Applied Thermal Engineering 120 (2017) 257–268

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Research Paper

Heat transfer enhancement in annular flow with outer grooved cylinder

and rotating inner cylinder: Review and experiments
A. Nouri-Borujerdi ⇑, M.E. Nakhchi
School of Mechanical Engineering, Sharif University of Technology, Tehran, Iran

h i g h l i g h t s

Among four aspect ratios, heat transfer with aspect ratio of b/c = 2 is the highest.

The heat transfer is increased up to 40–50% when the inner cylinder is slotted.

Heat transfer for outer grooved cylinder is 10% lower than that of the inner one.

Thermal entrance region reduces up to 17% in the presence of grooves. .

At low Taylor numbers Ta < 1700 the grooves have negligible effect on heat transfer.

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

Article history: This experimental work is an attempt to identify the most significant parameters influencing heat
Received 7 September 2016 transfer enhancement in annular flow with outer grooved cylinder and rotating inner cylinder. This type
Revised 3 March 2017 of flow known as Taylor-Couette flow has many industrial applications such as electrical power genera-
Accepted 20 March 2017
tors and rotating machineries. A comprehensive review of the previous works on flow regimes, heat
Available online 21 March 2017
transfer and pressure drop in grooved channels is provided. The experiments are conducted based on four
different factors including Taylor number, 0 < Ta < 8:36  106 ; groove aspect ratio, 0 < b=c < 2; number
Keywords:
of grooves, 0 < Ngr < 20 and wall temperature, 50 < T w < 90  C. Among four different test cases, the
Electric motor
Grooved channel
average Nusselt number of the model with aspect ratio b/c = 2 is higher than the other three cases, i.e.,
Response surface methodology b=c ¼ 0; 0:5; 1. It is found that in the presence of grooves, the thermal entrance length in rotating channel
Heat transfer enhancement is reduced about 17% due to perturbation and vorticity of the flow inside the channel. Finally, the results
are analyzed according to the principle of response surface methodology. It is found that, heat transfer
enhances significantly with increasing groove aspect ratio and rotational speed.
Ó 2017 Elsevier Ltd. All rights reserved.

1. Introduction and literature review
Many engineering applications involve rotating machinery
components including electrical motors, rotating heat pipes and
turbogenerators [1–4]. The fluid flow between two concentric
cylinders with one or both cylinders rotating is called Taylor–Cou-
ette (TC) flow and has been studied extensively over the years [5–
9]. Different types of flow regimes of Taylor–Couette flow was
investigated by Lasagna et al. [10] and Nemri el al. [11]. The effect
of large aspect ratios on heat transfer enhancement of Couette-
Taylor flow was numerically investigated by Kang et al. [12]. They
found that the radial heat transfer increases with Grashof number.
If an axial flow is superimposed on the rotation, it is named Taylor-
⇑ Corresponding author.
E-mail addresses: anouri@sharif.edu (A. Nouri-Borujerdi), erfanian@mech.sharif.
ir (M.E. Nakhchi).
Couette-Poiseuille (TCP) flow. This configuration is also popular in
cooling of electric machineries [13]. Neto et al. [14] numerically
investigated the turbulent flow in concentric annular channel with
rotating inner cylinder. They concluded that their numerical
approach for the axial and radial velocity profiles is in good agree-
ment with experimental results. Structure and heat transfer of
these kinds of flows, have been the subject of several reviews.
The convection heat transfer and fluid flow in annulus was com-
prehensively reviewed by Davood et al. [15] and Togun et al.[16].
Heat transfer between concentric rotating cylinders was reviewed
by Fenot el al. [17]. they also compared the results of many numer-
ical and experimental studies for different geometrical parameters
(air gap thickness and radius to length ratio) and physical param-
eters (rotational and axial velocities) on heat transfer enhancement
between rotating cylinders.
Temperature control is a key factor affecting the efficiency
and capacity of electric motors and generators [18–21]. Different

http://dx.doi.org/10.1016/j.applthermaleng.2017.03.095
1359-4311/Ó 2017 Elsevier Ltd. All rights reserved.
258 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268

Nomenclature

A surface area, mm2 Greek symbols

a groove pitch, mm a thermal diffusivity, m2 =s
b groove depth, mm b thermal expansion coefficient, 1=K
c groove width, mm m kinematic viscosity, m2 =s
D diameter, mm q density, kg=m3
e air gap between two cylinders, mm X rotational speed, rad/s or rpm
g gravitational acceleration, m=s2
h heat transfer coefficient, W=m2 K Subscripts
L length, mm 1 inner
N number 2 outer
Nu Nusselt number a axial, ambient
p pressure, N=m2 c convective heat transfer
Pr Prandtl number e effective
Q heat transfer rate, W gr groove
r radius coordinate h hydraulic
R cylinder radius, mm l lost
Re axial Reynolds number r radiation
T temperature, °C s surface, smooth
Ta Taylor number w wall
U axial velocity component, m/s
V peripheral velocity component, m/s
Y axial coordinate
z z-direction

cooling methods and the importance of heat transfer enhancement Most of the studies on rotating annular channels (especially
in different types of electric motors was investigated by Malumbres electric motors) have focused on determining the flow structures,
et al. [22] and Wen and Zheng [23]. Huai et al. [18] numerically heat transfer coefficient and pressure drop, however, the research
investigated the temperature rise in fan-cooled induction motors. on the effect of groove geometry and the number of grooves is
They show that temperature control is an important factor affecting rarely reported. Section 1.1 provides details regarding the previ-
the efficiency of the overall design. Kim et al. [24] performed a ous studies on fluid flow structures inside rotating grooved annu-
numerical analysis to analyze the effects of air-gap fan on heat lus. The convective heat transfer is discussed in Section 1.2,
transfer enhancement in high speed induction motors. They found including TC flow (Section 1.2.1) and TCP flow (Section 1.2.2).
that the heat transfer coefficient at the air gap was increased up to Pressure drop in the presence of axial flow is reviewed in
90% in the presence of axial air gap fans. One of the techniques that Section 1.3.
is used for saving energy and increasing heat transfer of air flow
inside electric motors is mounting axial grooves on one or both
1.1. Flow structures
cylinders [25,26]. Fenot et al. [5] conducted an experimental study
on the effect of axial flow on heat transfer enhancement inside four-
Few studies have focused on the fluid structures inside rotating
pole synchronous motors. Tzeng et al. [27] experimental investi-
grooved annulus and transition from laminar to wavy vortex and
gated the heat transfer enhancement inside high-speed rotating
turbulent flow, while most of them have concentrated on the influ-
machineries. They found that mounting different types of axial ribs
ence of axial flow on vortex appearance.
improves the heat transfer in rotating devices.
Tachibana and Fukui [33] have compared the critical Taylor
The effect of different types of grooves on heat transfer
number of smooth cylindrical gaps with the critical Taylor number
enhancement in grooved channels was investigated by different
at grooved annular gaps. The number of grooves on the inner
researchers. In the work of Hayase et al. [28] the authors conducted
cylinder and the diameter of the cylinders have not been specified
a numerical study on heat transfer inside grooved channel corre-
in their work. The groove geometries used in their analysis were
sponds to an electric motor. They show that heat transfer is raised
10  3 mm. The inner cylinder was heated while the outer one
by a factor of 1.2 in the case of inner grooved cylinder and by a fac-
remained unheated. The authors show that when slots are present,
tor of 1.1 in the case of outer grooved cylinder. Bilen et al. [29]
the transition toward flow with Taylor vortices is delayed:
experimentaly studied the effect of groove geometry on heat
TaC = 1680 if the rotor is smooth, and TaC = 6400 in grooved
transfer and pressure drop of a fully developed turbulent air flow
channel. Numerically, Hayase et al. [28] have worked with
in different grooved channels. Lorenzini-Gutierrez et al. [30]
grooved configurations, and they concluded that the higher the
numerically and experimentally investigated the effect of curved
rotation velocity, the more pronounced the interaction. Three-
flow deflectors on heat transfer enhancement in grooved channels.
dimensional flow fields with supercritical and subcritical Reynolds
Liu et al. [31] investigated the effect of rounded axial grooves on
numbers were also investigated.
heat transfer and pressure drop inside grooved channels. They con-
For the case of TC without axial flow, five different types of flow
cluded that pressure drop significantly decreases in the presence of
have been recorded by Lee et al. [34] as follows;
rounded axial grooves instead of square grooves in grooved
channels. Joo and Kim [32] optimized the geometry of vertically

circular Couette flow
oriented, internally grooved tubes in natural convection. Flow

Taylor vortex flow
structures and heat transfer in both smooth and grooved annular

wavy vortex flow
channels in the presence of rotation depend on Taylor number

modulated wavy vortex flow
Ta ¼ x2 R1 ðDh =2Þ3 =m2 .
turbulent Taylor vortex flow
 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268
The range of ReX varied from 52 to 2863 and the number of
grooves on the outer cylinder varied from 0 to18. The stator placed
in a constant temperature bath and the rotor remained unheated.
The groove geometries of their experimental work were
5  5 mm .They also found that the number of grooves significantly
reduces the rotational Reynolds number for transition from wavy
vortex flow to modulated wavy and turbulent vortex flow. In the
case of 18 grooves, transition from wavy vortex flow to turbulent
vortex flow occurs at ReX ¼ 143 while it occurs at ReX ¼ 1909 for
the case of smooth annular channel. In grooved channels, vortices
expanded into the axial groove space and vorticity became weaker
as the number of grooves decreased.
The effect of radial temperature gradient on Taylor–Couette
flow with axial wall grooves was experimentally investigated by
Liu et al. [35]. In their work, the number of grooves varied from
0 to 18. They found that the transition process of a larger number
of grooves (9 and 18) was accelerated due to the number of
grooves and when the temperature gradient became larger, the
critical Reynolds number of the transition process decreased. They
found 6 different flow regimes using a PIV technique: single vortex
flow regime; helical vortex flow regime; helical wavy vortex flow
regime; random helical vortex flow regime; wavy vortex flow
regime and turbulent Taylor vortex flow regime. Liu et al. [36]
numerically investigated the effect of different number of axial
grooves on flow regimes in rotating annular channels. Their analy-
sis had a good agreement with experimental results.
It should be noted that the effect of width to depth ratio of the
grooves and its effect on the flow transition have not been studied
in the above articles also the definition of critical Taylor number is
quite different from one author to another and so, it is difficult to
compare the different studies.
1.2. Convective heat transfer
1.2.1. Taylor-Couette flow
Improved heat transfer is very important for the energy effi-
ciency, higher performance, and sizereduction of a fluid flow in a
grooved channel [37]. Gardiner and Sabersky [38] have experimen-
tally investigated heat transfer in annular gap in the range of Tay-
lor numbers up to about 106, and the range of Reynolds numbers
based on the axial velocity components and the gap distance up
to 7000. Their experimental results showed that presence or
absence of the groove has no effect on heat transfer. Eiamsa-ard
and Promvonge [39] have reported that all groove geometries sig-
nificantly enhance the heat transfer rate in comparison with the
smooth channel. Zheng et al. [40] have investigated the effect of
groove inclination angle on heat transfer enhancement in inter-
nally grooved channels. They also proposed a new groove geome-
try for heat transfer enhancement in flow channels.
In another study, Bouafia et al. [41] have performed an experi-
mental study for heat transfer inside grooved channels. In their
work Nusselt number was defined based on temperature differ-
ence between rotor and stator. It was observed that Nusselt num-
ber increases slightly in the presence of grooves. They found the
following correlation:
Nu ¼ 0:132Ta0:3 for 6000 < Ta < 1:4  106
ð1a-bÞ
Nu ¼ 0:029Ta0:4 for 1:4  106 < Ta < 2  107
Hayase et al. [28] have numerically simulated heat transfer
inside rotating grooved channel with uniform wall temperature.
Their three-dimensional calculations revealed a spanwise-
dependent Taylor vortex flow in the annulus. These vortices inter-
acted with the recirculating flow in the cavity. They have also
noted in their study that Nusselt number is raised by a factor of
259
1.2 in the case of cavities embedded in the inner cylinder and by
a factor of 1.1 in the case of cavities embedded in the outer
cylinder.
According to Gazley [42], heat transfer in annular grooved chan-
nels reduces in comparison with smooth channels. (from 10% when
inner cylinder is grooved, to 20% when both cylinders are grooved).
In the case of turbulent flow inside grooved channel, a small rise of
the Nusselt number was observed. It is necessary to point out that
in the study of Gazley, the grooves were all insulated and the sur-
face area of heat transfer was small.
1.2.2. Taylor-Couette-Poiseuille flow
Several studies have been conducted on heat transfer in TCP
flows but their results have some conflicts with each other. Most
of researchers have observed heat transfer enhancement in the
presence of grooves.
Lee and Minkowycz [26] have performed an experimental inves-
tigation to determine the heat transfer coefficients in a Taylor–Cou
ette–Poiseuille problem in the range of Taylor numbers (Ta), from
103 to 2  107 with the axial Reynolds numbers Rea from 50 to
1000. In their tests, the concentric cylinders were either smooth
or one was smooth and the other was grooved. They found that
the Nusselt number increases with increasing axial Reynolds num-
ber at low Taylor numbers. They have also found that the presence
of grooves on the rotating inner cylinder significantly increases
Nusselt number, over the smooth cylinder. The groove geometries
of their experimental work were 10:3  10:3 mm, 10:3  7:8 mm
and 10:3  5:3 mm. Moreover, as shown by Gazley [42], effective
Reynolds number ðRee Þ is a good way to describe heat transfer
inside grooved channels in the presence of axial flow and rotation.
This parameter is a combination of radial and axial velocities as:
1=2
½U 2 þ 0:5ðR1 XÞ2  Dh
Ree ¼ ð2Þ
m
Numerical simulation was carried out by Sommerer and Lauriat
[43] in a study devoted to the heat transfer in an annulus with rect-
angular grooves embedded in the outer cylinder. They found that
increasing the number of grooves from 0 to 25 will significantly
increase the mean Nusselt number, but their analysis was
restricted to very low Taylor numbers ðTa < 1250Þ. Entry region
effects on heat transfer of annuli with rotating inner grooved cylin-
der was experimentally investigated by Fénot et al. [5]. They found
local heat transfer on both cylinders using thermography device for
2140 < Rea < 6425 and 104 < Ta < 4  106 . They found the follow-
ing correlation for rotor and stator as:

ReX 105 0:95
Y
Nus ¼ 0:09Re0:84
e
Dh
 0:44 ð3a-bÞ
Y
Nur ¼ 0:22Re0:62
e
Dh
Jeng et al. [25] have investigated the heat transfer enhancement
of Taylor–Couette–Poiseuille flow in an annulus by mounting four
longitudinal grooves on the rotating inner cylinder. The authors
have also investigated the effects of axial Reynolds numbers from
30 to 1200 and rotational Reynolds number from 0 to 2922 on
the convective heat transfer. The groove geometry was
10  3 mm and the Nusselt number of the system raised by a factor
of 1.4 in the case of mounted grooves. They presented the relation-
ship between the Nusselt number and the rotational Reynolds
number as:
NuX
¼ 1 þ aReXb ð4Þ
NuX¼0
where a and b are constants.
260 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268
According to Lancial et al. [44] who presented an experimental
investigation of convective heat transfer in an annular grooved
channel corresponding to a salient pole hydro generator, near
the entrance region, a better cooling can be observed. Moreover,
the cooling is higher in the groove leading edge, especially for
the higher Taylor numbers. Abou-Ziyan et al. [45] have carried
out experimental investigation on heat transfer and pressure drop
inside annulus with inner finned cylinder in Taylor–Couette–Poi
seuille flow. Their results showed that Nusselt number enhances
by a factor of 7.5 in the presence of helical fins at
Rea ¼ 1:5  105 and X ¼ 400 rpm. They also found the heat trans-
fer increases with increasing axial Reynolds number with low fin
spacing. A summary of the numerical and experimental
researches done on heat transfer inside rotating grooved annulus
is presented on Table 1.
To summarize the previous observation on the heat transfer of
Taylor-Couette (TC) and Taylor-Couette-Poiseuille (TCP) flows in
grooved annular channels;
i. Heat transfer enhancement is observed in the presence of
grooves inside rotating annular channels by nearly all the
previous works done by researchers.
ii. Most of studies are focused on mounting grooves on inner
cylinder and less attention is attracted to outer stationary
grooved cylinders.
iii. Nearly about 2/3 of experimental and numerical studies are
done on TCP flows and just 1/3 of them are focused on TC
flows
iv. The number of grooves varied from 3 to 48, but it is not yet
concluded that the number and aspect ratios of grooves how
effects the heat transfer exactly.
v. The range of flow parameter conditions are: 0 < Rea <
1:5  105 and 0 < Ta < 2:15  109 . All of researchers
conducted heat transfer increment by increasing axial and
rotational speed of fluid flow.
The main challenge in researching fluid flows inside grooved
annuli lies with the groove geometry optimization in order to
enhance heat transfer coefficient, thus, a need exists for a thorough
study to be conducted on such specially designed grooves covering
aspects of design and optimization, the fluid flow and pressure
drop using experimental methods.
1.3. Pressure drop
Some researchers have stated that the rotation effects increase
the pressure drop inside annular channels [26,38,48]. This param-
eter has an important effect on axial flow rate and cooling
efficiency of electric motors and generators. Consequently, the
pressure drop must be included in the TCP flow analysis.
In the work of Lee and Minkowycz [26], the authors conducted
an experimental work on pressure drop inside grooved channels.
Pressure drop measurements were made at different flow rates
between the atmosphere and the downstream location of the test
section in presence of axial flow. They found that the pressure drop
would increase by order of magnitude of 21.8 with axial Reynolds
number. It was observed that friction factor inside grooved channel
increases with increasing rotational speed (Taylor number) and
decreases with axial flow rate (axial Reynolds number).
Chong et al. [49] investigated the effects of rotation on pressure
drop of the flow passing through the rotor-stator gap of radial flux
electrical machine. They conducted a correlation for stationary
pressure loses as DP s ¼ 276392Q 2 þ 2229:7Q where Q is the inlet
flow rate. It is shown by the authors that the flow resistance caused
by rotation can be significant and ignoring rotational pressure loss
may lead to overestimation of the cooling performance achievable
by an electric machine.
Abou-Ziyan et al. [45] presented an experimental investigation
of pressure drop and heat transfer inside rotating annular channel
with interrupted helical fins. They found the following correlation
for friction factor
 0:932  0:301
H t
f ¼ 0:000665Re0:979
a Ta0:2 ð5Þ
s s
where H=s and t=S are dimensionless geometrical parameters. It can
be observed that friction factor increases with Reynolds number
and Taylor number.
In summary, there are plenty of researches about heat transfer,
pressure drop and fluid structures inside electric motors and turbo
generators, however, the research on the number of stator grooves
and groove aspect ratio is not reported yet. The motivation is to fill
the knowledge gap that has been identified through the review
work done on these kinds of flows inside grooved annuli. A series
of verification and validation experiments is presented to verify
the experimental results. The effect of rectangular groove geome-
try on thermal entrance region and local heat transfer coefficient
is presented. The most significant geometrical and physical param-
eters are also determined using response surface methodology.
2. Test apparatus and procedure
2.1. Apparatus and equipment for experiments
The experimental test section is made of inner rotating cylinder
placed in a stationary outer grooved cylinder. The rotor is made of
aluminum material with a length of 191.6 mm and a diameter of
88.2 mm. The outer grooved cylinder is made of 0.2 mm copper
sheet with axial grooves. The diameter of outer cylinder is
95 mm and a length of 191.6 mm. The outer grooved cylinder is
heated by hot water at constant temperature while the rotor
remains unheated. After increasing water temperature in reservoir
with heating elements, the hot water is pumped into test section
bath. The temperature of outer grooved cylinder is controlled by
a thermostat varying from 50 to 90  C. The experimental setup is
shown in Fig. 1. The surface temperature at 20 different points is
recorded using K-type surface mounted thermocouples. The shaft
is driven by a belt and pulley system connected to an electrical
motor. The rotation speed is controlled by an inverter which is
measured by a tachometer. For the thermocouples calibration,
two reference temperatures are maintained using ice bath and
boiling water [50]. Moreover, a mercury-based thermometer has
been immersed in reference temperature baths to ensure the accu-
racy of calibration. The outer surface of test section is thermally
insulated using a cellular rubber layer, which is 35 mm thick to
minimize the heat exchange to the environment.
The inlet ambient air temperature is measured by one thermo-
couple placed in the beginning of the entrance section, while the
outlet air temperature is measured by two K-type thermocouples
located in the test section exit. The local ambient air temperature
is calculated by 20 thermocouples inserted in the air gap between
two cylinders at equal distances in the axial direction. Every tem-
perature measuring point is fixed by thermal glue (HC-131). Fig. 2
shows the details of surface thermocouple measurement system.
The schematic of test section is shown in Fig. 3. The inner cylin-
der is rotating while the outer grooved cylinder is stationary with
constant wall temperature.
Table 2 presents the dimensions of four test cases used in this
study. The dimensions are chosen in a way that dimensionless
parameters are available in many engineering applications such
as a simple electrical motor.
Table 1
Summary of the numerical and experimental researches done on heat transfer inside rotating grooved annulus.

Authors Flow type Number of grooves Ranges Findings

Gardiner and Sabersky, [38] 1978, TC-TCP-Laminar- Inner grooves 30 L=Dh ¼ 610 Ta ¼ 0—106
The convective heat transfer is increased from 40 to 50%, when the rotor is
Exp. work Turbulent R1 =R2 ¼ 0:981 Rea ¼ 0—7000 slotted.
e=R1 ¼ 0:019 b=c ¼ 0:5
At low Taylor numbers before the onset of vortices the slots have little effect
and free convection still dominates the heat-transfer process.
Lee and Minkowycz [26] 1989, TCP-Turbulent Inner grooves 20 L=Dh ¼ 4:9; 9:48; 14:1
The presence of grooves on the rotating inner cylinder significantly increases
Exp.work Outer grooves 20 Ta ¼ 103  2  107 R1 =R2 ¼ 0:983 Nu, over the smooth cylinder.
Rea ¼ 50—1000
In grooved cylinders, the length of dynamic flow development is small and
b=c ¼ 0:51; 0:78; 1 negligible for high Taylor numbers.

A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268

Hayase et al. [28] 1992, Num. work
Bouafia et al. [41] 1997, Exp. work
Sommerer and Lauriat [43] 2001,
Num. work
Gilchrist et al. [46] 2005, Exp. work
Jeng et al. [25] 2007, Exp. work
Fenot et al. [5] 2013, Exp. work
Lancial et al. [44] 2014, exp.& num.
work
Zhu et al. [47] 2016, Num. work
Abou-Ziyan et al. [45] 2016, Exp.
work
TC-Laminar-Turbulent Inner grooves 12 L=Dh ¼ 5
Outer grooves 12 ðR2 þ bÞ=R2 ¼ 1:2
R1 =R2 ¼ 0:95
ReX ¼ 65—260
TC-TCP-Turbulent Outer grooves 48 L=Dh ¼ 49:2 R1 =R2 ¼ 0:965
e=R1 ¼ 0:036 b=c ¼ 1:81
Ta ¼ 103  4:8  107
Rea ¼ 4400  1:7  104
TCP-Laminar Outer grooves 3–25 L=Dh ¼ 5 Ta ¼ 100—1250
R1 =R2 ¼ 0:99—0:90
ðR2 þ bÞ=R2 ¼ 1:01—1:4
TCP-Turbulent Inner grooves 9 Ta ¼ 106  5  107
Rea ¼ 900—2100
R1 =R2 ¼ 0:833
e=R1 ¼ 0:2 b=c ¼ 1
TCP-Laminar-Turbulent Inner grooves 4 L=Dh ¼ 8:57 ReX ¼ 0—2922
R1 =R2 ¼ 0:895 Rea ¼ 30—1200
e=R1 ¼ 0:23 b=c ¼ 0:3
TCP-Turbulent Inner grooves 4 (slots and L=Dh ¼ 156:25 Ta ¼ 104  4  106
poles) R1 =R2 ¼ 0:989 Rea ¼ 2140  6425
e=R1 ¼ 0:01 c ¼ 53:2 mm
TCP-Laminar-Turbulent Inner grooves 10 L=Dh ¼ 9:3 Ta ¼ 0  1:7  107
R1 =R2 ¼ 0:922 Rea ¼ 0  14250
e=R1 ¼ 0:084  =c ¼ 0:36
TC-Laminar-Turbulent V-shaped grooves L=Dh ¼ 1 Ta ¼ 0  2:15  109
R1 =R2 ¼ 0:714
TCP-Laminar-Turbulent Inner grooves (Helical fins) L=Dh ¼ 40 Ta ¼ 0  1:7  107
R1 =R2 ¼ 1=3
Rea ¼ 8  104  1:5  105 e=R1 ¼ 2

For high Reynolds numbers, the flow in a groove interacts with Taylor vor-
tices in the annular space to enhance heat transfer.

Heat transfer enhancement in annular channel with inner grooved cylinder
is about 10% higher that outer grooved cylinder

Nusselt number is increased up to 14% in the presence of grooves.

A new correlation based on effective Reynolds number is proposed.

Correlation describing the influences on the groove geometry both on the
Nusselt number is derived by using a design of experiments.

Nusselt number is increased by decreasing the number of grooves

Heat transfer increases from 5 to 40% in the presence of grooves.

The effect of rotation speed on heat transfer is small and negligible for high
Taylor numbers.

Heat transfer increases by a factor of 1.4 in the case of ribs mounted on the
inner cylinder.

The Nusselt number for the rotating inner cylinder is almost minimal at the
inlet of the annular channel and then sharply rises in the axial direction.

Average heat transfer on the rotor increases until Ree ¼ 26; 270. At that
point, the average heat transfer becomes constant or decreases slightly.

Heat transfer decreases from the entrance for both rotor and stator.

It is observed that a hot spot is located near the downstream end of the rotor
notch.

The heat transfer coefficient on the pole face and in the notch region
decreases with increasing z due to the growth of thermal boundary layers

The boundary layer thickness becomes smaller than the groove height with
increasing Ta.

Effective scaling laws between the Nusselt number and Taylor number in
smooth and grooved cases are presented.

The heat transfer increases with Rea , Ta and low fin spacing.

The attachment of interrupted helical fins enhances Nusselt number by 2–3
folds than Nu for smooth channel.

261
262 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268

Fig. 1. Schematic diagram of the experimental setup.

2.2. Data reduction

The experimental results are presented in terms of four princi-

pal dimensionless parameters; Rotational Reynolds, Taylor, Gra-
shof and Nusselt number.
Rotational Reynolds number ðReX Þ and Taylor number (Ta) are
defined by [5]:
V r Dh XR1 Dh
ReX ¼ ¼ ð6Þ
mair mair
 3
X2 R1 Dh
Ta ¼ ð7Þ
m2air 2

The averaged convective heat transfer coefficient h can be cal-

Fig. 2. Schematic of surface temperature measurement system.
Fig. 3. Schematic of the test section.
culated based on the surface temperature and average bulk air
temperature as h ¼ Q c =AðT w  T a Þ where Q c is the convective heat
transfer rate and A ¼ N gr ð2b þ a þ cÞL is the surface area. The aver-
aged Nusselt number is Nu ¼ hDk h and for Q c ¼ Q t  Q r  Q l .
In which the total heat transfer rate Q t is calculated as
Q t ¼ mwater cp DT water . The water temperature at the inlet and outlet
of test section is measured via TT-T-30SLE high-precision thermo-
couples ð0:1  CÞ. The volumetric flow rate of water is obtained by
measuring the required time to fill a secondary reservoir tank at
the outlet of the test section. It should be mentioned that the water
temperature difference at the inlet and outlet of test section is less
than 3  C. Radiation heat transfer ðQ r Þ is negligible due to polished
surface of inner and outer cylinders. Lost heat transfer ðQ l Þ takes
into account heat transfer to the environment in both axial and
radial directions. It is assumed to be zero, due to insulation of test
section.
The local heat transfer coefficient hðzÞ and the local Nusselt
number NuðzÞ are evaluated by the following equations
respectively;
Q c ðzÞ
hðzÞ ¼ ð8Þ
AðT w ðzÞ  T a ðzÞÞ
 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268 263

Table 2
Dimensions of present experimental study.

Parameter L ðmmÞ N gr R1 ðmmÞ R2 ðmmÞ b ðmmÞ c ðmmÞ b=c X ðrpmÞ

Case I 191.6 0 44.1 47.5 0 0 0 0–1740
Case II 191.6 10 44.1 47.5 7 14 1/2 0–1740
Case III 191.6 20 44.1 47.5 7 7 1 0–1740
Case IV 191.6 20 44.1 47.5 14 7 2 0–1740
Case V 191.6 20 44.1 47.4 7 14 1/2 0–1740

hðzÞDh in uniform wall temperature condition. It is observed that Nusselt

NuðzÞ ¼ ð9Þ
k number increases with Taylor number.
A comparison of experimental data with averaged Nusselt num-
where z is the axial position in the air gap between two cylinders. ber predicted by [55] for the case of Ta = 0 is made at Fig. 5. In the
The axial distance between the bottom and the top of air gap is absence of rotation and axial fluid flow, natural convection is the
divided into 20 equal distances. only mechanism of heat transfer involved in grooved channels. It
can be observed that experimental results have a good agreement
with predicted values of averaged Nusselt number.
3. Results and discussions

3.1. Uncertainty analysis 3.3. The effect of groove geometry

Thermocouple calibration is achieved with a 0:1  C tempera-
ture offset. It should be noted that the temperature measurement
errors due to axial heat gradients are considered negligible at mid-
way of test section. The rotational speed measurement is measured
by a maximum error of 1%. The uncertainty of different parameter
values is evaluated following the procedure suggested by [51,52] .
The maximum experimental uncertainty of surface area is around
0.17%. The uncertainties in the Taylor number, heat rate, averaged
Nusselt number and local Nusselt number are estimated to be
8.34%, 5.25%, 9.28% and 16.58% respectively. All these uncertainty
values have been reported at a 95% confidence level.
3.2. Validation experiments for experimental setup
In order to make a check of the reliability of the experimental
apparatus, the first part of the results is focused on heat transfer
inside smooth annular channels with or without rotation. Fig. 4
illustrates a comparison between the results of the present exper-
imental work for round channel (b/c = 0) and data of Tachibana
et al. [33], Bjorklund and Kays [53] and Aoki et al. [54]. The com-
parison shows an excellent agreement between present study
and other literature for averaged Nusselt number of TC flow
between two smooth cylinders. The outer cylinder is maintained
One of the purposes of this study is to see the effect of axial
grooves on enhancement of natural convection in rotating grooved
annular channels. Fig. 6 shows a comparison between the present
results for mean Nusselt number ratio with experimental results
and correlation of Tzeng [56] for smooth annular channels with
respect to 1012 GrReX .The natural convection effects (Grashof
number) are included. It can be observed that Grashof number is
an effective parameter in TC flows. Fig. 6 also shows NuX =NuX¼0
for grooved channel. It is shown that the heat transfer ratio
increases with b/c. The surface area increases with increasing the
groove aspect ratio, therefore heat transfer enhances due to natural
convection and rotation of inner cylinder in the presence of axial
grooves.
The effect of groove aspect ratio on averaged heat transfer coef-
ficient with respect to Taylor number is depicted in Fig. 7. Heat
transfer slightly increases from b/c = 0 to b/c = 2. The range of Tay-
lor number is between 3:95  105 and 5:53  106 . The outer cylin-
der wall temperature is maintained at constant value of T w ¼ 70  C.
When the b/c increases from 0.5 to 2, the grooves become deeper
and the fluid flow inside the groove is less affected under the influ-
ence of the rotational flow. In other words, the formation of

Fig. 5. Comparison of experimental results for the case of natural convection

Fig. 4. Comparison between the results of the present study for b/c = 0 and results between two smooth concentric vertical cylinders ðTa ¼ 0; b=c ¼ 0Þ with predicted
of Aoki et al. [54], Tachibana et al. [33] and Bjorklund and Kays [53]. values of [55].
264 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268

Nugr Nugr
Fig. 8. The effect of groove aspect ratio on Nus
for Nus
and T w ¼ 70  C.

Fig. 6. A comparison between mean Nusselt number ratios with 1012 Gr ReX for
0:328
NuX
different groove aspect ratios with Nu ¼ 0:375ð1011 Gr ReX Þ for smooth
X¼0
channel [56].

Fig. 9. The effect of number of grooves ðN gr Þ on averaged heat transfer coefficient

with respect to Taylor number for b=c ¼ 0:5 and T w ¼ 70  C.
Fig. 7. The effect of groove aspect ratio on averaged heat transfer coefficient with
respect to Taylor number for N gr ¼ 20 and T w ¼ 70  C.

vortexes inside the groove may be weak or completely disappear

and the heat transfer negligibly increases. It will be shown in Sec-
tion 4 that the effect of wall temperature on heat transfer enhance-
ment inside rotating grooved channels is negligible.
Fig. 8 shows the effect of groove aspect ratio on heat transfer
effectiveness parameter ðNugr =Nus Þ for 3 different groove aspect
ratios. This parameter is a relation between the Nusselt numbers
of grooved channel to smooth channel. The outer cylinder wall
temperature is maintained at 70  C. The results indicate that
ðNugr =Nus Þ increases with b/c and exceeds one at nearly all Taylor
numbers. Mounting axial grooves on the stationary outer cylinder
normally increases heat transfer, especially for b/c = 2. the maxi-
mum value of effectiveness parameter is 1.33 for the case of b/
c = 2 which is higher than 1.1 observed by Hayase et el. [28]. It also
should be pointed out that the effect of natural convection is not
considered in numerical analysis of Hayase et al. [28]. It can be
observed that for Ta > 5:5  106 , Taylor number doesn’t have any
significant effect on Nugr =Nus due to increment of perturbation Fig. 10. Local heat transfer coefficient for N gr ¼ 20, Ta ¼ 1:2  106 and T w ¼ 50  C
and vorticites of air fluid flow between two cylinders. This finding with different groove aspect ratios ðb=cÞ.
 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268

Fig. 11. Local heat transfer coefficient for b=c ¼ 1, N gr ¼ 20 and T w ¼ 90 C with
different Taylor numbers.
can be explained as follows. The system in case b/c = 2, has the
deepest grooves on the outer cylinder. The cavities increase the
effective surface to dissipate heat. However, at strong radial flow,
heat accumulates in the cavities, reducing the effect of grooves
on heat transfer enhancement.
Fig. 9 shows the effect of number of grooves on the heat transfer
enhancement in rotating grooved channels. It can be observed that
the averaged Nusselt number generally increases with increasing
the number of grooves from 10 to 20. The main reason could be
increment of perturbations and vortex flows inside the air gap. It
can also be concluded that comparing with groove aspect ratio,
the number of grooves plays a less dominant role in enhancing
the heat transfer in grooved channels
3.4. Local heat transfer coefficient
The local Nusselt number for different values of b/c is presented
in Fig. 10. The results show that for the case of b/c = 0 the thermal
entrance region is about z=L ¼ 0:4. The entrance region reduces in
grooved channels to z=L ¼ 0:33. In high rotation speeds the grooves
causes perturbation in air flow and the entrance region becomes
shorter. The results show that increasing b/c will increase the local
Nusselt number in developed region. It can be observed that ther-
mal entrance region reduces up to 17% in the presence of grooves.
Air flow between two cylinders is affected with both rotation and
buoyancy forces. It can be concluded that thermal entrance region
Fig. 12. Generation of experimental points in a central composite design with three factors.
265
is not negligible, because the air gap is open-ended and air flow
exists in the axial direction due to ‘‘buoyancy” forces.
Fig. 11 shows the effect of Taylor number on local heat transfer
coefficient inside grooved channel with groove aspect ratio of 1.
Thermal entrance region in the presence of grooves is small and
negligible at high Taylor numbers. This is because of increasing
perturbation of air flow in axial grooves at higher rotation speeds.
The fluid flow becomes unstable and therefore thermal entrance
region reduces. Thermal entrance region at Ta ¼ 1:05  106 is
z ¼ 0:41 while it reduces to z ¼ 0:17 at Ta ¼ 7:21  106 . It also
can be observed that heat transfer is highly increased by Taylor
number. In the next section, it will be shown that Taylor number
is the most significant parameter on heat transfer enhancement
in rotating grooved channels.
4. Response surface methodology
To model the effective geometric and physical parameters on
the heat transfer and fluid flow between two cylinders with inner
one rotating and outer stationary grooved one, the design of exper-
iments is performed to analyze the experimental data. The method
is a collection of mathematical and statistical techniques for
empirical model building. The objective is to optimize heat transfer
which is influenced by different parameters such as groove aspect
ratio, rotation speed and wall temperature.
Response surface methodology (RSM) is a powerful statistical
method to analyze different experimental studies by using
limited number of experiments [57,58]. The number of
experiments required in the central composite RSM design is given
by 2k þ 2k þ n where k is the number of factors and n is the num-
ber of center points [59]. The objective is to see the effect of three
independent variables (Taylor number, geometry of grooves and
two cylinders) on heat transfer. The number of center points in this
analysis is assumed to be 6. Therefore 20 experiments
ð23 þ 2  3 þ 6Þ are needed in this analysis. It is assumed that a
central point for each factor is 0 and the design is symmetric
around this as shown in Fig. 12.
The experimental results are analyzed using Design-Expert 8.0.
Table 3 shows series of experiments according to the experimental
plan based on the central composite face design. The input vari-
ables are wall temperature, rotation speed and groove aspect ratio
and the response parameter is averaged Nusselt number ðNuÞ.
The degree of model accuracy can be appraised with analysis of
variance. Table 4 shows the analysis of variance for Nusselt num-
ber monitored. Degree of freedom, sum of squares, adjusted mean
squares, F-value and P-value are statistical estimators used in the
analysis of variance. The p-values are used as a tool to check the
266 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268

Table 3
Central composite design for 3 different test cases with 3 different levels.

Std. Wall Rotation speed Groove Response

temperature ratio
A T w (°C) B Ta C b=c Nu
1 1 50 1 4:83  105 1 0 11.23
2 1 90 1 4:83  105 1 0 11.42
3 1 50 1 8:36  106 1 0 17.71
4 1 90 1 8:36  106 1 0 17.83
5 1 50 1 4:83  105 1 2 13.35
6 1 90 1 4:83  105 1 2 13.41
7 1 50 1 8:36  106 1 2 20.69
8 1 90 1 8:36  106 1 2 20.81
9 1 50 0 4:42  106 0 1 16.11
10 1 90 0 4:42  106 0 1 16.29
11 0 70 1 4:83  105 0 1 13.03
Fig. 13. Residual histogram plot for NuL .
12 0 70 1 8:36  106 0 1 19.37
13 0 70 0 4:42  106 1 0 14.69
14 0 70 0 4:42  106 1 2 16.98
15 0 70 0 4:42  106 0 1 15.89
16 0 70 0 4:42  106 0 1 16.16
17 0 70 0 4:42  106 0 1 16.47
18 0 70 0 4:42  106 0 1 16.33
19 0 70 0 4:42  106 0 1 16.41
20 0 70 0 4:42  106 0 1 16.29

Table 4
Analysis of variance.

Source DOF Adj ss Adj ms F-Value P-Value

Model 9 132.049 14.672 390.65 0.000
A 1 0.045 0.045 1.20 0.300
B 1 115.396 115.396 3072.44 0.000
C 1 15.277 15.277 406.76 0.000
A2 1 0.002 0.002 0.05 0.826 Fig. 14. Response surface as a function of groove aspect ratio and Taylor number.
B2 1 0.002 0.002 0.05 0.826
C2 1 0.421 0.421 11.21 0.007
AB 1 0.000 0.000 0.00 0.986
AC 1 0.002 0.002 0.06 0.817
BC 1 0.428 0.428 11.39 0.007 Table 5
Lack of fit 5 0.156 0.031 0.71 0.640 Estimated regression coefficients based on experimental design.
Pure error 5 0.219 0.044 Term Coefficient SE coefficient T-Value P-value
Total 19 132.424
Constant 16.2468 0.0666 243.83 0.000
DOF = degree of freedom, Adj ss = adjusted sum of square, Adj ms = adjusted mean A 0.0670 0.0613 1.09 0.300
square. B 3.3970 0.0613 55.43 0.000
C 1.2360 0.0613 20.17 0.000
A2 0.026 0.117 0.23 0.826
B2 0.028 0.117 0.23 0.826
significance of different parameters. The smaller the value of p, the C2 0.391 0.117 3.35 0.007
more significant is the corresponding coefficient. The P-value AB 0.0013 0.0685 0.02 0.986
obtained from model is less than 0.001 and indicates that the AC 0.0162 0.0685 0.24 0.817
BC 0.2313 0.0685 3.37 0.007
regression is significant. If the F-value is greater than one, the anal-
R2 99.72%
ysis has more accuracy. In this study, the F-value of the model is R2-adjusted 99.46%
390.61 which are ample to satisfy the requirements.
In order to understand better and more efficient of experimen-
tal design, Fig. 13 shows the residual plot for averaged Nusselt
will be significant if the amount of P-value is very small
number. Residual plot illustrates the difference between the
ðP < 0:08Þ. According to this table, the estimated coefficients for
observed response and respective predicted response. As it is
B (Taylor number), C (groove aspect ratio), C2 and BC related to
shown, the maximum residual is around 0.4 for Nusselt number
Nusselt number are significant terms.
as a response.
According to Table 5, the regression coefficients of A, A2, B2, AB
The response surface 3-D plot for interactive effects of Taylor
and AC are small and negligible. The second-order polynomial
number and groove aspect ratio is illustrated in Fig. 14. As it is
shown, the averaged Nusselt number significantly increases with model for heat transfer can be expressed as Nu ¼ b0 þ b1 Bþ
Taylor number and groove aspect ratio. It can be observed that b2 C þ b3 C 2 þ b4 BC where b0 to b4 are constant coefficients and B
the heat transfer enhancement is increased by increasing the rota- and C corresponds to Taylor number and b/c, respectively. Based
tional speed and groove aspect ratio to higher values. on regression analysis of experimental results and with respect
Table 5 presents the estimated regression coefficients for Aver- to response surface methodology, the following correlation for
aged Nusselt number.as mentioned above, the coefficient effect average Nusselt number can be expressed as follows
 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268 267

Table 6
Twenty experiments required in the central composite design

Experimental results and their comparison with predicted values. are successfully conducted to study the individual and interac-
Design parameters Nusselt number, Nu tive effects of some geometrical and physical parameters on TC
b=c Ta Exp. Pred. Error% flow inside annular channel with grooves. It is found that Taylor
number is the most important parameter in heat transfer
0.5 1:6  106 12.98 12.580 3.08
enhancement. It is also concluded that increasing groove aspect
1 1:6  106 13.89 13.242 4.66
ratio will increase heat transfer inside grooved channels. The
2 1:6  106 14.32 13.979 2.38
effect of wall temperature is negligible in comparison with
0.5 9:6  105 12.56 12.036 4.17
other parameters.
1 9:6  105 13.11 12.679 3.29
2 9:6  105 13.36 13.397 0.28
0.5 5:14  106 15.42 15.589 1.09
1 5:14  106 16.76 16.355 1.93 References
2 5:14  106 17.48 17.299 1.03
0.5 7:89  106 18.44 17.927 2.78 [1] C.-C. Chang, Y.-F. Kuo, J.-C. Wang, S.-L. Chen, Air cooling for a large-scale motor,
Appl. Therm. Eng. 30 (2010) 1360–1368.
1 7:89  106 18.32 18.773 2.47
[2] W. Lian, W. Chang, Y. Xuan, Numerical investigation on flow and thermal
2 7:89  106 20.73 19.880 4.10
features of a rotating heat pipe, Appl. Therm. Eng. 101 (2016) 92–100.
[3] H. Wang, E. Dominguez-Ontiveros, Y.A. Hassan, Computational fluid dynamics
analysis of core bypass flow and crossflow in a prismatic very high
Nu ¼ 10:41 þ 0:08207  105 Ta þ 1:816b=c  0:391ðb=cÞ
2
temperature gas-cooled nuclear reactor based on a two-layer block model,
Nucl. Eng. Des. 268 (2014) 64–76.
þ 0:00587  105 Ta:b=c ð10Þ [4] W. Yoo, S. Jeon, C. Son, J. Yang, D. Ahn, S. Kim, K. Hwang, S. Ha, Full surface heat
transfer characteristics of rotor ventilation duct of a turbine generator, Appl.
The accuracy of proposed second order correlation based on Therm. Eng. 94 (2016) 385–394.
[5] M. Fénot, E. Dorignac, A. Giret, G. Lalizel, Convective heat transfer in the entry
regression analysis for Nusselt number is determined experimen- region of an annular channel with slotted rotating inner cylinder, Appl. Therm.
tally by applying 12 additional confirmation experiments. The Eng. 54 (2013) 345–358.
comparison of predicted values and the experimental results are [6] S. Grossmann, D. Lohse, C. Sun, High-Reynolds number Taylor-Couette
turbulence, Annu. Rev. Fluid Mech. 48 (2016) 53–80.
summarized in Table 6. The results indicated that the predicted
[7] W. Wang, B.-W. Li, Effects of optical parameters on fluid flow and heat transfer
values are in good agreement with the experimental data. The of participating fluid between two vertical coaxial rotating cylinders, Int. J.
maximum error between experimental results and Eq. (10) is Heat Mass Transfer 90 (2015) 167–179.
4.66%. [8] J.M. Lopez, F. Marques, M. Avila, Conductive and convective heat transfer in
fluid flows between differentially heated and rotating cylinders, Int. J. Heat
Mass Transfer 90 (2015) 959–967.
5. Conclusions [9] M. Monfared, E. Shirani, M.R. Salimpour, F. Aloui, Numerical and experimental
study on the flow history effects of axial flow on the Couette-Taylor flow, Acta
Mech. (2016) 1–12.
This study is divided into two main parts. In the first part, dif- [10] D. Lasagna, O. Tutty, S. Chernyshenko, Flow regimes in a simplified Taylor–
ferent flow regimes, heat transfer and pressure drop characteristics Couette-type flow model, Eur. J. Mech.-B/Fluids 57 (2016) 176–191.
[11] M. Nemri, S. Charton, E. Climent, Mixing and axial dispersion in Taylor-Couette
of fluid flow inside electric motors and generators in the presence
flows: the effect of the flow regime, Chem. Eng. Sci. 139 (2016) 109–124.
of grooves are reviewed. Investigations of previous works for dif- [12] C. Kang, K.-S. Yang, I. Mutabazi, Thermal effect on large-aspect-ratio Couette-
ferent groove geometries, gap thickness, axial and radial ratios Taylor system: numerical simulations, J. Fluid Mech. 771 (2015) 57–78.
[13] S. Poncet, S. Haddadi, S. Viazzo, Numerical modeling of fluid flow and heat
are summarized and compared. The effect of axial fluid flow
transfer in a narrow Taylor–Couette–Poiseuille system, Int. J. Heat Fluid Flow
ðRea Þ and rotational fluid flow ðTaÞ are also considered. 32 (2011) 128–144.
In the second part, the effect of groove aspect ratios, number of [14] J. Neto, A. Martins, A.S. Neto, C. Ataíde, M. Barrozo, CFD applied to turbulent
grooves, Taylor number and wall temperature are experimentally flows in concentric and eccentric annuli with inner shaft rotation, Can. J. Chem.
Eng. 89 (2011) 636–646.
investigated in a grooved annulus. Taylor number is varied from [15] H. Dawood, H. Mohammed, N.A.C. Sidik, K. Munisamy, M. Wahid, Forced,
0 to 8:31  106 . Number of grooves and groove aspect ratios are natural and mixed-convection heat transfer and fluid flow in annulus: a
review, Int. Commun. Heat Mass Transfer 62 (2015) 45–57.
varied from 0 to 20 and from 0 to 2, respectively. Based on the [16] H. Togun, T. Abdulrazzaq, S. Kazi, A. Badarudin, A.A.H. Kadhum, E.
reported results, the following conclusions may be drawn: Sadeghinezhad, A review of studies on forced, natural and mixed heat
transfer to fluid and nanofluid flow in an annular passage, Renew. Sustain.
Energy Rev. 39 (2014) 835–856.

The distributions of averaged Nusselt number for an annular
[17] M. Fénot, Y. Bertin, E. Dorignac, G. Lalizel, A review of heat transfer between
channel with grooves are generally higher than smooth annular concentric rotating cylinders with or without axial flow, Int. J. Therm. Sci. 50
channels. Among all test modes, the averaged Nusselt number (2011) 1138–1155.
of the model with b/c = 2 is higher than three other models [18] Y. Huai, R.V. Melnik, P.B. Thogersen, Computational analysis of temperature
rise phenomena in electric induction motors, Appl. Therm. Eng. 23 (2003)
ðb=c ¼ 0; 0:5; 1Þ. 779–795.

Buoyancy effects and Grashof number are considered in rotat- [19] J. Cleeton, R. Kavanagh, G. Parks, Blade cooling optimisation in humid-air and
ing annular channels with grooves for the first time. The ratio steam-injected gas turbines, Appl. Therm. Eng. 29 (2009) 3274–3283.
[20] J. Nerg, M. Rilla, J. Pyrhonen, Thermal analysis of radial-flux electrical
of Nusselt number ðNuÞ to stationary Nusselt number ðNuX¼0 Þ machines with a high power density, IEEE Trans. Industr. Electron. 55 (2008)
shows that increasing Gr ReX will increase heat transfer coeffi- 3543–3554.
[21] D.A. Howey, P.R. Childs, A.S. Holmes, Air-gap convection in rotating electrical
cient, especially at 1012 Gr ReX ¼ 19:963, Nu of grooved channel machines, IEEE Trans. Industr. Electron. 59 (2012) 1367–1375.
with b/c = 2 reaches around 2.23 times NuX¼0 . [22] J.A. Malumbres, M. Satrustegui, I. Elosegui, J.C. Ramos, M. Martínez-Iturralde,
Analysis of relevant aspects of thermal and hydraulic modeling of electric

The effectiveness parameter indicates the importance of heat
machines. Application in an open self ventilated machine, Appl. Therm. Eng. 75
transfer enhancement in the presence of axial grooves. The (2015) 277–288.
maximum value of effectiveness parameter is 1.33 for the case [23] J. Wen, J. Zheng, Numerical analysis of the external wind path for medium-size
high-voltage asynchronous motors, Appl. Therm. Eng. 90 (2015) 869–878.
of b/c = 2.
[24] C. Kim, K.-S. Lee, S.-J. Yook, Effect of air-gap fans on cooling of windings in a

Local heat transfer coefficient inside grooved channels and the large-capacity, high-speed induction motor, Appl. Therm. Eng. 100 (2016)
effect of groove aspect ratio on thermal entrance region is 658–667.
investigated experimentally. Thermal entrance region reduces [25] T.-M. Jeng, S.-C. Tzeng, C.-H. Lin, Heat transfer enhancement of Taylor–
Couette–Poiseuille flow in an annulus by mounting longitudinal ribs on the
up to 17% in the presence of grooves. rotating inner cylinder, Int. J. Heat Mass Transfer 50 (2007) 381–390.
268 A. Nouri-Borujerdi, M.E. Nakhchi / Applied Thermal Engineering 120 (2017) 257–268
[26] Y.N. Lee, W. Minkowycz, Heat transfer characteristics of the annulus of
twocoaxial cylinders with one cylinder rotating, Int. J. Heat Mass Transfer 32
(1989) 711–722.
[27] S.-C. Tzeng, W.-P. Ma, C.-W. Lin, W.-Y. Jywe, C.-H. Liu, Y.-C. Wang,
Experimental investigation of lubrication and cooling effects of high-speed
rotating machines, Int. J. Adv. Manuf. Technol. 35 (2007) 394–399.
[28] T. Hayase, J. Humphrey, R. Greif, Numerical calculation of convective heat
transfer between rotating coaxial cylinders with periodically embedded
cavities, J. Heat Transfer 114 (1992) 589–597.
[29] K. Bilen, M. Cetin, H. Gul, T. Balta, The investigation of groove geometry effect
on heat transfer for internally grooved tubes, Appl. Therm. Eng. 29 (2009) 753–
761.
[30] D. Lorenzini-Gutierrez, A. Hernandez-Guerrero, J.L. Luviano-Ortiz, J.C. Leon-
Conejo, Numerical and experimental analysis of heat transfer enhancement in
a grooved channel with curved flow deflectors, Appl. Therm. Eng. 75 (2015)
800–808.
[31] J. Liu, G. Xie, T.W. Simon, Turbulent flow and heat transfer enhancement in
rectangular channels with novel cylindrical grooves, Int. J. Heat Mass Transfer
81 (2015) 563–577.
[32] Y. Joo, S.J. Kim, Thermal optimization of vertically oriented, internally finned
tubes in natural convection, Int. J. Heat Mass Transfer 93 (2016) 991–999.
[33] F. Tachibana, S. Fukui, Convective heat transfer of the rotational and axial flow
between two concentric cylinders, Bull. JSME 7 (1964) 385–391.
[34] S.-H. Lee, H.-T. Chung, C.-W. Park, H.-B. Kim, Experimental investigation of the
effect of axial wall slits on Taylor? Couette flow, Fluid Dynam. Res. 41 (2009)
045502.
[35] D. Liu, I.-S. Kang, J.-E. Cha, H.-B. Kim, Experimental study on radial
temperature gradient effect of a Taylor-Couette flow with axial wall slits,
Exp. Therm. Fluid Sci. 35 (2011) 1282–1292.
[36] D. Liu, Y.-Z. Wang, W.-D. Shi, H.-B. Kim, A.-K. Tang, Slit wall and heat transfer
effect on the Taylor vortex flow, Energies 8 (2015) 1958–1974.
[37] N. Biswas, P.S. Mahapatra, N.K. Manna, Mixed convection heat transfer in a
grooved channel with injection, Numer. Heat Transfer Part A: Appl. 68 (2015)
663–685.
[38] S.R.M. Gardiner, R. Sabersky, Heat transfer in an annular gap, Int. J. Heat Mass
Transfer 21 (1978) 1459–1466.
[39] S. Eiamsa-Ard, P. Promvonge, Thermal characteristics of turbulent rib-grooved
channel flows, Int. Commun. Heat Mass Transfer 36 (2009) 705–711.
[40] N. Zheng, P. Liu, F. Shan, Z. Liu, W. Liu, Heat transfer enhancement in a novel
internally grooved tube by generating longitudinal swirl flows with multi-
vortexes, Appl. Therm. Eng. 95 (2016) 421–432.
[41] M. Bouafia, Y. Bertin, J. Saulnier, P. Ropert, Analyse expérimentale des
transferts de chaleur en espace annulaire étroit et rainuré avec cylindre
intérieur tournant, Int. J. Heat Mass Transfer 41 (1998) 1279–1291.
[42] C. Gazley, Heat transfer characteristics of the rotational and axial flow
between concentric cylinders, Trans. ASME 80 (1958) 79–90.
[43] Y. Sommerer, G. Lauriat, Numerical study of steady forced convection in a
grooved annulus using a design of experiments, J. Heat Transfer 123 (2001)
837–848.
[44] N. Lancial, F. Torriano, F. Beaubert, S. Harmand, G. Rolland, Study of a Taylor-
Couette-Poiseuille flow in an annular channel with a slotted rotor, in: 2014
International Conference on Electrical Machines (ICEM), IEEE, 2014, pp. 1422–
1429.
[45] H.Z. Abou-Ziyan, A.H.B. Helali, M.Y. Selim, Enhancement of forced convection
in wide cylindrical annular channel using rotating inner pipe with interrupted
helical fins, Int. J. Heat Mass Transfer 95 (2016) 996–1007.
[46] S. Gilchrist, C. Ching, D. Ewing, Heat transfer enhancement in axial Taylor-
Couette Flow, in: ASME 2005 Summer Heat Transfer Conference collocated
with the ASME 2005 Pacific Rim Technical Conference and Exhibition on
Integration and Packaging of MEMS, NEMS, and Electronic Systems, American
Society of Mechanical Engineers, 2005, pp. 227–233.
[47] X. Zhu, R. Ostilla-Mónico, R. Verzicco, D. Lohse, Direct numerical simulation of
Taylor-Couette flow with grooved walls: torque scaling and flow structure, J.
Fluid Mech. 794 (2016) 746–774.
[48] T. Hanagida, N. Kawasaki, Pressure drop and heat-transfer characteristics of
axial air flow through an annulus with a deep-slotted outer cylinder and a
rotating inner cylinder (second report-heat-transfer characteristics), Heat
Transfer Jpn. Res. 21 (1992) 292–304.
[49] Y. Chong, D. Staton, M. Mueller, J. Chick, Pressure loss measurement in rotor-
stator gap of radial flux electrical machines, in: 2014 International Conference
on Electrical Machines (ICEM), IEEE, 2014, pp. 2172–2178.
[50] K. Singh, R. Das, An experimental and multi-objective optimization study of a
forced draft cooling tower with different fills, Energy Convers. Manage. 111
(2016) 417–430.
[51] R.J. Moffat, Describing the uncertainties in experimental results, Exp. Therm.
Fluid Sci. 1 (1988) 3–17.
[52] H.W. Coleman, W.G. Steele, Experimentation, Validation, and Uncertainty
Analysis for Engineers, John Wiley & Sons, 2009.
[53] I. Bjorklund, W. Kays, Heat transfer between concentric rotating cylinders, J.
Heat Transfer 80 (1959) 79–90.
[54] H. Aoki, H. Nohira, H. Arai, Convective heat transfer in an annulus with an
inner rotating cylinder, Bull. JSME 10 (1967) 523–532.
[55] M.A. El-Shaarawi, A.-A. Sarhan, Developing laminar free convection in a heated
vertical open-ended concentric annulus, Ind. Eng. Chem. Fundam. 20 (1981)
388–394.
[56] S.-C. Tzeng, Heat transfer in a small gap between co-axial rotating cylinders,
Int. Commun. Heat Mass Transfer 33 (2006) 737–743.
[57] G.E. Box, K. Wilson, On the experimental attainment of optimum conditions,
in: Breakthroughs in Statistics, Springer, 1992, pp. 270–310.
[58] A. Nouri-Borujerdi, M.E. Nakhchi, Optimization of the heat transfer coefficient
and pressure drop of Taylor-Couette-Poiseuille flows between an inner
rotating cylinder and an outer grooved stationary cylinder, Int. J. Heat Mass
Transfer 108 (2017) 1449–1459.
[59] A. Nouri-Borujerdi, M. Bovand, S. Rashidi, K. Dincer, Geometric parameters and
response surface methodology on cooling performance of vortex tubes, Int. J.
Sustain. Energy (2016) 1–15.