You are on page 1of 58

Forced Convection Studies in Tube Bank Arrangement

Using CFD Technique

Dissertation submitted to
Visvesvaraya National Institute of Technology, Nagpur
In partial fulfillment of the requirements for the award of
The degree

Master of Technology
In
HEAT POWER ENGINEERING
By
SUDEESH KUMAR PATEL (MT16HPE022)

Under the guidance of


Dr. A.S. Dhoble

Department of Mechanical Engineering


Visvesvaraya National Institute of Technology
Nagpur 440010 (India)
2018
© Visvesvaraya National Institute of Technology (VNIT) 2018

1
DEPARTMENT OF MECHANICAL ENGINEERING

VISVESVARAYA NATIONAL INSTITUTE OF TECHNOLOGY,


NAGPUR

DECLARATION
I, Sudeesh Kumar Patel, hereby declare that this dissertation titled “Forced convection studies in
tube bank arrangement using CFD technique” is carried out by me in the Department of
Mechanical Engineering of Visvesvaraya National Institute of Technology, Nagpur. The work is
original and has not been submitted earlier whole or in part for the award of any degree/diploma
at this or any other Institution/University.

Sudeesh Kumar Patel

Date:

2
DEPARTMENT OF MECHANICAL ENGINEERING
VISVESVARAYA NATIONAL INSTITUTE OF TECHNOLOGY,
NAGPUR

CERTIFICATE

This is to certify that the dissertation titled “Forced convection studies in tube bank arrangement
using CFD technique” is submitted by Sudeesh Kumar Patel in partial fulfillment of the
requirements for the award of the degree of Master of Technology in Heat Power Engineering,
VNIT Nagpur. The work is comprehensive, complete and fit for final evaluation.

Dr. A.S. Dhoble


Asst. Professor ME, VNIT, Nagpur

Dr. V.R. Kalamkar

Head, Department of Mechanical Engineering


VNIT, Nagpur
Date:

3
ACKNOWLEDGEMENTS

I take this opportunity to express my deep sense of respect and heartfelt gratitude
towards my guide Dr.A.S. DHOBLE, who not only guided my project work, but also stood as a
mentor in realizing my potential. I thank him for introducing me to the field of computational
fluid dynamics. His invaluable advice, unwavering trust and unconditional support helped
immensely in timely and successful completion of the project.

I would like to dedicate this work to my loving parents as well as friends and family for always
being there for me.

Sudeesh Kumar Patel


MT16HPE022

4
ABSTRACT

Heat transfer to or from a bank (or bundles) of tubes in cross flow heat exchanger is relevant to
numerous industrial applications; such as, steam generation in a boiler or air cooling in the coil
of an air conditioner. The purpose of this research work is to find out the effect of horizontal
pitch on the nusselt number and friction factor of tube .The compact tube bank arrangement
experiences high convective resistance on air side, due to low convective heat transfer
coefficient. The arrangement of seven cylinders in triangular arrangement is made in such a way
that maximize the turbulence of air in the tube bank arrangement. The numerical analysis of this
tube bank arrangement with anasys fluent is carried out. This work consists of circular and
diamond shape tubes, with different value of pitch for maximum heat transfer. The increase in
turbulence will increase the convective heat transfer coefficient of air due to better mixing of air.
The circular tube & diamond shape tube bank have maximum nusselt no at horizontal pitch of
2.5D & 3D respectively. The value of coefficient of friction should be minimum so that there will
be less pressure drop. The circular & diamond shape tube have least values of coefficient of
friction at horizontal pitch of 3D

5
TABLE OF CONTENTS

CHAPTER 1..........................................................................................................................................12
INTRODUCTION...................................................................................................................................12
1.1 TYPES OF HEAT EXCHANGER...................................................................................................13
1.2 STEPS FOR PROJECT COMPLETION.........................................................................................14
1.3 ORGANIZATION OF THESIS.......................................................................................................15
CHAPTER 2............................................................................................................................................16
LITERATURE REVIEW.......................................................................................................................16
2.1 EXPERIMENTAL STUDIES..........................................................................................................16
2.2 NUMERICAL AND EXPERIMENTAL STUDIES........................................................................17
2.3 NUMERICAL STUDIES................................................................................................................18
2.4 AIM AND OBJECTIVE..................................................................................................................21
CHAPTER 3............................................................................................................................................22
METHODOLOGY..................................................................................................................................22
3.1 CIRCULAR SHAPE TRIANGULAR TUBE BANK ARRANGEMENT.......................................22
3.2 DIAMOND SHAPE TRIANGULAR TUBE BANK ARRANGEMENT........................................22
3.3 VALIDATION.................................................................................................................................24
3.3.1 MODEL VERIFICATION........................................................................................................24
3.3.2 ASSUMPTIONS...........................................................................................................................26
3.4 GOVERNING EQUATIONS..........................................................................................................27
3.5 METHODOLOGY FOR CIRCULAR SHAPE TUBE....................................................................28
3.5.1 GEOMETRY MODELLING....................................................................................................28
3.5.2 MESH GENERATION.............................................................................................................29
3.5.3 BOUNDARY CONDITION.....................................................................................................29
3.6 METHODOLOGY FOR DIAMOND SHAPE TUBE.....................................................................30
3.6.1 GEOMETRY MODELLING........................................................................................................30
3.6.2 GOVERNING EQUATIONS.......................................................................................................31
3.6.3 MESH GENERATION.................................................................................................................32
3.6.4 BOUNDARY CONDITION.........................................................................................................32

6
CHAPTER 4............................................................................................................................................34
RESULTS & DISCUSSION....................................................................................................................34
4.1 RESULT & DISCUSSION FOR CIRCULAR SHAPE TUBE........................................................34
4.1.1 ANALYSIS AT A HORIZONTAL PITCH OF 3D....................................................................35
4.1.2 ANALYSIS AT LONGITUDINAL PITCH OF 2.5D................................................................36
4.1.3 ANALYSIS AT PITCH OF 2D.................................................................................................38
4.2 ANALYSIS OF RESULT................................................................................................................39
4.3 OPTIMIZATION FOR CIRCULAR SHAPE TUBE.......................................................................42
4.4 RESULT AND ANAYSIS DIAMOND SHAPE TUBE...................................................................44
4.4.1 CFD SOLUTION......................................................................................................................44
4.4.2 ANALYSIS AT LONGITUDINAL PITCH OF 2.5D................................................................45
4.4.3 ANALYSIS AT LONGITUDINAL PITCH OF 2.5D................................................................46
4.4.4 ANALYSIS AT TRANSVERSE PITCH OF 2D.......................................................................48
4.5 STREAMLINE AND VELOCITY CONTOURS............................................................................49
4.6 OPTIMIZATION FOR DIAMOND SHAPE TUBE........................................................................52
CHAPTER 5............................................................................................................................................55
CONCLUSION........................................................................................................................................55
REFERENCES........................................................................................................................................56

7
LIST OF FIGURES
Figure 1 Circular shape triangular tube bank arrangement...........................................................13
Figure 2 Diamond shape triangular tube bank arrangement..........................................................14
Figure 3 Types of heat exchanger..................................................................................................15
Figure 4 Inline tube bank arrangement..........................................................................................25
Figure 6 Validation with k-€ model...............................................................................................26
Figure 7 Circular shaped triangular tube bank arrangement.........................................................28
Figure 8 Nusselt no. Vs Reynolds no............................................................................................32
Figure 9 Skin Friction coeff. Vs reynolds number........................................................................32
Figure 10 Nusselt number Vs. Reynolds number at pitch of 2.5D................................................33
Figure 11 Skin Friction coeff. Vs. Reynolds number at pitch of 2.5D..........................................34
Figure 12 Nusselt number Vs Reynolds number...........................................................................35
Figure 13 Skin friction coefficient vs. Reynolds number at pitch of 2D.......................................35
Figure 14 Comparison of velocity contour of 4m/s at pitch 3D....................................................36
Figure 15 Comparison of velocity contour of 4m/s at pitch 2.5D.................................................36
Figure 16 Comparison of velocity contour of 4m/s at pitch 2D....................................................37
Figure 17 Diamond shaped triangular tube bank arrangement......................................................38
Figure 18 Diamond shaped tube in 2D..........................................................................................39
Figure 19 Nusselt number Vs. Reynolds number for diamond shape tube at pitch of 3D............42
Figure 20 Skin Friction coeff. Vs. Reynolds number of diamond tube at pitch 3D......................43
Figure 21 Nusselt number Vs. Reynolds number for diamond shape tube at pitch of 2.5D.........44
Figure 22 Skin Friction coeff. Vs. Reynolds number for diamond shape tube at pitch 2.5D........44
Figure 23 Nusselt number Vs. Reynolds number for diamond shape tube at pitch of 2D............45
Figure 24 Skin Friction coeff. Vs. Reynolds number for diamond shape tube at pitch 2D..........45
Figure 25 Velocity contour at 4 m/s for longitudinal pitch of 3D for diamond tube.....................47
Figure 26 Velocity contour at 4 m/s for longitudinal pitch of 2.5D for diamond tube.................47
Figure 27 Velocity contour at 4 m/s for longitudinal pitch of 2D for diamond tube....................48
Figure 28 Contours of turbulence kinetic energy at 4m/s for diamond tube at pitch 2D.............48
Figure 29 Contours of turbulence kinetic energy at 4m/s for diamond tube at pitch 2.5D..........48
Figure 30 Contours of turbulence kinetic energy at 4m/s for diamond tube at pitch 3D.............49

8
Figure 31 Comparison of Nusselt number vs. Reynolds number for circular tube.......................52
Figure 32 Comparison of skin friction coeff. Vs. Re circular tube................................................52
Figure 33 Comparison of Nusselt number vs. Reynolds number for diamond tube.....................53
Figure 34 Comparison of skin friction coeff. Vs. Re diamond tube..............................................54

9
SYMBOLS & ABBREVIATIONS

A convection heat transfer area of duct, m2

AC cross-sectional area, m2

Dh hydraulic diameter of duct, m

SL longitudinal pitch

ST transverse pitch

f friction coefficient

h average heat transfer coif, W/m2K

K thermal conductivity of air, W/mK

Nu Nusselt number

Pr Prandtl number

Re Reynolds number

Q heat flux, W

T temperature, K

Umax max. Velocity

Prt turbulent prandtl number

P static pressure

Remax Reynolds number based on max.velocity

ρvd
Remax. =
μ

Κ turbulent kinetic energy (m2/s2)

Ү ratio of specific heats

10
Greek letters

Ʋ kinematics viscosity, m2/s

ρ density of air, kg/m3

μ dynamic viscosity, kg/ms

α thermal diffusivity

11
CHAPTER 1

INTRODUCTION

The heat transfer in tube bank arrangement in cross flow heat exchanger has wide range of
applications, such as steam generation in boiler or air cooling in the coil of an air conditioner. In
energy transfer related applications, heat exchanger performance is of great importance in
meeting today’s stringent energy efficiency standards with low cost and less environmental
impact. In the liquid-to-gas and phase-change heat exchangers, typical to many Heating,
Ventilating and Air Conditioning & Refrigeration (HVAC&R) systems, the gas-side thermal
resistance contributes heavily to the overall thermal resistance. Fin-and-tube heat exchangers
have been widely applied in lots of fields, such as energy, power, chemical, food, and
refrigeration. Improved heat transfer performance will have a significant impact on the energy
crisis. For the gas-to-liquid fin-and tube heat exchanger, a high thermal resistance exists because
of the poor thermo physical property of air. For example, the airside thermal resistance can
comprise 75% of the total thermal resistance in an evaporator and 95% in a condenser for typical
refrigeration systems. Researchers are devoted to developing the enhanced heat transfer surfaces,
especially on the air side. Generally in tube bank arrangement the fluid moves inside the tubes
while second fluid moves outside the tubes. So with the help of better mixing of fluids by
providing different tube bank arrangement the heat transfer can be enhanced. The spacing
between the rows of tubes have significant effect on the heat transfer and skin friction coefficient
of tubes. The prime motive of this project is to find out the effect of horizontal pitch on the
convective heat transfer coefficient as well as on the skin friction coefficient. So the development
of the higher effective heat exchanger requires a larger contact area for heat transfer with reduced
volume. Such larger contact area is easily achieved by passing one fluid through a number of
small channels or tubes. This process with an array of the cylinder leads towards the
development of a new class of heat exchanger, as tube banks. The tube bank is a special case of
the heat exchanger, where the heat interaction is between the hot fluids flowing through a
number of small tubes and cold fluid usually passes over the tube surface along the cross
direction. There are several techniques commonly used to increase the heat transfer rate by
improving thermal contact between the heat exchanger fluid andwall. Themost commonmethods

12
typicallymanipulate the surface, including its roughness, use of coiled tubes and vortex
generators.

1.1 TYPES OF HEAT EXCHANGER

HEAT
EXCHANGERS

DIRECT
REGENERATORS RECUPERATORS
CONTACT TYPE

PARALLEL FLOW CROSS FLOW COUNTER FLOW

UNMIXED FLOW MIXED FLOW

13
Figure 1 types of heat exchanger

1.2 STEPS FOR PROJECT COMPLETION

LITERATURE REVEIW

VALIDATION

DESIGN OF
TRIANGULAR TUBE
BANK ARRANGEMENT

CIRCULAR SHAPE TUBE DIAMOND SHAPE TUBE

GEOMETRY GEOMETRY

MESHING MESHING

CFD ANALYSIS CFD ANALYSIS

14
RESULT ANALYSIS RESULT ANALYSIS

CONCLUSIONS

15
1.3 ORGANIZATION OF THESIS
This thesis is organized as shown below, according to the logical sequence of the actions taken to
achieve the objectives.

Start

Introduction & literature survey

Validation of for inline arrangement


It includes the validation of methodology that is used for the numerical simulation of flow

Analysis at different values of longitudinal pitch and Reynolds number for circular
shaped tube
It includes the analysis at longitudinal pitch of 2D, 2.5D and 3D

Analysis at different values of longitudinal pitch and Reynolds number for diamond
shaped tube
It includes the analysis at longitudinal pitch of 2D, 2.5D and 3D

Conclusion
It marks the closure for thesis with summary of the work carried out and suggestions for
adaptation of shape of tube and longitudinal pitch for triangular tube bank arrangement

End

16
CHAPTER 2
LITERATURE REVIEW

2.1 EXPERIMENTAL STUDIES


Smith et al [ 2 ] have investigated experimentally the effect of Reynolds number on the nusselt
number and skin friction coefficient in tandem arrangement with uniform heat flux over the
tubes. The results presented by the author suggest that nusselt number have strong dependency
on cone angles, tail length ratio and Reynolds number. The heat transfer and friction factor will
increase with cone angle and reduction with tail length ratio due to augmentation in turbulence.
The heat transfer enhancement of 67%, 57% and 46% were obtained at the cone angle of 45
degree with tail length ratio of 1, 1.5 and 2.0 respectively.

Pongjet et al [ 3 ] investigated experimentally the heat transfer and friction factor in combined
twisted tape & winglet vortex generator for constant heat flux square duct. The author concluded
that the nusselt number & friction factor for the combined twisted tape and V- winglet increases
with increasing duct to height ratios & decreasing the pitch to tape width ratio. The maximum
value of heat transfer & friction factor is obtained at duct height ratio of 0.2, pitch to tape width
ratio of 2 and twist ratio of 4. The maximum thermal performance is obtained at duct height ratio
of 0.1, pitch to tape width ratio of 2 and twist ratio of 4. The thermal performance of combined
vortex flow deice gives 17% higher than twisted tape alone.

Bayat et al [ 4 ] had experimentally studied the fluid flow and heat transfer from cam-shaped
tube bank in staggered arrangement. Tubes were located in test section of an open loop wind
tunnel with two longitudinal pitch ratios 1.5 and 2. Reynolds number varies in range of 27,000
ReD 42,500 and tubes surface temperature is betweenn78 and 85 0C. Results show that both drag
coefficient and Nusselt number depends on position of tube in tube bank and Reynolds number.
Tubes in the first column have maximum value of drag coefficient, while its Nusselt number is
minimum compared to other tubes in tube bank. Moreover, pressure drop from this tube bank is
about 92–93% lower than circular tube bank and as a result thermal–hydraulic performance of
this tube bank is about 6 times greater than circular tube bank.

17
Zdanski et al [ 5 ] had performed an experimental study addressing the effects of delta winglet
vortex generators on convective heat transfer rate at in-line tube bank in external cross flow. The
main goal of the work is to evaluate a study assessing the influences of the following parameters
on convective heat transfer enhancement: the distance from the vortex generators to the tube
bank, the pitch and the incidence angle of the delta winglet, as well as the free-stream velocity
inside the wind tunnel (Reynolds number). The validation of the experimental methodology
employed was performed by comparing the present results with empiric correlations available in
the literature. The main results indicate that Nusselt number was enhanced when using
turbulence promoters, being the maximum increment around 30%.Otherwise, the pressure drop
through the tube bank was enhanced accordingly, being the maximum increment around 40%.
Finally, it is worth mentioning that all the experimental results for the Nusselt number were
condensed in a new empirical correlation for practical applications with good accuracy
(maximum error around ±6.0%).

2.2 NUMERICAL AND EXPERIMENTAL STUDIES

Hung-Yi Li et al [ 6 ] had studied, both experimentally and numerically investigated the


thermal-fluid characteristics of a flat-fin heat sink with a pair of vortex generators installed in a
cross flow channel. The effects of the distance between the trailing edges of the vortex
generators, the distance between each vortex generator and the heat sink, the attack angle of the
vortex generators, the height of the vortex generators, the configuration of the vortex generators,
and the Reynolds number on the thermal-fluid performance of the plate-fin heat sink are
elucidated. The performance of the heat sink with the vortex generators was compared to that
without. The results thus obtained revealed that when the distance between the trailing edges of
the vortex generators is too small, the thermal performance becomes worse because the air
cannot easily flow into the heat sink. Regarding the effect of the position of the vortex generators
on thermal performance, the best performance is achieve when the distance between the vortex
generator trailing edges equals the length of the heat sink and the distance between the trailing
edge of each vortex generator and the front end of the heat sink is zero. An attack angle of the
vortex generators of 30_ is preferred to optimize the thermal resistance and pressure difference.
Although heat transfer increases with the height of the vortex generators, the pressure difference
also increases. Regarding the effect of the configuration of the vortex generators on the

18
performance of the heat sink, a common-flow-up configuration outperforms a common- flow-
down configuration. The enhancement of the heat transfer of the heat sink by the use of vortex
generators is greater, and the increase of the pressure difference is lower, at lower Reynolds
numbers.

2.3 NUMERICAL STUDIES


Haolin et al [ 7 ] numerically studies the semicircle cross-sectioned tubes heat exchanger that
creates narrow slots in the stream wise direction. The effect of vortices on the wake of the wall &
near the exit of slots is observed. The conclusion of this study is that the slotted tube heat transfer
coefficient is 40% more than traditional non slotted tubes. The pressure drop across the slotted
tube is far greater than non-slotted tubes. The more turbulent flow diminishes the effect of slots.
Even though the cost of semi-circle tubes is much greater than simple tubes but still the slotted
design is feasible.

Zhang guan et al [ 8 ] min investigated numerically & experimentally that the heat flow around
an egg shaped tubes of semicircular geometry at upstream & semi elliptical at downstream. The
egg shaped tube has a higher value of pressure gradient at its front lower one at its back. Due to
this there will be delay in separation of flow around the tube. The performance evaluation
criterion of 2 is obtained at Reynolds number greater than 11952. The wake size decreases as the
performance evaluation criterion increases.
Leandro et al [ 9 ] investigated the optimization of winglet type vortex generator positions &
angles in plate fin compact heat exchanger by direct optimization and response surface
methodology. The numerical study was performed to observe the pressure loss and heat transfer
of a longitudinal vortex generator in fin tube heat exchanger. The best results are obtained when
the two vortex generators are not symmetric to each other.

Wejkowski [ 10 ] presented the heat transfer and pressure loss in convection heat exchanger
with triple finned tubes with initial evaluation with the help of LUSAS code. The investigation
over three fins with the same geometry with 300,450 between two symmetric fins and a vertical
line of symmetry and one fin and angle steel plate placed parallel to the line of symmetry of the
triple-finned tube. The fin with 45◦ between two symmetric fins and a vertical line of symmetry
will produce the better heat transfer as compared to the fins having fins with 30 0 between
symmetric fins and vertical line of symmetry. The ratios of Sherwood numbers is determined by
19
the concept naphthalene sublimation. The thermal performance of basic plain tube bank and
staggered tube bank arrangement is compared then results are compared with the computational
fluid dynamics codes. Later on he had compared the numerical results with the experimental
ones.

Anupam Sinha et al [ 11 ] has analytically studied the fluid flow through fin tube type heat
exchangers with rectangular winglet pairs of half the channel height as vortex generators. There
is increase in heat transfer due to nozzle like flow passages due to the winglet pair. The
performance is also increases due to CFD-CFU orientations of vortex generators, the
performance improves with increase in angle of attack up to a certain point & afterwards it is
going down.

Xue-Hong Wu [ 12 ] et al had numerically investigated the heat transfer and fluid flow
characteristics of a new type of fin with built-in interrupted delta winglets. In order to ensure
reliability of numerical model, plate fin with common flow- up delta winglets is firstly simulated.
The comparison of numerical and experimental results shows a maximum deviation of 11.4%
within the entire range of Reynolds number. The computational results show that heat transfer
capacity and overall performance increase by 35–60% and 19–64%, respectively. The flow field
visualization shows that the interrupted delta winglets can produce longitudinal vortices at the
rear of delta winglets and reduce the wake zone behind the tube, so the proposed fin can enhance
heat transfer accompanied by low pressure loss. The field synergy theory and entransy
dissipation extremum principle are employed on analyzing the mechanism of heat transfer
enhancement. The results indicate that enhancement heat transfer mechanism of interrupted delta
winglets can be explained as the result of the decrease of synergy angle and reduction of the
entransy dissipation.

Chidanand et al [ 13 ] ha numerically investigated various arrangements and along with


different geometries of the tube, so as to improve the thermal performance of the tube bank. In
the present study, the numerical investigation is carried out to determine the thermal performance
of cam shaped tube. The study is performed for the heat transfer and friction factor
characteristics in the Reynolds number range from 11,500 to 42,500. The result of the numerical
simulation indicates the superior thermal performance of the cam shaped tube banks over the

20
circular tubes. As the friction factor is reduced by 85–89% as compared to the circular tubes the
heat transfer by friction factor (Nu/f) is increased by 5 times as compared to circular tubes.
However, the efficiency of cam tubes with a pitch ratio of 1.5 is higher than that of the pitch ratio
of 2.0. Further, the area goodness factor is also around 9 times higher than circular tubes.

Zhi-Min Lin et al [ 14 ] had numerically studied the laminar flow and heat transfer
characteristics in a tube inserting a twisted tape having parallelogram winglet vortex generators.
The usage of wings or/and winglets vortex generators (VGs) could provide excellent heat
transfer performance with a minimal increase in pressure loss comparing with other VGs in a
wide range of Reynolds number. For classical twisted tape, the contact area between twisted tape
and working fluid is large, which would result in high friction drag. Motivated by this fact, a
newly designed twisted tape having parallelogram winglet VGs (PWVGs) is proposed to
enhance heat transfer. This twisted tape with winglet VGs is formed by cutting at the edge of the
straight tape with parallelogram shape to produce a parallelogram winglet as twisting around its
longitudinal axis. The laminar fluid flow and conjugate heat transfer characteristics in a circular
tube inserting the twisted tape with PWVGs were analyzed using numerical method under
uniform wall temperature condition, and four different attack angle a (a = 27.64_, 21.44_,17.44_
and 14.67_) and four different axial spacing St (St = 0.83D, 1.0D, 1.25D and 1.67D). The results
show that (1) the newly designed twisted tape has two ways to generate secondary flow,
including secondary flow generated by the twisted base tape and extra secondary flow generated
by PWVGs; (2) the combination of twisted base tape and PWVGs has evident effects on the
local flow and heat transfer characteristics; (3) compared with the reference empty tube, the
twisted tape with PWVGs can effectively enhance heat transfer under identical pumping power
condition, for studied case of a = 17.44_, St/D = 1.25 and Re ranging from 50 to 600, the average
Nusselt number and friction factor are increased by 76.4–190.9% and 179.9–289.1%,
respectively, and the corresponding thermal performance factor ranges from 1.25 to 1.85; and (4)
as Re ranges from 50 to 600, the helical angle a of PWVGs is a very important parameter in
design of the modified twisted tape, and when a is large and St is small the heat transfer
performance of the twisted tape with PWVGs is better.

Bender et al [ 15 ] numerical study addressing the hydrodynamic and thermal behavior of a new
trapezoidal tube bank arrangement composed of nine circular cylinders subjected to forced

21
convection. The pressure drop and Nusselt numbers are evaluated aiming to assess the influences
of the following parameters: the new trapezoidal factor, the longitudinal and transverse pitches
and the free stream velocity (Reynolds number). The second law analysis was also performed by
computing the entropy production as function of the trapezoidal factor. The validation of the
results obtained with the present methodology was conducted by comparing against empirical
correlations available in the literature. The main results indicate that entropy production in the
tube bank is very sensitive to the trapezoidal factor. Finally, all results for the global Nusselt
number for the new geometry were condensed in an empirical correlation for practical
applications with good accuracy (maximum error around± 1.15%)

As we have gone through the research papers related to tube bank arrangement in cross flow heat
exchanger. We can conclude there will be no paper related to triangular tube bank arrangement .
so I have selected the study of tube bank arrangement in cross flow heat exchanger. The
triangular arrangement provides the compact size tube bank arrangement. So we can perform the
study on triangular tube bank arrangement with circular and diamond shape tubes.

2.4 AIM AND OBJECTIVE


The main aim of this project are-

 To study the effect of Reynolds number on nusselt number & skin friction coefficient for
circular shaped tube with triangular arrangement by CFD analysis.
 To study the effect of Reynolds number on nusselt number & skin friction coefficient for
diamond shaped tube with triangular arrangement by CFD analysis.
 To compare the effect of shape of tube on nusselt number and skin friction coefficient

22
CHAPTER 3

METHODOLOGY
3.1 CIRCULAR SHAPE TRIANGULAR TUBE BANK ARRANGEMENT
In this project the analysis is done on circular shape tubes and diamond shape tubes. If we
consider the cylinder than we have a variation of pressure with increase in length. For the
analysis of cylinder the free stream fluid is brought to rest at forward stagnation point, with rise
in pressure as per Bernoulli equation. From stagnation point the pressure starts decreasing with
the motion of fluid in forward direction. So in this case the boundary layer develops under the
influence of favorable pressure gradient (dp/dx<0) and reaches to the minimum value [ 1 ] .
After the point of minimum pressure towards the rear end further development of hydrodynamic
boundary layer takes place under the influence of adverse pressure gradient (dp/dx>0). The
adverse pressure gradient will promote the turbulence which will enhance the heat transfer due to
better mixing.

3D 20D

3D

Figure 2 circular shape triangular tube bank arrangement

23
3.2 DIAMOND SHAPE TRIANGULAR TUBE BANK ARRANGEMENT
In this project we want to enhance the thermal performance by passive methods so that there will
be no increase in the cost of the project. The effect of diamond shape of tube of triangular tube
bank arrangement on the nusselt number and skin friction coefficient is analyzed. The variation
of skin friction coefficient and nusselt number with Reynolds number under the effect of certain
parameters are determined and presented. The diamond shape tube will produce a better
turbulent flow as compared to circular tube because of large pressure gradient at the tail end of
the tube. After reaching the value of minimum pressure the inverse pressure gradient has
developed and large amount of vortices formation takes place. These vortices will augment the
intensity of turbulence hence resulting into enhance thermal performance.

3D 20D

3D

3D

Figure 3 diamond shape triangular tube bank arrangement

In order to keep the focus on the discussions of the important physical aspects of the
incompressible flow over triangular tube bank arrangement, the governing equations and the
numerical scheme are only briefly discussed. The flow is modelled by the classical
incompressible Reynolds-averaged Navies-Stokes equations. For computation k- € model was
adopted. Furthermore, the numerical validation indicates that the k- € model offers a good
prediction for this class of flow. Besides, for determination of the turbulent thermal conductivity
we have adopted the definition of constant turbulent Prandtl number, i.e., Pr t = 0.9. The

24
numerical simulation was carried out with support of the commercial software ANSYS CFX,
which uses the cell vertex element based finite volume method.

3.3 VALIDATION
The numerical data is compared with the already available experimental data for inline
arrangement of tubes. The experimental data is taken from the already published research paper
for inline arrangement of tubes by A. Bender et al in international journal

3.3.1 MODEL VERIFICATION


Initially after doing the literature survey we have find out that SST model and turbulence K-
epsilon model are best suited for tube bank arrangement in cross flow heat exchanger. But after
running the simulation for both the models the K-epsilon model gives more accurate result as
compared to SST model. We had compared the results for nusselt number for inline arrangement
of tubes with SST model and turbulence K-epsilon model.

Selection of model is a big task for the simulation of triangular tube bank arrangement because
different authors have suggested different model and no paper has been found regarding CFD
analysis of triangular tube bank arrangement. For the selection of model the CFD analysis on
flow over inline tube bank arrangement is performed. Among all the turbulent model SST-K-
epsilon gives the best result which is explained below in various steps. The figure given below
shows the inline tube bank arrangement with circular shape tube. The bottom of the tube is at a
distance of 3D from the first row of tube and the top is at a distance of 3D from the last row of
tubes.

3D 20D

3D

25
3D
Figure 4 Inline tube bank arrangement
The inlet of these inline tubes is at a pitch 3D from the first column of tubes and exit is at
distance of 20D from the last column of tubes. The longitudinal pitch and transverse pitch both
have a constant value of 3D.

Object 35

Figure 5 Validation with k-€ model [ 14 ]

The bottom boundary of the tube is at a distance of 3D from the first row of tubes while the top
boundary is at distance of 3D from last row of tubes. The length of 20D from the last column of
tubes is provided for the flow settlement after the turbulence created due to inline arrangement of
tubes.
Transport equations for the SST K-epsilon model

The turbulence kinetic energy ‘k’ and specific dissipation rate ‘ω’ are obtained from the
following transport equations:

∂ ∂ ∂ ∂k
∂t
( ρk ) +
∂X
( ρku ) =( ) τ
∂X ∂X
+G k - Yk + Sk

26
∂ ∂ ∂ ∂k
∂t
( ρ ω )+
∂Y
( ρkv ) = ( )τ
∂Y ∂Y
+G ω - Yω + Sω

In these equations, Gk represents the generation of turbulent kinetic energy due to mean velocity
gradients. Gω represents the generation of ω. τ k and τ ω represents the effective diffusivity
of k and ω respectively. Yk and Yω represent the dissipation of k and ω due to turbulence
respectively. All of the above terms are calculated as described below. S k and Sω are user defined
source terms.

The effective diffusivities fort the k-ω model are given by

τk = μ + μt/σk

τ = μ + μt/ σω

Where σk and σω are the turbulent prandtl numbers for k and ω respectively.

3.3.2 ASSUMPTIONS
Numerical model was developed under the following assumptions

(1) The physical properties of the working fluid and the tube are constant.

(2) The incompressible flow is limited as laminar flow.

(3) The flow and heat transfer are in steady state.

(4) The viscous dissipation is negligible.

To broaden the application of the expected results, the governing equations describing the fluid
flow and heat transfer in computational domain should be in non-dimensional form. The fig 1
shows triangular tube bank arrangement with the main parameters considered in the main work.
The tube bank consists of seven cylinder arranged in triangular shape subjected to cross flow
conditions. The first case is having a circular shaped tube. The triangular arrangement can be
considered as new type of arrangement. It is compact arrangement taking up less space than the
inline and staggered grid arrangement. The inlet section is placed at a distance of 3D from the
inlet section. The upper and lower wall is also having the distance of 3D from the Centre of top
most and lower most tube respectively. The exit section is placed at a distance of 20D from the
last tube row. All the distance are measured from the tube Centre. In order to obtain the uniform

27
velocity distribution at the inlet and fully developed at the outlet, the computational domain is
extended to 3 times and 20 times the fin’s longitudinal length at the upstream and downstream,
respectively.
The boundary conditions are implemented as fallows;
1. At the entrance velocity is given with 3, 4, 6, 8, 10 m/s.
2. The tube is given wall as no slip condition with the heat generation of 25000 W.
3. Top and bottom walls are adiabatic with pressure outlet conditions.
4. The null value of pressure is applied at all the boundary conditions.
The various steps that are involved in project completion like designing of the tube bank
arrangement, selection of model, CFD analysis of the tube bank arrangement and further CFD
analysis at different parameters are involved. The steps that are followed to complete this project
and writing work of dissertation is explained in a sequence manner and represented by flow
chart.
The project is started with literature review about the heat exchangers. There will be so many
papers about the inline and staggered grid arrangement for enhancement in heat transfer and
reduction in skin friction coefficient. So I have validated my result for circular shape inline
arrangement. After validation I have generated the geometry of circular shape triangular tube
bank arrangement. The mesh is generated over the geometry. After the mesh generation the
analysis over circular shape triangular tube bank arrangement is performed. The similar steps are
performed for the diamond shape triangular tube bank arrangement. After that result analysis
over the circular and diamond shape triangular tube arrangement is performed. The conclusion
for the effect of longitudinal pitch on the nusselt number and skin friction coefficient is
calculated for diamond shape tube and circular shape tube is deduced.

3.4 GOVERNING EQUATIONS


The dimensionless equations for continuity, momentum and energy for this problem may be
expressed as [ 15 ] , [ 16 ]

∂u ∂v
+ =0
∂x ∂ y

X – Momentum equation

28
2 2
u
∂u
∂x
+v
∂ u −1 ∂ p
=
∂ y ρ ∂x
+ν (
∂ u ∂u
+
∂ x2 ∂ y 2 )
Y- Momentum equation
∂2 v ∂ 2 v
u
∂v
∂x
+v
∂ v −1 ∂ p
=
∂y ρ ∂y
+ν (+
∂ x2 ∂ y 2 )
Z – Momentum equation
∂2 ω ∂2 ω
u
∂ω
∂x
+v
∂ ω −1 ∂ p
∂y
=
ρ ∂y
+ν ( +
∂ x2 ∂ y 2 )
Energy equation
∂2 T ∂2 T
u
∂T
∂x
+v
∂T
∂y
=α ( +
∂ x2 ∂ y2 )
where u , v are the component velocities in the x - and y - directions, respectively, ʋ, ρ, α and T
are the kinematic viscosity, the density, the thermal diffusivity and the temperature of the air,
respectively. Above equations are used to describe the turbulent flow problems, so the symbols
u, v, p and T denote the instantaneous values of the corresponding pulsating physical quantities.
The hydraulic diameter of circular shape tube is 0.016m and maximum pitch between the tubes is
0.048. For the simulation taking 6 different tubes and arranged all in series at equal distance and
this whole arrangement is in triangular shape. Figure 2.1 shows the actual arrangement of
circular shape tube in two –dimension facing the incoming water. For the simulation analysis of
tubes, a rectangular domain is made around it and an inlet is given from which the fluid is
incoming. The inlet is having a length of 0.192m and the Centre point of inlet and middle
diamond shape tube are in the same horizontal line. The distance between the inlet and the
diamond shaped tube is 0.048m.

The effect of Reynolds number on triangular tube bank arrangement is numerically simulated
using the CFD theory. The simulation process consists of 4 steps geometry, modelling, mesh

29
generation, CFD solution and CFD analysis of the result. The whole process of simulation is
carried out in 2D.

3.5 METHODOLOGY FOR CIRCULAR SHAPE TUBE


3.5.1 GEOMETRY MODELLING
The setup of circular shaped triangular tube bank arrangement is shown in the figure. Below.

3D 20D

3D

3D

Figure 6 Circular shaped triangular tube bank arrangement

The analysis on triangular tube bank arrangement is never done earlier. So we are doing analysis
for different values of horizontal pitch like 3D, 2D, 2.5D

3.5.2 MESH GENERATION


Structured mesh is generated over the tube bank arrangement and the domain, and first cell
height of the mesh (dy) near to the wall of the tube is 0.0000026 according to the value of Y +
equal to one. The value of Y + indicates the distance up to which boundary layer will be
pronounced and it governs the production of kinetic energy.

Mesh details are as fallow, Y+ equal to one, a total number of nodes equal to 71517 and quality of
the mesh will be more than 80%. After mesh generation the mesh is solved numerically with
CFD solution. Fig.2.3 shows schematic drawing of the present computational domain. The
30
structured mesh of element size 71517 is developed with near wall mesh refinement of Y +=1.0
and the mesh quality is above 90%.

3.5.3 BOUNDARY CONDITION


In the present work, working fluid is air and it is passing through the domain having the
maximum velocity 2m/s,4m/s,6m/s,8m/s and 10m/s at Reynolds number 6477, 9729, 12972,
16215, 18127 respectively. The pressure based solver is used for the simulation with SST-K
epsilon model and flow is first order upwind. The diamond shaped tube surface is considered as
no slip boundary condition and domain boundaries are considered as slip condition. The outlet is
considered as pressure outlet at atmospheric pressure. The diamond shaped tube is also provided
with the heat generation of 25000W. The top and bottom boundary conditions are considered as
pressure outlet at atmospheric temperature and pressure. For the discretization of pressure the
PRESTO scheme is used. The temperature of air at inlet is taken as 300K.

The effect of Reynolds number on triangular tube bank arrangement is numerically simulated
using the CFD theory. The simulation process consists of 4 steps geometry, modelling, mesh
generation, CFD solution and CFD analysis of the result. The whole process of simulation is
carried out in 2D.

3.6 METHODOLOGY FOR DIAMOND SHAPE TUBE


3.6.1 GEOMETRY MODELLING
The setup of diamond shaped triangular tube bank arrangement is shown in the figure below.
The analysis on triangular tube bank arrangement is never done earlier. The inlet of triangular
tube bank arrangement is 3D from the first column of diamond shape tubes. The exit of the
section is at a distance of 20D from the last column of tubes so that proper settlement of flow
takes place. The bottom wall of rectangular domain is at a distance of 3D from the first row of
tubes. The top wall of rectangular domain is at a distance of 3D from the last row of tubes. The
longitudinal pitch of triangular tube bank arrangement is varied from 2D to 3D for the analysis.
The diagonal of square is taken as the hydraulic diameter of the tube for triangular tube bank
arrangement.
20D
3D

31

Figure 7diamond shaped triangular tube bank arrangement


3D

3D

Figure 8diamond shaped tube in 2D

3.6.2 GOVERNING EQUATIONS


The dimensionless equations for continuity, momentum and energy for this problem may be
expressed as [ 16 ]

∂u ∂v
+ =0
∂x ∂ y

X – Momentum equation

32
2 2
u
∂u
∂x
+v
∂ u −1 ∂ p
=
∂ y ρ ∂x
+ν (
∂ u ∂u
+
∂ x2 ∂ y 2 )
Y- Momentum equation
∂2 v ∂ 2 v
u
∂v
∂x
+v
∂ v −1 ∂ p
=
∂y ρ ∂y
+ν (+
∂ x2 ∂ y 2 )
Z – Momentum equation
∂2 ω ∂2 ω
u
∂ω
∂x
+v
∂ ω −1 ∂ p
∂y
=
ρ ∂y
+ν ( +
∂ x2 ∂ y 2 )
Energy equation
∂2 T ∂2 T
u
∂T
∂x
+v
∂T
∂y
=α ( +
∂ x2 ∂ y2 )
where u , v are the component velocities in the x - and y - directions, respectively, ʋ, ρ, α and T
are the kinematic viscosity, the density, the thermal diffusivity and the temperature of the air,
respectively. Above equations are used to describe the turbulent flow problems, so the symbols
u, v, p and T denote the instantaneous values of the corresponding pulsating physical quantities.
The hydraulic diameter of circular shape tube is 0.016m and maximum pitch between the
tubes is 0.048. For the simulation taking 6 different tubes and arranged all in series at equal
distance and this whole arrangement is in triangular shape. Figure 2.1 shows the actual
arrangement of circular shape tube in two –dimension facing the incoming water. For the
simulation analysis of tubes, a rectangular domain is made around it and an inlet is given from
which the fluid is incoming. The inlet is having a length of 0.192m and the Centre point of inlet
and middle diamond shape tube are in the same horizontal line. The distance between the inlet
and the diamond shaped tube is 0.048m.

33
3.6.3 MESH GENERATION
The structured mesh is generated around the diamond shaped tubes with triangular tube bank
arrangement in cross flow heat exchanger. The first cell height of the mesh (die) near to the wall
of the diamond shape tube is 0.0000026 accordingly Y+ equal to one.

Mesh details are as follow, y +equal to one, a total number of nodes equal to 64430 and quality of
the mesh is more than 70%. After the mesh generation system is solved numerically with CFD
solution

3.6.4 BOUNDARY CONDITION


In the present work, working fluid is air and it is passing through the domain having the
maximum velocity 2m/s,4m/s,6m/s,8m/s and 10m/s at Reynolds number 6477, 9729, 12972,
16215, 18127 respectively. The pressure based solver is used for the simulation with SST-K
epsilon model and flow is first order upwind. The diamond shaped tube surface is considered as
no slip boundary condition and domain boundaries are considered as slip condition. The outlet is
considered as pressure outlet at atmospheric pressure. The diamond shaped tube is also provided
with the heat generation of 25000W. The top and bottom boundary conditions are considered as
pressure outlet at atmospheric temperature and pressure. For the discretization of pressure the
PRESTO scheme is used. The temperature of air at inlet is taken as 300K.

34
CHAPTER 4
RESULTS & DISCUSSION
4.1 RESULT & DISCUSSION FOR CIRCULAR SHAPE TUBE
From the above analysis it is clear that the SST k-€ gives the good results among all turbulence
model and it is reliable further analysis of external flow over circular shape triangular tube bank
cross flow arrangement. Therefore by keeping model SST K-€ fixed, simulation is done at a
different type of solver and analysis of their result is explained in this section.

Parameters Input values


Analysis type Pressure based, steady state, Scheme-
SIMPLE
Turbulence model k- €(SST)
Material-Fluid Air (Density Constant)
Inlet Velocity inlet
Circular tube No slip
surface
Domain boundaries-
Top Pressure outlet
Bottom Pressure outlet
Inlet Wall (no shear)
Outlet Wall (no shear)
Simulation is done at pressure based solver and the scheme is SIMPLE and boundary condition
for given geometry of tube is defined in table 1. Solution initialization is standard with the
absolute reference frame.

35
4.1.1 ANALYSIS AT A HORIZONTAL PITCH OF 3D
Reynolds number Nusselt number Skin friction coefficient
6477 60.09 0.022184807
9729 67.70 0.016938647
12972 77.30 0.015632706
16215 87.46 0.014578183
18127 92.65 0.013093695

experimental
95 92.65

90 87.46
85
NUSSELT NUMBER

80 77.3
75
70 67.7

65
60.09
60
55
5800 7800 9800 11800 13800 15800 17800 19800
REYNOLDS NUMBER

Figure 9 Nusselt no. Vs Reynolds no.


As seen from the graph, the nusselt number is increased by increasing the value of Reynolds
number which is observed numerically in fluent data. The variation of nusselt number with
Reynolds number can be approximated as straight line as it can be seen from the graph. As we
know theoretically also that generally the nusselt number will increase with increase in Reynolds
number. The highest nusselt number of 92.65 is obtained with the Reynolds number of 18127 at
the longitudinal pitch of 3D

36
skin friction coeff. Vs Re
240
221.85
220
skin friction coefcient

200

180 169.39
156.23
160
145.78
141.27
140

120
6000 8000 10000 12000 14000 16000 18000 20000
reynolds number

Figure 10 skin Friction coeff. Vs. Reynolds number


The variation of skin friction coefficient with Reynolds number can be seen from the graph. The
value of skin friction coefficient decreases with increase in Reynolds number. At first there was
steep decrement in the value of skin friction coefficient from 0.022184 to 0.016938. Then further
there will be gradual decrement in the value of skin friction coefficient from 0.016938 to
0.014127. As we have studied in fluid mechanics that value of skin friction coefficient varies
inversely with respect to the Reynolds number.

4.1.2 ANALYSIS AT LONGITUDINAL PITCH OF 2.5D

Reynolds number
Nusselt number Skin friction coefficient
6477 69.93 0.028463176
9729 78.80 0.021723105
12972 86.78 0.019009365
16215 97.85 0.018471843
18127 103.49 0.018023364
As seen from the graph, the nusselt number is increased by increasing the value of Reynolds
number which is observed numerically in fluent data. If we compare the current value of nusselt
number with respect to the longitudinal pitch of 3D than there will be slight increment in the
value of nusselt number.

37
Nu Vs Re
110
103.49
105
100 97.85
NUSSELT NUMBER

95
90 86.78
85
78.8
80
75
69.93
70
65
6000 8000 10000 12000 14000 16000 18000 20000
REYNOLDS NUMBER

Figure 11Nusselt number Vs. Reynolds number at pitch of 2.5D


As we know theoretically also that generally the nusselt number will increase with increase in
Reynolds number. The highest nusselt number of 103.49 is obtained at Reynolds number of
18127 with horizontal pitch of 2.5D.

skin Friction coeff. Vs reynolds number

310
284.63
290
SKIN FRICTION COEFFICIENT

270
250
230 217.23
210
190.09 184.72
190 180.23
170
150
5800 7800 9800 11800 13800 15800 17800 19800
REYNOLDS NUMBER

Figure 12skin Friction coeff. Vs. Reynolds number at pitch of 2.5D


The variation of skin friction coefficient with Reynolds number can be seen from the graph. The
value of skin friction coefficient steeply decreases from 0.028463 to 0.021723. After that there

38
will be gradual decrement in the value of skin friction from 0.021723 to 0.018023 as the
Reynolds number will increase. As we have studied that the value of skin friction coefficient will
decrease with increase in Reynolds number.

4.1.3 ANALYSIS AT PITCH OF 2D


There will be approximately linear increment in the value of nusselt number with increase in the
value of Reynolds number. The highest value of nusselt number 102.61 is obtained at Reynolds
number of 18127 for horizontal pitch of 2D. If we compare the values of nusselt number for
respective Reynolds number for longitudinal pitch of 2.5D & 2D than the longitudinal pitch of
2.5D will provide better results as compared to 2D. The maximum value of nusselt number is
obtained at longitudinal pitch of 2.5D due to better mixing of fluid.

Nu Vs Re
105 102.61

100 97.35
95
NUSSELT NUMBER

90 86.98
85
79.81
80
75 70.6
70
65
5800 7800 9800 11800 13800 15800 17800 19800
REYNOLDS NUMBER

Figure 13 Nusselt number Vs. Reynolds number

39
The variation of skin friction coefficient vs. Reynolds number is given for longitudinal pitch of
2D. The value of skin friction coefficient of 0.0279588 is obtained at the Reynolds number
of6477 which is highest for the least value of Reynolds number. At first there will be steep
decrement in the value of skin friction coefficient from 0.0279588 to 0.022515 than there will be
gradual decrement in the value of skin friction coefficient for further values of Reynolds number.

skin Friction coeff. Vs reynolds number

290 279.59
SKIN FRICTION COEFFICIENT

270

250
225.15
230

210 196.84
187.61
190 182.6

170

150
5800 7800 9800 11800 13800 15800 17800 19800
REYNOLDS NUMBER

Figure 14skin friction coefficient vs. Reynolds number at pitch of 2D

4.2 ANALYSIS OF RESULT


The assumption is made that the flow is taking place from left hand side to right hand side and
flow through the circular tube. The inlet velocity is varied with the desired Reynolds number. As
from the contour we can see that there will be increase in velocity between the circular tubes
simultaneously there will be decrease in pressure between the tubes. The increase in velocity will
increase the turbulence in the flow region with the large number of eddies formation.

Comparison of velocity contour of 4m/s at different pitch

At longitudinal pitch of 3D

40
Figure 15Comparison of velocity contour of 4m/s at pitch 3D
At longitudinal pitch of 2.5D

Figure 16Comparison of velocity contour of 4m/s at pitch 2.5D


At longitudinal pitch of 2D

Figure 17Comparison of velocity contour of 4m/s at pitch 2D

41
The velocity of a fluid is varied from 4 m/s, 6 m/s, 8 m/s, 10 m/s and 11 m/s with temperature of
surrounding is considered as 300 K. the assumption is that air flows from left hand side to the
right hand side and pass over the triangular tube bank arrangement. The increase in velocity is
observed in between the tubes because of decrease in area between the tubes simultaneously
there will be decrease in the pressure between the tubes as per the Bernoulli equation.

The main objective of the present study is to determine the effects shape of the tube and their
arrangements on the flow structure and the overall heat transfer escalation for a fin-tube type
heat exchanger. Thus investigations with two different shape of fin-tube heat exchanger, one with
circular shape of tube and the other with diamond shape of tubes along with triangular
arrangement under different values of longitudinal pitch have been conducted. The longitudinal
pitch for circular tubes are varied as 2D, 2.5D and 3D. The present study has been carried out
with the objective of assessing the effects of the positions of the tubes on flow and heat transfer
characteristics in a representative periodic element of a fin-tube heat exchanger. The flow
structures and heat transfer behaviors have been examined at different cross-flow planes with the
help of velocity vectors, streamlines, temperature contours and Nusselt number distribution etc.
The performance of the effect of tube shape has been evaluated based on the comparison of
Nusselt number and friction factor.

In the case of heat transfer, understanding the distribution of the temperature of air close to the
tubes and development of the vortex within the column of tubes is studied. The distribution of
the temperature along the test array, determines the optimum spacing of the tubes within the
array. Additionally, the temperature contour provides an estimate of the thermal boundary layer
formation over the tube surface.

The temperature of the air in the cross flow across the tube bank increases, by gaining the heat
from the tube surface. The maximum temperature of 356 K, which is the boundary condition, is
observed at the wall boundary of the tubes. The intensity of the rise in the temperature of the air
increases as the air past the tube column. With the increase in the Reynolds number for the fluid
flow, temperature intensity surrounding the circular and diamond tubes gradually decreases. The
domain wall on above and below the tube array has no impact on the temperature distribution
over the tube surface.

42
4.3 OPTIMIZATION FOR CIRCULAR SHAPE TUBE
As we can see the horizontal pitch of 2.5D will provide the highest value of nusselt number so
the value of nusselt number increases by increasing the value of horizontal pitch from 2D due to
increase in turbulence which would result in large inter-mixing and hence better heat transfer. As
expected in the staggered case, there is very strong interactions between the vortices. Moreover,
both clockwise and counter clockwise vortices are interacting thus the temperature is more
dispersed. The same can be visualized from the temperature contours of different configurations
in different longitudinal pitch. It may be noted that the average temperature at the exit of the tube
bank arrangement cross flow heat exchanger with transverse pitch of 2.5D is the highest due to
better intermixing.

During the channel flows, the swirling motions generated by the staggered tube bank
arrangement disrupt the thermal boundary layer, intensify mixing and bring about enhancement
of heat transfer with relatively less pressure penalty. The fin surfaces are subjected to flow
impingement; accelerated and decelerated flow due to the presence of turbulence. All these
effects cause differential local distribution of heat transfer on the top and bottom channel walls.

The flow past over the tubes generate the turbulence in the fluid flow, across the tube bank. The
intensity of the turbulence increases with the increase in the Reynolds number for the fluid flow.
Further, the array of the tubes increases the intensity of the turbulence much higher. The
turbulence has direct impact on rate of heat transfer rate of the system.

It is seen that the intensity of the turbulent kinetic energy increase with the increase in the
Reynolds number for the fluid flow. The maximum intensity of the turbulence is observed at the
downstream of the last column of the tube bank. This may be due to the fact that the vortex
shedding phenomenon does not take place in the last column of the tube bank. The higher
intensity vortex increases the turbulence in the fluid flow. The high turbulence is observed after
the separation point, which may be due to the fluid recirculation within the tube bank, thereby
promoting better energy interaction. This may also result in the higher pressure drop within the
tube bank.

43
CIRCULAR SHAPE TUBE

Nu Vs Re
115

105

95
NUSSELT NUMBER

3D
85 2.5D
2D
75

65

55
6000 8000 10000 12000 14000 16000 18000 20000
REYNOLDS NUMBER

Figure 18 comparison of Nusselt number vs. Reynolds number for circular tube

 The increase in the pitch ratio of the tubes may increase the possibility of higher vortex
formation, which may enhance the heat transfer.

Skin friction coefcient vs reynolds number


300
SKIN FRICTION COEFFICIENT

280
260
240
220
200
180
160
140
120
100
6000 8000 10000 12000 14000 16000 18000 20000
REYNOLDS NUMBER

2D 3D 2.5D

Figure 19 comparison of skin friction coeff. Vs. Re circular tube

44
 For the heat exchangers with triangular tube bank arrangement, the performance of the
heat exchangers increases with decrease in longitudinal pitch from 3D to 2.5D although
with a pressure penalty. This is irrespective of the Re applicable for general purpose
industrial heat exchangers (Re varying from 6000 to 19000).
 But there will be decrement in the value of nusselt number with decrease in horizontal
pitch from 2.5D to 2D. Because the horizontal pitch of 2.5D will provide the better
turbulence and inter mixing of fluid as compared to the other values of horizontal pitch.
 All the combinations are prone to pressure loss penalty.

4.4 RESULT AND ANAYSIS DIAMOND SHAPE TUBE


4.4.1 CFD SOLUTION
In the fluent solution process, the solver is pressure based because flow is incompressible. The
energy equation is selected and for the turbulence theory, the SST K-€ model is selected. The
velocity of the tube is according to the respective Reynolds number. The temperature of the
surrounding is considered as 300 K. fluid properties is air having following properties, (1)
density is an ideal gas, (2) specific heat CP 1006.43 J/Kgk and viscosity is considered as
Sutherland.

The flux type is AUSM (Advection Upstream Splitting Method). First order upwind scheme is
used for the momentum equation, energy equation and for specific dissipation rate.

Parameters Input values


Analysis type Pressure based, steady state, Scheme-SIMPLE

Turbulence model k- €(SST)


Material-Fluid Air (Density Constant)
Inlet Velocity inlet
Circular tube surface No slip

45
Domain boundaries-
Top Pressure outlet
Bottom Pressure outlet
Inlet Wall (no shear)
Outlet Wall (no shear)

Considering the flow over the diamond shape triangular tube bank arrangement in cross flow
heat exchanger, using CFD theory, the numerical simulation of the influence of the Reynolds
number on nusselt number and skin friction coefficient is conducted. The numerical simulation
in fluent is converged after 2798 iterations. The CFD numerical simulation is analyzed, including
four parts analysis of velocity field, analysis of pressure field and analysis of turbulence field
over nusselt number and skin friction coefficient.

4.4.2 ANALYSIS AT LONGITUDINAL PITCH OF 3D

Nu Vs Re
115
107.92
105
93.19
95

85
Nu

77.62
75

65 61.54

55
6000 8000 10000 12000 14000 16000 18000
Re

Figure 20Nusselt number Vs. Reynolds number for diamond shape tube at pitch of 3D

As seen from the graph, the nusselt number is increased by increasing the value of Reynolds
number which is observed numerically in fluent data. The variation of nusselt number with

46
Reynolds number can be considered as straight line as it can be seen from the graph. As we
know theoretically also that generally the nusselt number will increase with increase in
Reynolds number. The highest value of nusselt number 107.92 is obtained at 18127 value of
Reynolds number

FRICTION COEFF.
190
180
170
160
150
Nu

140
130
120
110
100
5800 7800 9800 11800 13800 15800 17800
Re

Figure 21 skin Friction coeff. Vs. Reynolds number of diamond tube at pitch 3D

The variation of skin friction coefcient with Reynolds number can be seen from the graph. The
value of skin friction coefcient decreases with increase in Reynolds number. At first there will
be steep decrease in the skin friction coefcient than there will be gradual decrease in the skin
friction coefcient with the increase in Reynolds number.

4.4.3 ANALYSIS AT LONGITUDINAL PITCH OF 2.5D


As compared to the longitudinal pitch of 3D, the longitudinal pitch of 2.5D will produce less
turbulence resulting into the less value of nusselt number. The large turbulence in case of
longitudinal pitch of 3D will augment the intermixing of fluid. The highest value of nusselt
number 97.96 is obtained at the Reynolds number of 18127.

Variation of nusselt number with Reynolds number

47
Nu Vs Re
105
100 97.96

95
90 87.16
85
80
75.39
75
70
65 62.24
60
55
6000 8000 10000 12000 14000 16000 18000

Figure 22Nusselt number Vs. Reynolds number for diamond shape tube at pitch of 2.5D

As seen from the graph, the nusselt number is increased by increasing the value of Reynolds
number which is observed numerically in fluent data.

Variation of skin friction coefcient with Reynolds number

skin Friction coeff. Vs reynolds number


200
176.33
180
skin friction coeff.

160
138.41
140
116.16
120
102.69
100
80
5800 7800 9800 11800 13800 15800 17800
Re

Figure 23 skin Friction coeff. Vs. Reynolds number for diamond shape tube at pitch 2.5D

The variation of skin friction coefcient with Reynolds number can be seen from the graph. The
value of skin friction coefcient decreases with increase in Reynolds number. At first there will
be steep decrease in the skin friction coefcient than there will be gradual decrease in the skin
friction coefcient with the increase in Reynolds number.

48
4.4.4 ANALYSIS AT TRANSVERSE PITCH OF 2D

Nu Vs Re
115
103.81
105

95 92.21

85 79.46
Nu

75
63.94
65

55
6000 8000 10000 12000 14000 16000 18000
Re

Figure 24Nusselt number Vs. Reynolds number for diamond shape tube at pitch of 2D

As seen from the graph, the nusselt number is increased by increasing the value of Reynolds
number which is observed numerically in fluent data.

skin Friction coeff. Vs reynolds number


200 186.36

180
skin friction coeff.(10-4)

160 147.46
140
125.26
120 111.26

100

80
6000 8000 10000 12000 14000 16000 18000
Re

Figure 25 skin Friction coeff. Vs. Reynolds number for diamond shape tube at pitch 2D

The variation of nusselt number with Reynolds number can be considered as straight line as it
can be seen from the graph. As we know theoretically also that generally the nusselt number
will increase with increase in Reynolds number. As in this case the highest value of nusselt

49
number is somewhat similar to the value of nusselt number obtained in case of longitudinal
pitch of 3D

The variation of skin friction coefcient with Reynolds number can be seen from the graph. The
value of skin friction coefcient decreases with increase in Reynolds number. At first there will
be steep decrease in the skin friction coefcient than there will be gradual decrease in the skin

friction coefcient with the increase in Reynolds number.

4.5 STREAMLINE AND VELOCITY CONTOURS


The streamlines help to track the fluid particle path over the tube surface within the tube bank.
The streamline provides the details of the separation point in case of flow across a cylinder along
with the vortex formation within the tube columns. The streamline also indicates the intensity
and location of the vortex formation at the tube downstream. The placement of the next
successive column of the cylinder shifts the vortex close to the cylinder downstream. In the case
of the pitch ratio of 2, the smaller tube spacing leads to the formation of the weak vortex. The
rate of increase in the Nusselt number is much less with the increase in the Reynolds number, as
the flow is tangential to the wall surface at higher Reynolds number. This also reduces the
friction along with the heat transfer rate from the tube surface.

The velocity of the fluid within the column of the tube decreases, due to a higher intensity of the
vortex formation and intense fluid circulation as the pitch ratio increases. The reduced velocity of
the fluid increases the friction factor within the tube bank for the flow at both the pitch ratio of
1.5, and 2.0. However, there exists an uncertainty within velocity distribution across the tandem
cylinder, due to the formation of the vortex at the cylinder downstream.
It is seen that there is a sudden decrease in the intensity of the velocity of the fluid at the cylinder
downstream. The Reynolds number for the fluids critically governs
the decrease in the velocity of the flow over the test tubes. The intensity of the decrease in the
velocity of the fluids gradually increases along the tube column which is also the measure of the
intensity of the vortex formation can be seen from turbulent kinetic energy and streamline
contours. However, there is a reduction in the decrease in the velocity of the fluid at the tube
downstream as the Reynolds number increases. The reduction in the fluid velocity gives rise to
the better heat transfer enhancement from the subsequent column of the tubes. It is observed that

50
the last tube contributes to the maximum reduction in the velocity leading to the highest
enhancement in the rate of heat transfer.
The heat transfer is largely associated with the development of hydrodynamic boundary layer
over the surface of the tube. The boundary layer is controlled by the geometrical profile over
which flow development takes place. As in this case, the profile is circular and diamond shaped
and hence the development of flow over the tubes are more streamlines relatively to that of the
circular tubes.

Comparison of velocity contour at 4 m/s at different longitudinal pitches


At longitudinal pitch of 3D

Figure 26 velocity contour at 4 m/s for longitudinal pitch of 3D for diamond tube

At longitudinal pitch of 2.5D

Figure 27 velocity contour at 4 m/s for longitudinal pitch of 2.5D for diamond tube

At longitudinal pitch of 2D

51
Figure 28 velocity contour at 4 m/s for longitudinal pitch of 2D for diamond tube

CONTOURS OF TURBULENCE
At 2d

Figure 29 Contours of turbulence kinetic energy at 4m/s for diamond tube at pitch 2D
At 2.5D

Figure 30 Contours of turbulence kinetic energy at 4m/s for diamond tube at pitch 2.5D
At 3D

52
Figure 31 Contours of turbulence kinetic energy at 4m/s for diamond tube at pitch 3D

4.6 OPTIMIZATION FOR DIAMOND SHAPE TUBE

Nu Vs Re
115
107.92
103.81
105
97.96
93.19
92.21
95
Nusselt number

87.16
85
79.46
77.62
75.39
75

63.94
62.24
65 61.54

55
6000 8000 10000 12000 14000 16000 18000

Reynolds number

3D 2.5D 2D

Figure 32 comparison of Nusselt number vs. Reynolds number for diamond tube

53
In case of diamond shape tube initially the value of nusselt is maximum for longitudinal pitch of
2D up to the Reynolds number of 12000 but after that the longitudinal pitch of 3D will provide
maximum values of nusselt number.

skin friction coeff. Vs Re


200
186.36
182.56
176.33
180
SKIN FRIICTION COEFFICIENT

160
147.46
146.72
138.41
140
129.02
125.26
116.16 117.91
120 111.26
102.69
100

80
5800 7800 9800 11800 13800 15800 17800

REYNOLDS NUMBER

2D 3D 2.5D

Figure 33 comparison of skin friction coeff. Vs. Re diamond tube

 For the heat exchangers, the performance depends on the configuration as well as the Re.
Therefore based on the Re for a particular heat exchanger application, the optimum
configuration needs to be chosen.

The fluid flow is always associated with the friction coefficient over which the flow is
established. It is clear from the figure that the friction factor is maximum in case of the circular
tubes. The highest friction factor is reported for the circular tubes with a pitch ratio of 2. The
friction factor for the fluid decreases, with an increase in the Reynolds number for the flow. This
is usually due to the dominant pressure force, which reduces the friction.

54
The minimum value of skin friction coefficient will be the 186.36*10-4 at longitudinal pitch ratio
of 3 at Reynolds number of 6477 for diamond shape triangular tube bank arrangement. The value
of nusselt number will be maximum for longitudinal pitch ratio of 3 for diamond shaped tube.

 There will be around 35% decrease in skin friction coefficient from circular tube to
diamond shape tube triangular tube bank arrangement.

55
CHAPTER 5
CONCLUSION

1. There will be around 35% decrease in skin friction coefficient from circular tube to
diamond shape tube triangular tube bank arrangement.
2. The value of nusselt number is maximum for longitudinal pitch of 2.5D for all the values
of respective Reynolds numbers for circular shape triangular tube bank arrangement
3. The value of skin friction coefficient is maximum for longitudinal pitch of 2.5D for all
the values of respective Reynolds numbers for circular shape triangular tube bank
arrangement
4. In case of diamond shape tube initially the value of nusselt is maximum for longitudinal
pitch of 2D up to the Reynolds number of 12000 but after that the longitudinal pitch of
3D will provide maximum values of nusselt number.
5. In case of diamond shape tubes the minimum value of skin friction coefficient is obtained
with the longitudinal pitch of 2.5D

56
REFERENCES

1. Fundamentals of heat and mass transfer by Frank P. Incropera


2. Smith Eiamsa-ard, pongjet provonge , thermal characterization of turbulent tube flows
over diamond shaped elements in tandem, international journal of thermal sciences
49(2010) 1051-1062
3. Pongjet Promvonge , Supattarachai Suwannapan , Monsak Pimsarn, Chinaruk Thianpong
Experimental study on heat transfer in square duct with combined twisted-tape and
winglet vortex generators, International Communications in Heat and Mass Transfer 59
(2014) 158–165
4. Hamidreza Bayat , Arash Mirabdolah Lavasani , Taher Maarefdoost Experimental study
of thermal–hydraulic performance of cam-shaped tube bundle with staggered
arrangement, Energy Conversion and Management 85 (2014) 470–476
5. P.S.B. Zdanski , D. Pauli, F.A.L. Dauner Effects of delta winglet vortex generators on
flow of air over in-line tube bank: A new empirical correlation for heat transfer
prediction, International Communications in Heat and Mass Transfer 67 (2015) 89–96
6. Hung-Yi Li , Ci-Lei Chen, Shung-Ming Chao, Gu-Fan Liang Enhancing heat transfer in a
plate-fin heat sink using delta winglet vortex generators, International Journal of Heat
and Mass Transfer 67 (2013) 666–677
7. Haolin Ma, Dennis E. Oztekin, Seyfettin Bayraktar, Sedat Yayla, Alparslan Oztekin,
Computational Fluid Dynamics and Heat Transfer Analysis for a Novel Heat Exchanger,
http://heattransfer.asmedigitalcollection.asme.org/ on 05/28/2015
8. ZHANG Guan-min, LENG Xue-li, ZHOU Nai-xiang, SHI Yan-ping,LI Li-min, Flow and
heat transfer characteristics around egg-shaped tube, Journal of hydrodynamics,
2015,27(1):76-84 DOI: 10.1016/S1001-6058(15)60458-9
9. Leandro O. Salviano, Daniel J. Dezan, Jurandir I. Yanagihara, Optimization of winglet-
type vortex generator positions and angles in plate-fin compact heat exchanger: Response
Surface Methodology and Direct Optimization, International Journal of Heat and Mass
Transfer 82 (2015) 373–387

57
10. Robert Wejkowski, Heat Transfer and Pressure Loss In Combined Tube Banks With
Triple-Finned Tubes, http://www.tandfonline.com/loi/uhte20
11. Anupam Sinha, Himadri Chattopadhyay, Ashwin Kannan Iyengar , Gautam Biswas,
Enhancement of heat transfer in a fin-tube heat exchanger using rectangular winglet type
vortex generators, International Journal of Heat and Mass Transfer 101 (2016) 667–681
12. Xue-Hong Wu, Pei Yuan, Zhi-Ming Luo, Li-Xun Wang & Yan-Li Lu Heat Transfer and
Thermal Resistance Characteristics of Fin with Built-In Interrupted Delta Winglet Type,
ISSN: 0145-7632 (Print) 1521-0537 (Online) Journal homepage:
http://www.tandfonline.com/loi/uhte20
13. Chidanand K. Mangrulkar, Ashwinkumar S. Dhoble, Amey R. Deshmukh, Sachin A.
Mandavgane Numerical investigation of heat transfer and friction factor characteristics
from in-line cam shaped tube bank in cross flow, Applied Thermal Engineering 110
(2017) 521–538
14. Zhi-Min Lin , Liang-Bi Wang , Mei Lin , Wei Dang , Yong-Heng Zhang, Numerical study
of the laminar flow and heat transfer characteristics in a tube inserting a twisted tape
having parallelogram winglet vortex generators, Applied Thermal Engineering 115
(2017) 644–658
15. A. Bender, A.M. Meier, M. Vaz Jr., P.S.B. Zdanski⁎A numerical study of forced
convection in a new trapezoidal tube bank Arrangement International Communications in
Heat and Mass Transfer 91 (2018) 117–124
16. Heat and mass transfer by P.K. Nag
17. Heat transfer by J. P. Holman

58

You might also like