International Communications in Heat and Mass Transfer 74 (2016) 40–47

Contents lists available at ScienceDirect

International Communications in Heat and Mass Transfer

journal homepage: www.elsevier.com/locate/ichmt

Heat transfer enhancement in a micro-channel cooling system using

cylindrical vortex generators☆
Mushtaq T. Al-Asadi a,b,⁎, F.S. Alkasmoul a, M.C.T. Wilson a
a
Institute of Thermofluids, School of Mechanical Engineering, University of Leeds, UK
b
Refrigeration Department, Eng. Division, South Oil Company, Ministry of Oil, Basrah, Iraq

a r t i c l e i n f o a b s t r a c t

Available online 17 March 2016 Three-dimensional conjugate heat transfer under laminar flow conditions within a micro-channel is analysed nu-
merically to explore the impact of a new design of vortex generator positioned at intervals along the base of the
Keywords: channel. The vortex generators are cylindrical with quarter-circle and half-circle cross sections, with variants
Vortex generators spanning the whole width of the channel or parts of the channel. Micro-channels with Reynolds number ranging
Micro-channel from 100 to 2300 are subjected to a uniform heat flux relevant to microelectronics cooling. To ensure the accuracy
Hydraulic thermal performance
of the results, validations against previous microchannel studies were conducted and found to be in good agree-
Micro scale cooling system
Heat transfer enhancement
ment, before the new vortex generators with radii up to 400 μm were analysed. Using a thermal–hydraulic
performance parameter expressed in a new way, the VGs described here are shown to offer significant potential
in combatting the challenges of heat transfer in the technological drive towards lower weight/smaller volume
electrical and electronic devices.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction transfer in different geometries such as circular and non-circular ducts

under turbulent flow [15–17]. They have also been used in laminar
Recently, developments in electronic and electrical devices have led flow [18], with flat plate-fins in rectangular channels [19–21], tube
to the reduction in their volumes and weights, and managing the heat heat exchangers [22], heat sinks [18,23] and rectangular narrow
generated is becoming a real challenge to thermal system researchers channels [24,25].
[1]. Therefore there is a need to improve cooling systems by decreasing The rectangular micro-channel was the best geometry based on the
their size and weight to micro- and mini-scale systems, such as micro- numerical investigation of Xia et al. [26], who considered various
channel heat exchangers and heat sinks [2], whilst simultaneously microchannel shapes. They also investigated the distribution of flow
increasing the efficiency to meet this development [3]. Mini- and through a collection of 30 microchannels forming a heat sink, consider-
micro-channels are different from traditional channels, and can be ing different header chamber shapes and inlet/outlet positions.
classified according to their associated hydraulic diameters, Dh, [4–6] Ebrahimi et al. [13] used a finite volume based numerical analysis to
as presented in Table 1. study the impact of using vortex generators (VGs) with various orienta-
The term ‘micro-channel’ first appeared in 1981 [7]. The concept had tions on laminar fluid flow and heat transfer regimes in a micro-
a great influence in thermal science as it decreased the hydraulic diam- channel. The results showed that the Nusselt number rose from 2 to
eter and enhanced the heat transfer. In the revolution of advanced 25% when the Reynolds number ranged from 100 to 1100, whilst the
manufacturing processes, many experimental and numerical studies in- maximum increase of friction factor was 30% when using the VGs. Var-
vestigated the heat transfer and fluid flow performance of various mod- ious other recent investigations have also indicated potential benefits of
ified geometries such as micro-channels with grooves and ribs [8–11]. using VGs with laminar flow at different Reynolds number [24,27,28].
The effect of vortex generators (VGs) on heat transfer and fluid flow However, there remains a need for deeper understanding of VG perfor-
characteristics were investigated experimentally in 1969 [12]. They mance over a wider range of laminar flow, for example to minimise the
can take various forms such as protrusions, wings, inclined blocks, pressure drop penalty resulting from the introduction of VGs.
winglets, fins, and ribs [13,14], and have been used to enhance heat Cheraghi et al. [3] studied a 2D numerical smooth channel system
with fixed heat flux applied to the wall sides and considered the effect
of adding an adiabatic cylinder aligned perpendicular to the laminar
☆ Communicated by W.J. Minkowycz.
⁎ Corresponding author at: Institute of Thermofluids, School of Mechanical Engineering,
flow direction and at different positions across the channel. The
University of Leeds, UK. Reynolds number was 100 and the Prandtl number ranged from 0.1 to
E-mail addresses: ml13mtka@leeds.ac.uk, M.Wilson@leeds.ac.uk (M.T. Al-Asadi). 1, and it was found that the maximum heat transfer enhancement

http://dx.doi.org/10.1016/j.icheatmasstransfer.2016.03.002
0735-1933/© 2016 Elsevier Ltd. All rights reserved.
 M.T. Al-Asadi et al. / International Communications in Heat and Mass Transfer 74 (2016) 40–47 41

The main results are discussed in Section 5, and conclusions are drawn
Nomenclatures
in Section 6.
As Surface area of the whole heat sink (μm2)
2. Geometry description
CFD Computational fluid dynamics
Cp Specific heat, J/kg.K
The base geometry considered is a single micro-channel with rectan-
D Diameter, μm
gular cross-section, as shown in Fig. 1(a). Such channels are common in
FEM Finite element method
the heat sinks designed for CPUs (Fig. 1b), where they form the gaps
FVM Finite volume method
between the parallel fins of the heat sink, and that is the application
K Thermal conductivity, W/m.K
considered here, with the base area of the heat sink taken as As =
L Channel length, μm
6.27 × 108 μm2. Within the channel, a number of cylindrical vortex gen-
VGs Vortex generators
erators are equally distributed along the base. These have cross-sections
P Pressure, N/m2
that are either a quarter-circle or a half-circle, as shown in Fig. 1(c) and
q Uniform heat flux, W/cm2
(d) respectively, and a variable radius, r, ranging up to 400 μm. The
Re Reynolds number
micro-channel dimensions are given in Table 2. The table also shows
T Temperature, K
the three different spanwise configurations considered, namely: ‘full-
X Axial distance, μm
span’, where the VG occupies the full width of the micro-channel;
‘centred’, where the VG is shorter than the channel width and is
Greek Symbols
centred; and ‘split’, which is the same as the centred configuration ex-
μ Viscosity, kg/ms
cept that the VG is split into two equal parts, with a gap of 50–100 μm
Ө Thermal resistance
between them. Table 2 includes a view of the VGs looking along the
ρ Densities, kg/m3
channel from the inlet.
Subscript
3. Mathematical modelling
ave Average
In Inlet
3.1. Governing equations and key parameters
max Maximum
Out Outlet
The water flow in the micro-channel is considered to be laminar,
S Surface
steady, incompressible and Newtonian, with gravitational and viscous
L Liquid
dissipation effects neglected. With u = (u, v, w) representing the liquid
velocity in (x,y,z) Cartesian coordinates, and p, ρ and μ denoting the liq-
uid pressure, density and viscosity respectively, the (dimensional)
governing equations for the flow are the usual continuity and Navier–
occurred when the cylinder was fixed in the halfway position. The re- Stokes equations:
sults also showed that the low Prandtl number had a positive effect on
heat transfer enhancement. ∇
u¼0 ð1Þ
A modified channel having cylindrical vortex generators inside a
uniform channel under turbulent flow with Reynolds number of 3745 ρðu
∇Þu ¼ ∇p þ μ∇2 u: ð2Þ
has also been investigated numerically [29]. It was found that utilizing
a cylindrical vortex generator enhanced the heat transfer by 1.18 The energy equation for the liquid phase in the micro-channel is
times compared to the uniform channel.
Based on the published literature and to the authors' knowledge, no ρC p u
∇T L ¼ k∇2 T L ð3Þ
previous study has investigated the influence of the diameter of half-
cylindrical vortex generators placed along the base of the channel. where Cp, TL, and k are respectively the specific heat, temperature,
This study therefore proposes and explores new geometry designs not and thermal conductivity of the liquid. Conduction in the solid is
previously considered: specifically VGs based on cylinders with half- captured by
circular and quarter-circular cross-sections are introduced into rectan-
gular micro-channel heat sinks. Several lateral variations of these VGs ∇
ðkS ∇T S Þ ¼ 0 ð4Þ
are also considered, namely full-span cylinders, shorter centred cylin-
ders, and split/separated cylinders. The optimal radii of the VGs are where TS and kS are respectively the temperature and thermal conduc-
also established. Section 2 of this paper describes the geometry in tivity of the solid.
more detail, after which Section 3 discusses the mathematical model The Reynolds number is here defined in terms of the inlet velocity,
and Section 4 discusses the numerical approach and validation process. uin, and hydraulic diameter as

ρuin Dh
Re ¼ : ð5Þ
Table 1 μ
Channel classification by Mehendale et al. [5] and Kandlikar and Grande [6].
The heat transfer performance is quantified by the thermal resis-
Mehendale et al. [5]. Kandlikar and Grande [6].
tance, defined as
Conventional Dh N 6 mm Conventional Dh N 3 mm
channels channels T ave  T in
Compact Ѳ¼ ; ð6Þ
1 mm b Dh ≤ 6 mm Minichannels 1 μm b Dh ≤ 3 mm As q
passages
Meso-channels 100 mm b Dh ≤ 1 mm Microchannels 10 μm b Dh ≤ 200 μm
Transitional where Tave is the average temperature of the base, Tin is the inlet temper-
0.1 μm b Dh ≤ 10 μm
Micro-channels 1 μm b Dh ≤ 100 μm
channels ature, and q is the heat flux through the base of the heat sink. To give a
Molecular balanced assessment of the effective heat transfer enhancement provid-
Dh ≤ 0.1 μm
nanochannels
ed by VGs, taking into account the penalty paid in terms of the pressure
42 M.T. Al-Asadi et al. / International Communications in Heat and Mass Transfer 74 (2016) 40–47

Outlet
Inlet

c) d)

b)

a) e) f)
Inlet
4000 µm Symmetry plane
xin Hc
Ht
r

Ww /2 wc /2 Heat flux

Fig. 1. Geometry description: (a) rectangular micro-channel containing vortex generators; (b) top view of a heat sink comprised of a series of micro-channels; (c) side-view cross-section
of a quarter-circle vortex generator; (d) side-view cross-section of a half-circle vortex generator; (e) view along the channel showing the definition of parameters governing the
dimensions of the geometry; (f) boundary conditions applied.

drop, a thermal–hydraulic performance evaluation criteria (PEC) index 4. Numerical method, mesh, and code validation
[8,9] is used, which in this study is defined as:
Eqs. (1)–(4) were solved simultaneously using the finite element
software COMSOL Multiphysics version 4.4. A grid independence test
Ѳ=Ѳs was applied on both smooth channel and a VG-enhanced channel to as-
PEC ¼ ð7Þ
ðΔP=ΔP s Þ1=3 sess the density of mesh required. For the smooth channel, five meshes
were used and labelled as ‘coarser’, ‘coarse’, ‘normal’, ‘fine’, and ‘finer’,
with number of elements 53554, 145869, 283944, 788230, and
where ΔP and Ѳ are the pressure drop and thermal resistance in a 2831904 respectively, and Fig. 2 shows the average temperature calcu-
microchannel containing VGs and ΔPs and Ѳs are the same quantities lated using each mesh. As can be seen, the ‘fine’ mesh provided the best
in the corresponding smooth (i.e. uniform) micro-channel. run time whilst ensuring mesh independence of the solution. Therefore,
the standard fine ‘physics-controlled’ mesh was used as the base mesh.
3.2. Boundary conditions This was then modified to accommodate the VG micro-channel models,
and mesh sensitivity checked again with a sequence of increasingly
As indicated in Fig. 1(f), symmetry conditions were applied at the finer meshes ‘M1’ to ‘M5’. Using the maximum temperature as a
left- and right-hand outer boundaries of the domain, whilst a uniform measure, Fig. 3 confirms the suitability of mesh M1.
heat flux was applied at the bottom boundary. The top boundary was To ensure the accuracy of the results, two simulation models were
considered as adiabatic. At the inlet, the velocity of the flow was set in compared with previous numerical studies. The first validation model
terms of the specified Reynolds number using (5), and the inlet temper- was compared with Shkarah et al. [30], who modelled a graphene
ature was fixed at 293.15 K. At the outlet, the pressure was set to zero, micro-channel within a heat sink using the finite volume-based soft-
and on the micro-channel walls the no slip condition was applied. ware FLUENT. Figs. 4 and 5 show a comparison of the present

Table 2
Dimensions of the micro-channel and VGs models.

Micro-channel dimensions (μm)

L 25,000 Ht 900 Hc 700

Ww 300 Wc 500 xin 4000

VGs dimensions (μm)

Models Dimensions Front view of the

geometry

Full-span 500

Centred 300

Split 300 overall,

Fig. 2. Grid independence test for a smooth channel, showing the maximum temperature
gap 50–100
calculated using increasingly refined meshes, with q = 100 W/cm2 and Re = 600.
 M.T. Al-Asadi et al. / International Communications in Heat and Mass Transfer 74 (2016) 40–47 43

Table 3
Code validation

Angle of VG 45° 45°

Re 233 350
Heat flux 1 kW/m2 1 kW/m2
T (K) present 307 305
T (K) [31] 306.362 305.617

resistance was decreased as the Reynolds number increased due to

the increase of the velocity which leads to an enhanced heat transfer
rate, however the price paid is an increase in the pressure drop to
drive the faster flow.
Fig. 3. Grid independence test for micro-channel with VGs; q = 100 W/cm2, Re = 600. The results presented below focus on the particular value of
100 W/cm 2 for the heat flux, because this corresponds to the upper
limit for operation of electronic devices [17,32–34]. Before discussing
in detail the effects of the vortex generators, Fig. 6 shows a typical
plot of thermal resistance and pressure drop versus Reynolds number
for different heat flux values. As can be seen, the same trends are
observed in each case. Increasing the heat flux results in a very
slight improvement in the thermal resistance and a larger improvement
(i.e. reduction) in the pressure drop. This is typical of the results
found, so in what follows attention will be restricted to the case of
q= 100 W/cm2.

5.1. Quarter-circle VG models

Vortex generators with cross section shown in Fig. 1(c) and

spanwise configuration shown in Table 2 were inserted into the
Fig. 4. Code validation by comparison of calculated temperature distribution along a
graphene micro-channel with heat flux of 181 W/cm2 and flow rate of 8.6 cm3/s [30]. microchannel and the performance was analysed over a range of radii
and Reynolds numbers. The same trends were observed for the ‘full-
span’, ‘centred’ and ‘split’ VGs (see Table 2 for definitions), so two
COMSOL simulations with the published calculations, with the inlet models are considered here to present the effect of the VG radius on
temperature set at 300 K, the flow rate at 8.6 cm3/s, and the heat flux the conjugate heat transfer for various values of Reynolds number.
at two values: 181 and 277 W/cm2. Good agreement is seen between Fig. 7 shows the influence of radius ranging from 0 to 400 μm on the
the two approaches. The second validation was with Abdollahi and thermal resistance and pressure drop using full-span VGs at three rela-
Shams [31]. They investigated the impact on conjugate heat transfer of tively low Reynolds numbers. Introducing the VGs with radii r =
using rectangular vortex generators with various orientations and 100 μm and 200 μm raises the thermal resistance slightly, hence
their inclination in a channel, with water as the working fluid. The val- degrading performance, at all three values of Re. Increasing the radius
idation was done with the present study and showed good agreement, to 300 μm then reduces the thermal resistance slightly, but it is only at
using an angle for the VG of 45°, a heat flux of 1 kW/m2, and values Re = 500 that θ is reduced below its value for the uniform channel
for Reynolds number of 233 and 350 as shown in Table 3. with no VGs. Interestingly, a further increase of radius to 400 μm pro-
duces another slight increase in θ at Re = 700 and 900 and the same
5. Results and discussion performance at Re = 500, suggesting that there is a local minimum in
θ at around r = 300 μm. In all cases, the pressure drop increases with
Some 500 three-dimensional laminar flow simulations were con- VG radius, consistent with the increasing constriction of the flow.
ducted with Reynolds number in the range 100–2300 and heat flux Fig. 8(a) shows the impact of radius on the thermal resistance and
ranging from 100 to 300 W/cm2 to assess the impact of the vortex gen- the pressure drop at high Reynolds number ranging from 1100 to
erators described in Section 2 on the conjugate heat transfer, with water 2300 using ‘split’ VGs (see Table 2). Both the pressure drop and the
as the working fluid. As to be expected, it was found that the thermal

Fig. 5. Code validation by comparison of calculated temperature distribution along a Fig. 6. Thermal resistance and pressure drop as a function of Reynolds numbers for a
graphene micro-channel with heat flux of 277 W/cm2 and flow rate of 8.6 cm3/s [30]. micro-channel including half-circle VGs.
44 M.T. Al-Asadi et al. / International Communications in Heat and Mass Transfer 74 (2016) 40–47

Fig. 7. Thermal resistance and pressure drop for full-span quarter-circle VGs as a function
of radius at low Reynolds numbers.

thermal resistance increase as the radius increases. This can be attribut-

ed to hot spots which are behind each VG because of low velocity at
these areas, see example contour plots in Fig. 8(b). Hence it can be con-
cluded that VGs with a quarter-circle cross-section are not beneficial.

5.2. Half-circle VGs

The other VG geometry considered in this work is that of cylinders

with a half-circle cross-section as shown in Fig. 1(d). Again the three dif-
Fig. 9. Thermal resistance and pressure drop as a function of VG radius for full-span half-
ferent spanwise configurations shown in Table 2 were considered. Fig. 9 circle VGs: (a) Re = 100 to 900; (b) Re = 1100 to 1500.
presents the impact of the radius of full-span VGs on the thermal resis-
tance and pressure drop for a range of Reynolds numbers, which is split
into two plots for the sake of clarity. Unlike the quarter-circle VGs, for ΔP as a function of the VG radius for the centred VG, with the range of
the half-circle case, the thermal resistance decreases monotonically Reynolds number again split for clarity. Interestingly, there is a change
with radius for all Reynolds numbers above 100. The thermal resistance in behaviour as the VG radius is increased above 200 μm. For the
also decreases as Re increases, but of course the pressure drop increases. lower Re range, the rate of decrease of θ with r slows before increasing
Having seen that the full-span half-circle VGs appear to offer bene- again, producing something of a ‘kink’ in the plot at 200 μm. For the
fits, modifications to the VGs are now considered. First, the span of the higher Re range, there is also a noticeable change in the rate at which
VG is reduced to 300 μm and it is positioned centrally across the gap, θ changes with r, but in contrast this involves a more significant reduc-
creating the ‘centred’ VG shown in Table 2, which has a gap between tion in the thermal resistance above r = 200 μm. Comparing Fig. 10 with
the VG and the side walls of the micro-channel. Fig. 10 shows θ and Fig. 9, it is clear that the thermal resistance is lower for the centred VG at

a)

b)
r = 0µm

r = 100µm

r = 200µm

r = 300µm

r = 400µm

Fig. 8. Effect of using split quarter-circle VGs: (a) thermal resistance and pressure drop at high Reynolds numbers; (b) temperature contours at Re = 1100.
 M.T. Al-Asadi et al. / International Communications in Heat and Mass Transfer 74 (2016) 40–47 45

Fig. 12. Comparison of the thermal resistance and pressure drop characteristics of all
variants of half-circle VG considered, with radius equals to 200 μm.

Fig. 12 compares the thermal resistance and pressure drop observed

using each of the three half-circle VG designs, along with those of the
smooth, uniform channel with no VGs present. The VG radius in each
case is 200 μm. It is found that the lowest thermal resistance is achieved
with the centred VG, then the split, then the full-span VGs, and all the
VGs produce a lower thermal resistance than the uniform channel
with no VGs present. On the other hand, all VGs resulted in higher
pressure drops than the uniform channel, as is to be expected. The low-
est pressure drop with VGs present is seen with the centred, then the
split, and then the full-span VGs — i.e. the same ranking as for the
Fig. 10. Thermal resistance and pressure drop as a function of VG radius for ‘centred’ half- thermal resistance.
circle VGs: (a) Re = 100 to 900; (b) Re = 1100 to 2300.

high Re and larger radius. This is probably because the gap between the 5.2.1. Thermal–hydraulic performance
end of the VG and side wall of the channel allows heat transfer to To assess the real practical potential for improving the efficiency of
continue along the whole length of the side wall. micro-channel heat exchangers using VGs, it is important to combine
A second modification of the VG is to split the ‘centred’ VG in the the heat transfer performance with an appreciation of the fluid flow
middle, thus introducing an additional central gap as shown in the penalty resulting from the constriction of the channel. Clearly there is
‘split’ geometry row of Table 2. Fig. 11 shows the corresponding plots a huge design space available for the optimization of cylindrical VGs,
of thermal resistance and pressure drop against radius. In terms of ther- however a full optimization study is beyond the scope of this paper. In-
mal resistance, the introduction of the central gap negates the benefits stead, to illustrate the potential of such VGs, calculation of the perfor-
seen with the centred VG, and the behaviour is similar to that of the mance evaluation criteria (PEC) index, Eq. (7), is made just for the VG
full-span VG. However, the central gap does produce a reduction in type considered here that showed the lowest thermal resistance, i.e.
the pressure drop compared to the full-span VGs under equivalent the centred half-circle VG.
conditions. Fig. 13 plots the PEC index as a function of the radius of the centred
half-circle VG at Re = 200. It is found that the best radius is 30 μm under
those conditions, with the PEC reducing to unity at about r = 175 μm.
Fig. 14 extends the data to include a wider range of Reynolds number,
confirming that small-radius VGs offer good potential for improving
the efficiency of micro-channels operating at low Reynolds number,
with r = 30 μm appearing to give the best performance over the widest
range of Re. Though not shown in the figure, as either the VG radius or
the Reynolds number increases, eventually the increase in the pressure
drop due to the constriction of the channel starts to dominate and the

Fig. 11. Thermal resistance and pressure drop as a function of VG radius for split half-circle Fig. 13. Performance evaluation criteria index determined for the centred half-circle VGs
VGS: (a) Re = 100 to 900; (b) Re = 1100 to 2300. with different radii and Re = 200.
46 M.T. Al-Asadi et al. / International Communications in Heat and Mass Transfer 74 (2016) 40–47

Fig. 14. Variation of PEC index with Reynolds number for centred half-circle VGs of various radii.

PEC index falls below unity, indicating that the VGs are no longer bene- References
ficial in terms of the overall efficiency. [1] H.-Y. Li, et al., Enhancing heat transfer in a plate-fin heat sink using delta winglet
vortex generators, Int. J. Heat Mass Transf. 67 (2013) 666–677.
6. Conclusions [2] J. Albadr, S. Tayal, M. Alasadi, Heat transfer through heat exchanger using Al2O3
nanofluid at different concentrations, Case Stud. Therm. Eng. 1 (1) (2013) 38–44.
[3] M. Cheraghi, M. Raisee, M. Moghaddami, Effect of cylinder proximity to the wall on
Vortex generators (VGs) based on cylinders with quarter-circle and channel flow heat transfer enhancement, C.R. Mec. 342 (2) (2014) 63–72.
half-circle cross-sections attached to the base of a microchannel have [4] T. Dixit, I. Ghosh, Review of micro- and mini-channel heat sinks and heat ex-
changers for single phase fluids, Renew. Sust. Energ. Rev. 41 (2015) 1298–1311.
been explored as a means of enhancing the efficiency of micro-
[5] S.S. Mehendale, A.M. Jacobi, R.K. Shah, Fluid flow and heat transfer at micro- and
channel heat exchangers/heat sinks operating under thermal loads rel- meso-scales with application to heat exchanger design, Appl. Mech. Rev. 53 (7)
evant to microelectronics cooling. The effect of the VGs was quantified (2000) 175–193.
[6] S.G. Kandlikar, W.J. Grande, Evolution of microchannel flow passages: thermohydraulic
in terms of the thermal resistance, pressure drop along the channel,
performance and fabrication technology, Proceedings of the ASME International Me-
and a combination of these forming a performance evaluation criteria chanical Engineering Congress Exposition, 2002.
(PEC) index. The results showed that: [7] D.B. Tuckerman, R.F.W. Pease, High-performance heat sinking for VLSI, IEEE Electron
Device Lett. 2 (5) (1981) 126–129.
[8] O. Manca, S. Nardini, D. Ricci, A numerical study of nanofluid forced convection in
• The cylinders with a quarter-circle cross-section were not effective in ribbed channels, Appl. Therm. Eng. 37 (2012) 280–292.
reducing the thermal resistance, except marginally so when the VG [9] M.A. Ahmed, et al., Effect of corrugation profile on the thermal–hydraulic perfor-
radius is relatively large and the Reynolds number is relatively small. mance of corrugated channels using CuO–water nanofluid, Case Stud. Therm. Eng.
4 (2014) 65–75.
Under these conditions the increased pressure drop dominates, [10] C.-C. Wang, J.-S. Liaw, Air-side performance of herringbone wavy fin-and-tube heat
making the quarter-circle VGs infeasible. exchangers under dehumidifying condition — Data with larger diameter tube, Int. J.
• In contrast, for the VGs with a half-circle cross-section spanning the Heat Mass Transf. 55 (11–12) (2012) 3054–3060.
[11] L. Li, et al., Numerical simulation on flow and heat transfer of fin-and-tube heat
whole width of the microchannel, the thermal resistance was shown exchanger with vortex generators, Int. J. Therm. Sci. 92 (2015) 85–96.
to decrease monotonically with the VG radius. However, as expected [12] T. Johnson, P. Joubert, The influence of vortex generators on the drag and heat trans-
the pressure drop increased with radius as a result of the increasing fer from a circular cylinder normal to an airstream, J. Heat Transf. 91 (1) (1969)
91–99.
constriction in the flow.
[13] A. Ebrahimi, E. Roohi, S. Kheradmand, Numerical study of liquid flow and heat
• A modification of the full-span half-circle VG to produce a shorter, transfer in rectangular microchannel with vortex generators, Appl. Therm. Eng. 78
centred VG showed a greater reduction in the thermal resistance, (2015) 576–583.
[14] H.E. Ahmed, H.A. Mohammed, M.Z. Yusoff, An overview on heat transfer augmenta-
whilst also showing a smaller increase in the pressure drop, compared
tion using vortex generators and nanofluids: approaches and applications, Renew.
to the full-span VG. This is attributed to the gap between the ends of Sust. Energ. Rev. 16 (8) (2012) 5951–5993.
the VG cylinder and the side walls of the micro-channel, which allows [15] C. Min, et al., Experimental study of rectangular channel with modified rectangular
heat transfer to continue along the whole length of the side walls, and vortex generators, Int. J. Heat Mass Transf. 53 (15–16) (2010) 3023–3029.
[16] C. Habchi, et al., Enhancing heat transfer in vortex generator-type multifunctional
reduces the constriction of the flow. heat exchangers, Appl. Therm. Eng. 38 (2012) 14–25.
• Further modification to introduce a central gap in the VG negated the [17] A. Al-Damook, et al., An experimental and computational investigation of thermal
above benefit of the continuous centred VG, showing similar thermal air flows through perforated pin heat sinks, Appl. Therm. Eng. 89 (2015) 365–376.
[18] Y. Kai-Shing, et al., On the heat transfer characteristics of heat sinks: with and
resistance to the full-span VG. without vortex generators, IEEE Trans. Compon. Packag. Technol. 33 (2) (2010)
• Based on the PEC index, which offsets the improvement in thermal re- 391–397.
sistance against the pressure penalty, small-radius centred VGs offer [19] J.-S. Leu, Y.-H. Wu, J.-Y. Jang, Heat transfer and fluid flow analysis in plate-fin and
tube heat exchangers with a pair of block shape vortex generators, Int. J. Heat
good potential for improving the efficiency of micro-channels, partic- Mass Transf. 47 (19–20) (2004) 4327–4338.
ularly those operating at lower Reynolds number, e.g. Re b 600. The [20] J.M. Wu, W.Q. Tao, Effect of vortex generator on heat transfer in rectangular chan-
maximum PEC was found for a radius of 30 μm, which may prove chal- nels, Appl. Therm. Eng. 37 (2012) 67–72.
[21] M. Khoshvaght-Aliabadi, S. Zangouei, F. Hormozi, Performance of a plate-fin heat
lenging for manufacture, but good potential is also seen for radii up to
exchanger with vortex-generator channels: 3D-CFD simulation and experimental
150 μm in the cases considered here. validation, Int. J. Therm. Sci. 88 (2015) 180–192.
[22] J. Li, et al., Numerical study on a slit fin-and-tube heat exchanger with vortex
generators, Int. J. Heat Mass Transf. 54 (9–10) (2011) 1743–1751.
A full, formal optimization study has not been conducted here, but
[23] S. Chomdee, T. Kiatsiriroat, Enhancement of air cooling in staggered array of
the base-mounted cylindrical vortex generators offer a wide design electronic modules by integrating delta winglet vortex generators, Int. Commun.
space for further exploration and optimization. Heat Mass Transfer 33 (5) (2006) 618–626.
 M.T. Al-Asadi et al. / International Communications in Heat and Mass Transfer 74 (2016) 40–47 47

[24] C. Liu, et al., Experimental investigations on liquid flow and heat transfer in rectan- [30] A.J. Shkarah, et al., A 3D numerical study of heat transfer in a single-phase micro-
gular microchannel with vortex generators, Int. J. Heat Mass Transf. 54 (13–14) channel heat sink using graphene, aluminum and silicon as substrates, Int.
(2011) 3069–3080. Commun. Heat Mass Transfer 48 (2013) 108–115.
[25] J. Ma, et al., Experimental investigations on single-phase heat transfer enhancement [31] A. Abdollahi, M. Shams, Optimization of shape and angle of attack of winglet vortex
with longitudinal vortices in narrow rectangular channel, Nucl. Eng. Des. 240 (1) generator in a rectangular channel for heat transfer enhancement, Appl. Therm. Eng.
(2010) 92–102. 81 (2015) 376–387.
[26] G.D. Xia, et al., Effects of different geometric structures on fluid flow and heat trans- [32] X. Yu, et al., Development of a plate-pin fin heat sink and its performance compar-
fer performance in microchannel heat sinks, Int. J. Heat Mass Transf. 80 (2015) isons with a plate fin heat sink, Appl. Therm. Eng. 25 (2–3) (2005) 173–182.
439–447. [33] S.P. Gurrum, et al., Thermal issues in next-generation integrated circuits, IEEE Trans.
[27] C. Chen, et al., A study on fluid flow and heat transfer in rectangular microchannels Device Mater. Reliab. 4 (4) (2004) 709–714.
with various vortex generators, Int. J. Heat Mass Transf. 69 (2014) 203–214. [34] W. Yuan, et al., Numerical simulation of the thermal hydraulic performance of a
[28] H. Mirzaee, et al., Heat transfer enhancement in microchannels using an elastic plate pin fin heat sink, Appl. Therm. Eng. 48 (2012) 81–88.
vortex generator, J. Enhanc. Heat Transf. 19 (3) (2012) 199–211.
[29] J. Wang, Y. Zhao, Heat and fluid flow characteristics of a rectangular channel with a
small diameter circular cylinder as vortex generator, Int. J. Therm. Sci. 92 (2015)
1–13.