International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016

ISSN: 2395-3470
www.ijseas.com

Designing and Performance Evaluation of a Shell and Tube Heat Exchanger using Ansys
(Computational Fluid Dynamics)
Digvendra singh1, Narayan Das Pal2
1 Post graduate student, Bansal Institute of Research and Technology, Bhopal, India.
2 Assistant Professor, Bansal Institute of Research and Technology, Bhopal, India

ABSTRACT exchanger.CFD simulations are used for the

This work is based designing and performance optimum positioning of the inlet nozzle which
Evaluation of a Shell and Tube Heat Exchanger could be proposed from the uniform distribution
using Ansys (Computational Fluid Dynamics). of the liquid methanol and the uniform velocity
Heat exchanger has a variety of applications in distribution though each and every tube. The
different industries and in this study one such main objective of this research project is to
heat exchanger is taken into account. The heat verify the heat exchanger designed with the use
exchanger is designed as per the commercial of the Kern’s technique, by the use of
needs of the industry. In current manufacturing Commercial Computational Fluid Dynamics
process in the steel industries water is being (CFD) software. For the simulation, purpose a
used both as shell and tube side in heat symmetric view of the simplified geometry of
exchanger. The shell side fluid is water exit from the heat exchanger is made using Solidworks
the heat exchanger and goes to cooling tower for software. In the present study, CFD simulation
cooling of fluid but water is not sufficient cooled is used to study the temperature and velocity
by the cooling tower for again using in heat profiles through the tubes and the shell. It is
exchanger. found that the heat transfer through the length of
The new designing dimension of a shell and tube the tube is not uniform.
heat exchanger having 21 tubes, 170mm shell Keywords: heat exchanger, kern’s technique,
diameter and 610 long in this design does not CFD, solidworks software
include the type of the header to be used, so I
1. INTRODUCTION
have taken three types of header for analysis
Heat exchangers are devices used to transfer
which can provide a uniform velocity in the inlet
heat energy from one fluid to another. Typical
of each tube. Modifying the heat exchanger the
heat exchangers experienced by us in our daily
effectiveness of heat exchanger is increased by
lives include condensers and evaporators used in
changing the shell side fluid (from water to
air conditioning units and refrigerators. Boilers
methanol). Methanol provide sufficient cooling
and condensers in thermal power plants are
to as compared water in heat exchanger and
examples of large industrial heat exchangers.
increase the effectiveness of heat

427
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
There are heat exchangers in our automobiles in counter flow arrangement of Figure 1.1b, the
the form of radiators and oil coolers. Heat fluids enter at opposite ends, flow in opposite
exchangers are also abundant in chemical and directions, and leave at opposite ends.
process industries. Alternatively, the fluids may move in cross flow
Different heat exchangers are named according (perpendicular to each other), as shown by the
to their applications. For example, heat finned and without fin tubular heat exchangers
exchangers being used to condense are known as of Figure 1.2. The two configurations are
condensers; similarly heat exchangers for typically differentiated by an idealization that
boiling purposes are called boilers. Performance treats fluid motion over the tubes as unmixed or
and efficiency of heat exchangers are measured mixed. In Figure 1.2a, the fluid is said to be
through the amount of heat transferred using unmixed because the fins inhibit motion in a
least area of heat transfer and pressure drop. A direction (y) that is transverse to the main-flow
better presentation of its efficiency is done by direction (x). In this case the fluid temperature
calculating over all heat transfer coefficient. varies with x and y. In contrast, to the unfinned
Pressure drop and area required for a certain tube bundle of two configurations are typically
amount of heat transfer, provide an insight about differentiated by an idealization that treats fluid
the capital cost and power requirements motion over the tubes as unmixed or mixed. In
(Running cost) of a heat exchanger. Usually, Figure 1.2a, the fluid is said to be unmixed
there are lots of literature and theories to design because the fins inhibit motion in a direction (y)
a heat exchanger according to the requirements. that is transverse to the main-flow direction (x).
A good design is referred to a heat exchanger In this case the fluid temperature varies with x
with least possible area and pressure drop to and y. In contrast, to the unfinned tube bundle of
fulfil the heat transfer requirements [1]. Figure 1.2b, fluid motion, hence mixing, in the
1.1 Classification of Heat Exchanger transverse direction is possible, and temperature
variations are primarily in the main flow
Heat exchangers are typically classified
direction. Since the tube flow is unmixed, both
according to flow arrangement and type of
fluids are unmixed in the finned exchanger,
construction. The simplest heat exchanger is one
while one fluid is mixed and the other unmixed
for which the hot and cold fluids move in the
in the unfinned exchanger. The nature of the
same or opposite directions in a concentric tube
mixing condition can significantly influence heat
(or double-pipe) construction. In the parallel
exchanger performance. Another common
flow arrangement of Figure 1.1a, the hot and
configuration is the shell-and-tube heat
cold fluids enter at the same end, flow in the
exchanger. Specific forms differ according to the
same direction, and leave at the same end. In the

428
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
number of shell-and-tube passes, and the achieve a very large (≥ 400 m²/m³ for liquids
simplest form, involves single pass. and ≥ 700 m²/m³ for gases) heat transfer surface
area per unit volume. Termed compact heat
exchangers, these devices have dense arrays of
finned tubes or plates and are typically used
when at least one of the fluids is a gas, and is
hence characterized by a small convection
coefficient. The tubes may be flat or circular, as
in Figures 1.5a and 1.5b, c, respectively, and the
fins may be plate or circular, as in Figures 1.5a,
b and 1.5c, respectively. Parallel-plate heat
Figure:-1.1 Concentric tube heat exchangers. (a)
exchangers may be finned or corrugated and
Parallel flow. (b) Counter flow. [2]
may be used in single-pass (Figure 1.5d) or
multipass (Figure 1.5e) modes of operation.
Flow passages associated with compact heat
exchangers are typically small, and the flow is
usually laminar.

Figure:-1.2 Concentric tube heat exchangers. (a)

Tube Parallel flow.
and shell (b) Counter
passes, flow.
is shown in [2]
Figure 1.3.
Baffles are usually installed to increase the
convection coefficient of the shell-side fluid by
Figure-1.3 Shell-and-tube heat exchanger with
inducing turbulence and a cross-flow velocity one shell pass and one tube pass (cross-counter
component. In addition, the baffles physically flow mode of operation) [2]
support the tubes, reducing flow-induced tube
vibration. Baffled heat exchangers with one shell
pass and two tubes passes and with two shell
passes and four tube passes are shown in Figures
1.4a and 1.4b, respectively. A special and
important class of heat exchangers is used to

429
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
employed for heating district systems, largely
being used now days. Air conditioners and
refrigerators also install the heat exchangers to
condense or evaporate the fluid. Moreover, these
are also being used in milk processing units for
the sake of pasteurization. The more detailed
applications of the heat exchangers can be found
in Table 1.1 with respect to different industries
[3].

Figure-1.4 Shell-and-tube heat exchangers. (a) One shell 2. PROBLEM IDENTIFICATIONS

pass and two tube passes. (b) Two shell passes and four In current manufacturing process in the steel
tube passes. [2]
industries water is being used both as shell and
tube side in heat exchanger. The shell side fluid
is water exit from the heat exchanger and goes to
cooling tower for cooling of fluid but water is
not sufficient cooled by the cooling tower for
again using in heat exchanger. After modifying
the heat exchanger the effectiveness of heat
exchanger is increased by changing the shell
side fluid (from water to methanol). Methanol
provides sufficient cooling to as compared water
in heat exchanger. A Shell and Tube Heat
Figure 1.5 Compact heat exchanger cores. (a)
Fin–tube (flat tubes, continuous plate fins). Exchanger to coolant Methanol from 298 K to
(b) Fin–tube (circular tubes, continuous plate
331.22 K. Flow-rate of Methanol is 0.055036
fins). (c) Fin–tube (circular tubes, circular fins).
(d) Plate–fin (single pass). (e) Plate–fin kg/sec. Water will be used as the tube side, with
(multipass). [2] a temperature from 373 K to 354.31 K. In
current paper, a shell and tube heat exchanger is
1.2 Applications of Heat exchangers
selected from steel industries. The proposed
Applications of heat exchangers is a very vast
design will be based on the Kern’s method. The
topic and would require a separate thorough
design obtained from the Kern’s technique is
study to cover each aspect. Among the common
analysed and evaluated using the CFD software
applications is their use in process industry,
which will give us a detailed view of the
mechanical equipment’s industry and home
temperature, velocity and pressure profile. These
appliances. Heat exchangers can be found

430
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
profiles will have us understand the flow inside • For the identical tube pitch and flow
the shell and as well as tubes. The flow across rates, the tube layouts in decreasing
the baffle plates will also matter as they provide order of shell-side heat transfer
the turbulence according to the tubes and helps coefficient and pressure drop are: 30°,
in better heat transfer. 45°, 60°, 90°.

3. METHODOLOGY • The 90° layout will have the low heat

transfer coefficient and the low pressure

3.1 Kern’s technique of Shell-And-Tube Heat drop.

Exchanger • The square pitch (90° or 45°) is used

Kern’s technique was based on real industrial when jet or mechanical cleaning is

work on commercial heat exchanger. necessary on the shell side.

3.1.1 Shell and Tube Heat Exchanger

The number of tubes and size of tubes in heat
exchanger depends up on the
• Fluid flow rates.
• Pressure drop.
• The number of tubes and size of tubes
selected. Figure 3.1 Tube Layout and
Flow Scales [4]
• The lower velocity limit corresponds to
limiting the fouling factor. 3.2 Basic Design Procedure and Theory

• The upper velocity limit corresponds to As the objective of this present project is to
limiting the rate of erosion. evaluate the Kern’s technique which is the best
technique and by far the most used method of
3.1.2 Tube Layout the shell and tube heat exchanger design, below
• o
Triangular pitch (30 layout) is better for
P P
I have mentioned some of the most useful
heat transfer and surface area per unit equations for the designing of the shell and tube
length (greatest tube density.) heat exchanger using Kern’s technique.
• Square pitch (45° & 90° layouts) is
The general equation for heat transfer across a
needed for cleaning.
surface is:
• The 30°, 45° and 60° are staggered, and
90° is in line.
⟹ 𝑄 = 𝑈𝐴𝐴∆𝑇𝑚 [1] ................(4.1)

431
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
Where, available. The overall coefficient is the
Q = Heat transferred per unit time, W, reciprocal of the overall resistance to heat
U = the overall heat transfer coefficient, transfer, which is the sum of several individual
W/m²°C, resistances. For heat exchange across a typical
A = Heat-transfer area, m² heat exchanger tube the relationship between the
∆𝑇𝑚 = The mean temperature difference, the overall coefficient and the individual
temperature driving force, °C. coefficients, which are the reciprocals of the
Collect physical properties and Heat individual resistances, is given by:
Exchanger specifications: ⟹
Physical properties P

dn
d o ln
1 1 1 di + do × 1 + do × 1
= + +
Table 3.1 Physical properties of fluids U o hs ho d 2k w di hi di hi
P

(Methanol & Water) [1] (4.2)

Where,
Physical properties Methanol Water
𝑈𝑜 = the overall coefficient based on the outside
Cold fluid Hot fluid
area of the tube, W/m² ºC,
3
Density (kg/m ) P P 785 998.2 ℎ𝑜 = outside fluid film coefficient, W/m² ºC,
ℎ𝑖 = inside fluid film coefficient, W/m² ºC,
ℎ𝑜 𝑑 = outside dirt coefficient (fouling factor),
Specific heat C p R 2534 4182
(J/kg-K)
R W/m² ºC,
ℎ𝑖 𝑑 = inside dirt coefficient, W/m² ºC,
𝑘𝑤 = thermal conductivity of the tube wall
Heat Exchanger specifications
material, W/m ºC,
 Sea water is corrosive, so assign to tube-
𝑑𝑖 = tube inside diameter, m,
side.
𝑑𝑜 = tube outside diameter, m.
 Use one shell pass and two tube passes.
The magnitude of the individual coefficients will
 At shell side, fluid (methanol) is
depend on the nature of the heat transfer process
relatively clean. So, use 1.25 triangular
(conduction, convection, condensation, boiling
pitch.
or radiation), on the physical properties of the
(Pitch: Distance between tube centres)
fluids, on the fluid flow-rates, and on the
The prime objective in the design of an
physical arrangement of the heat-transfer
exchanger is to determine the surface area
surface. As the physical layout of the exchanger
required for the specified duty (rate of heat
cannot be determined until the area is known the
transfer) using the temperature differences
design of an exchanger is of necessity a trial and

432
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
error procedure. The steps in a typical design Where,
procedure are given below:
Ws= fluid flow-rate on the shell-side, kg/s,
1. Calculate the area for cross-flow As for 𝜌𝜌=⁡shell-side fluid density, kg/m3.
the hypothetical row of tubes at the shell
3. Calculate the shell-side equivalent
equator, given by:
diameter (hydraulic diameter), Figure 1.
(𝑃𝑡− 𝑑𝑜 )𝐷𝑠 𝑙𝑠
⟹ 𝐴𝐴𝑠 = [1] (4.3) For a square pitch arrangement:
𝑃𝑡

1.27 2
=de ( pt − 0.785d 02 )
do [1]

Fig.3.2 Arrangement of Tubes in Heat

Exchanger [4]
Figure 3.3 Equivalent diameter, cross-
Where, sectional areas and wetted perimeters [5]

𝑃𝑡 = tube pitch,
4. Calculate the shell-side Reynolds number,
𝑑𝑜 = tube outside diameter,
given by:
𝐷𝑠 = shell inside diameter, m, Gsde
Re =
𝑙𝑠 = baffle spacing, m. µ [1]

(𝑃𝑡− 𝑑𝑜 )
5. For the calculated Reynolds number, read the
The term is the ratio of the clearance
𝑃𝑡 value of jh from Figure 2 for the selected baffle
between tubes and the total distance between cut and tube arrangement, and calculate the
tube centres. shell-side heat transfer coefficient hs from:

2. Calculate the shell-side mass velocity 0.14

hs d e  µ 
Gs and the linear velocity us: =
Nu = jh Re Pr1/3  
kf  µw  [1]
𝐺𝐺𝑠𝑠=𝑊𝑊𝑠𝑠/𝐴𝐴𝑠𝑠 (4.4)
⟹
𝑢𝑢𝑠𝑠=𝐺𝐺𝑠𝑠/𝜌𝜌 (4.5) Where,
⟹ Nu = Nusselt number =
ℎ 𝑖 𝑑𝑒
𝑘𝑓

433
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com

Gsde Q = 4.301kw
Re = Reynolds number =
µ
Heat capacity methanol, C pc = 2.534 kJ/Kg k R R

𝐶𝑝 𝜇 The cold and hot stream heat loads are equal so,
Pr = Prandtl number =
𝑘𝑓
cooling methanol flow rate is calculated as
ℎ𝑖 = inside coefficient, W/m² ºC,
follow
𝑑𝑒 = equivalent (or hydraulic mean) diameter,
Cooling methanol flow,
m
𝑈𝑡 = fluid velocity, m/s, Q = m c .C pc (t c2 -t c1 )
R R R R R R R R [1]
𝑘𝑓 = fluid thermal conductivity, W/m ºC,
Q
mc =
𝐺𝐺𝑠 = mass velocity, mass flow per unit area,
(
C pc tc1 − tc2 )
R R

kg/m²s,
𝜇𝜇 = fluid viscosity at the bulk fluid 4.301
=
temperature, Ns/m², 2.534(58.22 − 25)
𝜇𝜇 = fluid viscosity at the wall,
mc = 0.0511 kg/s
𝐶𝑝 = fluid specific heat, heat capacity, J/kg ºC
R R

The well known "logarithmic mean" temperate

different (LMTD or θ m ) is calculated by
3.3 CALCULATIONS
∆TLMTD =
(
th1 − tc2 − th2 − tc1) ( )
On the basis of the given problem we will be  th − tc2 
ln  1
 th − tc 
calculating the values of the basic dimensions of  2 1 

the shell and tube heat exchanger .Only the θ −θ

θm = 2 1
thermal design will be considered. The Kern’s ln (θ 2 / θ1 )
method will be employed for the calculation. (100 − 58.22) − (81.31 − 25)
θm =
As this process is an iterative process, I’m  100 − 58.22 
ln  
only representing the final iteration of the  81.31 − 25 
calculation. As coolant is corrosive, so it is
θ m = 48.68o C [1]
assigned to the tube-side.
Choosing one shell pass and one tube pass for
Counter Flow Heat Exchanger the designing purpose.
Heat capacity water, Cp h = 4.182 Kj/kg k
R R
Assuming,
U= 140W/mºc
Heat load, Q = m h . c ph (t h1 -t h2 ) [1]
Provisional Area,
R R R R R R R R

= 0.055036 X 4.182 X (100-81.31) Q = U.A.Qm

A= Q/U.Qm

434
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com

4.301×103 An estimate of the bundle diameter Db can be

A=
140 × 48.68 obtained from equation below which is an
empirical equation based on standard tube
A = 0.6310m2 P

layouts. The constants for use in this equation

Choosing, outer diameter (𝑑𝑜 ) 20mm, inner
for triangular and square patterns.
diameter (𝑑𝑖 ) 16mm, length of the tube to be 1/ n1
N 
610mm and Aluminium as the material of the Db = d 0  t 
tube. The selection of the tube is based on the  K1  [1]
most easily available size of tube whereas
Where,
selection of the tube length is arbitrary.
K 1 & n 1 = Constant
R R R R

L = 610mm
K1 = 0.215
= π DL
R R

Surface Area of one tube A

n1
R = 2.207
R

= 3.14 X 16 X 10-3 X 0.61

Bundle diameter
P P

= 0.030 m2 P

1
 21  2.207
Numbers of tube Nt
R =
R

Db = 20  
Total outside surface of tubes (Provisional area)  0.215 
Outside surface area of one tube
Db R = 145mm
R

Bundle clearance = 25mm

0.6310
=
0.030 Use a spilt ring floating head type
From figure 3.3
= 21.033
Bundle diametrical clearance = 25mm

NtR =
R 21 Nos. Shell diameter (D s ) R R

Ds= Db+ clearance

Tube Pattern Applications Ds= 145+25
 The triangular and rotated square Ds = 170mm
patterns give higher heat transfer rates, Tube side coefficient
but at the expense of a higher pressure Now we have to calculate the heat transfer
drop than, the square pattern. coefficient i.e. h i for the tube side.
R R

 A Square or rotated square arrangement 100 + 81.31

Mean water temperature =
is used for heavy fouling of fluids, 2
where it is necessary to mechanically = 90.65oC
P P

clean the outside of the tubes.

π
 As the shell side fluid is relatively clean Tube cross sectional area a= d2
4
use 1.25 square pitch.

435
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
π The equation can also be calculated using
= × 162 =
201mm2
4 equation below; this is done to illustrate use of
this method.
Total flow area = 21 X (201 X 10-6) P P

0.14
hi di  µ 
= 4.221 X 10-3 m2 = jh Re Pr 0.33  
kf  µw 
P P P

[1]
0.055036
Water mass velocity (G t ) = Where
4.221× 10−3
R R

hi = inside coeff. (W/m2oC)

G t = 13.03 kg/sm2
P P

R R P

jh
R R = heat transfer factor (dimensionless)
Density of water,( ρ water )
R
R = 998.2 kg/m³ Re = Reynolds number (dimensionless)
Pr = Prandtl number (dimensionless)
Water linear velocity,
µ = Viscosity of water (N s/m2) P P

Gt
ut = µw = Viscosity of water at wall temperature
ρ water
R R

(Ns/m2) P P

ut
R R =
13.03 Viscosity of water ( µ ) = 0.8 m N s/m2 P

998.2 Thermal conductivity (k f ) = 0.59 W/mºC

R R

ρµ di 998.2 X 0.013 X 16 X 10−3

ut = 0.013 m/s =
Re =
R R

µ 0.8 ×10−3

Coefficient for water can also be calculated by

Re = 260
using below equation
C p µ 4.182 × 0.8 ×10−3
4200(1.35 + 0.02 t) u 0.8 =
Pr =
hi = t Kf 0.59
di0.2
Pr =0.0056
The above equation has been adapted from data
given by eagle and ferquson (1930) Neglect µ / µ w ≈ 1
Where, Put this value: -
0.14
kf  µ 
hi= inside coefficient for water, W/m2oC hi = jh Re ( Pr )
0.33
 
P P

T avgR = t = water mean temperature oC

R P P
di  µw 
4200(1.35 + 0.02 × 90.65) × (0.013)0.8 Coefficient j h = 3.9 X 10-3
hi = R R P

(16)0.2

hi = 236.41 W/m2oC P P

436
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
hi= R R Shell equivalent Diameter (hydraulic diameter),
0.59
−3
× 3.9 ×10−3 × 260 × (0.0056)0.33 ×1)0.14 = 1.27 2
( pt − 0.785d 02 )
16 ×10 de
do
hi R = R 6.755 W/m2oC P P

Shell side coefficient 1.27

d=
e (252 − 0.785 × 202 )
20
Now the shell side heat transfer coefficient is d e = 19.7 m
calculated i.e. h s R

A close baffle spacing will give higher heat Mean shell side temperature,
58.22 + 25
transfer coefficients but at the expense of higher =
2
pressure drop.
= 41.61
l b = Choosing baffle spacing = Ds/5
= 42o c
R R

= 170/5 =34mm
Methanol density( ρ )= 785 kg/m3 P

Tube pitch, Pt = 1.25 do

Viscosity, µ= 0.34 mNs/m2 P

= 1.25 X 20 Heat Capacity , Cpc= 2.534 KJ/Kgk

Thermal Conductivity , K f = 0.19 W/m ºC
R R P P

Pt = 25mm
Gsde
Re =
Area of cross flow, µ

As =
( Pt − do ) Dslb
Pt 20.50 ×19.7 ×10−3
=
(25 − 20) ×170 × 34 ×10−6 0.34 ×10−3
As =
25
−3 2
Re= 1187.79
=
As 1.156 ×10 m
Re= 1188
Mass velocity,
Gs =
Ws C pc × µ 2.534 ×103 × 0.34 ×10−3
As Pr= =
Kf 0.19

0.023675
Gs = −3
kg / sec m 2 Pr= 4.5
1.156 ×10 Crossing 15% baffle cut
Heat transfer factor
Gs = 20.48
i h = 4.6 X 10-2
R R P

Gs= 20.50 kg/m2 P Without the viscosity correction term,

Where
Ws= methanol flow rate on the shell side kg/sec

437
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
0.14 Taking the fouling coefficients from table 3.1
hs d e  µ 
=
Nu = jh Re Pr1/3   Wm-2oC-1,
 µw 
kf methanol (light Organic) 5000 P P P P

brackish water (sea water), take as highest value,

µ 3000 Wm-2oC-1 P P P

Neglect , =1
µw d o ln
dn
1 1 1 di + do × 1 + do × 1
= + +
0.19 U o hs ho d 2k w di hi di hi
=hs −3
× 4.6 ×10−2 ×1188 × (4.5)1/3
19.7 ×10
1 1 1 20 ×10−3 ln(20 /16) 20 1 20
= + + + × +
ℎ𝑠 = 870.16W/m2 0C P P U o 870.10 5000 2 × 50 16 3000 16

Estimate wall temperature,

Mean temperature difference, U o =140.88 w/m2oc
R R P P

Across all resistance = 90.65 -42 Above assumed value of 140 W/m oC P P

o
= 48.65 C P P

Dimensions of Shell and Tube Heat

U
Across methanol film = ∆T Exchanger
hs
Tubes:
140 1. Outer dia. = 20 mm
= × 48.6
870.16 2. Inner dia. = 16 mm
3. Length of the tubes = 610 mm
= 7.8oC = 8oC
P P P P

4. No. of the tubes = 21 tubes

Mean wall temperature =42-8 5. Bundle tubes = 145 mm
= 34oC P P

µw =0.37m N s/m2 P
6. Bundle dia. Clearance = 25 mm
Shell:
0.14
 µ  1. Inner dia. = 170 mm
  = 0.988
 µw  2. Baffle spacing = 34 mm
= 0.99 3. Baffle thickness = 2 mm
4. No. of baffle plates = 25
Which shows that the correction for a low
viscosity fluid is not significant 4. RESULTS & DISCUSSIONS

In beginning of the project we calculated the

Overall Coefficient
design values for our heat exchanger. Those
Thermal Conductivity of aluminium alloys =
calculated values were used in the CFD
237 W/m oC P P

simulation for the analysis of our heat

exchanger, now to study our design through

438
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
Ansys we studied the temperature pressure and
velocity profiles for our heat exchanger.

Figure 4.1 Project Schematic

4.1 Design Modeler
This is Ansys design modular environment. Here
the physical geometry of the heat exchanger can
be drawn or imported from another CAD
software like CREO, CATIA, SOLIDWORKS,
UNIGRAPHICS, etc. The heat exchanger
modelled here contains 21 tubes each 610 mm
long. We have also modelled internal and
external fluids. Since this is a symmetrical type
model we have shown symmetrically cut section
model and this will reduce computation time
without affecting the results.

(a)

Figure 4.2 Design Modeler

4.2 Meshing

439
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com

Figure 4.4 the mesh generated for

the geometry of the shell and tube
Heat Exchanger

This is Ansys meshing environment. Here the

CAD model of heat exchanger is divided into
several numbers of nodes and elements.
Meshing is required to get the solution at
different locations of the heat exchanger. The
solver will calculate for each node (using
(c) (d)
iterative method) which is created by meshing.
Figure 4.3 Meshing Process
This mesh was generated by Ansys meshing. It
Figure 4.3 (a) shows us the meshing
will be same for all cases and only boundary
tool window. From figure b and c we can
conditions will vary. From the Figure 5.4 it can
see that we can add the method we want to
be observed that the mesh contains 75192 nodes
use in the meshing process which in my
and 183967 elements. The finer is the mesh, the
case is the sweep method. Figure d shows
smaller is size of elements. This will result in
us the bunch of other options which are also
more computation time also. This mesh is an
very useful but they need very good
unstructured one and was tested for grid
configuration computers as they make a
independency. We have found that refining of
very fine quality of mesh and it results in
mesh to more smaller divisions are not required
higher no. of nodes and elements resulting
for this study as further refining will not
in a better and complex solution. Figure 4.4
improve solution to more than 5%. We have also
shows us the generated mesh designed by
found that meshing the component to coarse
using this process.
mesh would cost accuracy of the solution. So
current mesh can be acceptable. The mesh
metric skewness is also less than 0.9. It is

440
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
calculated by the Ansys software, if metric
skewness is more than 0.9 than, meshing are not
possible.

Figure 4.7 Fluent setup (General)

This is the Ansys Fluent solver environment.

Figure 4.5 the mesh generated for the Here solution method and boundary conditions
geometry of the shell and tube Heat
Exchanger are specified. Before beginning with solution
steps, the general settings are to be addressed.
4.3 Solution
Press check button in mesher section to check
The setup in fluent is used to generate the
the mesh produced by ansys mesher. The quality
solution using the boundary conditions.
of incoming mesh is obtained by pressing report
Here is a pictorial representation of the
quality button in mesher section.
process used in setting up the fluent setup.
We also need to set the units of operating
variables like length, temperature, velocity,
pressure, density, heat flux, etc.

Figure 4.8 Fluent setup (Models)

Figure 4.6 Fluent Launcher
Figure 4.6, 4.7 and 4.8 shows us the starting
windows of the fluent. As we can see there are
all the options available for the CFD simulation
in those windows. We start from the energy
equations selection from the above window as it
is applicable in our heat exchanger model. Then

441
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
next comes the turbulence model section. As I
have gone through different papers which
suggested the k-epsilon model to be the best and
effective method for a heat exchanger
evaluation. So I have also selected this model
due to its greater accuracy in heat exchanger
cases from model settings, turn on the energy
equation. Also enable the viscous settings to k-
Figure 4.10 Material Selection (tubes)
epsilon realizable settings and enhance wall
Now we need to define cell zone conditions.
functions.
Here the solid/ fluid behaviour of material is

After the equations selection comes the material defined.

selection based on the problem which in our

case is methanol water and aluminium.

Figure 5.11 Material Selection (outer fluid)

Boundary Conditions:

Boundary conditions are used

Figure 4.9 Material Selection according to the need of the model. The inlet

Here materials are assigned to the parts temperatures and velocities are given to the

modelled in cad software. Material assigned to setup.

cold fluid is methanol. It is shell side fluid.

While material assigned for hot fluid is water. It
is tube side fluid. Both of these materials are
present in Ansys Fluent’s material library.

442
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
Figure 4.12 Boundary condition of the Figure 4.13 Inlet temperature and velocity
fluids condition of the fluids
Now boundary conditions are defined. This Convergence of solution is defined at Residual
analysis will be carried out for counter flow monitor’s settings. The degree of solution
conditions. We have used 7 different boundary accuracy will be defined here.
conditions for our heat exchanger which are
tabulated below:

Table 4.1 Boundary Conditions

Inlet
Temp of Inlet
Inlet Velocity Inlet Temp
hot fluid Velocity of
Case of hot of cold fluid
(water) cold fluid
. No. fluid(water)(i (in (methanol)(i
(methanol)(i
n m/s) Kelvin’s n Kelvin’s)
n m/s)
)
1 0.156 368 0.3 298

2 0.1 373 0.3 298

Figure 4.14 Solution is initialized
3 0.156 373 0.3 293

4 0.156 373 0.3 298 Then the solution is initialized.

5 0.2 373 0.3 298

6 0.3 373 0.3 298

7 0.3 373 0.5 298

Ansys Fluent Solution will be used to simulate

solutions for all these conditions. Figure 4.15 Run calculation
Now solution methods are set up. Here second Now the solution is calculated. Numerical
order upwind scheme is chosen in spatial iterations are performed in this section till
discretization section for momentum, pressure, acceptable solution is reached.
turbulent kinetic energy, energy and turbulent 4.4 Contour Plots
dissipation rate. Solution for case 1
Conditions are:
• Inlet Velocity of hot fluid (water) =
0.156 m/s
• Inlet temperature of hot fluid (water) =
368 K
• Inlet Velocity of cold fluid (methanol) =
0.3 m/s

443
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
• Inlet temperature of cold fluid on that location with colour of colour scale.
(methanol) = 298 k

Figure 4.16.3: Velocity contour plot at

symmetrical plane

These are the contours of velocity for case 1.

Figure 4.16.1 Pressure contour plot at Velocity of any fluid on any location can be
symmetrical plane
obtained in m/s by matching the colour on that
These are the contours of pressure for case 1. location with colour of colour scale.
Pressure on any location can be obtained in
Pascal by matching the colour on that location
with colour of colour scale.

Figure 4.16.4 Velocity vectors contour plot at

Figure 4.16.2: Temperature contour plot at symmetrical plane
symmetrical plane
These are the contours of temperature for case 1.
Temperature of fluid/tubes on any location can Table 4.2 CFD Results for temperature and

be obtained in Kelvin's by matching the colour the overall effectiveness

Simulation Inlet Velocity of Inlet Temp of Outlet Temp of Inlet Velocity of Inlet Temp of Outlet Temp of Effectiveness (%)
Case. No. hot fluid (water) hot fluid hot fluid cold cold fluid cold fluid
(in m/s) (water)(in (water)(in fluid(methanol) (methanol)(in (methanol)(in
Kelvin's) Kelvin's) (in m/s) Kelvin's) Kelvin's)
T h,max
R T h,min
R T c, min
R

1 0.156 368 350.56552 0.3 298 328.99466 95.583

2 0.1 373 346.65433 0.3 298 330.34497 86.403
3 0.156 373 353.70276 0.3 293 330.74741 92.546
4 0.156 373 354.31256 0.3 298 331.22256 95.595
5 0.2 373 359.0162 0.3 298 332.51306 91.707
6 0.3 373 364.45981 0.3 298 334.12851 84.0116
444
7 0.3 373 359.61053 0.5 298 328.44592 79.028
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
These are the velocity vectors for case 1. These In current manufacturing process in the
vectors are coloured by temperature. steel industries water is being used both
Temperature of fluid molecules on different as shell and tube side in heat exchanger.
locations is obtained by matching the colour on The shell side fluid is water exit from
that location with colour of colour scale. the heat exchanger and goes to cooling
tower for cooling of fluid but water is
not sufficient cooled by the cooling
tower for again using in heat exchanger.
The new designing dimension of a shell
and tube heat exchanger having 21
tubes, 170mm shell diameter and 610
long in this design does not include the
type of the header to be used.
After modifying the heat exchanger the

Figure 4.16.5 Average temperatures effectiveness of heat exchanger is

calculated increased by changing the shell side

These are the average temperatures calculated fluid (from water to methanol).

for outlet zones of hot and cold fluids. Outlet 1 Methanol provide sufficient cooling to

represents temperature of hot fluid when it as compared water in heat exchanger

leaves the heat exchanger. Outlet 2 represents and increase the effectiveness of heat

temperature of cold fluid when it leaves the heat exchanger up to 95.5%.

exchanger. 5. Conclusion
From above figure we can clearly see significant Kern’s method is the most used and
changes like temperature drop of hot fluid and effective method for the designing
temperature rise of cold fluid at exit zones. In a procedure. In the research analysis shell and
similar way exit temperatures will be calculated tube heat exchanger is designed and
for all 7 cases. simulated by using Computational Fluid
Results: Dynamic (CFD).The header selection for
The given heat exchanger analysis was carried the heat exchanger has also been based on
out under counter flow conditions. The the Computational Fluid Dynamic (CFD)
simulated results are tabulated below for 7 simulation. I can see that the uniform flow
different conditions which were mentioned in in tubes can be achieved using a suitable
previous table also. header. The nozzle placement normal to the
plane of tubes and also eccentric to the

445
 International Journal of Scientific Engineering and Applied Science (IJSEAS) – Volume-2, Issue-3,March 2016
ISSN: 2395-3470
www.ijseas.com
head side of the headers has been the most cooling to as compared water in heat
effective. The simplified geometry of the exchanger and increase the effectiveness of
shell and tube heat exchanger is used. The heat exchanger up to 95.5%.
assumption of plane symmetry works for
REFERENCES
most of the length of heat exchanger except
[1] P.K Nag, Heat and Mass
the outlet and inlet regions where the rapid
Transfer,McGraw Hill education
mixing and change in flow direction takes
(India) Private Limited,New Delhi.
place. Thus improvement is expected if
[2] Frank P. Incropera, David P. Dewitt,
complete geometry is modeled.
Theodore L. Bergman, Adrienne S.
Furthermore, the enhanced wall functions are
Lavine “Fundamentals Of Heat And
not used in this project due to convergence
Mass Transfer”, John Wiley & Sons
issues, but they can be very useful with
SIXTH EDITION, 111 River Street,
k−epsilon models. The heat transfer is found
Hoboken, NJ 07030-5774 2007.
to be poor because the most of the shell side
[3] http://www.wcr-regasketing.com,
fluid by-passes the tube bundle without
“Heat exchanger applications”
interaction. Thus the design can be modified
http://www.wcr-regasketing.com/heat-
in order to achieve the better heat transfer in
exchanger-applications.htm, 2010.
two ways. Either, the shell diameter is
[4] P M V Subbarao "Figure:
reduced to keep the outer fluid mass flux
Arrangement of tubes, Tube Layout
lower or tube spacing can be increased to
and Flow Scales" mel709 tutorial,
enhance the inner fluid mass flux. Just doing
Professor mechanical Engineering
this might not be enough, because it is seen
Department I I T Delhi.
that the shell side fluid after 550 mm doesn’t
[5] Majid Hayati "Kern technique of
transfer heat efficiently.
Shell-And-Tube Heat
Then, change the length of tube up
Exchanger".University of Kashan,
to 610 mm from 550 mm.The new designing
Kashan, I.R. IRAN 2014.
dimension of a shell and tube heat exchanger
having 21 tubes, 170mm shell diameter and
610 long in this design does not include the
type of the header to be used. After
modifying the heat exchanger the
effectiveness of heat exchanger is increased
by changing the shell side fluid (from water
to methanol). Methanol provide sufficient

446