine fin: Scaled data of Eckels and Rabas
[1985]
(2). Figure 6.2b slit fin: Mori and Nakayama [1980] data
(3). Figure 6.2c wavy fin: Webb [1990] wavy fin correlation
(4). Figure 6.2d CLF: Hatada and Senshu [1984] data
(5). Figure 6.20b OSF: Wieting [1975] correlation
(6). Figure 6.20c louver fin: Davenport [1984] correlation

32
63/72
The Kandlikar [1987] correlation was used to
predict the tube-side heat transfer coefficient for
vaporization of R-22 in plain tubes and flat tubes
(using the hydraulic diameter).
Figure 6.21 (a) ho vs. air velocity, (b) air Δp vs. air velocity.
64/72
The calculation methodology used the VG-1
criterion described in Chapter 3. The objective is to
reduce the heat exchanger size and weight. The case
of an R-22 evaporator is used to compare the
performance of the various geometries.
The evaporator operates at the following conditions:
heat duty (2408 W), airflow rate (0.53 m3/s),
refrigeration saturation temperature (2.77°C), inlet
air temperature (10°C), fins/m for the reference
spine fin (787), air Δp (0.74 mm w.g.). The air
frontal velocity was allowed to vary from 0.76 to
1.78 m/s, and the fin/m is allowed to vary from 314
to 867 for each air-side surface geometry.

33
65/72
The calculation procedure used to size the heat
exchanger for a specified frontal velocity is as
follows:
(1). The required kA is calculated for the specified
airflow rate, heat duty, inlet air temperature, and
refrigerant saturation temperature.
(2). Set the fins per meter at the lowest value (314
fins/m) and the number of tube rows in the airflow
direction (N) to one.
(3). Calculate the heat exchanger frontal area, and
the air-side and tube-side heat transfer coefficients
for the specified air frontal velocity.
66/72
(4). Calculate the available kA from.
fr fr fr
i i o o
A N A N A N
kA h A h A

 
(5). Calculate the air pressure drop.
(6). If the available kA is less than the required value,
increase the fins/m value by 20 fins/m, and repeat
steps 2 through 4. If the
available kA≥required kA, then the pressure drop is
checked. If this is within the specified 0.74 mm w.g.
limit, an acceptable solution exists. If no design is
obtained, the number of rows is increased by one
and the calculations are repeated from step 2.

34
67/72
This methodology results in selecting the heat
exchanger having the minimum number of tube
rows, which should yield the minimum material cost.
3.10.3 Calculated Results
Table 6.2 compares the various performance
parameters of the six heat exchanger configurations.
The Figure 6.18b spine fin is taken as the reference
for comparison. Each design listed in Table 6.3
provides the same heat duty and operates at the
same air-side pressure drop and airflow rate.
68/72
Table 6.2 Comparison of All-Aluminum Heat
Exchangers (Plain Tubes with Two Refrigerant Circuits)
Spine Wavy Slit CLF OSF Louver
Row 1 1 1 1 0.59 0.59
Fins/m 728 433 590 433 866 866
ufr (m/s) 0.96 1.07 1.32 1.19 1.39 1.44
ho (W/m2
K) 64.8 81.7 77.7 93.1 130.0 133.1
 0.93 0.91 0.90 0.89 0.83 0.83
Gref (kg/m2
s) 336 336 336 336 1127 1107
hi (W/m2
K) 3531 3554 3690 3667 10288 9732
wfin (kg) 1.96 1.58 1.85 1.49 0.97 0.93
wtube (kg) 1.23 1.19 0.96 0.99 0.84 0.84
wtot (kg) 3.19 2.77 2.81 2.48 1.81 1.77

35
69/72
3.11 Conclusions
Enhanced surfaces are routinely used for application
to gases. The dominant enhancement types are wavy
or some form of interrupted strip fin. When used on
circular tubes, a heat transfer enhancement of 80 to
100% is practically achieved.
Some advances were made on analytical or
numerical models to predict the heat transfer
performance of banks of high finned tubes. Power-
law empirical correlations have been developed for
plain fins.
70/72
Although some work has been done to develop
correlations for enhanced fins on circular tubes,
they are not sufficiently general to account for the
many geometric variables involved. The problem is
complicated by the row effect of inline and
staggered tube arrangements.
Both the individually finned-tube and plate fin-and-
tube geometries are used. Enhanced surface
geometries are available for both geometries.
Preference of the geometry type is dependent on
application and manufacturing cost considerations,
rather than performance.

36
71/72
Higher performance can be obtained from oval or
flat tubes. Oval tubes may not be practical for high
tube-side design pressure. If aluminum is an
acceptable material, extruded aluminum tubes
having internal membranes offer potential for
significant performance improvement, relative to
round tube designs. High tube-side design pressures
can be met. The resulting brazed aluminum heat
exchangers offer many advanced technology
possibilities. High-performance fin geometries are
quite adaptable to this heat exchanger concept.
72/72
Thank you for your attention!
Prof. Min Zeng
Key Laboratory of Thermo-Fluid Science and
Engineering, Ministry of Education, Xi'an
Jiaotong University
Xi'an, Shaanxi, 710049, P.R. China
Email: zengmin@mail.xjtu.edu.cn
http://gr.xjtu.edu.cn/web/zengmin
Tel : +86-29-82665581

Techniques of heat transfer enhancement and their application chapter 6

More Related Content

What's hot (20)

Similar to Techniques of heat transfer enhancement and their application chapter 6 (20)

More from ssusercf6d0e (20)

Recently uploaded (20)

Techniques of heat transfer enhancement and their application chapter 6