Journal Pre-proofs

Heat transfer enhancement of a fin-and-tube compact heat exchanger by em-

ploying magnetite ferrofluid flow and an external magnetic field

Mojtaba Bezaatpour, Hadi Rostamzadeh

PII: S1359-4311(19)33948-1
DOI: https://doi.org/10.1016/j.applthermaleng.2019.114462
Reference: ATE 114462

To appear in: Applied Thermal Engineering

Received Date: 9 June 2019

Revised Date: 28 August 2019
Accepted Date: 28 September 2019

Please cite this article as: M. Bezaatpour, H. Rostamzadeh, Heat transfer enhancement of a fin-and-tube compact
heat exchanger by employing magnetite ferrofluid flow and an external magnetic field, Applied Thermal
Engineering (2019), doi: https://doi.org/10.1016/j.applthermaleng.2019.114462

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Elsevier Ltd. All rights reserved.

 Heat transfer enhancement of a fin-and-tube compact heat

exchanger by employing magnetite ferrofluid flow and an external

magnetic field

Applied Thermal Engineering

Title Page

Mojtaba Bezaatpour1,a and Hadi Rostamzadehb

aDepartment of Mechanical Engineering, Sahand University of

Technology, Tabriz, Iran

bDepartment of Aerospace Engineering, Sharif University of Technology,

Azadi Ave., Tehran, Iran

Signature (on behalf of all co-authors (if any))

Corresponding author

Name: Mojtaba Bezaatpour

1Corresponding author email address: [email protected] (M. Bezaatpour),

Co-author email address: [email protected] (H. Rostamzadeh)

1
 Affiliation: Department of Mechanical Engineering, Sahand University of Technology, Tabriz, Iran
E-mail address: [email protected]

Submission date: 10/1/2019

Abstract

Compact heat exchangers as modern industrial devices are designed to improve heat recovery and

saving energy processes in restricted spaces. In the current study, effect of a uniform external

magnetic field with Fe3O4/water nanofluid for heat transfer enhancement of a fin-and-tube

compact heat exchanger is numerically investigated. The obtained results are verified by the

available experimental data to demonstrate accuracy of the present simulation. The results

indicated that the local and average heat transfer coefficients increase around the tubes in the

presence of an external magnetic field due to the vortex formation behind the tubes as well as the

flow pattern alteration in the heat exchanger, no mention that the pressure drop increment is subtle

through this variation. Also, it is figured out that employing an external magnetic field at low

Reynolds numbers (approximately 𝑅𝑒𝐷 < 30) is much more appropriate since the heat exchanger

effectiveness increases exponentially. A maximum heat transfer enhancement of 8.7% was

obtained by employing 2%Vol. magnetite nanoparticles in deionized water as a coolant. This value

surges up to 52.4% by applying an external magnetic field in the compact heat exchanger. Based

on these results, employing an external magnetic field for heat transfer enhancement of compact

heat exchangers and heat sinks designed for restricted spaces and low Reynolds numbers can be

2
more efficient in comparison with other available methods since no pumping power for mass flow

rate augmentation is required in the present approach.

Keywords: Compact heat exchanger; Magnetic field; Magnetite ferrofluid; Heat transfer; Vortex

generator.

Nomenclatur
e
Samples
A overall heat transfer area (𝑚2 Greek
) letters
𝐵 magnetic flux density (𝑁.𝐴 ―1. 𝜑 volume fraction

𝑚 ―1)
𝐶𝑝 specific heat (𝑘𝐽.𝑘𝑔 ―1.𝐾 ―1) 𝜌 density (𝑘𝑔.𝑚 ―3)
𝐷ℎ hydraulic diameter (𝑚) 𝜇 dynamic viscosity (𝑁.𝑠.
𝑚 ―2)

𝑓 friction factor 𝜇0 permeability of free

space (4𝜋 × 10 ―7 𝑁.𝐴 ―2)
𝐹𝑀 Kelvin body force (𝑁.𝑚 ―3) 𝜏𝑖𝑗 stress tensor (𝑃𝑎)

ℎ convection heat transfer 𝛿𝑖𝑗 Kronecker delta

coefficient (𝑊.𝑚 ―2.𝐾 ―1)

𝐻 magnetic field intensity ( 𝜒𝑚 magnetic susceptibility

𝐴.𝑚 ―1)

𝑘 thermal conductivity (𝑊.𝑚 ―1. 𝜒0 differential magnetic

𝐾 ―1) susceptibility
LMTD logarithmic mean temperature 𝛽 fraction of the liquid
difference (K) volume (𝑘 ―1)

𝑀 magnetization (𝐴.𝑚 ―1) 𝜂 heat exchanger

effectiveness

𝑚 mass flow rate (𝑘𝑔.𝑠 ―1) 𝜀 heat exchanger efficiency

3
 𝑛 normal direction

𝑁𝑢 Nusselt number subscripts

P pressure (𝑃𝑎) avg average

𝑄 overall heat transfer (𝑊) f fluid

𝑞′′ wall heat flux (𝑊.𝑚 ―2) in inlet

𝑅𝑒 Reynolds number mag magnetic

𝑇 temperature (𝐾) max maximum

𝑢 velocity in x-direction ( np nanoparticle

𝑚.𝑠 ―1)

𝑣 velocity in y-direction ( out outlet

𝑚.𝑠 ―1)

𝑉 overall velocity (𝑚.𝑠 ―1) s solid

𝑤 velocity in z-direction ( w isothermal surface

𝑚.𝑠 ―1)

𝑥,𝑦,𝑧 directions 0 reference state (@ 300K)

Acronyms
CHE compact heat exchanger
FTCHE fin-and-tube CHE

1. Introduction

In comparison with the conventional heat exchangers, compact heat exchangers (CHEs) are mainly

characterized by their high heat transfer area per volume. Broadly speaking, proposing smaller and

lighter devices can be a promising solution for mini-scale applications, including but not limited

to sensor technologies [1], cryogenics components [2], automotive industry [3, 4], aerospace

engineering [5, 6], nuclear technologies [7], etc. Regarding this aim, CHEs are proposed which

encompass several channels each of which is separated by a fin in order to increase contact surface

area. For miniature devices operating at low Reynolds numbers, CHEs experience high pressure

4
drops through the channel due to the abrupt expansions/contractions at its exit/entrance as well as

the small length of the channel [8, 9].

In recent years, performance enhancement of CHEs has been deliberated as a top priority by

scholars due to the high cost associated with the manufacturing of CHEs [10]. From the first

proposal of using mini-channels in performance improvement of heat sinks by Tuckerman and

Pease [11] in 1981 onward, different scholars have devised various strategies to further enhance

heat transfer characteristics of compact channels. As a result of these investigations, it was

recommended to use several mini channels with small cross section in form of a compact heat

exchanger to provide high heat transfer rate between cold and hot streams [11]. Numerous scholars

have used passive methods to improve heat transfer and flow characteristics of CHEs in recent

years. For instance, Nascimento and Garcia [12] used shallow square dimples on the walls of flat

tubes in CHEs and experimentally demonstrated that the heat transfer factor augments within range

of 1.37-2.28. Mortean et al. [10] presented four new CHEs by sandwiching or stacking the flat-

plates interfiled with water jet machined plates. They used analytical, numerical, and experimental

tools and demonstrated that stacking flat-plates (called cut-plate heat exchanger) can be the most

promising technology to be fabricated. Zeeshan et al. [13] optimized thermal performance of three

tubes with different configurations of circular, oval and flat in CHEs and compared their

performance with each other. They achieved 13.99% enhancement in heat transfer coefficient at

low Reynolds number (Re=400) and 4.99% at high Reynolds number (Re=900) in the optimum

scenario, i.e., oval tube. Mortean et al. [9] used more accurate method to model thermal and

hydrodynamic trends of a cross-flow CHE and demonstrated that their results are analogous to

those of experiment. They also figured out that the headers of a CHE attribute over 50% of the

total pressure drop. Gorobets et al. [14] suggested the use of no gaps and shifting in the transversal

5
direction of the fluid flow between tubes of the CHEs. They showed that their designed CHEs are

smaller (1.7-2 times) and lighter (10-15%) than the conventional ones with the same heat power.

Aragon National Research Laboratory (ANRL) first recommended the use of nanofluids in thermal

energy systems to increase thermal conductivity of mixture which is ascribed to the increase of the

thermal conductivity of the base fluid, the Brownian motion, and thermophoresis [15]. However,

the pressure drop as a results of the presence of nanoparticles augments due to the friction forces

created by the movement of the base fluid layers [16]. Various studies have recorded merits and

demerits of using nanofluid in CHEs and have reported positive insight towards such deliberation.

For instance, Sarafraz and Arjomandi [17] experimentally investigated impact of using copper

oxide nanoparticles dispersed in the liquid indium. They found that for the nanoparticle

concentrations of up to 8% heat transfer coefficient improves, while decreases for the nanoparticle

concentrations of >8%. In another studies by these scholars reported in Ref. [18, 19], the potential

of using a liquid metal enriched with aluminum oxide nanoparticles in a CHE is investigated. Their

findings indicated that using gallium nano-suspension leads to a substantial improvement in

cooling process of the problem. Sarafraz et al. [20] found a small pressure drop, friction factor,

and fouling thermal resistance when silver nanoparticles were dispersed into the deionized water.

They reported a maximum thermal efficiency enhancement of 37% at Reynolds number of 1400

and nanoparticle concentration of 5%.

Meantime, applications of ferrofluid (defined as a strongly magnetized liquid exposed to a

magnetic field) as a coolant in heat transfer enhancement of heat exchangers and electronics

devices in small and large scales are received a great deal of attentions [21]. Generally speaking,

thermal conductivity specifications of ferrofluid are appreciably improved in the presence of an

external magnetic field [22]. For instance, using ferrofluid in a LED (light emitting diode) cooling

6
system increases its illuminance at specific elapsed time since illuminance efficiency of a LED

cooling system increases with the presence of magnet by around 7% compared to that without the

magnet [23]. Possibility of controlling the fluid flow of ferrofluids by applying magnetic field is

another efficient option in heat transport equipment [22]. Numerous investigations on performance

enhancement and impact of ferrofluid applied to various problems under external magnetic field

are presented in recent years which show the superiority of the topic. For example, Ghofrani et al.

[24] experimentally investigated convection heat transfer of a circular copper tube with ferrofluid

exposed to a varying magnetic field under laminar flow and uniform heat flux conditions. They

measured the maximum convection heat transfer improvement of 27.6% in low Reynolds numbers.

Goharkhah and Ashjaee [25] used ferrofluid exposed to a non-uniform magnetic field in a 2-D

channel under a constant heat supply from the bottom and top surfaces. They reported heat transfer

improvement of approximately 14% at Reynolds number of 2,000 and frequency of 20 Hz with a

minimum pressure drop of 6% at this Reynolds number and frequency of 5 Hz. In 2015, Ashjaee

et al. [26] investigated impact of a magnetic field on the pressure drop and forced convection heat

transfer rate of ferrofluid in a mini heat sink with five circular channels. They recorded heat

transfer augmentation of 38% when 1200 G magnetic force is applied to the ferrofluid. Bahiraei

and Hangi [27] numerically studied flow and heat transfer behaviors of the magnetic nanofluid in

a toroidal loop for proposing a self-powered cooling system. They used two magnets and

discovered that the fluid circulation velocity increases by 100%. More recently, Bezaatpour and

Goharkhah [28] numerically studied impact of a uniform external magnetic field on heat transfer

and hydrodynamic characteristics of a porous fin heat sink, using Fe3O4/water nanofluid. They

found that using ferrofluid instead of pure water can improve heat transfer rate of the problem by

13% (when solid fins are used) and by 35% (when porous fins are used).

7
Considering the above-reviewed open literature, it can be figured out that the application of

magnetic field in compact heat exchangers for heat transfer improvement purposes is not

investigated up to yet. However, recent investigations have explicitly demonstrated that the

applications of ferrofluid as coolant in mini or micro-scale devices for heat transfer augmentation

result in encouraging outcomes. Due to the widespread use of FTCHEs in small-scale equipment,

using ferrofluid in FTCHEs in presence of an external magnetic field can increase local heat

transfer coefficient, and hence the devised idea can be a viable solution for high-tech CHE

industrial applications. In fact, since CHEs are mainly designed for low Reynolds number

applications due to their mini and micro scales characteristics, it is highly imperative to devise

highly efficient methods to increase heat transfer behavior of these heat exchangers without

creating considerable pressure drops through the flow by consuming a pumping power proposed

in literature. On the other hand, literature shows that the effect of magnetic field on flow pattern is

more pronounced at low Reynolds numbers in which flow momentum is weak enough and Kelvin

body force is able to overcome the viscous force. Therefore, the main objective of present study is

to introduce the magnetic field as a promising solution for performance enhancement of the

FTCHEs by portraying variations of flow and heat transfer parameters through the heat exchanger

domain. It will be demonstrated that using magnetic field in the CHEs will improve the mixing

characteristics of the fluid flow without any need for employing complex devices such as different

vortex generators or bumps to achieve this aim.

The rest of the present report is arranged in the following order. In the second section, the proposed

problem modeled in the study is expounded in more detail. In the subsequent section, the governing

equations, boundary conditions, thermophysical properties of the selected nanofluid, and

numerical settings are presented. In section 4, the developed numerical simulation is verified by

8
accessible experimental data and the simulation results are presented thereafter. Finally, a

conclusion is drawn based on the obtained results.

2. Problem statement

Schematic of the reckoned mini fin-and-tube compact heat exchanger (FTCHE) for numerical

simulation is presented in Fig. 1. Accordingly, the heat exchanger is designed with length, width,

and height of 30 mm, 30 mm, and 10 mm, respectively. Three tubes with uniform temperature of

340 K and an equal distance of 8 mm from each other are symmetrically located at the middle of

the heat exchanger. Six fins are embedded within the heat exchanger in parallel with the fluid flow.

The details of geometrical specifications of the designed FTCHE are listed in Table 1. Ferrofluid

with 2% concentration and temperature of 300 K in the base fluid (water) is entered to the heat

exchanger to transfer heat via three tubes. The physical properties of magnetite

nanoparticle and heat exchanger material have been summarized in Table 2. The

uniform magnetic field (in the y-direction) generated in the center of a C-

shape electromagnet is applied to the ferrofluid flow. More details about the

electromagnet can be found in Ref. [26]. In addition, Fig. 2 illustrates

distribution of the constructed grid in different sections of the developed

model.

9
Figure 1: Schematic of the utilized mini fin-and-tube compact heat exchanger exposed to a C-shape electromagnet.

10
 Figure 2: Grid distribution in different sections of the configuration.

Table 1: Dimensions of the designed FTCHE.

Dimensions (mm) Fins Tubes
Length 30 30
Distance 5 8
Thickness 1 -
Diameter - 2

Table 2: Properties of the heat exchanger material and 𝐹𝑒3𝑂4 particle [29].
Material Diameter(mm) k (W/m.K) Cp(J/kg.K) 𝜌 (kg/m3)
Aluminum - 202.4 871 2719
Fe3O4 0.00002 7 640 4950

3. Governing equations and boundary conditions

3.1. Governing equations

11
The continuity, momentum, and energy equations under homogeneous and steady state

presumptions can be expressed respectively as follows [30]:

∇.(𝜌𝑽) = 0 (1)

∇.(𝜌𝑽𝑽) = ―∇𝑃 + ∇.(𝜏𝑖𝑗) + 𝑭𝑀 (2)

∇.(𝜌𝑽𝐶𝑝𝑇) = ∇.(𝑘∇𝑇) (3)

where,

2 (4)
𝜏𝑖𝑗 = 𝜇(∂𝑣𝑖 ∂𝑥𝑗 + ∂𝑣𝑗 ∂𝑥𝑖) ― 𝜇𝛿𝑖𝑗(∂𝑣𝑖 ∂𝑥𝑖)
3

𝑭𝑀 = (𝑴.∇)𝑩 (5)

The term 𝑭𝑀 is known as Kelvin body force (or the magnetic volume force) which

is explained in the next sub-section.

3.2. Governing equations

According to the Maxwell’s equations in the static form, the magnetic field

can be articulated as follows [31]:

∇.𝑩 = 0 (6)

∇×𝑯= 0 (7)

where, 𝐁 can be related to 𝐌 and 𝐇 as follows [32]:

𝑩 = 𝜇0(𝑴 + 𝑯) (8)

In addition, magnetization vector can be related to the magnetic field intensity

via magnetic susceptibility (𝜒𝑚) as:

12
𝑴 = 𝜒𝑚 𝑯 (9)

where, χm is defined as [30]:

𝜒0 (10)
𝜒𝑚(𝑇) =
1 + 𝛽(𝑇 ― 𝑇0)

Kelvin body force is a force per unit volume that a magnetic fluid experiences

in the presence of a magnetic field due to the affected magnetic particles in

the fluid domain. Kelvin body force appears in the momentum equation to predict

magnetic fluid behavior under the effect of a magnetic field. Substituting Eqs.

(8) and (9) into Eq. (5), magnetic body force can be rewritten as follow [30]:

1 (11)
𝑭𝑀 = 𝜇0𝜒𝑚(1 + 𝜒𝑚)∇(𝑯.𝑯) + 𝜇0𝜒𝑚𝑯((𝑯.∇)𝜒𝑚)
2

where, the first term in Eq. (11) is related to the magnetostatic pressure term.

Eq. (11) can further be simplified by assuming that the FTCHE is located at the

center of the C-shape electromagnet and exposed to a uniform electromagnetic

component in y-direction (𝑯 = 𝐻𝒋) [28]:

― 𝜒0𝛽 ∂𝑇 (12)
𝑭𝑀 = 𝜇0𝜒𝑚𝐻2 𝒋
[1 + 𝛽(𝑇 ― 𝑇0)]2∂𝑦

3.3. Boundary conditions

The inlet temperature of the ferrofluid is presumed 300 K and the tubes are at

340 K under isothermal condition. The bottom and top surfaces of the geometry

are also insulated. Nanofluid leaves the channel at the atmospheric pressure.

Additionally, the no-slip condition is set at all solid-fluid interfaces:

13
 ∂𝑇𝑓 ∂𝑇𝑠 (13)
𝑢 = 𝑣 = 𝑤 = 0, 𝑇𝑓 = 𝑇𝑠, ― 𝑘𝑓 = ― 𝑘𝑠
∂𝑛 ∂𝑛

For the insulated wall, we have:

∂𝑇 (14)
=0
∂𝑛

3.4. Thermophysical properties of nanofluid

The thermophysical properties of the selected nanofluid (nf) including specific

heat, density, and viscosity are computed in terms of the nanoparticle (np) and

base fluid (f) properties respectively as follows [33]:

(𝜌𝐶𝑝)𝑛𝑓 = (1 ― 𝜑)(𝜌𝐶𝑝)𝑓 + 𝜑(𝜌𝐶𝑝)𝑛𝑝 (15)

𝜌𝑛𝑓 = (1 ― 𝜑)𝜌𝑓 + 𝜑𝜌𝑛𝑝 (16)

𝜇𝑛𝑓 = 𝜇𝑓(1 + 2.5𝜑) (17)

The thermal conductivity can be written as [34]:

𝑘𝑛𝑓 = 𝑘𝑆𝑡𝑎𝑡𝑖𝑐 + 𝑘𝐵𝑟𝑜𝑤𝑛𝑖𝑎𝑛 (18)

(𝑘𝑛𝑝 + 2𝑘𝑓) ― 2𝜑(𝑘𝑓 ― 𝑘𝑛𝑝) (19)

𝑘𝑆𝑡𝑎𝑡𝑖𝑐 = 𝑘𝑓[ ]
(𝑘𝑛𝑝 + 2𝑘𝑓) + 𝜑(𝑘𝑓 ― 𝑘𝑛𝑝)

𝑘𝑇 (20)
𝑘𝐵𝑟𝑜𝑤𝑛𝑖𝑎𝑛 = 5 × 104𝛽𝜑𝜌𝑓𝐶𝑝,𝑓 𝑔(𝜑,𝑇)
𝜌𝑛𝑝𝐷𝑛𝑝

here, 𝛽 is the volumetric fraction of the liquid that travels with a particle

and 𝑘 = 1.3807 × 10 ―23J/K. Also, 𝑔(𝜑,𝑇) is expressed as:

𝑔(𝜑,𝑇) = ( ―6.04𝜑 + 0.4705)𝑇 + 1722.3𝜑 ― 134.63 (21)

For pure water, it is presumed that the thermophysical properties depend on

temperature as follows [35]:

14
 𝜌𝑓 = 2446 ― 20.674 𝑇 + 0.11576 𝑇2 ― 3.12895 × 10 ―4𝑇3 + 4.0505 × 10 ―7𝑇4 ― 2.0546 × 10 ―10𝑇5 (22)

𝜇𝑓

= 2.414 × 10 ―5
× 10
[ 247.8
𝑇 ― 140 ]
(23)

𝑘𝑓 = ―1.13 + 9.71 × 10 ―3 𝑇 ― 1.31 × 10 ―5𝑇2 (24)

The local heat transfer coefficient is attained in terms of the surface heat flux (𝑞′′), outlet

temperature of nanofluid (𝑇𝑜𝑢𝑡), and surface temperature (𝑇𝑤 = 340 𝐾) as follows:

𝑞′′
ℎ𝑙𝑜𝑐𝑎𝑙 = (25)
(𝑇𝑤 ― 𝑇𝑜𝑢𝑡)

The heat transferred through the process is expressed as follows:

𝑄 = 𝑚.𝐶𝑝.(𝑇𝑜𝑢𝑡 ― 𝑇𝑖𝑛) (26)

where, 𝑚, 𝑇𝑜𝑢𝑡, and 𝑇𝑖𝑛 are the mass flow rate, outlet temperature and inlet temperature of

nanofluid, respectively.

The average heat transfer coefficient is written as:

𝑄
ℎ𝑎𝑣𝑔 = (27)
𝐴.𝐿𝑀𝑇𝐷

where, 𝐴 is the overall heat transfer area associated with the three tubes and LMTD stands for the

logarithmic mean temperature difference and is computed as follows:

(𝑇𝑤 ― 𝑇𝑜𝑢𝑡) ― (𝑇𝑤 ― 𝑇𝑖𝑛)

𝐿𝑀𝑇𝐷 = (28)
(𝑇𝑤 ― 𝑇𝑜𝑢𝑡)
𝑙𝑛
(𝑇𝑤 ― 𝑇𝑖𝑛)

The heat exchanger effectiveness and efficiency can be articulated respectively as follows:

15
 𝑁𝑢 𝑁𝑢
𝐵=0
𝜂= 1
(29)
𝑓 3
( 𝑓𝐵 = 0)

𝑄
ε= (30)
𝑄𝑚𝑎𝑥

where,

ℎ𝐷ℎ
𝑁𝑢 = (31)
𝑘𝑛𝑓

𝑄𝑚𝑎𝑥 = 𝑚.𝐶𝑝.(𝑇𝑤 ― 𝑇𝑖𝑛) (32)

In Eq. (31), 𝐷ℎ is the hydraulic diameter and 𝑘𝑛𝑓 is the thermal conductivity of the nanofluid.

3.5. Numerical method

Due to the periodic characteristics of the FTCHE geometry, the region outlined

by the dashed lines (Fig. 1) is selected as the computational domain. As shown

in Fig. 2, a structured and uniform grid is used at all regions and the grid

points are clustered near the tube wall. The generated grid near the fin-tube

intersection, where the velocity and temperature gradients are large enough, is much finer

than elsewhere. The second-order upwind scheme is used to discretize the domain of the

momentum and energy equations. SIMPLE algorithm is applied to couple velocity and pressure.

In order to secure the preciseness of the numerical simulation, the residuals considered in iteration

process are set on as low value as 10 ―6 for continuity and energy equations. The number of nodes

in the selected grid is chosen 289110 which can be justified in the following grid resolution study.

Three grids with total nodes number of 150320, 289110, and 410522 are selected and variation of

16
the friction factor and local heat transfer coefficient through the first tube at Reynolds number of

14.33 is illustrated in Fig. 3. As Fig. 3 indicates, among all investigated grids, a grid with 289110

nodes can be more appropriate in terms of both computational efficiency and cost. Considering

the magnified plot in Fig. 3, it can be argued that the results of the grids 289110 and 410522 are

overlapped with each other. Therefore, increasing the number of grid from 289110 will lead to no

further preciseness, while the run-time cost increases since more computational memory is

required to carry out such computation.

Figure 3: Variation of the friction factor and local heat transfer coefficient through the first tube for grid

independency study at 𝑅𝑒 = 14.33.

4. Results and discussion

This section is arranged into two sub-sections. In the first part, a valuable experimental study is

selected and the accuracy of the simulated numerical method is demonstrated by comparing the

results of present study with those of the selected literature. In the second part, the simulation

results of the reckoned problem (explained in Fig. 1) are deliberated and further explained.

4.1. Model verification

17
As mentioned before, the effect of magnetic field on ferrofluid can be simulated by adding the

Kelvin body force as an additional source in momentum equation. In order to evaluate accuracy of

the employed numerical method of present study and demonstrate on what preciseness the

developed numerical simulation can imitate the results of the experimental study for ferrofluid in

the presence of a uniform magnetic field, two comparisons between numerical results of present

study and those of experimental have been carried out (Fig. 4). As a first comparison, the effect of

the localized magnetic field on heat transfer in rectangular duct flow of a magnetic fluid conducted

by Motozawa et al. [22] is numerically investigated. In this simulation, the effects of radiation and

free convection due to the dominant impact of the flow momentum and magnetic force (with

relative computational order of 107) are neglected. Also, notwithstanding the large volumetric

ratio of nanofluid, it is presumed that the ferrofluid has non-conductive characteristics. Boundary

conditions and nanofluid properties are treated similar to the study of Motozawa et al. [36].

Simulations are conducted under single-phase and solid-fluid equilibrium temperature conditions

for two magnetic strengths of 100 and 300 mT. A localized uniform magnetic field is applied to

only one part of the channel (as expressed in Ref. [36]). The wall and ferrofluid temperatures and

local heat transfer coefficient are calculated at 50 different cross-sections of the channel and the

results are illustrated in Fig. 4(a). According to Fig. 4(a), it can be said that the obtained results via

the performed numerical simulation well concur with those of Motozawa et al. [36] with relative

difference of below 10%.

In the second comparison, the experimental results of Ashjaee et al. [26] for a miniature heat sink

with magnetite ferrofluid as a coolant under the effect of a magnetic field is considered to validate

the current numerical results. Fig. 4(b) shows variation of the pressure drop along the heat sink

18
length for different Reynolds numbers at φ=2% and B=800G. The maximum deviation from the

experimental results has been found less than 8% in this case.

Figure 4: Model verification between present study and (a) Motozawa et al. [36] (b) Ashjaee et al. [26].

4.2. Simulation results

4.2.1. Magnetic force distribution

It can be inferred from Eq. (12) that the resultant magnetic force has a single component in y-

direction due to the vertical temperature gradient. Therefore, employing a uniform magnetic field

results in a magnetic force in vertical direction as ferrofluid contacts with surface of the tubes and

a temperature gradient is created thereafter. Based on this concept, distribution of the magnetic

force around each tube is calculated and illustrated in Fig. 5. Since distribution of this force in the

upper and lower sides of the tubes is symmetrical, only distribution of this force on the upper half

of the tubes is plotted. As Fig. 5 depicts, the magnetic force is zero at 0 and 180 degrees, where

the plane is symmetric and no temperature gradient exists in the y-direction. As 𝜃 increases, the

vertical temperature gradient increases, and hence the magnetic force augments and reaches its

maximum value. A comparative study between graphs of the three tubes indicates that location of
19
the maximum force in each tube is different and the peak value occurs at low locations for the first

tube and moves forward for the subsequent tubes. Also, the strength of the magnetic field

influences location of the maximum magnetic force, especially for the third tube. A quantitative

investigation has revealed that at 𝐵 = 250 𝐺, the maximum magnetic force of the first, second,

and third tubes are calculated approximately 9, 4.21, and 1.97 GN/m3 at 𝜃 = 55.53, 65.82, and

92.91 degrees, respectively. Interestingly enough, two peak points are observed for magnetic force

of the first tube at high magnetic intensities. This phenomenon can be justified as follows. At high

magnetic intensities, the nanofluid flow separation on the tube (attributed by high magnetic force)

decreases flow contact with tube surface, and hence temperature gradient decreases drastically and

magnetic force decreases subsequently. The trend is reversed at 𝜃 > 90 degrees and the magnetic

force slightly augments onward since the temperature gradient is not as high as the beginning.

In order to have a better comparison between three tubes, Fig. 6 is illustrated, where the magnetic

force for all three tubes at two magnetic intensities of 100 and 200 G is presented. As Fig. 6

indicates, the location of the peak magnetic force moves forward for the tubes in the flow direction

(x-direction). Also, the magnetic force around the first tube is substantially higher than that around

the second and third tubes since the temperature difference between the tube and nanofluid is high

around the first tube.

In addition, the magnetic force contours for the case of Fig. 6 are illustrated in Fig. 7. Fig. 7(a)

shows magnetic force distribution on all tubes. According to Fig. 7(a), the effect of magnetic field

on nanofluid flow near the fins is less than the central regions since the velocity is small around

this region. Fig. 7(b) demonstrates magnetic force appearance and its strength in the heat exchanger

around the isothermal tubes at symmetric plane. As shown in Fig. 7(b), the magnetic force can be

present even further away from the tubes at high strengths.

20
Figure 5: Magnetic force distribution through the tubes on symmetric boundary at 𝑅𝑒 = 14.33 and different

magnetic field intensities for the: (a) first, (b) second, and (c) third tubes.

21
 Figure 6: Comparison of the magnetic force distribution through the tubes on symmetric boundary at 𝑅𝑒 = 14.33

for 𝐵 = 100 𝑎𝑛𝑑 200 𝐺.

Figure 7: Magnetic force contours (a) on the tubes (b) at symmetric plane at 𝑅𝑒 = 14.33 for 𝐵 = 100 𝑎𝑛𝑑 200 𝐺.

Fig. 8 shows the 3D velocity streamlines through the tubes inside a channel of FTCHE at five

magnetic intensities of 0, 100, 150, 200, and 250 G. Fig. 8 shows that the above-mentioned

magnetic force appeared around the tubes changes the flow pattern and separates the nanofluid

streams from the surface of tubes. Consequently, the nanofluid velocity on the tubes increases and

22
a vortex is formed behind each tube. As the magnetic field strength increases, a small vortex is

first created behind the first tube, and then this vortex becomes larger, and hence the second vortex

with smaller size is generated behind the second tube. With increasing the magnetic field strength,

these two vortices also become larger, and finally, at 𝐵 = 250 𝐺 a strong repulsive force on the

first tube leads to appearance of the third vortex near the fins. It is expected that these vortices can

play a paramount role in increasing the heat transfer rate via changing the nanofluid flow direction,

its turbulence, or combining the hot stream with the cold bulk stream to increase their convection

heat transfer coefficients (the hot stream around the tubes is replaced by cold stream due to the

convection effect). It can be concluded that using magnetic field in the fin-and-tube compact heat

exchangers improves the mixing characteristics of the fluid flow without any need for employing

complex devices such as different vortex generators or bumps to achieve this aim.

23
24
 Figure 8: Velocity streamlines through the tubes under the influence of different magnetic field intensities for

𝑅𝑒 = 14.33.

4.2.2. Hydrothermal coefficients distribution

Fig. 9 illustrates variation of the friction factor through the tubes at different magnetic field

intensities (𝐵 = 0, 100, 150, 200, 𝑎𝑛𝑑 250 𝐺). According to Fig. 9, in the absence of a magnetic

field, the friction factor through the tubes increases at first and then decreases, resulting in a peak

value (similar to the magnetic force). In the presence of an external magnetic field, the variation

rate of the friction factor is boosted, especially in the first and second tubes. Exerting a magnetic

force in the vertical direction at the initial points of the tubes increases momentum of the flow,

resulting in the increase of the friction factor at these points since the magnetic force becomes

more tangential to the tube’s surface. This augmentation continues up to the point where the

friction factor is maximized and decreases onward since presence of the magnetic force decreases

flow momentum by increasing boundary layer on the surface. As flow moves through the tubes

after the peak point, the angle between the surface of tubes and magnetic force at each point

increases; and consequently, the magnetic force decreases the contact surface and the friction

factor decreases thereafter. Of course these variations are appreciable around the first tube due to

a stronger magnetic force. It is noteworthy to pinpoint that two peak points are observed for the

friction factor for the first and second tubes (mainly at high magnetic intensities), which is mainly

due to the fact that the contact surface of the nanofluid with the tubes decreases as flow separation

phenomenon occurs. Obviously, since the momentum of the fluid stream decreases through the

channel as it passes over the first tube, it was predicted to observe the second pinnacle value of the

friction factor at lower values (compared to the first maximum point). Another significant

indication of Fig. 9(c) is that the friction factor variation trend around the third tube is different in

25
comparison with that of two other tubes. In other words, the friction factor is decreased as the

magnetic strength augments and experiences a sinusoidal variation at high magnetic intensities.

This phenomenon can be expounded as follows. At high magnetic intensities, the flow streamlines

are stretched through the third tube due to the appearance of vortices behind the first and second

tubes (see Fig. 8), leading to a drop in the flow momentum through the third tube and the friction

factor thereafter.

Figure 9: Friction factor variation through the tubes on symmetric boundary at 𝑅𝑒 = 14.33 and different magnetic

field intensities for the: (a) first, (b) second, and (c) third tubes.

26
To present a better comparison between the friction factors of different tubes, Fig. 10 is depicted.

In Fig. 10, the friction factor distribution and its contours around each tube at two magnetic

intensities of 0 and 150 G are presented. Fig. 10 substantiates the previous findings illustrated in

Fig. 9 and explicitly indicates that the friction factor is lower near the fins compared with that in

the middle region since the velocity and the magnetic force are small around this region.

Figure 10: Comparison of friction factor distribution and its contours through the tubes at 𝑅𝑒 = 14.33 for

𝐵 = 0 𝑎𝑛𝑑 150 𝐺.

It is also expected that the temperature distribution and heat transfer characteristics would be

influenced by the presence of the external magnetic field due to the observed changes in

characteristics of the fluid flow. In order to better understand impact of the external magnetic field

on heat transfer characteristics of the problem, it is imperative to depict temperature contours

around the tubes and inside the heat exchanger. Fig. 11 illustrates temperature distribution around

the tubes and inside the heat exchanger at symmetric plane. As indicated in Fig. 11, the heated

nanofluid engendered by the presence of the magnetic field near the isothermal tubes penetrates

smoothly upon the tubes into the cold flow due to the previous mentioned changes which increases

27
the average temperature at the outlet of channel and then increases the heat transfer rate through

the channel as well. It is also observed that the thermal boundary layer thickness decreases at the

initial points of tubes under the influence of the magnetic field due to the presence of the vertical

force and the momentum augmentation around the tubes (especially around the first tube).

Figure 11: Temperature contours through the tubes at symmetric plane for Re=14.33 in the presence of different

magnetic field intensities.

To further support the discussed findings of present study in terms of heat transfer, variations of

the local convection heat transfer coefficient around each tube at different magnetic intensities (0,

100, 150, 200, and 250 G) are presented in Fig. 12. As it can be understood from Fig. 12, the

highest local convection heat transfer coefficient occurs at 𝜃 = 0, where the highest temperature

difference is seen and the nanofluid contacts with the first tube with constant temperature. When

no magnetic field is employed, the local convection heat transfer coefficient decreases

28
continuously through the tubes. It is shown that the local heat transfer coefficient increases

significantly at the initial points under the influence of the magnetic field due to the decrease of

the thermal boundary layer thickness. A minimum point is also observed for the local convection

heat transfer coefficient (especially around the first tube). The local heat transfer coefficient is

decreased rapidly at higher magnetic intensities due to the formation of vortices behind the first

and second tubes at low 𝜃, while increasing rapidly as flow re-contacts with the tube surface. For

the third tube, however, the local convection heat transfer coefficient is nearly decreased at all

magnetic field intensities. As justified before, the flow density and velocity around the third tube

are decreased, while the nanofluid temperature passing through the first and second tubes is

increased. Consequently, the temperature difference between the nanofluid and the third tube

declines since the nanofluid loses specific amount of its heating capacity via direct contact with

the first and second tubes, resulting in decrement of the local convection heat transfer coefficient.

29
Figure 12: Variation of the local convection heat transfer coefficient through the tubes on the symmetric boundary

at 𝑅𝑒 = 14.33 and different magnetic field intensities around the: (a) first, (b) second, and (c) third tubes.

To present a better comparison between the local convection heat transfer coefficients of different

tubes, Fig. 13 is depicted. In Fig. 13, the local heat transfer coefficient distribution and its contours

around each tube at two magnetic intensities of 0 and 150 G are presented. Fig. 13 substantiates

the previous findings illustrated in Fig. 12 and explicitly indicates that the local convection heat

transfer coefficient is lower near the fins since the velocity and magnetic force are small around

this region. Fig. 13 also delineates the fact that the local heat transfer coefficient is minimum near

the fins and endpoints of the tubes, while it is maximum at their initial points in the middle region.

30
 Figure 13: Comparison of the local convection heat transfer coefficient distribution and its contours through the

tubes at 𝐵 = 0 𝑎𝑛𝑑 150 𝐺.

To find out the effect of the magnetic field on the hydrothermal performance of the heat exchanger,

the simulation results for pressure drop and overall convection heat transfer coefficient in the

presence of the magnetic field are investigated. The results of this simulation for the average

convection heat transfer coefficient at different Reynolds numbers and magnetic intensities are

shown in Fig. 14. According to Fig. 14, it is observed that the average convection heat transfer

coefficient is increased with employing an external magnetic field, and hence the heat transfer rate

increases. Of course, the average convection heat transfer coefficient augmentation rate is more

appreciable at high magnetic intensities and low Reynolds numbers (due to the low momentum of

the flow and more influences of magnetic field as well), where the same is true about the heat

transfer rate. In the absence of the external magnetic field, the average heat transfer coefficient

increases with increasing of the Reynolds number since the nanofluid velocity increases at high

Reynolds numbers and it has less time for transferring heat with the hot tubes, resulting in a high

temperature difference between the nanofluid and tubes. It is worthy to mention that the increment

rate of the average heat transfer coefficient with the Reynolds number at low magnetic intensities

31
is nearly linear and continuous. However, at high magnetic intensities, a nadir point is observed.

This significant trend is preeminently due to the mutual impact of the temperature difference

increment and the magnetic field decrement with an increase in the Reynolds number. At low

Reynolds numbers (𝑅𝑒𝐷 < 20 for 𝐵 = 200𝐺 and 𝑅𝑒𝐷 < 24 for 𝐵 = 250G), the impact of the

magnetic field decrement rate is dominant on the increment rate of the temperature difference due

to the low momentum of the flow, while the trend is reversed at 𝑅𝑒𝐷 > 20 𝑎𝑛𝑑 25. According to

the obtained results, it can be prognosticated that the value of the minimum local heat transfer

coefficient is expected to occur at high Reynolds numbers as magnetic intensity further augments.

Figure 14: Variation of the average convection heat transfer coefficient versus of the Reynolds number at different

magnetic field intensities.

Fig. 15 is plotted to show variation of the pressure drop through the tubes in the presence of the

magnetic field. As expected before, the pressure drop increases as the Reynolds number rises at

all magnetic intensities. Since the studied device works in low range of Reynolds numbers (14-

44), the relation between Reynolds number and pressure drop seems linear. However, it should be

noticed that a nonlinear relation (can be quadratic relation) is drawn between these two parameters

32
at high ranges of Reynolds numbers, as justified in our previous work [37]. Additionally, it is

discovered that the external magnetic field has a subtle impact on the pressure drop. Considering

heat transfer enhancement due to the presence of an external magnetic field (explained through

Figs. 11-14) and subtle impact of the magnetic field presence on the pressure drop, it can be

deduced that the proposed method to enhance heat transfer characteristics of a FTCHE can be

highly encouraging.

Figure 15: Pressure drop variation versus of the Reynolds number at different magnetic field intensities.

4.2.3. Heat exchanger performance

In order to investigate effect of the employed external magnetic field on heat transfer and pressure

drop simultaneously, variations of the heat exchanger effectiveness (η) versus of the Reynolds

number at different magnetic intensities are presented in Fig. 16. As Fig. 16 reveals, the heat

exchanger effectiveness value at all examined intensities is above 1, underlining that the heat

transfer augmentation rate is substantially higher than that of the pressure drop. Increasing the

magnetic field strength also improves the effectiveness of the heat exchanger since η augments

through this variation. Additionally, Fig. 16 delineates that employing an external magnetic field

33
at low Reynolds numbers is more appropriate since the heat exchanger effectiveness increases

exponentially.

Figure 16: Variation of the heat exchanger effectiveness versus of the Reynolds number at different magnetic field

intensities.

Fig. 17 portrays variation of the heat exchanger efficiency with Reynolds number at different

magnetic intensities. Similar to Fig. 16, the heat exchanger efficiency is augmented as the magnetic

field strength increases due to the same reason.

34
 Figure 17: Variation of the heat exchanger efficiency versus Reynolds number at different magnetic field

intensities.

5. Concluding remarks

Feasibility investigation of using a uniform external magnetic field with Fe3O4/water nanofluid in

a fin-and-tube compact heat exchanger (FTCHE) to enhance heat transfer rate of the package was

numerically investigated in this study. Effects of the Reynolds number and magnetic field strength

on hydrodynamic and heat transfer characteristics of the heat exchanger were studied. Overall, it

was concluded that employing an external magnetic field for heat transfer enhancement of a CHE

can be a promising solution since the heat transfer augmentation rate was substantially higher than

that of the pressure drop. Based on the obtained results, employing an external magnetic field for

heat transfer enhancement of equipment designed for restricted spaces and low Reynolds numbers

such as CHEs and heat sinks can be more efficient in comparison with other available methods

since no pumping power for mass flow rate augmentation is required in the present approach.

Following central concluding remarks are drawn from the present study:

 Magnetic force appeared around the tubes in the presence of a magnetic field and formed

vortices behind each tube which improved mixing characteristics of the flow.

 The local and average heat transfer coefficients were increased around all tubes (especially

around the first and second tubes) as the magnetic field strength increases, while the

pressure drop was increased subtly.

 Employing an external magnetic field at low Reynolds numbers is more appropriate since

the heat exchanger effectiveness was increased.

 The local convection heat transfer coefficient was lower near the fins since velocity and

the magnetic force were small around this region.

35
  A maximum heat transfer enhancement of 8.7% was obtained by employing 2%Vol.

magnetite nanoparticles in deionized water as a coolant. The heat transfer rate was

increased up to 52.4% by applying an external magnetic field as well.

References

[1] B. Torregrosa-Jaime, J. M. Corberán, J. Payá, and J. Delamarche, "Thermal characterisation of

compact heat exchangers for air heating and cooling in electric vehicles," Applied Thermal
Engineering, vol. 115, pp. 774-781, 2017.
[2] S. Baek, C. Lee, and S. Jeong, "Effect of flow maldistribution and axial conduction on compact
microchannel heat exchanger," Cryogenics, vol. 60, pp. 49-61, 2014.
[3] M. S. Yadav, S. A. Giri, and V. C. Momale, "Sizing analysis of louvered fin flat tube compact heat
exchanger by genetic algorithm," Applied Thermal Engineering, vol. 125, pp. 1426-1436, 2017.
[4] G. Junjanna, N. Kulasekharan, and H. Purushotham, "Performance improvement of a louver-
finned automobile radiator using conjugate thermal CFD analysis," International Journal of
Engineering Research & Technology, vol. 1, 2012.
[5] X. Xu, X. Zhang, P. Ke, C. Wang, H. Yang, and C. Yang, "Study on the heat transfer characteristic of
compact heat exchanger based on experimental data," Procedia Engineering, vol. 121, pp. 293-
299, 2015.
[6] J. Wen, H. Huang, H. Li, G. Xu, and Y. Fu, "Thermal and hydraulic performance of a compact plate
finned tube air-fuel heat exchanger for aero-engine," Applied Thermal Engineering, vol. 126, pp.
920-928, 2017.
[7] N. Bartel, M. Chen, V. Utgikar, X. Sun, I.-H. Kim, R. Christensen, et al., "Comparative analysis of
compact heat exchangers for application as the intermediate heat exchanger for advanced
nuclear reactors," Annals of Nuclear Energy, vol. 81, pp. 143-149, 2015.
[8] P.-S. Lee and S. V. Garimella, "Thermally developing flow and heat transfer in rectangular
microchannels of different aspect ratios," international journal of heat and mass transfer, vol. 49,
pp. 3060-3067, 2006.
[9] M. Mortean, L. Cisterna, K. Paiva, and M. Mantelli, "Thermal and hydrodynamic analysis of a cross-
flow compact heat exchanger," Applied Thermal Engineering, vol. 150, pp. 750-761, 2019.
[10] M. Mortean, L. Cisterna, K. Paiva, and M. Mantelli, "Development of diffusion welded compact
heat exchanger technology," Applied Thermal Engineering, vol. 93, pp. 995-1005, 2016.
[11] T. Dixit and I. Ghosh, "Review of micro-and mini-channel heat sinks and heat exchangers for single
phase fluids," Renewable and Sustainable Energy Reviews, vol. 41, pp. 1298-1311, 2015.
[12] I. P. Nascimento and E. C. Garcia, "Heat transfer performance enhancement in compact heat
exchangers by using shallow square dimples in flat tubes," Applied Thermal Engineering, vol. 96,
pp. 659-670, 2016.
[13] M. Zeeshan, S. Nath, and D. Bhanja, "Numerical study to predict optimal configuration of fin and
tube compact heat exchanger with various tube shapes and spatial arrangements," Energy
Conversion and Management, vol. 148, pp. 737-752, 2017.
[14] V. Gorobets, Y. Bohdan, V. Trokhaniak, and I. Antypov, "Investigations of heat transfer and
hydrodynamics in heat exchangers with compact arrangements of tubes," Applied Thermal
Engineering, vol. 151, pp. 46-54, 2019.

36
[15] M. Sarafraz, B. Yang, O. Pourmehran, M. Arjomandi, and R. Ghomashchi, "Fluid and heat transfer
characteristics of aqueous graphene nanoplatelet (GNP) nanofluid in a microchannel,"
International Communications in Heat and Mass Transfer, vol. 107, pp. 24-33, 2019.
[16] M. Sarafraz, J. Hart, E. Shrestha, H. Arya, and M. Arjomandi, "Experimental thermal energy
assessment of a liquid metal eutectic in a microchannel heat exchanger equipped with a (10 Hz/50
Hz) resonator," Applied Thermal Engineering, vol. 148, pp. 578-590, 2019.
[17] M. Sarafraz and M. Arjomandi, "Thermal performance analysis of a microchannel heat sink cooling
with Copper Oxide-Indium (CuO/In) nano-suspensions at high-temperatures," Applied Thermal
Engineering, vol. 137, pp. 700-709, 2018.
[18] M. Sarafraz and M. Arjomandi, "Demonstration of plausible application of gallium nano-
suspension in microchannel solar thermal receiver: experimental assessment of thermo-hydraulic
performance of microchannel," International Communications in Heat and Mass Transfer, vol. 94,
pp. 39-46, 2018.
[19] M. Sarafraz, H. Arya, and M. Arjomandi, "Thermal and hydraulic analysis of a rectangular
microchannel with gallium-copper oxide nano-suspension," Journal of Molecular Liquids, vol. 263,
pp. 382-389, 2018.
[20] M. Sarafraz, V. Nikkhah, M. Nakhjavani, and A. Arya, "Thermal performance of a heat sink
microchannel working with biologically produced silver-water nanofluid: experimental
assessment," Experimental Thermal and Fluid Science, vol. 91, pp. 509-519, 2018.
[21] M. Bezaatpour and M. Goharkhah, "Three dimensional simulation of hydrodynamic and heat
transfer behavior of magnetite nanofluid flow in circular and rectangular channel heat sinks filled
with porous media," Powder technology, vol. 344, pp. 68-78, 2019.
[22] M. Bahiraei and M. Hangi, "Investigating the efficacy of magnetic nanofluid as a coolant in double-
pipe heat exchanger in the presence of magnetic field," Energy Conversion and Management, vol.
76, pp. 1125-1133, 2013.
[23] J.-H. Seo and M.-Y. Lee, "Illuminance and heat transfer characteristics of high power LED cooling
system with heat sink filled with ferrofluid," Applied Thermal Engineering, vol. 143, pp. 438-449,
2018.
[24] A. Ghofrani, M. Dibaei, A. H. Sima, and M. Shafii, "Experimental investigation on laminar forced
convection heat transfer of ferrofluids under an alternating magnetic field," Experimental Thermal
and Fluid Science, vol. 49, pp. 193-200, 2013.
[25] M. Goharkhah and M. Ashjaee, "Effect of an alternating nonuniform magnetic field on ferrofluid
flow and heat transfer in a channel," Journal of Magnetism and Magnetic Materials, vol. 362, pp.
80-89, 2014.
[26] M. Ashjaee, M. Goharkhah, L. A. Khadem, and R. Ahmadi, "Effect of magnetic field on the forced
convection heat transfer and pressure drop of a magnetic nanofluid in a miniature heat sink,"
Heat and Mass Transfer, vol. 51, pp. 953-964, 2015.
[27] M. Bahiraei and M. Hangi, "Automatic cooling by means of thermomagnetic phenomenon of
magnetic nanofluid in a toroidal loop," Applied Thermal Engineering, vol. 107, pp. 700-708, 2016.
[28] M. Bezaatpour and M. Goharkhah, "Effect of magnetic field on the hydrodynamic and heat
transfer of magnetite ferrofluid flow in a porous fin heat sink," Journal of Magnetism and
Magnetic Materials, vol. 476, pp. 506-515, 2019.
[29] T. Sawada, H. Kikura, A. Saito, and T. Tanahashi, "Natural convection of a magnetic fluid in
concentric horizontal annuli under nonuniform magnetic fields," Experimental thermal and fluid
science, vol. 7, pp. 212-220, 1993.
[30] R. Ganguly, S. Sen, and I. K. Puri, "Heat transfer augmentation using a magnetic fluid under the
influence of a line dipole," Journal of Magnetism and Magnetic Materials, vol. 271, pp. 63-73,
2004.

37
[31] D. J. Griffiths, "Introduction to electrodynamics," ed: AAPT, 2005.
[32] D. K. Cheng, "Fundamentals of engineering electromagnetics," 1993.
[33] Y. Xuan and W. Roetzel, "Conceptions for heat transfer correlation of nanofluids," International
Journal of heat and Mass transfer, vol. 43, pp. 3701-3707, 2000.
[34] J. Koo and C. Kleinstreuer, "A new thermal conductivity model for nanofluids," Journal of
Nanoparticle Research, vol. 6, pp. 577-588, 2004.
[35] F. P. Incropera, A. S. Lavine, T. L. Bergman, and D. P. DeWitt, Fundamentals of heat and mass
transfer: Wiley, 2007.
[36] M. Motozawa, J. Chang, T. Sawada, and Y. Kawaguchi, "Effect of magnetic field on heat transfer
in rectangular duct flow of a magnetic fluid," Physics Procedia, vol. 9, pp. 190-193, 2010.
[37] M. Bezaatpour and M. Goharkhah, "A novel heat sink design for simultaneous heat transfer
enhancement and pressure drop reduction utilizing porous fins and magnetite ferrofluid,"
International Journal of Numerical Methods for Heat & Fluid Flow, 2019.

38
Highlights:

 Effect of an external magnetic field with Fe3O4/water nanofluid on a compact heat exchanger is

studied.

 Local and average heat transfer coefficients are increased around the tubes due to the vortex

formation behind the tubes.

 Heat transfer rate is enhanced by 8.7% as nanoparticles with 2%Vol. are employed.

 Heat transfer rate is enhanced by 52.4% as an external magnetic field is applied to the compact heat

exchanger.

39