This action might not be possible to undo. Are you sure you want to continue?
Su Thet Mon Than, Khin Aung Lin, Mi Sandar Mon
Abstract—This paper is intended to assist anyone with some
general technical experience, but perhaps limited specific
knowledge of heat transfer equipment. A characteristic of heat
exchanger design is the procedure of specifying a design, heat
transfer area and pressure drops and checking whether the
assumed design satisfies all requirements or not. The purpose
of this paper is how to design the oil cooler (heat exchanger)
especially for shellandtube heat exchanger which is the
majority type of liquidtoliquid heat exchanger. General
design considerations and design procedure are also illustrated
in this paper and a flow diagram is provided as an aid of
design procedure. In design calculation, the MatLAB and
AutoCAD software are used. Fundamental heat transfer
concepts and complex relationships involved in such
exchanger are also presented in this paper. The primary aim of
this design is to obtain a high heat transfer rate without
exceeding the allowable pressure drop. This computer
program is highly useful to design the shellandtube type heat
exchanger and to modify existing deign.
Keywords—ShellandTube Heat Exchanger, MatLAB and
AutoCAD
I. INTRODUCTION
HE process of heat exchange between two fluids that are
at different temperatures and separated by a solid wall
occurs in many engineering applications. The device used to
implement this exchange is termed a heat exchanger, and
specific applications may be found in space heating and air
conditioning, power production, waste heat recovery, and
chemical processing.
The problem of heat transmission is encountered in many
industries and because of the diversity in the fields of
application there exist countless difference in detail. However,
the calculation principle underlying the problem to design a
heat exchanger are everywhere the same, and it is the purpose
of this paper to present design calculations rather than to deal
with the details of individual problems and special cases [1].
A computer program in MatLAB has been written for shell
and tube heat exchanger design. Normally, the heat exchanger
design calculations are seems to be complex and repeated
calculations are required with assumed and/or fixed data. So,
many thermodynamic and fluid dynamic parameters are
encountered into designs procedure and therefore this
condition will satisfy and save the time by using a computer.
Su Thet Mon Than is with Department of Mechanical Engineering, Pathein
Technological University, Myanmar, (corresponding author to provide Phone:
098569035; Fax: 954224102; Email: suthetmonthan @gmail.com).
Khin Aung Lin is with Department of Mechanical Engineering, Thanlyin
Technological University, Myanmar.
Mi Sandar Mon is with Department of Mechanical Engineering, Yangon
Technological University, Myanmar.
With the help of this computer program, not only the shell
and tube heat exchanger can be designed without knowing the
detail design calculation but also plant engineer can modify
the currently used heat exchanger for some necessary changes,
such as blocking the blown out tubes without much effects to
operation, changes in fluid flow rate according to other
process, etc.
It is hoped that this paper will aid in classifying the many
details questions that arise during design calculation and this
will support practicing engineers to apply the formal
backgrounds in fluid flow and heat transfer to the practical
problems posed by the design, selection, testing, or installation
of the shell and tube heat exchanger.
II. DESIGN CONSIDERATION
The designer must consider several factors that influence
the shellside heat transfer coefficient that, in turn, determine
the rate of heat transfer in the shellside.
1. When baffles are provided, the system directs the
shellfluid from axial flow to toptobottom flow or
sidetoside flow with the effect that the heat transfer
coefficient is higher than for undisturbed flow along
the axes of the tubes [2].
2. Patterns of tube layout influence turbulence and
hence heat transfer coefficient e.g. triangular pitch
gives greater turbulence than square pitch. And under
comparable conditions of flow and tube size the heat
transfer coefficient for triangular pitch are roughly
25% greater than for square pitch [3].
3. Closer the baffle spacing, greater is the number of
times the shellfluid is to change its direction
resulting in greater turbulence [4].
4. Shellside coefficient is also affected by tube size,
clearance and fluidflow characteristics [5]
5. Shellside flow area varies across the bundle diameter
with the different number of tube clearances in each
longitudinal row of tubes. That’s why there is no true
shellside flow area by which the mass velocity of the
shellfluid can be computed. The correlation obtained
for fluids flowing in tubes is obviously not applicable
to fluids flowing over the tube bundles punctuated
with segmental baffles [3].
6. There are some terms used in heat exchanger
specification problems and their solutions, which are
often confused. These are ‘rating’, ‘design’ and
‘selection’. ‘Rating’ defines as the computational
process in which the inlet flow rates and
temperatures, the fluid properties, and the heat
exchanger parameters are taken as input and the
outlet temperatures and thermal duty (if the
exchanger length is specified) or the required length
of the heat exchanger are calculated as output [2].
T
World Academy of Science, Engineering and Technology 46 2008
604
‘Design’ defines as the process of determining all
essential constructional dimensions of an exchanger that must
perform a given heat duty and respect limitations on shellside
and tubeside pressure drop. ‘Selection’ defines choosing a
heat exchanger from among a number of units already existing
[2].
Most designers employ empirical relations with a cutand
try approach that depends on their judgment and experience
for convergence on a new design by extrapolation from tested
units. However, an analytical approach is easier to follow for
the less experienced designer, since it shows the basic
relationships. Our primary concern in this paper shall be
thermal analysis based on analytical approach developed by
Wolverine.
III. INITIAL CONDITIONS AND REQUIREMENTS
In general, the design of most heat exchangers involved
initial conditions in which the following variables are known
and assumed data:
1. flow rate of fluids
2. temperature range of fluids
3. length/tube and arrangement of tubes
With this information, it must prepare a design for the
optimum exchanger that will meet the required process
conditions. Ordinarily, the following results must be
determined.
1. Number of tubes and shell diameter
2. Heat transfer rate
3. Overall heat transfer coefficient
4. Tube side and shell side pressure drop
IV. DESIGN PROCEDURE
This computer program was divided into three main steps.
1. Calculation of total number of tubes
2. Calculation and Checking of heat load
3. Calculation and Checking of pressure drop for tube and
shell side
A. Calculations of Total Number of Tubes
The suitable tube size, tube diameter (D
t
) and tube material
are chosen first. The tube configuration has to be decided. In
our program the equilateral triangular pitch geometry is set
since it is mostly used. No other configuration is available in
our program. Total number of tubes must be calculated by
using following equations.
2
t
2
s
t
4CLP
πD
CTP N = == =
(1)
2
t
2
otl
t
4CLP
πD
CTP N = == =
(2)
Fig.1 reproduced from Tabordk defines the principal heat
exchanger dimensions. D
otl
is the outer tube limit diameter and
D
ctl
is the centerline tube limit diameter (D
ctl
= D
otl
 D
t
) where
D
t
is the outside diameter of the tubes). The baffle cut height
is shown as a height L
bch
; the value of the baffle cut B
c
is
(L
bch
/D
s
)
x100%.
− −− −
100
B
2 1 D
c
s
Fig. 1 Baffle and Tube Bundle Geometry [6]
The diametral clearance between the shell internal diameter
D
s
and outer tube limit diameter D
otl
is L
bb
. One half of L
bb
is
the width of this bypass channel. A pass partition lane is
shown with a width of L
p
. The diametral clearance between
the shell internal diameter D
s
and the diameter of the baffle D
b
is L
sb
, where the gap is equal to L
sb
/2.
2
L
sb
100 .
D
L
B
s
bch
c
= == =
Fig. 2 Baffle Cut and Clearance [6]
The dimensions D
s
, D
otl
, baffle cut (% of D
s
), and L
bb
and
L
sb
, shown in Fig. 2, can be obtained from a tube layout
drawing of the heat exchanger [6].
B. Calculation and Checking of Heat Load for UTube
Design procedures of heat load check with calculation of
heat transfer coefficient of shell and tube side are as shown in
Fig. 3.
1. Calculation of Tubeside Heat Transfer Coefficient
Because of total number of tube and total tubeside flow
rate are known, tube side mass flow rate (G
1
) can be solved as
total flow rate divided by the tubeside flow passage area per
pass. In our program, removable inserts are installed inside the
tubes. Use of tube inserts (wire mesh or twisted tape types) is
highly effective in laminar flows inside tubes. Installation of
inserts on the tubeside of heat recovery units to increase
energy recovery via a larger overall heat transfer coefficient
and smaller temperature approaches. To overcome the fact that
laminar flow heat transfer coefficients are almost independent
of fluid velocity and performance is difficult to improve using
plain tubes even resorting to large pressure drops, either
inserts or external low fins may be the simple solution.
Properly designed units with tube inserts normally are much
World Academy of Science, Engineering and Technology 46 2008
605
smaller in size and have smaller or equal pressure drops as
conventional plain tube units [7].
Fig. 3 Flow Chart for Heat Load
Fig. 4 Tube with Insert Plate
According to the Fig. 4, use the following equation for the
tubeside flow passage area.
a
t
= πd
i
2
/4
(3)
a
f
= [ (l
f
n
f
) + ( π d
f
x 3/2 ) ] t
f
(4)
a
n
= a
t
a
f
(5)
Where, a
t
is tube cross flow area, a
f
is tip area of plate fin,
a
n
is the net flow area for one tube. To calculate tubeside flow
passage area for one pass, a
n
is multiplied by N
t
/one pass. And
then, calculate tube side mass flow rate G
1
.
After that, Reynold number of the flow inside tube can then
be calculated by the following equations.
For Laminar flow,
0.14
3 / 1
φ
L
d
Re.Pr 86 . 1
k
hd
Nu   
¹ ¹¹ ¹
  
\ \\ \
  
= == = = == =
(6)
For Transistion flow,
  
¹ ¹¹ ¹
  
\ \\ \
  
+ ++ + − −− − = == = = == =
3 / 2
0.14 3 / 1 2/3
L
d
1 φ Pr ) 125 Re ( 116 . 0
k
hd
Nu
(7)
For Turbulent flow,
14 . 0 3 / 1 80 . 0
φ Pr . Re 023 . 0
k
hd
Nu = == = = == =
(8)
The criterion of distinguishing between laminar and
turbulent flow is the observed mixing action. Nusselt number
(Nu) is a function of Reynolds number (Re) and Prandtl
number (Pr). Nu of the flow inside tube can then be calculated
by above Reynold number equation. TubeSide Heat Transfer
Coefficient is also calculated by above Reynold number
equation [7].
2. Calculation of Shellside Heat Transfer Coefficient
The stream analysis shellside heat transfer coefficient for
singlephase flow h
as
(α
ss
) is used by following equation.
α
ss
= (J
C
J
L
J
B
J
R
J
S
Jµ ) α
1
(9)
In this expression, α
1
is the ideal tube bank heat transfer
coefficient calculated for all the flow across the tube bundle
and J
C
, J
L
, J
B
, J
R
and J
S
are the correction factor. J
C
is
calculated by using (10).
J
C
= 0.55+0.72 F
C
(10)
F
C
= 12F
W
(11)
In this equation, the fraction of the crosssectional area
occupied by the window (F
w
) is calculated from (12).
2π
θ Sin

360
θ
= F
ctl ctl
W
(12)
For a welldesigned unit, J
C
typically ranges in value from
0.65 to 1.175. The maximum value of Baffle leakage
correction J
L
is 1.0. To calculate J
L
, (13) is used.
J
L
=0.44(1r
s
)+[10.44(1r
s
)]exp(2.2r
lm
)
(13)
From this equation r
s
and r
lm
are calculated from 14 and 15.
tb sb
sb
s
S + S
S
= r
(14)
m
sb
lm
S
S
= r
tb
S + ++ +
(15)
After that, J
L
is calculated.
Fig. 5 Singlesegmental Shell and Tube Heat Exchanger Showing Baffle
Spacing [8]
World Academy of Science, Engineering and Technology 46 2008
606
Fig. 6 Tube Length Definition for UTube [8]
Fig. 5 depicts a singlesegmental shellandtube bundle
geometry with fixed tube sheets at both heads in which the
shellside flow makes one shell pass from one end of the tube
bundle to the other with the flow directed across the tube
bundle by the baffles. The inlet, central and outlet baffle
spacing are shown and are identifies as L
bi
, L
bc
and L
bo
,
respectively. By changing N
b
, baffle spacings L
bi
, L
bc
, L
bo
are
change.
L
bi
and L
bo
, are often equal in length to L
bc
, expect when
the first and last baffle compartments must be enlarged to
allow for the placement of the respective shellside nozzles.
The baffle layout is determined from the inlet, central, and
outlet baffle spacing and the effective tube length. The
effective tube length L
ta
is equal to the total tube length less
the combined thickness of the two tube sheets. The number of
baffles (an integer) and baffle spacing can be determined from
these values. The effective length for determining the baffle
spacing for Utube exchanger includes the straight length of
the tube plus D
s
/2, where D
s
is the shell internal diameter.
Thus the baffle spacing at the UBend, shown in Fig.6, should
include the tube straight length in this compartment plus
(D
s
/2) or 0.3 D
otl
[8].
By using the following Equation, baffle spacing are
calculated.
L
bi
= {1.25/N
b
+1}+0.005 (16)
L
bo
= L
bi
+ 0.3 D
otl
(17)
L
bc
= {1.252L
bi
}/{N
b
1} (18)
The maximum limit of Bundle by pass correction
factor J
B
is 1.0 at r
ss
≥ ½. To calculate J
B
, use F
sbp
.
)] 2r  (1 F exp[C = J
3
ss sbp bh B
(19)
The empirical factor C
bh
=1.35 for laminar flow (100 ≥ ≥≥ ≥ Re)
and C
bh
=1.25 for transition and turbulent flows (Re>100). To
evaluate this expression, one requires the ratio of the bypass to
the cross flow area F
sbp
, and the ratio r
ss
of the number of
sealing strips N
ss
(number of pairs if any) passed by the flow
to the number of tube rows crossed between baffle tips in one
baffle section N
tcc
. To calculate F
sbp
, use the following
equation.
m
b
sbp
S
S
= F (20)
S
b
= L
bc
[(D
s
D
otl
)+L
pl
]
(21)
In above expressions, S
b
is the bypass area; L
pp
represents
the width of the bypass lane between tubes. For situations
without a pass partition lane or for such a lane normal to the
flow direction, set L
pl
=0 while for a pass partition lane parallel
to the flow direction L
pl
is equal to ½ the actual dimension of
the lane or can be assumed to be equal to a tube diameter D
t .
tcc
ss
ss
N
N
= r (22)
  
¹ ¹¹ ¹
  
\ \\ \
  
− −− − = == =
100
B
2 1
L
D
N
c
pp
s
tcc
(23)
Where L
pp
=0.866L
tp
for a 30˚ layout, L
pp
=L
tp
for a
90˚ layout and L
pp
=0.707L
tp
for a 45˚ layout. This expression
has a maximum limit of J
B
=1 at r
ss
≥ ≥≥ ≥ 1/2 [8].
The unequal baffle spacing correction factor J
S
accounts for
the adverse effect of an inlet baffle spacing L
bi
and/or outlet
baffle spacing L
bo
larger than the central baffle spacing L
bc
.
Some exchangers have larger baffle spacing in the inlet and
outlet nozzle compartments compared to the central baffle
spacing, allowing placement of the shellside nozzles without
interference with the body flanges and without overlapping the
first baffle. The flow velocity in these compartments is thus
lowered and has an adverse influence on heat transfer. The
correction factor J
S
<1.0 for larger inlet and outlet spacing than
the central baffle spacing. For inlet and outlet baffle spacing
equal to the central baffle spacing, no correction is required
and J
S
=1.0. The value for J
S
is determined directly from the
effect on the flow velocity and is given by the following
expression.
) /L (L + ) /L (L + 1)  (N
) /L (L + ) /L (L + 1)  (N
= J
bc bo bc bi b
n  1
bc bo
n  1
bc bi b
s
(24)
Where n=0.6 for turbulent flow and n=1/3 for laminar flow.
The number of baffle compartments N
b
is determined from the
effective tube length and the baffle spacing [6][8].
In laminar flows, heat transfer is reduced by the adverse
temperature gradient formed in the boundary layer as the flow
thermally develops along the flow channel. The laminar flow
correction factor J
R
accounts for this effect. For laminar shell
side flow J
R
<1.0 (i.e. for 100 ≥ ≥≥ ≥ Re) while for Re>100, no
correction is needed and J
R
=1.0.
J
R
=(J
R
)
20
=
18 . 0
c
N
10
  
  
¹ ¹¹ ¹
  
\ \\ \
  
(For20 ≥ ≥≥ ≥ Re) (25)
N
c
= (N
tcc
+ N
tcw
)(N
b
+1) (26)
Where N
c
is the total number of tube rows crossed by the
flow in the entire heat exchanger and N
tcw
is the number of
tube rows crossed in the window area. For Re>20 but Re>100,
the value is prorated in (27). The minimum value of J
R
in all
cases is 0.4.
( (( ( ) )) )       1 J
80
Re  20
) (J J
20 R 20 R R
− −− −   
¹ ¹¹ ¹
  
\ \\ \
  
+ ++ + = == = (27)
For laminar shellside flow J
R
<1.0 while for Re>100, no
correction is needed. The correction factor J
µ
is greater than
1.0 for heating the shellside fluid and viceversa for cooling
the shellside fluid.
World Academy of Science, Engineering and Technology 46 2008
607
To calculate ideal tube bank heat transfer coefficient h
s
(α
1
),
use the following equation.
1
α = j
1
c
p
m˚Pr
2/3
(28)
2
Re
D
L
1.33
a j
t
tp
1 1
a
a
  
  
  
  
¹ ¹¹ ¹
  
\ \\ \
  
= == =
(29)
4
a
3
0.14Re 1
a
a
+ ++ +
= == = (30)
For heat load check, the overall heat transfer
coefficient, U, can be achieved from (31).
o so i
o
w
o
si i i
o
o
h
1
h
1
r
r
ln
k
r
h
1
h
1
r
r
1
U
+ +


¹

\

+


¹

\

+
=
(31)
In this equation, fouling resistance (factor) associated with
fluid outside tube R
fo
(1/h
so
)
and fouling resistance (factor)
associated with fluid inside tube R
fi
(1/h
si
).
K is wall thermal
conductivity. And then, calculate overall heat transfer
coefficient.
The heat transfer area A is calculated by the following
equation.
( (( ( ) )) )
t 1
f
f 1 o
N L
2
x3 d
nL 2 L d A
) )) )
` `` `
¹ ¹¹ ¹
¹ ¹¹ ¹
´ ´´ ´
¦ ¦¦ ¦
+ ++ + + ++ + = == =
π
π
(32)
In this equation, next term is represented for insert plate
heat transfer area. F
T
is the correction factor. Log Mean
Temperature Difference is calculated by following equation
[7].
∆T
m
=
  
  
¹ ¹¹ ¹
  
\ \\ \
  
− −− −
= == =
1
2
1 2
lm
∆T
∆T
ln
∆T ∆T
∆T (33)
Where, ∆T
2
= T
ho
– T
co
subscript i = inlet
∆T
1
= T
hi
– T
ci
subscripto = outlet
The heat load is computed from (34) and comparing with
given heat load.
q= U A F
T
∆T
lm
(34)
If the calculated heat load is greater than the given heat load,
it can be said that the design is satisfied. If not, by increasing
TN, the total number of tubes, and the whole procedure may
be repeated till the above condition is satisfied.
C. Calculation and Checking of Pressure Drop
Design procedures of Pressure Drop check for shell and
tube side are as shown in Fig .7.
The pressure drop for tube side is computed from the
summation of (35) and (36) and comparing with allowable
pressure drop. For the nozzle losses, it is usually sufficient to
calculate the loss for each nozzle at about three times the
velocity head in the nozzle [12].
c
2
noz
noz
2g
ρV
3 = ∆p
(35)
Where
,
v
noz
is calculated at the smallest crosssection area
for flow (i.e. highest velocity) in the nozzle. The combined
header and tube entrance losses are estimated in a similar way,
but using the velocity inside the tube, V
i
.
Fig. 7 Flow Chart for Pressure Drop on Shell and Tube Side
  
  
¹ ¹¹ ¹
  
\ \\ \
  
= == =
c
2
i
ent
2g
ρV
3 ∆P (36)
The pressure drop for shellside flow is equal to the sum of
the inlet nozzle pressure drop, the bundle pressure drop and
the outlet pressure drop. The inlet and outlet nozzle pressure
drops can be approximated as being equal to two velocity
heads each.
The bundle pressure drop is equal to the sum of the cross
flow pressure drops ∆p
c
, the window pressure drops ∆p
w
, and
the two end zone pressure drops (first and last baffle
compartments) ∆p
e
as illustrated in Fig. 8 [6][8].
So, the bundle pressure drop for shell side is computed from
using (37) and comparing with allowable pressure drop.
∆p
total
= ∆p
c
+ ∆p
w
+ ∆p
e
(37)
∆p
c
= ∆p
bl
(N
b
1) R
B
R
L
(38)
World Academy of Science, Engineering and Technology 46 2008
608
∆p
bl
= 0.002 f
1
N
tcc
m˙
2
/ρ R
µ
(39)
2
Re
/D L
1.33
b f
t tp
1 1
b
b
  
  
¹ ¹¹ ¹
  
\ \\ \
  
= == =
(40)
Fig. 8 Pressure Drop Regions in Shellside Flow [8]
4
b
3
0.14Re + 1
b
= b
(41)
m
w
µ
µ
µ
R
= == =
(42)
)] 2r  (1 F exp[C = R
3
ss sbp bp B
(43)
[ ]
p
lm s L
)r r + 1.33(1  exp = R
(44)
P = 0.15(1+r
s
) + 0.8
(45)
Where ∆p
bl
is the ideal bundle pressure drop for one baffle
compartment of the N
b
compartments. f
1
is the friction factor
and the empirical constants. R
µ
is the viscosity correction
factor and R
B
is the bypass correction factor. The limit of R
B
is
1 for r
ss
≥ ≥≥ ≥1/2. Use C
bp
is 4.5 for laminar flow (100 ≥ ≥≥ ≥ Re) and
C
bh
is 3.7 for transition and turbulent flows (Re>100). R
L
is
the leakage correction factor.
w m
w
S S
M
= mɺ (46)
( (( ( ) )) )
µ L
2
w
tcw b w
R R
2ρ
m 0.001
0.6N 2 N ∆p
+ ++ + = == =
ɺ
(47)
Where M is the shellside flow rate in kg/s and ∆p
w
and m˚
w
are the pressure drop and mass velocity in all N
b
window
zones for turbulent flow (Re>100).
S B
tcc
tcw
bl e
R R
N
N
1 ∆p ∆p
+ ++ + = == =
(48)
n 2
bi
L
bc
L
n 2
bo
L
bc
L
s
R
− −− −
+ ++ +
− −− −
= == =
(49)
∆p
e
is the pressure drop in the two end zones of the tube
bundle and R
s
is the pressure drop correction for unequal
baffle spacing at the inlet and/or outlet with respect to the
central baffle spacing. For all baffle spacing of equal length,
R
s
is 2.0. ‘n’ is 1 for laminar flow and n is 0.2 for turbulent
flow [6][8].
V. CASE STUDY
A transmission oil cooler is designed according to the
MatLAB program discussed above.
TABLE I
INPUT DATA FOR TRANSMISSION OIL COOLER FOR DIESEL
HYDRAULIC LOCOMOTIVE
Tube side Shell side
Fluid Hydraulic
oil
water
Inlet Temperature
(˚C)
105 77
Outlet Temperature
(˚C)
80.7 79.9
Flow rate
(m
3
/hr)
27.7 84
Heat load
(kW)
282 282
Pressure drop
(bar)
0.6 0.3
Max working
pressure (bar)
15 2
The following figures are obtained by using the MatLAB
Program.
R
e
y
n
o
l
d
s
N
u
m
b
e
r
i
n
T
u
b
e
6000
7000
8000
9000
10000
11000
12000
180 200 220 240 260 280 300 320 340
Number of Tube
Fig. 9 A fact of Reynolds Number on Number of Tubes
World Academy of Science, Engineering and Technology 46 2008
609
H
e
a
t
T
r
a
n
s
f
e
r
C
o
e
f
f
i
c
i
e
n
t
i
n
T
u
b
e
(
W
/
m
2
K
)
900
1000
1100
1200
1300
1400
1500
180 200 220 240 260 280 300 320 340
Number of Tube
Fig. 10 A fact of Heat Transfer Coefficient on Number of Tube
It can be seen that in Fig. 9 and Fig. 10 shows that the Re and
h
o
are gradually decreased corresponding to as high as N
t
.
Because of mass flow rate are constant, velocity are increase
and Re, h
o
are also decrease. Between total number of tube
220 and 240, there are a slight rise in heat transfer coefficient
because of changing from turbulent flow to transition flow.
R
e
y
n
o
l
d
s
N
u
m
b
e
r
i
n
S
h
e
l
l
Tube Length (mm)
14000
16000
18000
20000
22000
24000
26000
28000
30000
32000
1335 1340 1345 1350 1355 1360 1365
Nb=2
Nb=3
Nb=4
Fig. 11 A fact of Reynolds Number on Number of Baffles and Length
of Tube
The decreasing pattern of curves of Reynold Number and
heat transfer coefficient shown in Fig. 11 and Fig. 12 shows
that the Re and h are gradually decreased corresponding as
high as Tube Effective Length (L
1
). However, this graph also
describes, due to Number of Baffles increase, Reynold
Number increase.
The result table for Heat Load is shown in Fig. 13. Between
shell diameter 320m and 330m, there are also a slight rise in
heat transfer coefficient because of changing from turbulent
flow to transition flow.
H
e
a
t
T
r
a
n
s
f
e
r
C
o
e
f
f
i
c
i
e
n
t
i
n
S
h
e
l
l
(
W
/
m
2
K
)
1000
2000
3000
4000
5000
6000
7000
8000
9000
10000
11000
1335 1340 1345 1350 1355 1360 1365
Nb=2
Nb=3
Nb=4
Length of Tube (mm)
Fig. 12 A fact of Heat Transfer Coefficient on Number of Baffles and
Length of Tube
200
210
220
230
240
250
260
270
280
290
300 310 320 330 340 350 360 370 380 390 400
Nb=2
Nb=3
Nb=4
H
e
a
t
L
o
a
d
(
k
W
)
Shell Diameter (mm)
Fig. 13 A fact of Heat Load on Shell Diameter
TABLE II COMPARISON WITH INDUSTRIAL DATA
Result Industrial
Data
Calculated
Data
Shell diameter,
m
0.385 0.39
Tube length, m 1.25 1.25
Total no. of
tubes
248 318
Tube diameter,
m
0.014 0.0127
Number of
Baffle
4 4
Heat Load, kW 282 283.02
Tube side
pressure drop
0.6 0.4032
Shell side
pressure drop
0.3 0.1652
V. DISCUSSION AND CONCLUSION
There are few limitations to be considered in program. To
reduce size and cost of heat exchanger, UBend Exchangers
World Academy of Science, Engineering and Technology 46 2008
610
are chosen. The arrangements of tubes are only for a widely
used triangular pattern which permits the use of more tubes.
The limitation of shell diameter is within the minimum 0.3 m
and maximum 0.4 m. It is still assumed that there is no
advantage in varying less than the allowable pressure drop and
baffles will be employed within the minimum (one fifth of the
shell diameter) and maximum (inside diameter of the shell)
spacing. These limitations stem from the fact that at wider
spacing the flow tends to be axial rather than across the bundle
and at closer spacing there is excessive leakage between
baffles and the shell. If the shell side pressure drop is higher
than the acceptable limits, the baffles spacing is increased
slightly, the shell side pressure drop will be satisfactory. If the
tube side pressure is not within acceptable limits, the
calculation can be repeated with a revised value of total
number of tube because of the pressure drop depends on the
square of the velocity.
There are many types heat baffle. Among them, segmental
baffle is selected in this design. This type of baffle is probably
the most popular. It is a circle of near shell diameter from
which a horizontal or vertical portion has been cut. The cut
out portion which represents the free flow area for the shell
side fluid is usually from 20 to 50 percent of the open shell
area. In this design, 25 percent baffle cut is used according to
the method of Wolverine. The limitation of number of baffle is
2, 3 and 4.
Shell diameter for calculated design is nearly the same the
existing design. But total numbers of tubes are more than
existing design because of tube diameter changes.
Furthermore, use simple insert flat plate in tubes. By using of
tube inserts (flat plate) are highly effective in laminar flow
inside tubes. Tube inserts flat plate in existing design is more
confused than calculated flat plate. In this program, tube
diameter is limited to 12.7 mm, so this diameter is less than
existing tube diameter 14mm. Because of these facts, increase
total number of tube, heat transfer area and heat load is more
than limited heat load (283.2 kW). The tube and shell pressure
drop for acceptable limits is 0.6 and 0.3 bar. The calculated
pressure drop for tube and shell are 0.4032 and 0.1652 bar.
The design is satisfied because the pressure drop for both sides
is lower than the limited pressure drop. So this computer
program is highly useful to design the liquidtoliquid shell
and tube type heat exchanger.
REFERENCES
[1] Frass, A.P and M.Necatic Ozisik, 1965, Heat Exchanger Design, John
Wiley and Sons Inc.
[2] Max S.Peters and Klaus D. Timmerhuaus, ©1958, Plant Design and
Economics for Chemical Engineers, 4
th
ed, McGrawHill Book
Company.
[3] TEMA, 1999, Standards of TEMA, 8
th
ed, Turbular Exchanger
Manufacturers Association, New York.
[4] Kays, W.M and A.L.London, 1998, Compact Heat Exchangers, 3
rd
ed,
Krieger Publishing Malabar, FL.
[5] E.A.Krasnoshchekov and A.S.Sukomel, 1977, Problems in Heat
Transfer, MIR Publishers, Moscow.
[6] Professor John R.Thom, © 2004, “Wolverine Tube Heat Transfer Data
Book III”, Wolverine Tube Inc, www.Wolverine.com.
[7] J.P. Holman, 1963, Heat Transfer, 8
th
ed, McGrawHill Book Company.
[8] Dr. K. J. Bell and Dr. A.C.Muller, ©1984, “Wolverine Tube Heat
Transfer Data Book II”, Wolverine Division of UOP Inc,
www.Wolverine.com.
World Academy of Science, Engineering and Technology 46 2008
611
The diametral clearance between the shell internal diameter Ds and the diameter of the baffle Db is Lsb. Engineering and Technology 46 2008 ‘Design’ defines as the process of determining all essential constructional dimensions of an exchanger that must perform a given heat duty and respect limitations on shellside and tubeside pressure drop. The dimensions Ds. removable inserts are installed inside the tubes. 3. Properly designed units with tube inserts normally are much N t = CTP N t = CTP 2 πD s 4CLPt2 (1) 2 πD otl 4CLPt2 (2) Fig.1 reproduced from Tabordk defines the principal heat exchanger dimensions. In our program the equilateral triangular pitch geometry is set since it is mostly used. The baffle cut height is shown as a height Lbch. Use of tube inserts (wire mesh or twisted tape types) is highly effective in laminar flows inside tubes. baffle cut (% of Ds). can be obtained from a tube layout drawing of the heat exchanger [6]. the following results must be determined. L B c = bch . DESIGN PROCEDURE B D s 1 − 2 c 100 Fig. 605 . Calculation and Checking of pressure drop for tube and shell side A. Most designers employ empirical relations with a cutandtry approach that depends on their judgment and experience for convergence on a new design by extrapolation from tested units. Overall heat transfer coefficient 4. temperature range of fluids 3. B. Number of tubes and shell diameter 2. it must prepare a design for the optimum exchanger that will meet the required process conditions. Total number of tubes must be calculated by using following equations. since it shows the basic relationships. either inserts or external low fins may be the simple solution. In our program. shown in Fig. tube diameter (Dt) and tube material are chosen first. One half of Lbb is the width of this bypass channel. 1. Calculations of Total Number of Tubes The suitable tube size. Ordinarily. Our primary concern in this paper shall be thermal analysis based on analytical approach developed by Wolverine. ‘Selection’ defines choosing a heat exchanger from among a number of units already existing [2]. 1 Baffle and Tube Bundle Geometry [6] The diametral clearance between the shell internal diameter Ds and outer tube limit diameter Dotl is Lbb. flow rate of fluids 2. Calculation and Checking of Heat Load for UTube Design procedures of heat load check with calculation of heat transfer coefficient of shell and tube side are as shown in Fig. A pass partition lane is shown with a width of Lp.100 Ds L sb 2 Fig. Dotl. where the gap is equal to Lsb/2. INITIAL CONDITIONS AND REQUIREMENTS In general. Calculation and Checking of heat load 3. No other configuration is available in our program. Tube side and shell side pressure drop IV. Dotl is the outer tube limit diameter and Dctl is the centerline tube limit diameter (Dctl = Dotl . and Lbb and Lsb. Heat transfer rate 3. To overcome the fact that laminar flow heat transfer coefficients are almost independent of fluid velocity and performance is difficult to improve using plain tubes even resorting to large pressure drops. length/tube and arrangement of tubes With this information. 2 Baffle Cut and Clearance [6] This computer program was divided into three main steps. 1. tube side mass flow rate (G1) can be solved as total flow rate divided by the tubeside flow passage area per pass. an analytical approach is easier to follow for the less experienced designer. 2. the value of the baffle cut Bc is (Lbch/Ds) x100%. Calculation of total number of tubes 2. The tube configuration has to be decided. Installation of inserts on the tubeside of heat recovery units to increase energy recovery via a larger overall heat transfer coefficient and smaller temperature approaches. III. 1.Dt) where Dt is the outside diameter of the tubes). Calculation of Tubeside Heat Transfer Coefficient Because of total number of tube and total tubeside flow rate are known. the design of most heat exchangers involved initial conditions in which the following variables are known and assumed data: 1.World Academy of Science. However.
And then.175. Nu = hd = 0.World Academy of Science. To calculate tubeside flow passage area for one pass. 5 Singlesegmental Shell and Tube Heat Exchanger Showing Baffle Spacing [8] 606 .14 1 + k L 1/ 3 (6) (7) For Turbulent flow. JL. The maximum value of Baffle leakage correction JL is 1. (9) αss = (JC JL JB JR JS Jµ ) α1 In this expression.86 Re. Engineering and Technology 46 2008 smaller in size and have smaller or equal pressure drops as conventional plain tube units [7]. To calculate JL. JB. JL=0. the fraction of the crosssectional area occupied by the window (Fw) is calculated from (12).65 to 1.023 Re 0. 2. at = πdi2/4 (3) af = [ (lf nf) + ( π df x 3/2 ) ] tf (4) an = ata f (5) Where. 3 Flow Chart for Heat Load For a welldesigned unit.Pr φ 0.0. an is multiplied by Nt/one pass. JR and JS are the correction factor. Nusselt number (Nu) is a function of Reynolds number (Re) and Prandtl number (Pr). Fig.72 FC (10) FC = 12FW (11) In this equation. Pr 1/ 3 φ 0. After that. af is tip area of plate fin. (13) is used.44(1rs)+[10. Nu = d 2/3 hd = 0. hd d Nu = = 1. use the following equation for the tubeside flow passage area. Calculation of Shellside Heat Transfer Coefficient The stream analysis shellside heat transfer coefficient for singlephase flow has (αss) is used by following equation. 4. (14) (15) According to the Fig. TubeSide Heat Transfer Coefficient is also calculated by above Reynold number equation [7]. α1 is the ideal tube bank heat transfer coefficient calculated for all the flow across the tube bundle and JC.14 k L For Transistion flow.55+0. JL is calculated. 4 Tube with Insert Plate Ssb Ssb + Stb S + S tb rlm = sb Sm rs = After that. For Laminar flow.80 .2rlm) (13) From this equation rs and rlm are calculated from 14 and 15. Nu of the flow inside tube can then be calculated by above Reynold number equation. JC = 0. Fig. calculate tube side mass flow rate G1. JC typically ranges in value from 0.14 k (8) The criterion of distinguishing between laminar and turbulent flow is the observed mixing action.116(Re2/3 − 125) Pr1/ 3 φ 0. an is the net flow area for one tube. FW = θ ctl Sin θ ctl 360 2π (12) Fig. at is tube cross flow area. Reynold number of the flow inside tube can then be calculated by the following equations.44(1rs)]exp(2. JC is calculated by using (10).
expect when the first and last baffle compartments must be enlarged to allow for the placement of the respective shellside nozzles. By changing Nb. The unequal baffle spacing correction factor JS accounts for the adverse effect of an inlet baffle spacing Lbi and/or outlet baffle spacing Lbo larger than the central baffle spacing Lbc. For inlet and outlet baffle spacing equal to the central baffle spacing. The laminar flow correction factor JR accounts for this effect. The correction factor JS<1. should include the tube straight length in this compartment plus (Ds/2) or 0. Engineering and Technology 46 2008 to the flow direction Lpl is equal to ½ the actual dimension of the lane or can be assumed to be equal to a tube diameter Dt . Thus the baffle spacing at the UBend. The value for JS is determined directly from the effect on the flow velocity and is given by the following expression. 6 Tube Length Definition for UTube [8] (23) Fig. The inlet. 607 . baffle spacings Lbi. shown in Fig.0. Lpp represents the width of the bypass lane between tubes.005 (16) Lbo = Lbi + 0.25/Nb+1}+0. Sb is the bypass area.0 while for Re>100. set Lpl=0 while for a pass partition lane parallel For laminar shellside flow JR<1.3 Dotl [8].e.3 2rss )] (19) The empirical factor Cbh=1. For laminar shellside flow JR<1. the value is prorated in (27). respectively.3 Dotl Lbc = {1. use the following equation. no correction is needed. For situations without a pass partition lane or for such a lane normal to the flow direction.1) + (L bi /L bc )1n + (L bo /L bc )1n (N b . In laminar flows. The effective tube length Lta is equal to the total tube length less the combined thickness of the two tube sheets. The minimum value of JR in all cases is 0. are often equal in length to Lbc. The correction factor Jµ is greater than 1. The flow velocity in these compartments is thus lowered and has an adverse influence on heat transfer.0 (i. For Re>20 but Re>100. and the ratio rss of the number of sealing strips Nss (number of pairs if any) passed by the flow to the number of tube rows crossed between baffle tips in one baffle section Ntcc. To calculate Fsbp. The baffle layout is determined from the inlet. To evaluate this expression. Lbo are change.866Ltp for a 30˚ layout. To calculate JB. Lbc. Lbi = {1.0.25 for transition and turbulent flows (Re>100). Js = (N b .6.18 (For20 ≥ Re) (25) (26) Nc = (Ntcc + Ntcw)(Nb+1) Fsbp = Sb Sm Where Nc is the total number of tube rows crossed by the flow in the entire heat exchanger and Ntcw is the number of tube rows crossed in the window area. no correction is required and JS=1. Lpp=Ltp for a 90˚ layout and Lpp=0. no correction is needed and JR=1. one requires the ratio of the bypass to the cross flow area Fsbp.6 for turbulent flow and n=1/3 for laminar flow. Some exchangers have larger baffle spacing in the inlet and outlet nozzle compartments compared to the central baffle spacing.0 at rss ≥ ½.0 for larger inlet and outlet spacing than the central baffle spacing. (20) (21) Sb = Lbc[(DsDotl)+Lpl] 20 . Where n=0.World Academy of Science. rss = N ss N tcc Ds L pp Bc 1 − 2 100 (22) N tcc = Fig. heat transfer is reduced by the adverse temperature gradient formed in the boundary layer as the flow thermally develops along the flow channel. The effective length for determining the baffle spacing for Utube exchanger includes the straight length of the tube plus Ds/2. central and outlet baffle spacing are shown and are identifies as Lbi. and outlet baffle spacing and the effective tube length. The number of baffles (an integer) and baffle spacing can be determined from these values.1) + (L bi /L bc ) + (L bo /L bc ) (24) The maximum limit of Bundle by pass correction factor JB is 1.252Lbi}/{Nb1} (17) (18) Where Lpp=0.707Ltp for a 45˚ layout. allowing placement of the shellside nozzles without interference with the body flanges and without overlapping the first baffle. 5 depicts a singlesegmental shellandtube bundle geometry with fixed tube sheets at both heads in which the shellside flow makes one shell pass from one end of the tube bundle to the other with the flow directed across the tube bundle by the baffles. This expression has a maximum limit of JB=1 at rss ≥ 1/2 [8].0 for heating the shellside fluid and viceversa for cooling the shellside fluid. central. baffle spacing are calculated. where Ds is the shell internal diameter.Re J R = (J R ) 20 + [(J R )20 − 1] 80 (27) In above expressions. J B = exp[Cbh Fsbp (1 .4. use Fsbp. By using the following Equation. for 100 ≥ Re) while for Re>100. Lbi and Lbo.35 for laminar flow (100 ≥ Re) and Cbh=1. Lbc and Lbo. The number of baffle compartments Nb is determined from the effective tube length and the baffle spacing [6][8]. 10 JR=(JR)20= N c 0.
For the nozzle losses. The pressure drop for shellside flow is equal to the sum of the inlet nozzle pressure drop. Calculation and Checking of Pressure Drop Design procedures of Pressure Drop check for shell and tube side are as shown in Fig . (37) ∆ptotal = ∆pc + ∆pw + ∆pe ∆pc = ∆pbl (Nb1) RB RL (38) 608 . vnoz is calculated at the smallest crosssection area for flow (i. highest velocity) in the nozzle. fouling resistance (factor) associated with fluid outside tube Rfo (1/hso) and fouling resistance (factor) associated with fluid inside tube Rfi (1/hsi). Vi. a= a3 1 + 0.33 j1 = a 1 L tp Dt Re a2 a ∆p noz = 3 2 ρVnoz 2g c (35) (29) Where. And then. but using the velocity inside the tube. calculate overall heat transfer coefficient. use the following equation. the total number of tubes. q= U A FT ∆Tlm (34) If the calculated heat load is greater than the given heat load. the bundle pressure drop and the outlet pressure drop. Log Mean Temperature Difference is calculated by following equation [7]. ∆T2 = Tho – Tco subscript i = inlet ∆T1 = Thi – Tci subscripto = outlet The heat load is computed from (34) and comparing with given heat load. the bundle pressure drop for shell side is computed from using (37) and comparing with allowable pressure drop. ∆Tm = ∆Tlm = ∆T2 − ∆T1 ∆T ln 2 ∆T 1 (33) Where. The inlet and outlet nozzle pressure drops can be approximated as being equal to two velocity heads each. 8 [6][8]. K is wall thermal conductivity. FT is the correction factor. it is usually sufficient to calculate the loss for each nozzle at about three times the velocity head in the nozzle [12].e. by increasing TN.World Academy of Science. the window pressure drops ∆pw . U. 7 Flow Chart for Pressure Drop on Shell and Tube Side ρV 2 ∆Pent = 3 i 2g c (36) C. can be achieved from (31). The pressure drop for tube side is computed from the summation of (35) and (36) and comparing with allowable pressure drop. If not.14Re a 4 (30) For heat load check. the overall heat transfer coefficient. Engineering and Technology 46 2008 To calculate ideal tube bank heat transfer coefficient hs (α1). The bundle pressure drop is equal to the sum of the crossflow pressure drops ∆pc. The heat transfer area A is calculated by the following equation.7. and the two end zone pressure drops (first and last baffle compartments) ∆pe as illustrated in Fig. it can be said that the design is satisfied. So. (28) α1 = j1cpm˚Pr2/3 1. next term is represented for insert plate heat transfer area. The combined header and tube entrance losses are estimated in a similar way. Fig. πd x3 (32) A = (πd o L1 ) + 2 nL f + f L1 N t 2 In this equation. (31) 1 Uo = ro 1 1 ro ro 1 1 + ln + + + r i h i h si k w ri h so h o In this equation. and the whole procedure may be repeated till the above condition is satisfied.
9 A fact of Reynolds Number on Number of Tubes N ∆p e = ∆p bl 1 + tcw R B R S N tcc (48) 609 .3 2rss )] R L = exp . Rs is 2. f1 is the friction factor and the empirical constants.14Re b4 µ Rµ = µw m (41) (42) (43) (44) (45) R B = exp[Cbp Fsbp (1 . RL is the leakage correction factor.9 84 282 0. CASE STUDY A transmission oil cooler is designed according to the MatLAB program discussed above. 11000 10000 9000 M ɺ mw = Sm S w 8000 (46) 7000 ɺw 0.3 2 Fig.2 for turbulent flow [6][8].33(1 + rs )r P = 0. Rµ is the viscosity correction factor and RB is the bypass correction factor.7 282 0.33 f 1 = b1 L /D tp t Re b2 b (40) L R = bc s L bo 2− n L + bc L bi 2−n (49) ∆pe is the pressure drop in the two end zones of the tube bundle and Rs is the pressure drop correction for unequal baffle spacing at the inlet and/or outlet with respect to the central baffle spacing.7 for transition and turbulent flows (Re>100).6N tcw ) R L R µ 2ρ Where M is the shellside flow rate in kg/s and ∆pw and m˚w are the pressure drop and mass velocity in all Nb window zones for turbulent flow (Re>100). ‘n’ is 1 for laminar flow and n is 0.World Academy of Science. For all baffle spacing of equal length.0. 12000 R eynolds N ber in Tube um Where ∆pbl is the ideal bundle pressure drop for one baffle compartment of the Nb compartments.001m 2 (47) ∆p w = N b (2 + 0. Engineering and Technology 46 2008 ∆pbl = 0.15(1+rs) + 0. TABLE I INPUT DATA FOR TRANSMISSION OIL COOLER FOR DIESEL HYDRAULIC LOCOMOTIVE Fluid Inlet Temperature (˚C) Outlet Temperature (˚C) Flow rate (m3/hr) Heat load (kW) Pressure drop (bar) Max working pressure (bar) Tube side Hydraulic oil 105 80.002 f1 Ntcc m˙2/ρ Rµ (39) 1. The limit of RB is 1 for rss ≥ 1/2.7 27.5 for laminar flow (100 ≥ Re) and Cbh is 3. 8 Pressure Drop Regions in Shellside Flow [8] b3 b= 1 + 0.6 15 Shell side water 77 79.1. V. Use Cbp is 4. 6000 180 200 220 240 260 Number of Tube 280 300 320 340 Fig.8 [ p lm ] The following figures are obtained by using the MatLAB Program.
0127 4 283. m Total no. this graph also describes. V. Because of mass flow rate are constant. ho are also decrease. 11 and Fig. 9 and Fig.World Academy of Science. Reynold Number increase. 13 A fact of Heat Load on Shell Diameter TABLE II COMPARISON WITH INDUSTRIAL DATA 1340 1345 1350 1355 1360 1365 Result Shell diameter.4032 0. 11 A fact of Reynolds Number on Number of Baffles and Length of Tube Industrial Data 0. Engineering and Technology 46 2008 1500 H Transfer Coefficient in Tube (W K eat /m ) 2 11000 10000 1400 H T eat ransfer C oefficient in Shell (W K /m ) 9000 8000 7000 6000 5000 4000 3000 2000 1000 1335 1340 1345 1350 1355 1360 1365 Length of Tube (mm) Nb=2 Nb=3 Nb=4 1300 1200 1100 1000 900 180 200 220 240 260 280 Number of Tube 300 320 340 Fig. due to Number of Baffles increase.02 0. 10 shows that the Re and ho are gradually decreased corresponding to as high as Nt. there are also a slight rise in heat transfer coefficient because of changing from turbulent flow to transition flow.25 248 0. 12 A fact of Heat Transfer Coefficient on Number of Baffles and Length of Tube 290 280 270 260 30000 28000 Reynolds Number in Shell Nb=2 Nb=3 Nb=4 H t L ad (k ) ea o W 250 240 230 220 Nb=2 26000 24000 22000 20000 18000 16000 14000 1335 210 200 300 310 320 330 340 350 360 370 380 Nb=3 Nb=4 390 400 Shell Diameter (mm) Fig. 13. The result table for Heat Load is shown in Fig.3 Calculated Data 0. 32000 2 Fig. 12 shows that the Re and h are gradually decreased corresponding as high as Tube Effective Length (L1). 10 A fact of Heat Transfer Coefficient on Number of Tube It can be seen that in Fig.6 0. m Number of Baffle Heat Load.1652 The decreasing pattern of curves of Reynold Number and heat transfer coefficient shown in Fig. of tubes Tube diameter.39 1. Between total number of tube 220 and 240. However. there are a slight rise in heat transfer coefficient because of changing from turbulent flow to transition flow. To reduce size and cost of heat exchanger. velocity are increase and Re. Between shell diameter 320m and 330m.014 4 282 0. kW Tube side pressure drop Shell side pressure drop Tube Length (mm) Fig.25 318 0.385 1. UBend Exchangers 610 . m Tube length. DISCUSSION AND CONCLUSION There are few limitations to be considered in program.
Krasnoshchekov and A. 3rd ed. Professor John R. Wolverine Tube Inc. John Wiley and Sons Inc. FL. McGrawHill Book Company. E. But total numbers of tubes are more than existing design because of tube diameter changes.P and M. A. Turbular Exchanger Manufacturers Association. In this design. The tube and shell pressure drop for acceptable limits is 0. The arrangements of tubes are only for a widely used triangular pattern which permits the use of more tubes. so this diameter is less than existing tube diameter 14mm.Peters and Klaus D.S. In this program. segmental baffle is selected in this design. ©1984. www.4032 and 0. McGrawHill Book Company. It is still assumed that there is no advantage in varying less than the allowable pressure drop and baffles will be employed within the minimum (one fifth of the shell diameter) and maximum (inside diameter of the shell) spacing. Because of these facts. It is a circle of near shell diameter from which a horizontal or vertical portion has been cut. the calculation can be repeated with a revised value of total number of tube because of the pressure drop depends on the square of the velocity. This type of baffle is probably the most popular. REFERENCES [1] [2] Frass. 4th ed. “Wolverine Tube Heat Transfer Data Book II”. Among them. W. These limitations stem from the fact that at wider spacing the flow tends to be axial rather than across the bundle and at closer spacing there is excessive leakage between baffles and the shell.P.7 mm.6 and 0. 8th ed. the shell side pressure drop will be satisfactory. Engineering and Technology 46 2008 are chosen. Dr. K. Problems in Heat Transfer. MIR Publishers.C. Tube inserts flat plate in existing design is more confused than calculated flat plate. The calculated pressure drop for tube and shell are 0.com.L. Bell and Dr. 1963.Wolverine. Standards of TEMA. J. ©1958. Moscow.Wolverine.com. Heat Exchanger Design. use simple insert flat plate in tubes. 1998. 1965. TEMA. Furthermore.Sukomel. Heat Transfer. Wolverine Division of UOP Inc. “Wolverine Tube Heat Transfer Data Book III”. www. 25 percent baffle cut is used according to the method of Wolverine. tube diameter is limited to 12. There are many types heat baffle. J. Krieger Publishing Malabar. Timmerhuaus. Kays. increase total number of tube.2 kW).1652 bar. If the tube side pressure is not within acceptable limits.4 m. The limitation of shell diameter is within the minimum 0. 3 and 4. So this computer program is highly useful to design the liquidtoliquid shell and tube type heat exchanger.3 bar.M and A.3 m and maximum 0.London. The design is satisfied because the pressure drop for both sides is lower than the limited pressure drop. 1977. By using of tube inserts (flat plate) are highly effective in laminar flow inside tubes. The limitation of number of baffle is 2. Compact Heat Exchangers. New York. If the shell side pressure drop is higher than the acceptable limits. Shell diameter for calculated design is nearly the same the existing design. A.Muller. © 2004. 1999.A. 8th ed. the baffles spacing is increased slightly. heat transfer area and heat load is more than limited heat load (283. [3] [4] [5] [6] [7] [8] 611 .Thom. Max S. Holman.World Academy of Science. The cutout portion which represents the free flow area for the shell side fluid is usually from 20 to 50 percent of the open shell area. Plant Design and Economics for Chemical Engineers.Necatic Ozisik.