M. Bošnjaković i dr.

Razvoj novog tipa lamelnog izmjenjivača topline

ISSN 1330-3651 (Print), ISSN 1848-6339 (Online)

https://doi.org/10.17559/TV-20171011071711

DEVELOPMENT OF A NEW TYPE OF FINNED HEAT EXCHANGER

Mladen Bošnjaković, Ante Čikić, Simon Muhič, Marinko Stojkov

Original scientific paper

A new geometric form of fins (needle fins) for finned heat exchangers of gas-liquid type is described. It achieves a higher heat transfer coefficient and
lower mass of the heat exchange surface relative to the heat exchangers with circular fins which in this case served as a basis for comparison. For the
analysis of the heat transfer, the computational fluid dynamics analysis with ANSYS Fluent software was applied. Numerical simulation was performed
for the round fins and fins with a new, innovative geometric shape at air flow speeds of 1 m/s to 5 m/s. The results of numerical simulations, with regard
to the exchanged heat and pressure drop were verified by the correlations available in the literature and compared to the simulation results for tubes with
innovative fins. Results show an increase in the amount of Nu from 20 % to 30 % for 2300 >Re > 12000 while reducing mass of heat exchange surfaces of
23,8 %.

Keywords: finned heat exchanger; heat exchange; heat exchanger efficiency; needle fins; numerical simulation

Razvoj novog tipa lamelnog izmjenjivača topline

Izvorni znanstveni članak

Opisan je novi geometrijski oblik lamele (igličaste lamele) za orebrene cijevne izmjenjivače topline plin – tekućina pomoću kojeg se postiže veći
koeficijent prijelaza topline, a time i manja masa izmjenjivačkih površina u odnosu na izmjenjivače topline s okruglim lamelama koji su u ovom slučaju
poslužili kao baza za usporedbu. Za analizu izmjene topline i pada tlaka primijenjena je računalna dinamika fluida s programom ANSYS Fluent.
Numerička simulacija provedena je za slučaj zagrijavanja zraka koji struji oko cijevi s lamelama novog geometrijskog oblika te oko cijevi s lamelama
konvencionalnog kružnog oblika brzinama od 1 m/s do 5 m/s. Rezultati numeričke simulacije obzirom na izmijenjenu toplinu i pad tlaka za cijevi s
okruglim lamelama provjereni su pomoću korelacija dostupnih u literaturi te su uspoređeni s rezultatima simulacije za cijevi s inovativnim lamelama.
Rezultati pokazuju povećanje iznosa Nu značajke od 20 % do 30 % za 2300 > Re > 12000 uz smanjenje mase izmjenjivačkih površina od 23,8 %.

Ključne riječi: efikasnost izmjenjivača topline; igličaste lamele; lamelni izmjenjivač topline; numerička simulacija; prijelaz topline

1 Introduction In practical application, weight and size of the heat

exchanger are often very important and contribute to
Heat exchangers are used in different technical areas proper and economical functioning of unit or the entire
and very often in energy and thermal engineering, system (e.g., the space for car access is limited).
petroleum, chemical, food, automotive industry and in However, the mass of the heat exchanger is also very
various processes where heat exchange is required. important (less mass means less total car mass and lower
Caused by intensity of heat change and the adaptable fuel consumption).
construction and fabrication, finned tube exchangers with
circular fins are often used. In finned tube exchangers, 1.1 Literature review
heating fluid is usually flowing through the tubes and the
gas flow across finned tubes in staggered or in-line The analytical approach to solving heat transfer for
configuration. Heat transfer resistance on the gas side simple geometry of finned surfaces dates from the last
makes 80 % of the total heat transfer resistance, so century when basic functions were defined for
determining the improvement of the heat transfer on the temperature profiles and fin efficiency for different fin
gas side for different cases is a significant contribution to shapes. Hashizume et al. [1] analysed the thermal
increasing the effectiveness of the finned heat exchangers. efficiency of rectangular cross sectional circular fins and
Increased heat exchange can be achieved by performed an analytical solution in the form of a modified
increasing the heat transfer surface, increasing the heat Bessel function.
transfer coefficient, or increasing the thermal efficiency of Kays and London [2] provided experimental
the finned surface. The increase in the exchange surface correlations for a wide range of configurations adopted in
directly increases the cost of the material, which may compact extended surface heat exchangers and presented
amount to up to 80 % of the total price of the exchanger, test data on finned circular tubes.
so researches are focused on the other two parameters that Ward and Young [3] developed heat transfer and pressure
are interconnected. drop correlations for plain finned tubes with triangular
Increase in the heat transfer coefficient can be pitch. They compared their pressure drop data with a
achieved by applying different fins shapes that increase correlation from literature.
the turbulence of the flow, so the thermal boundary layer Kearney and Jacobi [4] investigated experimentally
is periodically interrupted which results in more efficient local heat transfer behaviour in staggered and in-line
heat exchange. However, the innovative design of fins arrangements with the help of optical adaptation of the
with the increase in the heat transfer coefficient causes naphthalene sublimation technique in order to evaluate
increase of pressure drop in the gas flow across tube the analytical fin efficiency.
bundle or increases the fan power to transport the gas Experimental investigations of round (solid) and serrated
through the heat exchanger. finned-tubes have been studied by [5÷9]. Taborek [10]
and Frasz [6] compared the varying influences of round

Tehnički vjesnik 24, 6(2017), 1785-1796 1785

Development of a new type of finned heat exchanger M. Bošnjaković et al.

and segmented fins. Weierman [11] investigated the dynamics can be used as a flow visualization tool. Due to
performance of in-line and staggered tube arrangement of certain deviations in the results between certain models it
segmented finned tubes. is necessary to compare the results with experimental
For crossflow over finned tube banks, a fairly large research.
set of experimental data can be found in the Russian
literature (Zhukauskas, [12]) and the recommended 2 Research object, material and method
correlations for tubes with circular and helical fins.
Stasiulevicius et al. [13] developed correlations of the The geometric model of the tubular heat exchanger
convective heat transfer coefficient and resistance of on which its thermal efficiency and thermal effectiveness
finned-tube bundles in a cross flow including the effects will be estimated in relation to the reference model,
of geometric parameters of fins and tube arrangement consists of an insulated casing inside which is a bundle of
within the bundle. Mon and Gross [14] investigated the tubes with needle shaped fins. Water as a heating medium
effects of the fin spacing on four-row round-finned tube is introduced into the inlet chamber, passes through the
bundles in staggered and in-line arrangements by three tubes and goes to the outlet chamber. Across the finned
dimensional numerical study. tubes flows colder air. The new and innovative shape of
Ali Shakir Al-Jaberi et al. [15] performed the needle-shaped fins enables achievement of higher
experiments on circular fins with slanted blades attached turbulence levels of the fluid flowing around the fins, and
on the copper tube surface in order to reduce the thermal thus higher heat exchange.
boundary layer. They examined effects of number of
slanted blades and Reynolds number on the heat transfer 2.1 Physical model of heat exchanger with needle-shaped
characteristics. Results show 20 % to 27,5 % higher fins
Nusselt number for slanted fins than for circular fins, but
they did not examine pressure drop in tube bundle. For research implementation, the tube ∅20×1,5 is
A. Lemouedda et al. [16] presented the results of selected. As a reference geometry, a round fin ∅40/∅20
numerical calculations of spiral finned tubes and with thickness 0,5 mm was selected (Fig. 1a). The new
compared them with the results for bare tubes. Naess [17] needle type fin is shown in Fig. 1b and has eight vertices
experimentally tested heat transfer and pressure drop for spaced uniformly along the circumference.
L-fins and in this case gave correlations for heat transfer
and pressure drop.
Martinez at al. [18] conducted a comparative study of
heat transfer and pressure drop for spiral serrated fins.
Cléirigh and Smith [19] studied three spiral shapes of
fins: full fin; partially segmented and fully segmented.
B. Anoop, C. Balaji, and K. Velusamy [20]
investigated the heat transfer on one tube with serrated
fins combining experimental approach and numerical
calculation. The depth of the segment, the pitch of the Figure 1 Geometry of round fin and needle type fin
fins, fin height and thickness were investigated.
Rene Hofmann et al. [10] experimentally investigated Different materials are used for making heat
heat exchanger with finned tubes in order to optimize it. exchangers. Steel, copper and aluminium are the most
U and L shaped fins were full and serrated. Fin height, common. For the implementation of comprehensive
thickness, fin spacing and the width of the fin segment research, the selected material of the tubes and fins is
were varied. They found an average increase in Nusselt's stainless steel. Stainless steel is selected because of a
number by 22 % when using serrated fins. wider range of application, higher temperature resistance,
The complex turbulent flow pattern across the fin and resistance to corrosion and resistance to fouling, and
tube surface made the theoretical prediction of pressure facility to easy clean the fin surface even though it has
drop very difficult and therefore most of the published smaller heat conductivity coefficient than aluminium and
correlations were related to the experiments. All of them copper.
use the fluid velocity at the minimum flow area of tube Initial diameter, thickness and pitch of the fins were
bundle. One group of correlations is based on tube outside selected according to the recommendations of the authors
diameter (d0) or fin collar diameter as the characteristic dealing with the problem of finned tubular heat
length for calculations of Re and Nu number. exchangers [4, 12, 13, 14, 22, 28, 29, 30, 31].
Widely cited correlations for pressure drop are
authorized by Briggs and Young [22], ESDU [35], 2.2 Numerical analysis
Schmidt [23], Ward and Young [24] and Nir [25].
Numerous methods used in computational fluid The numerical analysis is focused on air flow across
dynamics have been developed today. A review of all the tubes because the air side resistance of the heat exchange
methods applied today in computer fluid dynamics was makes up 80 % of the total resistance to the heat
given by Md Lokman Hosain and Rebei Bel Fdhil [26]. exchange.
Hossain Nemati and Mohammad Moghimi [27] analysed Certain assumptions and simplifications were adopted
the influence of choice of different models of turbulence for the mathematical description of the thermal
at the flow simulation in a four-row tubular tube bundle. phenomena within the heat exchanger.
Detailed analysis shows that computational fluid

1786 Technical Gazette 24, 6(2017), 1785-1796

M. Bošnjaković i dr. Razvoj novog tipa lamelnog izmjenjivača topline

The most common assumption that is applied when conductivity in the layer direction and in the direction
defining fluid flow is fluid continuity. The characteristic normal to the layer.
dimensions of the analysed innovative exchanger are Air is considered as homogeneous mixture of gases
large enough and the conditions of the continuum and is treated as a one-component fluid (dry air). There is
hypothesis are fulfilled. no fluid leakage through the walls of the finned exchanger
The assumptions of homogeneity and isotropy were and the heat exchange with the environment is negligible.
also introduced. Homogeneity implies that the physical The fins are completely and tightly attached to the tube
properties of the substance are the same at all points. and there is no contact resistance between the tube and the
Thus, the air is considered a homogeneous mixture of fin surface. Steady state analysis is performed.
gases because the proportion of individual gases that The numerical domain was chosen to perform the
compose the air is equal at all points of the fluid. Isotropy numerical analysis of the tube heat exchanger. For a heat
implies that the physical properties of the observed exchanger with a uniform fluid flow field at the inlet, a
substance are the same in all directions. It is assumed that typical repeating section in the heat exchanger is selected
the thermal conductivity of the fin is equal in all for the domain (Fig. 2), and the solutions obtained are
directions, though, for example, the fin composed of assumed valid for the entire heat exchanger.
different materials layers does not have the same thermal

Figure 2 Mathematical model of computational domains

• The gauge pressure pout,c at the outlet of channel is set

to zero. This corresponds to atmospheric conditions
p a.
• Hydraulically smooth walls were defined for outer
tube and the fin walls.
• Symmetry condition is set for top, bottom, left and
right side of computational domain. Symmetry
condition is applied due to simplifying the calculation
where it is possible to reduce computational domain.
The basic requirement is that the geometry is
symmetric and also that the expected solution of the
Figure 3 3D model of computational domain temperature field and the fluid flow fields are
symmetric. At the symmetry planes a zero heat flux is
Boundary and initial conditions are determined when assumed. The normal velocity component at the
defining the computational model and the physical symmetry plane equals zero, i.e. no convective flux
properties of fluids and materials of the exchange across that symmetry plane exists. Thus, the
surfaces. temperature gradients and tangential components of
the velocity gradients in normal direction are set to
Boundary and initial conditions include defining: zero.
• The air entering the computational domain is assumed • Physical properties of air are defined as polynomial
to have uniform velocity over the cross section and function of temperature (and pressure) and thus set in
ranges from 1 m/s to 5 m/s, and turbulence intensity 5 Ansys Fluent software. Density of the air is based on
%. incompressible ideal gas law.
• The air temperature at the inlet to the heat exchanger • The physical properties of the fin and tube material are
is 288 K. set constant (λt=16,2 W/(m·K), ρt=7860 kg/m3).
• Hot water at the inlet to tubes has temperature 353 K.
Because water has a high thermal capacity, it is 2.3 Properties of materials
assumed that the temperature of the tube inner wall is
constant and equal to the water temperature. The definition of physical properties of used
• At the sides of the computational domain it is set materials includes the determination of thermal
symmetry boundary condition. conductivity, dynamic viscosity, density, specific heat
capacity of fluid flowing across the tube bundle (dry air)

Tehnički vjesnik 24, 6(2017), 1785-1796 1787

Development of a new type of finned heat exchanger M. Bošnjaković et al.

and the material of the fins and tubes. The material of the α 0 ⋅ d0
tubes and fins is stainless steel, the surfaces are clean Nu = (5)
lbl
without any additional heat resistance.
The calculation of non-dimensional numbers Re, Nu,
Pr and Eu is based on the physical properties of the where α0 is actual average gas-side heat transfer
working fluid (air) that are generally dependent on coefficient.
temperature and pressure. For a precise calculation of the For pressure drop correlations, the Euler number (Eu)
above non-dimensional features it is necessary to clearly is defined as
define the temperature at which the physical properties of
∆p
the fluid are taken. The fluid temperature in the thermal Eu = (6)
boundary layer varies from Tw at the surface to free- N l ⋅ ρ av ⋅ uff2
stream temperatures at the outer edge of the boundary.
The fluid properties also vary with temperature, and thus 2.4 Governing equations
with position across the boundary layer. In order to
account for the variation of the properties with The mathematical model used to describe a physical
temperature, the fluid properties are usually evaluated at problem is a set of differential equations and constitutive
the so-called film temperature, defined as relations and initial and boundary conditions. The basic
equations of fluid dynamics are derived from:
Tbl = (Tw + Tav ) / 2 (1) Conservation of Mass:

which is the arithmetic average of the surface and the ∂ρ ∂ ( ρ u j )

average free-stream temperatures. Average free-stream + =0 (7)
∂t ∂x j
temperature is defined as

Tav = (Tin + Tout ) / 2 (2) Conservation of Momentum:

∂( ρ u j ) ∂ ( ρ u j ui ) ∂σ ji
The fluid properties are then assumed to remain + = ρ fi + (8)
constant at those values during the entire flow. ∂t ∂x j ∂x j
An alternative way of accounting for the variation of
properties with temperature is to evaluate all properties at Conservation of Energy:
the average free stream temperature Tav and to multiply
the Nusselt number relation by (Prav/Prw)r or (μav/μw)r,
∂ ( ρ e) ∂ ( ρ e u j ) ∂ (σ ji ui ) ∂q j
where r is appropriate coefficient [32]. + = − ρ f i ui + − (9)
In this paper, the temperature of the tube outer wall is ∂t ∂x j ∂x j ∂x j
assumed approximately equal to the inner wall
temperature. The established mathematical model is solved by
Reynolds number (Eq. 3) is a measure of the relative using the finite volume method that is based on dividing
importance between the momentum flux by advection and the computational domain into small volumes and
by diffusion in the same direction. It is important that integration of equations of conservation at these volumes
velocity (uff) and length (L) are taken in the same thus obtaining a system of discrete algebraic equations
direction. For calculating the Reynolds number, the that are then solved iterative. Computational domain is
physical properties (density, dynamic viscosity) are taken meshed by ANSYS software.
at mean air temperature Tav. Characteristic dimension (L)
is based on the outer diameter of the tube (d0), and the
mass flux, i.e. velocity through the narrowest free flow
area within the tube bundle. This is consistent with the
relevant literature

ρ av ⋅ u ff ⋅ L m ⋅ d0
Re = = (3)
m av m av

For the geometries considered in the present study, the

narrowest free flow area is the area between two tubes in
the transversal direction, and can be defined as

Aff = sf ⋅ ( sl − d 0 ) − 2 ⋅ hf ⋅ t f (4)

For heat transfer correlations, the Nusselt number (Nu) is Figure 4 Part of mesh with round fins
defined as
Meshing was performed by using a hybrid mesh
where most of the volume is structured mesh, and the

1788 Technical Gazette 24, 6(2017), 1785-1796

M. Bošnjaković i dr. Razvoj novog tipa lamelnog izmjenjivača topline

smaller part around the fins is unstructured mesh. The Table 2 Characteristic dimensions of tube and fins
area along the fins and tubes where the convection heat Tube and fin material - - Stainless steel
transfer is performed is covered with eight boundary Tube outside diameter d0 mm 20
layers and the thickness of the first layer is 0,025 mm Tube inside diameter di mm 17
Tube rows configuration - - staggered
(Fig. 4). By successive refining of the mesh, we found the
Transverse tube pitch st mm 50
solution, independent of the computational mesh. In the
Longitudinal tube pitch sl mm 40
final analysis we used the mesh with 10,9 million of the
Fin height hf mm 10
finite volumes (elements). Fin thickness tf mm 0,5
Numerical calculation was performed with ANSYS Fin pitch sf mm 4,5
Fluent software. A steady state model of heat exchange is Number of rows Nl - 5
assumed. Two models of turbulence were used: k-ε
Realizable with Enhanced Wall Treatment (EWT) method Table 3 Boundary conditions
and k-ω SST. The calculation results of both models were Inlet air temperature Tin,c K 288
used for the computation of Nu and Eu, which were Air velocity at the inlet of
uin,c m/s 1; 2,4; 5
compared with the results of the literature with the aim of the heat exchanger
selecting a turbulence model whose results coincide more Inlet turbulence intensity - % 5
with the results of the literature. Internal tube wall
Tw K 353
temperature
pout,c
2.5 Solution methods and convergence of the solution Outlet pressure Pa 101325
Hydraulically
Wall condition
smooth wall
An important aspect when performing numerical
simulations is to judge whether the simulation is complete Table 4 Output data from Ansys Fluent for k-ε Realizable with EWT
and converged, or not. There are several different ways to turbulence model
check this and ensure convergence. The most used uin,c= uin,c= uin,c =
Item
method is to look at the residuals for each variable being 1,0 m/s 2,4 m/s 5,0 m/s
solved. In ANSYS Fluent software the default Outlet air temp. from
Tout K 327,3 315,4 307,1
convergence criterion is defined using globally scaled tube bundle
Pressure at the tube
residuals. The default convergence criterion in ANSYS pin Pa 9,30 36,98 115,1
Fluent states that needs to be reduced to 10−3 for all bundle inlet
Pressure at the tube
solved equations, except the energy equation for which pout Pa −1,35 −8,11 −31,13
bundle outlet
the criterion is set to 10−6. Scaled residuals may not give
the full picture. A good indicator to judge convergence is Table 5 Output data from Ansys Fluent for k-ω SST turbulence model
to look at how the residuals develop with the iterations. A uin,c= uin,c= uin,c =
Item
good indication that convergence is reached is if the 1,0 m/s 2,4 m/s 5,0 m/s
residuals continue to decrease or remain low for several Air temp.at tube
Tout K 327,6 314,8 306,5
iterations. Overview of solution methods and convergence bundle outlet
Pressure at the tube
criteria is visible in Tab. 1. pin Pa 9,30 40,55 149,2
bundle inlet
Pressure at the tube
Table 1 Solutions methods and convergence criteria pout Pa −1,85 −11,28 −49,85
Solutions methods / bundle outlet
Item
convergence criterion
Pressure velocity coupling SIMPLE 3.2 The results for needle type fins
Momentum Second order upwind
Turbulent kinetic energy Second order upwind The input and output data of the heat calculations are
Turbulent dissipation rate Second order upwind summarized in the tables below.
Continuity, k, ω, velocity for x,
Convergence judgement
y, z: 1×10−4 Table 6 Characteristic dimensions of tube and fins
conditions
Energy: 1×10−9 Tube and fin material - - Stainless steel
Tube outside diameter d0 mm 20
Solution is converged after about 11000 iterations. Tube inside diameter di mm 17
The convergence error is verified on the amount of Tube rows configuration - - staggered
heat delivered by the inner wall of the tube through the Transverse tube pitch st mm 50
outer surface of the tube and the fins and the amount of Longitudinal tube pitch sl mm 40
heat received by the air. The maximum deviation is Fin height hf mm 10
0,077%. Needle root diameter - mm 27
Number of needle - - 8
Fin thickness tf mm 0,5
3 The results of numerical calculation
Fin pitch sf mm 4,5
3.1 The results for round fins
Number of rows Nl - 5
The input and output data of the heat calculations are
Boundary conditions for needle type fins are the same
summarized in the tables below.
as for round fins.
For the calculation of the necessary flow variables at
boundaries (for example at inlet and outlet of tube
bundle), mass flow averaging was used.

Tehnički vjesnik 24, 6(2017), 1785-1796 1789

Development of a new type of finned heat exchanger M. Bošnjaković et al.

Table 7 Output data from Ansys Fluent for k-ω SST turbulence model transfer rate is enhanced. In practice, the fin is thin and
uin,c = uin,c = uin,c = temperature changes in the transverse direction within the
Item
1,0 m/s 2,4 m/s 5,0 m/s fin are small compared with the temperature difference
Air temp.at tube between the fin and the environment (Fig. 5b).
Tout K 321,8 311,1 304,3
bundle outlet
Ideally, the fin material should have a large thermal
Pressure at the tube
pin Pa 8,10 37,48 142,84 conductivity to minimize temperature variations from its
bundle inlet
Pressure at the tube base to its tip. In the limit of infinite thermal conductivity,
pout Pa −2,10 −12,28 −52,52 the entire fin would be at the temperature of the base
bundle outlet
surface, thereby providing the maximum possible heat
4 Data Reduction and interpretation transfer enhancement. However, since a finite conduction
resistance characterizes any fin, a temperature gradient
The heat duty is calculated from the calculated mass must exist along the fin. In our case, fin material is
flow rates and overall temperature changes. stainless steel with relative small thermal conductivity 16
W/(m·K) and therefore temperature variation in
m = ρin ⋅ Atot ⋅ uin (10) longitudinal direction (Fig. 5b) will be important causing
heat transfer less than maximum possible.
Qair = m ⋅ cp,av ⋅ (Tout − Tiu ) (11) In addition, temperature field across fin area is not
uniform (Fig. 5a) and temperature of fin tip is different
To calculate the overall heat transfer coefficient (U), for each needle. All these parameters are taken in account
the log-mean (LMTD) method is used. by fin efficiency.
The logarithmic mean temperature difference for
counter-current flow is chosen. According to Næss [17]
this method has proven to be accurate to within 0,1 % of
the actual cross-counter-current arrangement mean
temperature difference.

Tin − Tout
ΔTln = (12)
T − Tw
ln in
Tout − Tw

The overall heat transfer coefficient is then calculated

Qair
U= (13)
A ⋅ ∆Tln

The effective/apparent air-side heat transfer

coefficient is

1
αe = (14)
 d 
 ln 0  Figure 5 Temperature over fins in 3rd row for air velocity 5 m/s
 1 −A di A
− tot 
U tot
2 π Lt l Ai α i 
  Round fins efficiency is calculated according to
  Schmidt

Lt is tube length that is equal to model width. In the 2α 0

case of air temperatures lower than 300 °C, the value of m= (16)
λf ⋅ tf
radiation heat transfer coefficient could be negligible and
then its value may be taken as zero. Outside and inside tf
he = h + (17)
fouling factors are also taken as zero. The effective heat 2
transfer coefficient in Eq. (14) is the apparent heat
transfer coefficient including fin efficiency. In order to Corrected fin length he is based on assumption of
calculate the actual average gas-side heat transfer equivalence between heat transfer from the actual fin with
coefficient αo, the fin efficiency ηf is needed. tip convection and heat transfer from a longer,
hypothetical fin with an adiabatic tip.
α e ⋅ Atot
α0 = (15)
( At + η f ⋅ Af ) tanh(ψ ⋅ m ⋅ he )
h th, f = (18)
ψ ⋅ m ⋅ he
The thermal conductivity of the fin material has a  he 
strong effect on the temperature distribution along the fin ψ = 1,0 + 0,35 ⋅ ln1,0 + 2,0 ⋅  (19)
and therefore influences the degree to which the heat  d 0 

1790 Technical Gazette 24, 6(2017), 1785-1796

M. Bošnjaković i dr. Razvoj novog tipa lamelnog izmjenjivača topline

E = 0,76 + 0,24 ⋅h th, f (20) ∆p = pin − pout (28)

hf = E ⋅h th, f (21)
where pin and pout is the mass weighted average pressure
For new type of fins proposed in the article, there is at inlet and outlet of tube bundle (see Fig. 2).
no expression for efficiency in literature. As serrated fins
are similar to new type of fins expression for serrated fins 5 Validation with literature results for round fins
will be used. This means some error in calculation and
has to be examined in future. Flow through tube banks is studied experimentally
The appropriate theoretical fin efficiency for serrated since it is too complex to be treated analytically.
fins, which compensates for the finite conductance in the Engineers are primarily interested in the average heat
fins, under the assumption of a uniform air side heat transfer coefficient for the entire tube bank, which
transfer coefficient, can be expressed as [1] depends on the number of tube rows along the flow as
well as the arrangement and the geometry of the tubes and
fins. Several correlations from literature, all based on
tanh(m ⋅ he )
h th,f = (22) experimental data, have been proposed for the average
m ⋅ he Nusselt number for cross flow over tube banks with round
fins.
where the parameter me is given by For the validation of selected turbulence models, a
comparison of the Nu and Eu obtained by heat calculation
2α 0 for the described model and calculation by correlation
me = (23) with Briggs [9], ESDU [35], VDI-Warmeatlas [33], D. J.
λf ⋅ te
Ward and E. H. Young [3], T. E. Schmidt [23], PFR [36]
tf and Feten [37].
te = (24)
tf It should be noted that in most cases the correlations
1+
ws that define Nu number do not specify whether the
physical properties of the fluid should be selected for the
d r,f ⋅ π
ws = (25) boundary layer or for the bulk flow. For the evaluation of
2 ⋅ Nn Nu number in this article, physical properties of the fluid
boundary layer are taken. This can potentially make a
where ws is average width of needle fin and te is certain error in the results. In Fig. 6 to Fig. 9 there is a
equivalence thickness of fin, dr,f is root diameter of fin comparison of the calculation results of the analysed
needle. model with round fins and the results obtained by
Due to the formation of vortices and boundary layers, correlations from the literature.
the assumption of uniformity of heat transfer coefficient
across the fin does not hold in actual situations.
Experiments have shown lower heat transfer coefficients
than predicted theoretically. In the present analysis for
needle type fins an empirical correction factor to the
theoretical fin efficiency proposed by Weierman [5] for
serrated fins is used.

E = 0,9 + 0,1 ⋅h th, f (26)

Efficiency of serrated fins is defined acc. to Eq. 21.

Recall that fins are used to increase the heat transfer
from a surface by increasing the effective surface area.
However, the fin itself represents a conduction resistance
to heat transfer from the base surface. For this reason,
there is no assurance that the heat transfer rate will be
increased by the use of fins. An assessment of this matter
Figure 6 Calculated Nu and the results from literature for the round fins
may be done by evaluating the fin effectiveness εƒ.
Both the fin efficiency and fin effectiveness are
The values of Nu numbers obtained on the SST k-ω
related to the performance of the fin, but they are different
turbulence model show max. deviation of 13 % according
quantities. However, they are related to each other by
to the reference correlations which can be evaluated as
very good (Fig. 7).
Af Eu values for the SST k-ω turbulence model show
εf = ⋅η f (27)
At slightly greater deviation from literature. The deviations
are higher for lower Re values and up to 29% while for
Numerical pressure drop in tube bundle is calculated higher values of Re deviations are up to 16% (Fig. 9).
from the next expression Also, the values of Eu are less than by correlations.
Interestingly, the k-ε Realizable model shows a better
trend of results than the SST k-ω model in the observed

Tehnički vjesnik 24, 6(2017), 1785-1796 1791

Development of a new type of finned heat exchanger M. Bošnjaković et al.

range of Re compared to the literature although the In accordance with the previously discussed for
deviations are higher. further comparison of the round fins and the new type of
needle fins, the SST k-ω turbulence model will be
selected.

6 Comparison of calculation results for round fins and

needle type fins
6.1 Flow characteristic

The global velocity distribution in the bundle is

essential for understanding the local flow and heat
transfer phenomena. Fig. 10 shows velocity fields in a
cross-section passing through the middle of fin thickness.
Figs. 9, 10, 11 and 12 show the local characteristics of the
temperature, pressure, turbulent kinetic energy and the air
speed of the needle type finned exchanger and round
finned exchanger for inlet air velocity of 5,0 m/s. Cross
flow over a tube exhibits complex flow patterns. The fluid
approaching the tube branches out and encircles the tube,
Figure 7 The difference between the computed Nu and the results from forming a boundary layer that wraps around the tube. The
the literature fluid particles on the mid-plane strike the tube at the
stagnation point, bringing the fluid to a complete stop and
thus raising the pressure at that point. The pressure
decreases in the flow direction while the fluid velocity
increases. Fluid embraces the tube on the frontal side, but
it is too fast to remain attached to the surface as it
approaches the top of the tube. As a result, the boundary
layer detaches from the surface at approximately 120°,
forming a separation region behind the tube. Flow in the
wake region and especially after the 5th row of tubes is
characterized by vortex formation and pressures much
lower than the stagnation point pressure. Caused by the
slower air flow, higher temperatures persisted in wake
areas.

Figure 8 Calculated Eu and the results from literature for the round fins

Figure 10 Temperature fields comparison of round and needle type fins

Figure 11 Pressure fields comparison of round and needle type fins

Figure 9 The difference between the computed Eu and the results from Figure 12 Velocity fields comparison of round and needle type fins
the literature
The difference in the air velocity field (Fig. 12) and
From the previous diagrams, it can be seen that the
the level of turbulence, and thus the difference in heat
SST k-ω model of turbulence in the observed area of Re
transfer coefficient increases with row number because of
numbers shows a better match to the results of literature
the combined effects of upstream rows. In the case of a
than the k-ε Realizable model.
needle shaped fins, the formation of the vortex is more

1792 Technical Gazette 24, 6(2017), 1785-1796

M. Bošnjaković i dr. Razvoj novog tipa lamelnog izmjenjivača topline

emphasized than in the case of round fins. It is also [24] stated the average increase in the Nu number of 22 %
apparent that the amount of turbulence energy is greater relative to the finned exchanger with circular round fins.
for the needle shape fins (Fig. 13). In our case, increase in Nu number is higher and it is
around 26 %.

Figure 13 Turbulence kinetic energy fields comparison of round and

needle type fins

Figure 14 Density fields comparison of round and needle type fins Figure 16 Computational Nu results for round and needle type fins

To ensure accuracy in calculating the thermal

boundary layer, the low-Re approach to modelling the
near-wall flow was used. Consequently, the near-wall
mesh resolution places nodes within the viscous sublayer,
avoiding reliance on a wall function. Accordingly, on
fluid-side the wall-adjacent mesh spacing (Δy) was
sufficiently small so that y+ < 1 at the fin and tube
surfaces. For inlet air velocity 1 m/s and 2,4 m/s max.
value of y+ is 1,0. For nominal velocity 5 m/s on the
largest part of the surface y+ is less than 1 and in a small
part of the surface reaches a value of 1,8 (Fig.15). SST k-
ω model switches automatically from a wall function to a
low-Re formulation, based on the grid density provided by
the user. It ensures therefore an optimal accuracy of the Figure 17 Nu difference between round and needle type fins
CFD solution on a given grid and allows the use of
coarser near-wall grids, which also benefits the cell aspect
ratio.

Figure 15 y+ values of round and needle type fins

Figure 18 Computational Eu results for round and needle type fins
In general, it can be concluded that the two types of
fins have substantially similar flow characteristics but
turbulence is somewhat higher for needle type fins that
results in better heat transfer.

6.2 Non-dimensional numbers Nu and Eu

The numerical results of the calculation based on the

data given in Tab. 5 and Tab. 7 can be graphically
displayed.
As can be seen from Fig. 16 and Fig. 17, Nu number
is greater than 20 % to 30 % for needle type fins
compared to Nu number for round fins. The difference is
more significant for higher Re numbers. In the analysis of
finned heat exchanger with segmented fins R. Hofmann Figure 19 Eu difference between round and needle type fins

Tehnički vjesnik 24, 6(2017), 1785-1796 1793

Development of a new type of finned heat exchanger M. Bošnjaković et al.

Fig. 18 and Fig. 19 show that Eu number is Specific fan power is about 10 % greater for needle
approximately equal for both fin types. Even for the type fins than round fins.
needle type fin Eu decreases a little with its value.
Although the fin shape varies the net free area in a tube
row is nearly the same in both cases. Thus, a significant
difference among pressure drop is not expected.

6.3 Performance assessment

The performance rating of the two observed heat

exchangers will be analysed through multiple parameters.
The efficiency and effectiveness of round fin and needle
type fin are analysed first. It is evident that round fin has a
much greater effectiveness of needle type fin
proportionally larger surface. The efficiency of the round
fin is also greater than needle type fin efficiency, meaning
that the temperature difference is smaller between the fin Figure 22 Exchanged heat per unit of the exchanger surface
root and fin tip. For a given material and flow conditions
it actually means lower heat exchange which is evident
from Fig. 22 and Fig. 23.

Figure 23 Exchanged heat per unit of mass of exchanging surfaces

Figure 20 Fin efficiency

Figure 24 Specific fan power

7 Conclusion
Figure 21 Fin effectiveness
Numerical simulations of heat exchanger with needle
The exchanged heat per unit of heat exchanger type fins and round fins were performed in the same way,
surface area is higher for the needle type fins by 35 % at under the same initial assumptions and parameters
Re = 2340 and 39,3 % at Re = 12000. applying two turbulence models. Comparing data
The power required to keep a fluid flowing through obtained from CFD and data from literature, SST k-ω
the tube bank (and thus the operating cost) is proportional turbulence model was selected. For the selected
to the pressure drop. Therefore, the benefits of enhancing turbulence model and air input speeds of 1,0; 2,4 and 5,0
heat transfer via fins should be weighed against the cost m/s, a comparison was made between the Nu and Eu
of fan power requirements. Specific fan power can be numbers for round fins and needle fins. The needle fins
defined as showed better results in terms of increasing the Nu
number to 30 % while simultaneously reducing the weight
of the exchange surfaces by 23,8 %. Increase in the
m ⋅ ∆p
es, fan = (29) required specific fan power for the operation of exchanger
Qair ⋅ rin ⋅ηfan with needle type is about 10 %. On the other side, round

1794 Technical Gazette 24, 6(2017), 1785-1796

M. Bošnjaković i dr. Razvoj novog tipa lamelnog izmjenjivača topline

fins are more effective, primarily because of the larger U overall heat transfer coefficient W/(m2·K)
surface for heat exchange. uff velocity at minimum flow area m/s
The variance of the fin geometry, the fin pitch and the uin air velocity at the inlet of tube bundle m/s
tube pitch will be the subject of the next research to give a uin,c air velocity at the inlet of the heat m/s
more complete picture of the new needle type fin. exchanger
ws average width of needle fin mm
Nomenclature α0 actual average gas-side heat W/(m2·K)
transfer coefficient
A surface area m² αe effective heat transfer coefficient W/(m2·K)
Af fin surface area m² based on total fin side surface area
Aff min. free flow area in a tube row m² ΔTln logarithmic mean temperature K
At tube outside surface area except m² difference (LMTD)
fins εf fin effectiveness -
Atot total heat transfer area on air side m² ηf,th theoretical fin efficiency -
cp,av air specific heat capacity for Tav J/(kg·K) λbl thermal conductivity of boundary W/(m·K)
di inside tube diameter mm layer at Tbl
do Outside tube diameter mm λf thermal conductivity of fin W/(m·K)
dr,f root diameter of fin needle mm material
e total specific energy J/kg λt thermal conductivity of tube W/(m·K)
E empirical correction factor to the - material
theoretical fin efficiency for μav air average kinematic viscosity m2/s
serrated fins ρav air average density (at Tav) kg/m3
Eu Euler number - ρin air density at inlet (at Tin) kg/m3
ESDU Engineering Sciences Data Unit, - σij tensor of stress in fluid Pa
United Kingdom
fi vector of mass density of the N/kg 8 References
external forces
hf fin height mm [1] Hashizume, K.; Matsue, T. Fin Efficiency of Serrated Fins:
he fin height mm Part 1. Analysis of Theoretical Fin Efficiency and
Lt tube length (equal to model width) mm Experimental Results. // Heat Transfer—Asian Research.
28, 6(1999), pp. 528-540. https://doi.org/10.1002/(SICI)1523-
m fin effectiveness parameter for -
1496(1999)28:6<528::AID-HTJ9>3.0.CO;2-B
round fins [2] Kays, W. M.; London, A. L. Compact Heat Exchangers,
me fin effectiveness parameter for - McGraw-Hill, New York, 1964.
serrated fins [3] Ward, D. J.; Young, E. H. Heat transfer and pressure drop
𝑚𝑚̇ air mass flow kg/s of air in forced convection across triangular pitch banks of
Nl number of tubes in the flow - finned tubes. // Chem. Eng. Prog. Symp. Ser. 55, 29(1959),
direction pp. 37-44.
Nn number of needles of fin - [4] Kearney, S. P.; Jacobi, A. M. Local Convective Behavior
Nu Nusselt number - and Fin Efficiency in Shallow Banks of Inline and
Staggered, Solidly Finned Tubes. // ASME Journal of Heat
pin mass weighted average pressure at Pa
Transfer. 118, (1996), pp. 317-326.
inlet of tube bundle https://doi.org/10.1115/1.2825847
pout mass weighted average pressure Pa [5] Weierman, C. Correlations ease the selection of finned
outlet of tube bundle tubes. // The Oil and Gas Journal. 74, 36(1976), pp. 94-100.
pout,c mass weighted average pressure Pa [6] Frasz, F. Waermeuebertragung in Rippenrohrwaerme -
outlet of channel austauschern.// Waermeaustauscher, Energieeinsparung
∆p pressure drop Pa durch Optimierung von Waermeprozessen, Vulkan-Verlag
Pr Prandtl number - Essen. 2(1994), pp. 70-76.
Q heat flow rate W [7] Kawaguchi, K. Heat transfer and pressure drop
characteristics of finned tube banks in forced convection. //
q heat flux vector W Journal of Enhanced Heat Transfer. 12, 1(2005), pp. 1-20.
Re Reynolds number - https://doi.org/10.1615/JEnhHeatTransf.v12.i1.10
sf fin pitch mm [8] Halici, F.; Taymaz, I. Experimental study of the airside
sl longitudinal tube pitch mm performance of tube row spacing in finned tube heat
st transverse tube pitch mm exchangers. // Heat and Mass Transfer. 42(2006), pp. 817-
Tav mean air temperature in tube K 822. https://doi.org/10.1007/s00231-005-0042-1
bundle [9] Briggs, D. E.; Young, E. H. Convection Heat Transfer and
Tbl film temperature – boundary layer K Pressure Drop of Air Flowing Across Triangular Pitch
temperature Banks of Finned Tubes. // Chem.-Ing. Technik. 33,
6(1963), pp. 431-438.
te equivalence thickness of fin mm [10] Reid, D. R.; Taborek, J. Selection criteria for plain and
tf fin thickness mm segmented finned tubes for heat recovery systems. //
t time s Transactions of the ASME. 116 (1994), pp. 406-410.
Tin air inlet temperature K https://doi.org/10.1115/1.2906835
Tout air outlet temperature K [11] Weierman, C., Taborek, J.; Marner, W. J. Comparison of
Tw tube wall temperature K the performance of in-line and staggered banks of tubes

Tehnički vjesnik 24, 6(2017), 1785-1796 1795

Development of a new type of finned heat exchanger M. Bošnjaković et al.

with segmented fins. // The American Institute of Chem. https://doi.org/10.1080/01457638708962793

Engineers. 74, 174(1978), pp. 39-46. [29] Kuntysh, V. B.; Iokhvedov, F. M. Heat Transfer. // Soviet
[12] Chapter 14. // High-Performance Single-Phase Heat Research. 3(1971), pp. 50-61.
Exchangers // Zhukauskas, A. New York: Hemisphere [30] Pongsoi, P.; Wongwises, S. Determination of fin pitches for
Publishing, 1989. maximum performance index of l-footed spiral fin-and-tube
[13] Stasiulevičius, J.; Skrinska, A. Heat Transfer of Finned heat exchangers. // Journal of Thermal Engineering, Yildiz
Tube Bundles in Crossflow, New York: Hemisphere Technical University Press, Istanbul. 1, Special Issue
Publishing, 1988, ISBN 3-540-18211-x 1(2015), pp. 251-262.
[14] Sandar M. M.; Gross, U. Numerical study of fin-spacing [31] Halici, F.; Taymaz, I. Experimental study of the airside
effects in solid-finned tube heat exchangers. // International performance of tube row spacing in finned tube heat
Journal of Heat and Mass Transfer. 47(2004), pp. 1953- exchangers. // Heat and Mass Transfer, 42(2006), pp. 817-
1964. https://doi.org/10.1016/j.ijheatmasstransfer.2003.09.034 822. https://doi.org/10.1007/s00231-005-0042-1
[15] Ali Shakir A. J.; Majid H. M.; Bassam A. S. Circular fins [32] Incropera, F. P.; DeWitt, D. P.; Bergman, T. L.; Lavine, A.
with slanted blades attached on the copper pipe: Uniform S. Fundamentals of Heat and Mass Transfer. John Wiley &
heat flux and isothermal processes. // International Journal Sons, Inc., 2007.
of Mechanical Engineering and Technology (IJMET). 5, [33] Verein Deutscher Ingenieure, VDI-Wärmeatlas,
5(2014), pp. 133-143. Berechnungsblätter feur den Wärmeübergang, 8. Aufl.
[16] Lemouedda, A.; Schmid, A.; Franz, E.; Breuer, M.; Berlin u.a., Springer, 2000.
Delgado, A. Numerical investigations for the optimization [34] Yunus, Ç. A. Heat Transfer: A Practical Approach. 2nd ed,
of serrated finned-tube heat exchangers. // Applied Thermal New York: McGraw-Hill, 2004.
Engineering. 31(2011), pp. 1393-1401. [35] ESDU 1986. 86022: High-fin Staggered Tube Banks: Heat
https://doi.org/10.1016/j.applthermaleng.2010.12.035 Transfer and Pressure Drop for Turbulent Single Phase Gas
[17] Næss, E. Experimental investigation of heat transfer and Flow. Engineering Sciences Data Unit, United Kingdom.
pressure drop in serrated-fin tube bundles with staggered [36] PFR Engineering Systems, Inc.: Heat Transfer and Pressure
tube layouts. // Applied Thermal Engineering. 30(2010), pp. Drop Characteristics of Dry Tower Extended Surfaces.
1531-1537. https://doi.org/10.1016/j.applthermaleng.2010.02.019 Battelle Memorial Institute, 1976.
[18] Martinez, E.; Vicente, W.; Soto, G.; Salinas, M. [37] Feten, T. G. F. Multivariate analysis of heat transfer and
Comparative analysis of heat transfer and pressure drop in pressure drop in finned tube bundles. Master Thesis, 2013.
helically segmented finned tube heat exchangers. // Applied
Thermal Engineering. 30(2010), pp. 1470-1476.
https://doi.org/10.1016/j.applthermaleng.2010.03.007 Authors’ addresses
[19] Cathal Cléirigh, T. Ó.; Smith, W. J. Can CFD accurately
predict the heat-transfer and pressure-drop performance of Mladen Bošnjaković, M.Sc.
finned-tube bundles? // Applied Thermal Engineering. 73, College of Slavonski Brod,
Dr. M. Budaka 1,
1(2014), pp. 681-690.
35000 Slavonski Brod, Croatia
https://doi.org/10.1016/j.applthermaleng.2014.08.019 [email protected]
[20] Anoop, B.; Balaji, C.; Velusamy, K. A characteristic
correlation for heat transfer over serrated finned tubes. // Ante Čikić, Ph.D. Associate Prof.
Annals of Nuclear Energy. 85(2015), pp. 1052-1065. Technical College in Bjelovar,
https://doi.org/10.1016/j.anucene.2015.07.025 Trg Eugena Kvaternika 4,
[21] Hofmann, R.; Frasz, F.; Ponweiser, K. Heat Transfer and 43 000 Bjelovar, Croatia
Pressure Drop Performance Comparison of Finned-Tube [email protected]
Bundles in Forced Convection. // Wseas transactions on
Simon Muhič, Ph.D., Associate Prof.
heat and mass transfer. 4, 2(2007), ISSN: 1790-5044 Faculty of Technologies and Systems,
[22] Briggs, D. E.; Young E. H. Convection heat transfer and Na Loko 2,
pressure drop of air flowing across triangular pitch banks 8000 Novo mesto, Slovenia
of finned tubes. // Chem. Eng. Prog. Symp. Ser. 59, No. 41, [email protected]
1963, pp. 1-10.
[23] Schmidt, T. E. Verbesserte Methoden zur Bestimmung des Marinko Stojkov, Ph.D., Full Prof.
Wärmeaustausches an berippten Flächen. // Kältetechnik- Mechanical Engineering Faculty in Slavonski Brod,
Trg Ivane Brlić Mažuranić 2,
Klimatisierung. 18(1966), pp. 135-138.
35000 Slavonski Brod, Croatia
[24] Ward D. J.; Young, E. H. Heat transfer and pressure drop of [email protected]
air in forced convection across triangular pitch banks of
finned tubes. // Chem. Eng. Prog. Symp. Ser. 55, 29(1959),
pp. 37-44.
[25] Nir, A. Heat Transfer and Friction Factor Correlations for
Cross Flow Over Staggered Finned Tube Banks. // Heat
Transfer Engineering. 12, 1(1991), pp. 43-58
https://doi.org/10.1080/01457639108939746
[26] Lokman, H. M.; Fdhila, R. B. Literature Review of
accelerated CFD Simulation Methods towards Online
Application. // Energy Procedia 75(2015), pp. 3307-3314.
https://doi.org/10.1016/j.egypro.2015.07.714
[27] Hossain, N.; Moghimi, M. Numerical Study of Flow Over
Solid-Finned Tube Heat Exchangers by Different Turbulent
Models. // CFD Letters, 6, 3(2014).
[28] Rabas, T. J.; Taborek, J. Survey of Turbulent Forced-
Convection Heat Transfer and pressure Drop
Characteristics of Low-Finned Tube Banks in Cross Flow.
// Heat Transfer Engineering. 8, 2(1987), pp. 49-62.

1796 Technical Gazette 24, 6(2017), 1785-1796