International Journal of Thermal Sciences 130 (2018) 264–277

Contents lists available at ScienceDirect

International Journal of Thermal Sciences

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

Thermal performance augmentation in the cooling zone of brick tunnel kiln T

with two types of guide vanes
H.A. Refaey∗, Ali A. Abdel-Aziz, M.R. Salem, H.E. Abdelrahman, M.W. Al-Dosoky
Department of Mechanical Engineering, Faculty of Engineering at Shoubra, Benha University, 11629 Cairo, Egypt

A R T I C LE I N FO A B S T R A C T

Keywords: The energy consumption in bricks and ceramics industry is of great importance. Therefore, the present work
Tunnel kiln introduces an augmentation technique to improve the thermal performance during the cooling process in bricks
Bricks tunnel kiln. A test rig by scale 1:4 has been designed and fabricated to simulate the cooling zone of tunnel kiln.
Experiments Two types of guide vanes (side wall (SV) and U-shape (UV)) have been designed with attack angles (θ= 120o ,
Thermal performance
135o , and 150o ) in flow direction. Seven different settings within Reynolds number range of
Guide vanes
13,609 ≤ Re ≤ 27,634 are tested in a suction-type tunnel kiln. Effect of brick setting arrangement, Reynolds
number, guide vane type, and attack angle on the thermal performance is studied. The results reveal that the UV
guide vanes influence the thermal performance more than SV guide vanes. The maximum enhancement in heat
transfer rate to pumping power ratio (Qavg/PP) is about 25 for setting 7 in absence of vanes and 23 for (UV) with
attack angles 150° and 135° at Re = 14,000. The present study introduces two correlations for average Nusselt
number. Settings 4 and 1 provide a moderate production time with highest productivity while setting 2 has
lowest productivity with small production time.

1. Introduction Dugwell and Oakley [4] studied the heat transfer process along the
kiln. The brick column was represented as a solid block. Therefore, the
Tunnel kilns are of great importance in manufacturing ceramics and actual hydrodynamic pattern around the bricks was ignored in their
bricks. They are composed of a collection of attached opposite directed study. Riedel [5] concluded that convection is the key factor for tunnel
heat exchangers with the solids on the kiln car that move counter- kiln. Number of features such as setting pattern and kiln roof was
current with the air flow. There are three main temperature zones in the suggested to promote cross convection. Almeida et al. [6] provided a
kiln; preheating, firing, and cooling zone. Tunnel kilns are long struc- mathematical and numerical study to dry hollow bricks in a tunnel
ture kilns in which the green products are heated up in the preheating dryer. Refaey and Specht [7] presented three-dimensional analysis to
zone, and then to the sintering temperature which is different from one simulate the burning of Sanitaryware products. The effect of nozzle
product to another. After that the product goes through the cooling axial velocity and nozzles arrangement was presented. The results re-
zone, where it is cooled down to a temperature near the ambient vealed that the radial velocity produced by the burner and/or nozzles
temperature [1]. Fig. 1 shows a schematic representation of tempera- was essential to increase the heat transfer. Mancuhan et al. [8] devel-
ture profile along the tunnel kiln. oped one-dimensional model for the preheating zone of the tunnel kiln.
Energy conservation management pushed many researchers to The model described the following; gas flow, heat transfer between
consider the thermal trouble inside the tunnel kilns to reduce the in- bricks and gases, and water evaporation. Ambient air was fed into the
tensive energy consumption. A worthy mathematical, numerical and preheating zone by two different profiles and vent locations. The results
optimization studies have been performed on tunnel kilns. These stu- revealed that the gas temperature reached 350 °C at the entrance of the
dies included many aspects on the separate zones or on the whole kiln preheating zone when there was no air fed.
such as; temperature profile, fuel distribution, flow field, and heat Kaya et al. [9] developed a mathematical model to compute the
transfer. Boming [2] established a dynamic model for tunnel kiln. The mass flow rate of air and temperature profile along the cooling zone of
results showed that the model could describe the process and could help tunnel kiln. The results revealed that, to minimize the pressure drop,
in kiln design. Tehzeeb et al. [3] used the computational fluid dynamics the kiln cooling zone should be composed of four regions, two of them
to simulate the temperature profile in bricks tunnel kiln. with a suction flow and the others of blowing type. Nicolau and Dadam

∗
Corresponding author.
E-mail address: hassanein.refaey@feng.bu.edu.eg (H.A. Refaey).

https://doi.org/10.1016/j.ijthermalsci.2018.04.027
Received 14 November 2017; Received in revised form 19 April 2018; Accepted 19 April 2018
Available online 04 May 2018
1290-0729/ © 2018 Elsevier Masson SAS. All rights reserved.
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Nomenclature Vd Duct volume, m3

Vf Volume of flow, m3, Vf = Vd − Vb
A Brick length, mm Vi Voltage drop, Volt
Ab Brick surface area, m2
Aw Wet area, m2, Aw = Ab,w + A d,w Greek letters
Ab,w Bricks wet area, m2
Ad,w Duct wet area, m2 θ Attack angle
B Brick width, mm ε Void fraction, ε= Vf /Vd
C Brick height, mm ρ Density, kg/m3
Cp Specific heat, J/kg.K ρs Setting density = Vb/Vd
D Diameter, m ω Uncertainty
4Vf
Dh Hydraulic diameter, m, Dh = A η Performance criteria
w
F Friction factor ν Kinematic viscosity, m2/s
G Gap between cover plate and top layer
H Tunnel height Subscript
h Convective heat transfer coefficient, W/m2.K
I Electric current, Amp avg average
K Fluid thermal conductivity, W/m.K i Local value
L Length of brick setting, mm mv Middle with vanes
M Mass flux, kg/s.m2 w,nv Wall with no vanes
Nu Nusselt number
P Pressure, Pa Abbreviations
Q Heat transfer rate, W, Q input = VIi cos ∅
Qinput Input heat from variac, W LM Longitudinal middle
S Spacing between columns, m LW Longitudinal wall
U Superficial velocity, m/s PP Pumping power
U Interstitial velocity, m/s , U= ε
u SV Side wall guide vanes
Re Reynolds number TM Transversal middle
Ta,b Air bulk temperature, K TW Transversal wall
Ts,i Local brick surface temperature, K UV U-shape guide vanes
Vb Bricks volume, m3

[10] presented three-dimensional numerical study to show the tem- influence of combined heating & power system (CHPS), the number of
perature distribution through load, gas, and walls. In addition, an ex- CHPS modules, and the structure of the needed purchased-power
perimental thermal analysis of a tunnel kiln was presented. Naccache component on the economic efficiency. Recently, Soussi et al. [17]
et al. [11] numerically investigated heat transfer and fluid flow of optimized the recovered air mass flow from the cooling zone to the
combustion gases inside a tunnel kiln. The numerical results were firing zone to reduce the natural gas consumption during the manu-
compared with experimental results from literature. The results showed facturing of hollow bricks in a Tunisian tunnel kiln. The results showed
that natural gas can be used in tunnel kiln instead of sawdust. that the existence of an optimal value of the recovered air mass flow
Essenhigh [12] performed an analysis of the energy equation to that could reduce the actual daily consumption of the natural gas up to
determine the relation between input and output energies. Santos [13] 4.6%. Durakovic and Delalic [18] established a mathematical model to
provided a numerical formulation to get the thermal behavior of a analyze and check the stationary temperature field in brick and kiln.
tunnel kiln. A good agreement was obtained between the numerical and Other researchers made experimental investigations on tunnel kilns.
experimental results of sawdust as a fuel. In addition, other simulations Karaush et al. [19] studied the heat absorption from the ceramic kiln
were performed by using natural gas as a fuel in this kind of kiln. Refaey radiating walls experimentally. They concluded that there was an op-
et al. [1&14] developed one-dimensional mathematical model by using timal spacing between the ceramic pieces and there was no increase in
MATLAB program to predict temperature profile along tunnel kiln. the heat absorption rate if this space increased. Abou-Ziyan [20] ex-
Furthermore, the influence of fuel distribution along the firing zone was perimentally studied the thermal performance in the cooling zone for
studied. Dugwell and Oakley [15] built up a laboratory model of a six different brick settings. The results showed that the setting ar-
tunnel kiln. The results presented a correlation to calculate convective rangement affects the pressure drop and convective heat transfer
heat transfer rates for the firing of refractory. Roth [16] described the coefficient. Correlations for the friction factor and Nusselt number were
presented. Ros-Dosdá et al. [21] provided a study of the environmental
life cycle assessment of different Porcelain stoneware tile to identify the
hotspots and choose the Environmental Product Declaration (EPD)
program.
Recently, Refaey et al. [22] presented an augmentation technique
using guide vanes with different attack angles attached to the kiln side
walls. The heat transfer and the pressure drop for ten different brick
settings were experimentally investigated. The results were obtained for
a wide range of Reynolds number from 11,867 to 25,821. The results
revealed that both of convective heat transfer and pressure drop were
strongly depend on the brick settings arrangement. The attack angle has
a great influence on the average Nusselt number. The results revealed
Fig. 1. Schematic diagram of temperature distribution along tunnel kiln [1]. that a maximum enhancement of about 94.5% was obtained for

265
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

longitudinal brick at middle column in setting 2 at θ= 135o and the blower (4.1 kW) and is delivered to a variable gate area to adjust the
Re = 22,407. Moreover, an empirical correlation for the average Nus- flow rate at a constant value. To reduce the air swirl and develop a
selt number as a function of Reynolds number, attack angle, and brick uniform flow to the test section, bell mouth intake and straightener are
setting dimensionless groups within a specified range for the in- located inside the main duct before the test section. A mild steel sheet
dependent factors was obtained. with 1.5 mm thickness is used to fabricate the test section. The heat
Redemann and Specht [23] used a mathematical model to simulate losses from the test section to the ambient are minimized by glass wool
the roof tiles tunnel kiln process. The results showed that the required layer insulating material with 25 mm thickness. To replace the brick
mean flow velocity through the roof tiles setting should be in the range model setting and the guide vanes, a movable cover of mild steel is
of 20 m/s. Araújo et al. [24] conducted a numerical simulation to used. During each test run, measurements of mass flow rate, heat input
predict the temperature and moisture content in hollow ceramic brick to the typical brick models, surface temperatures of brick models, inlet
during drying process. The results stated that the smaller thickness of and outlet temperatures of air flow, and pressure drop across the test
brick provides higher temperatures and lower moisture content (heat section are carried out.
transfer and dry become fast) than the others. Milani et al. [25] con- The transition section was used between the unheated exit duct and
ducted a numerical model to analyze the performance of industrial the blower circular tube as shown in Fig. 2. Because the test section
ceramic kiln. The revealed that the numerical model could be used to with side guide vanes (SV) still needs a high enhancement and a
evaluate the different kiln configurations with new burners after vali- compound turbulence generator must be provided to indicate whether
dation with the experimental measurements. Furthermore, fuel saving the present new guide vane U-shape (UV) can do better performance
could be attained by about 10% for that new heat recovery burner. compared to the (SV). Guide vanes are fixed with two different posi-
The present experimental investigation shows the effect of guide tions through the test section to direct the flow to the confined zone
vanes type on the heat transfer and pressure drop in the brick tunnel behind columns. The guide vanes induce vortices in front of them, so
kiln. Seven brick settings were conducted and tested. The effect of they can provide good mixing between the main flow between columns
setting arrangement, guide vane type, wall effect, bricks spacing and and the fluid flow close to the brick models. These vortices can enhance
columns spacing on the heat transfer and pressure drop are studied. heat transfer with little increase in the friction factor.
Three different guide vanes attack angles (150o , 135o , and 120o ) with two Two different guide vanes are used in the present work; side wall
different positions within Reynolds number ranged from 13,609 to guide vanes (SV) are mounted on the two side-vertical walls of the test
27,634. Furthermore, the study provides Nusselt number correlations as section and U-shape guide vanes (UV) which are mounted on the three
a function of longitudinal and transversal spacing ratios, vane attack sides of the test section. Figs. (2 and 3a) demonstrate two views of
angle, and Reynolds number. setting 2. The side view of setting 2 without vanes is shown in Fig. 2
whereas, Fig. 3a represents the top view with side guide vanes (SV)
through the test section. In addition, Fig. 3b presents the isometric and
2. Experimental test rig side view of the U-shape guide vane. A mild steel sheet is used to
fabricate the two types of guide vanes with three attack angles; 120°,
A tunnel kiln experimental test rig is designed and fabricated to 135°, and 150° to the flow direction. The guide vane has a length of
investigate the heat transfer characteristics around bricks model with 250 mm equal to the tunnel height, base width 350 mm equal to tunnel
different settings. The experiments have been conducted on a scaled width, and depth 20 mm in stream-wise direction. Furthermore, Fig. 3c
test section to simulate the cooling zone of the brick tunnel kiln. The shows the construction of the heating element for a typical brick model.
cooling zone was experimentally tested by Refaey et al. [22] and the The heating element consists of a nickel-chromium resistance wire of
schematic diagram of the test rig is shown in Fig. 2. A suction tunnel 0.2 mm diameter which is wounded helically (10 Ω/m) and inserted in
kiln has a test section with cross section 350 × 250 mm and length of a stainless-steel tube that is filled with MgO material and then inserted
1.50 m. The test rig of the tunnel kiln consists of bell-mouth intake, in each typical brick model to attain a uniform heat flux. Four heating
straightener, main entrance duct, test section provided with brick elements are used for typical brick models which made of refractory
model settings, transition duct, a centrifugal blower, and measuring brick and have the same dimension as the bricks in the settings.
instruments as shown in Fig. 2. Air flow is sucked to the test section by

Fig. 2. Schematic diagram of the experimental test rig.

266
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Fig. 3. (a) Top view details of setting 2 with guide vane angle 120°. (b) Isometric and side view of the U-shape guide vane (c) Cross section in a heating brick model.

There are three main parameters, layers, rows, and columns, iden- investigate characteristic turbulent flow heat transfer through bricks
tify each setting. Layers are the number of bricks in the vertical direc- tunnel kiln with two guide vanes types for Reynolds number ranged
tion and there are 7-layers for each setting in the present work. Rows from 13,609 to 27,634. The roof air gap (distance between the movable
are in the flow direction with 30 mm constant spacing as shown in cover plate and the surface of the bricks setting top layer as shown in
Fig. 3a. Columns are perpendicular to flow direction in the span-wise Fig. 2 is investigated. The brick arrangement with 8-layers has a small
direction and the space between two columns is (S) as shown in Fig. 3a. roof air gap distance in comparison with 7-layers. Two values of roof air
Fig. 4 illustrates one row from each setting in the seven used bricks gap ratio (G/H) of 0.104 and 0.216 were studied for 8-layers and 7-
arrangement settings. Table 1 introduces the characteristic dimensions layers bricks arrangement, respectively.
of the seven configurations. The experiments were conducted to Fourteen calibrated K-type thermocouples (wires of 0.2 mm

Fig. 4. Schematic representation of one row from each pattern of the seven different investigated settings.

267
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Table 1
Characteristics of the present tested settings.
Setting No. of Bricks ε Dh (mm) Aw,b (m2) Aw,b/Aw,b1 (%) ρs = Vb/Vd S (mm) S/a S/b

1 504 0.6995 43.17 2.2936 100.0 0.3005 19.3 0.3333 1.2083

2 378 0.7746 59.54 1.7202 75.0 0.2254 58.0 1.0000 3.6250
3 252 0.8497 81.31 1.2574 54.8 0.1503 58.0 1.0000 3.6250
4 504 0.6995 43.97 2.2418 97.7 0.3005 26.5 0.4569 1.6563
5 462 0.7245 48.49 2.0679 90.2 0.2755 37.0 0.6379 2.3125
6 420 0.7496 53.20 1.9137 83.4 0.2504 37.0 0.6379 2.3125
7 336 0.7996 65.77 1.5660 68.3 0.2004 58.0 1.0000 3.6250

*The brick dimensions are a = 58 mm, b = 16 mm and c = 28 mm ρs is the setting density.

diameter) are used to measure the air inlet and outlet temperatures and average convective heat transfer coefficients hi and have , respectively
the heating brick surfaces temperatures. The surface temperatures of are calculated from the following equations [22]:
the four heating bricks were measured by eight K-type thermocouples Q input
(wires of 0.2 mm diameter). Two thermocouples were used to measure hi =
Ab (Ts,i − Ta, b) (1)
the temperature of the side surfaces for each heating brick. Six ther-
mocouples were used to measure the inlet and exit temperatures of the ∑ hi Ai
flow stream through the test section. A mercury thermometer was used havg =
∑ Ai (2)
to calibrate the thermocouples with ± 0.5 K accuracy. During all ex-
periments a data acquisition system was used to record the thermo- Note, the index, i, refer to either longitudinal or transversal brick in
couple temperatures. A digital differential pressure transducer (Dwyer® the middle or near the wall.
series WWDP with an accuracy of ± 2% of full scale) was used to Moreover, the Reynolds number which is based on interstitial ve-
measure the static pressure drop across the setting. The measurements locity and the hydraulic diameter of the duct is calculated as follows;
are taken in the fifth raw as mentioned in Refs. [16,20]. The heating UDh
elements are located in the third and six layers of the longitudinal and Re =
ν (3)
transversal direction for middle and near wall columns.
where, Dh , is the hydraulic diameter of the duct and U is the interstitial
velocity.
3. Experimental procedures and calculations
Then the local, Nui , and average, Nuavg , Nusselt numbers can be
obtained as follows;
The experimental procedures are initiated after assembling all parts
of the test rig. Then the bricks are loaded according to the tested setting hi Dh
Nui =
on the test section part. Furthermore, the guide vanes are installed k (4)
according to the experiments type i.e. side vanes (SV) experiments or U- havg Dh
shape vane (UV) experiments. There is a 1 kW AC voltage regulator to Nuavg =
k (5)
control the consumed power by the heating element. The voltage drop
and the electrical current are fixed at 8 ± 0.1 V and 2.9 ± 0.1 Amp, The friction factor and heat transfer are measured at the same time
respectively. Formerly, the air flow rate is adjusted by regulating the air to show the effect of geometrical parameters of the studied settings for
gate variable area controller. Brick model surfaces temperature and the different operating conditions. The Darcy friction coefficient for the air
inlet and exit flow temperatures are taken every 1 s until reaching the in circulation inside the duct is calculated from the following equation;
thermal steady state condition that depends on the Reynolds number. 2ΔP Dh
All measured values (mass flow rate, heat input, surface temperatures, f=
L ρ U 2 (6)
inlet and exit air flow temperatures, and pressure drop) are fed into
Excel sheets on Laptop via the data acquisition system. All outputs are
recorded to calculate the required parameters; Reynolds number, heat 3.2. Uncertainty analyses
transfer coefficient, Nusselt number, and friction coefficient. In the
present study there are 294 experiments were conducted on the seven The uncertainties (ω ) in the present study were calculated based
bricks settings; 126 runs for the settings with side guide vanes, 126 runs upon the root sum square combination of the effects of each of the
for the settings with U-shape guide vanes and 42 runs for settings individual inputs as introduced by Kline and McClintock [26]. In ad-
without vanes. All the experiments were conducted under a steady-state dition, the bricks and duct dimensions measurements were assumed to
condition when the stable fluid inlet and outlet temperatures are ob- be ± 0.5 mm and the uncertainty applied to the thermal properties of
tained with a maximum variation of 0.5 K for each thermocouple air is assumed to be ±0.1%. The uncertainty of the parameters is cal-
reading. culated. For example, the uncertainty for the average Nu was estimated
as follows;
3.1. Heat transfer and friction factor calculations ωNuave ωh 2
ωD 2 ω 2
= ± ⎛ ave ⎞ + ⎛ h ⎞ + ⎛ k ⎞ = ± 2.4 %
⎜ ⎟ ⎜ ⎟

Nuave ⎝ h ave ⎠ ⎝ D h ⎠ ⎝ k ⎠ (7)

In the present study, the heat transfer and pressure drop were cal-
culated with the same procedures as represented by Refaey et al. [22]. Herein, Table 2 represents the average uncertainties in the main
Therefore, an Excel sheets were organized for the two guide vanes (SV parameters for all experimental runs.
and UV).
From the primary measurements of flow rate, voltage drop, current, 4. Results and discussion
heating bricks surface temperatures, and inlet and outlet air tempera-
tures are fed to heat transfer calculations. The heat losses are neglected The experimental setup has been validated with Abo-Ziyan [20]
and the heat transfer rate to the flowing air is equal to the power with maximum error does not exceed 10.5% as introduced before by
consumed by the heaters. From this energy balance, the local and Refaey et al. [22]. The achieved low values of the average uncertainties

268
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Table 2 Fig. 5 for different settings in absence of vanes and in presence of SV

Average uncertainties in main parameters. with different angles. The local Nusselt number ratio (NuLM7/NuLM8) is
Parameter Uncertainty (ω ) presented to show the roof air gap effect on the heat transfer in all
settings. For all values of Re and for all different settings the values of
have ± 1.48% (NuLM7/NuLM8) are less than unity, these figures show an enhancement
Nuave ± 2.4%
in local Nusselt number for 8-layers in comparison with 7-layers. The
u ± 2.25%
U ± 2.72%
lowest values of local Nusselt number ratio (the maximum enhance-
Re ± 3.31% ment of heat transfer for 8-layers in comparison to 7-layers) is more
ΔP ± 4.26% likely to occur for setting 7 and 2 at Re = 25,600 and Re = 24,900,
f ± 5.48% respectively while the minimum enhancement in heat transfer is about
Q ± 2.12%
2% for setting 5 at Re = 15000 in absence of vanes. It seems that the
combined effect of both Re and bricks arrangement setting on (NuLM7/
and the worthy agreement with former studies show the confidence in NuLM8) is noticed significantly in absence of vanes compared to the side
the experimental setup and the used measurement techniques. The U- vanes with different attack angles. Also, it is clear that the SV with
shape guide vanes (UV) results are presented below to show their effect angles 120° and 135° reduces the values of (NuLM7/NuLM8) compared to
on the heat transfer and pressure drop over the side vanes (SV) tech- the angle 150° as noticed in Fig. 5. The maximum enhancement in heat
nique for all studied settings. The results include the average Nusselt transfer for settings 7 and 2 with 8-layers is obtained due to the forced
number (Nuavg), average heat transfer rate (Qavg), and performance convection with high turbulence generation and this enhancement
criteria (η). Also, the maximum enhancement percentage of normalized changes slightly with Re. For 8-layers with small roof air gap ratio, the
Nusselt number for brick and column spacing changes are listed in ta- trapped air between the roof and the top layer of setting is subjected to
bles. rapid velocities and good mixing associated with high turbulence gen-
erated by side wall vanes with angles 135° and 120°.

4.1. Effect of roof air gap ratio on local Nusselt number

4.2. Local Nusselt number
The roof air gap ratio effects on the local longitudinal Nusselt
number only will be presented firstly in presence of side vanes (SV). The The combined effects of side wall, Reynolds number, and attack
results are presented in a normalized form (NuLM7/NuLM8) where NuLM7 angle on the local Nusselt number heat transfer is presented in Fig. 6 for
and NuLM78 are the local longitudinal Nu for the middle column for 7- longitudinal and transversal bricks for setting 7 in the absence and
layers and 8-layers for the same brick arrangement setting, respectively. presence of U-shape guide vanes (UV). The local and average Nusselt
The experimental results of local Nusselt number ratio along the long- number increase as the Reynolds number increases with all values of
itudinal brick on the middle against Reynolds number are shown in attack angle (θ = 120°, 135°, and 150°) in the direction of the main flow

Fig. 5. Effect of roof air gap ratio on longitudinal Nu for different settings in presence of side vanes (SV) with different angles.

269
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Fig. 6. Nusselt number at fifth row for three columns in setting 7.

with and without UV. In absence of vanes, the rate of increasing of local enhancement percentage that obtained in absence of guide vanes is
Nusselt number decreases for high values of Reynolds number. For all lower than that in the presence of UV for all attack angles. Hence, the
plots, it is observed that the maximum values of local Nu were obtained use of U-shape guide vanes can improve the flow characteristics of
for longitudinal brick at middle column (LM) compared with different longitudinal brick in middle relative to wall columns and the heat
brick positions (LW, TW, and TM) where the values of local and average transfer rate increased due to the turbulence generation with
Nu are closed to each other in the absence of vanes. The maximum accelerated flow. The U-shape guide vanes with θ= 120o generate a
values of local Nu were obtained for LM brick at θ = 135° and large-scale vortex with high turbulence level along their trailing edges
Re = 26,000 and reaches about 226 compared to the other angles. The much more likely than those for other angles. This is attributed to the
maximum local Nu was obtained for LM brick in the absence of vanes good turbulence mixing between columns especially for θ= 120o due to
because of the highly-turbulent nature with high flow velocity on both the trailing-vortex formation towards longitudinal middle brick.
sides of LM brick than that of TW and TM where a stagnation zone is The transversal and longitudinal brick column near tunnel wall has
existed behind it. the minimum values of [(NuLM,UV − Nu LW,NV)/NuLW,NV] when com-
pared with that in the middle column due to the flow separation and
4.2.1. Setting characteristics effects side wall effect. The flow separation at side wall of the longitudinal
4.2.1.1. Wall effect. The present experimental investigation has been brick near tunnel wall is responsible for the quite substantial trailing
carried out to study the side wall effect on the enhancement ratio for vortex wake zone formation, which has a minimum heat transfer en-
four different settings (1, 2, 3, and 4) in the cooling zone of brick tunnel hancement compared to the middle column. The maximum enhance-
kiln. The local Nusselt number results in presence of U-shape guide ment in heat transfer for transverse brick in middle was about 58.01%
vanes are normalized by the local Nu in absence of vanes in the form of (for setting 4, attack angle 135° and Re = 27,336) while the minimum
enhancement ratio percentage. This ratio can be called the relative local enhancement was about 2.45% for setting 2 at Re = 20,989 in the ab-
Nusselt number ratio of longitudinal middle position [(NuLM,UV − Nu sence of vanes as shown in Table 4. The large value of [(NuLM,UV − Nu
LW,NV)/NuLW,NV] where NuLM,UV is the local Nusselt number of LW,NV)/NuLW,NV] for transverse middle brick in setting 4 might be at-

longitudinal brick in middle in presence of UV and Nu LW,NV is the tributed to the combined effect of the high turbulence generated by UV
local Nusselt number of longitudinal brick nearest to side wall in the upstream, the transverse middle brick and the boundary layer distortion
absence of vanes. The maximum enhancement ratio of the relative local
Nusselt numbers [(NuLM,UV − Nu LW,NV)/NuLW,NV] for longitudinal Table 3
middle position are listed in Table 3 in terms of the attack angles and A maximum enhancement percentage obtained for relative local Nu of long-
Reynolds number for the four settings (1, 2, 3, and 4) to show their itudinal brick in middle relative to wall columns (UV).
effects on heat transfer. Heat transfer enhancement is obtained for all
Setting 1 Setting 2 Setting 3 Setting 4
values of both attack angle and Reynolds number. The brick position,
brick setting, and Reynolds number significantly affects the location of % Re % Re % Re % Re
maximum enhancement ratio. The maximum enhancement in relative
No vanes 13.33 23,345 17.32 24,473 15.03 16,088 7.75 27,336
local Nusselt number [(NuLM,UV − Nu LW,NV)/NuLW,NV] of about 68.79%
θ = 150o 42.94 23,345 45.85 22,462 53.84 23,560 34.74 27,336
is obtained for setting 3 in presence of UV with θ= 120o and θ = 135o 45.57 23,345 54.18 22,462 65.88 25,271 48.93 27,336
Re = 23,560 while the minimum enhancement of about 7.75% is θ = 120o 41.41 14,392 68.52 24,473 68.79 23,560 48.93 27,336
obtained for setting 4 in absence of vanes at Re = 27,336. The

270
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Table 4 enhances the heat transfer when located in the middle column. Good
A maximum enhancement percentage obtained for relative local Nu of trans- mixing is provided between the main flow in the middle column and the
verse brick in middle relative to wall columns (UV). fluid flow between columns because the U-shape vanes generate vor-
Setting 1 Setting 2 Setting 3 Setting 4 tices which turn the flow field perpendicular to the main flow direction.
The settings 4, 1, and 3 cause a lowest value for longitudinal Nu, this
% Re % Re % Re % Re indicates that for UV with angles of attack 150°, 135°, and 120° the
blockage and distortion of the flow is more significant as listed in
No vanes 6.56 21,897 2.45 20,989 5.08 16,088 2.52 27,336
θ = 150o 24.06 21,897 21.97 20,989 23.60 23,560 44.27 27,336 Table 1. The maximum increase in the local longitudinal middle Nu for
θ = 135o 27.18 21,897 27.32 20,989 25.76 23,560 58.01 27,336 setting 7 reaches about 1.4 times of setting 3 value at Re = 26,000 and
θ = 120o 27.37 25,480 29.27 22,462 34.97 21,565 55.06 27,336 θ = 135° although there is more void fraction in setting 3. An im-
provement in heat transfer in longitudinal middle positions was ob-
tained for all settings in presence of UV with different attack angles as
due to the flow accelerated between bricks column with large trans- shown in Fig. 7. Furthermore, it is found that the enhancement in the
verse spacing. local longitudinal middle Nu for setting 7 at angle of attack (135°) is
The variations of the local Nusselt number for the longitudinal brick higher than that at angles (120° and 150°). This is attributed to the
in middle column with Reynolds number is shown in Fig. 7 in the ab- turbulence generation with high accelerated flow toward the middle
sence of vanes and in the presence of UV with the three attack angles columns. The maximum local enhancement in longitudinal middle Nu
(150°, 135°, &120°) for different brick arrangement settings. The heat for setting 7 is about 66.7% higher when using UV with at
transfer enhancement for longitudinal middle brick that obtained for θ= 135o comparison with no vanes.
settings 7 and 2 is greater than that for the other settings in absence of
vanes. Furthermore, setting 3 introduces the lowest value of Reynolds
number as shown in Fig. 7. 4.2.1.2. Bricks spacing effect. The effect of bricks spacing for the middle
The rate of change of local Nusselt number versus the Reynolds columns can be obtained from two different pairs of settings (2 & 7, and
number through the longitudinal middle brick is very high for setting 7 5 & 6) as demonstrated in Fig. 4. The results are represented in the form
with U-shape guide vanes at θ = 135°. This increasing in the heat [(NuLM,7 − Nu LM,2)/NuLM,2] where NuLM,7 and NuLM,2 are the local
transfer rate is obtained because setting 7 has a middle column with Nusselt numbers of longitudinal (or transversal) bricks in middle for
large space between bricks as shown in Fig. 4 which increases the settings 7 and 2, respectively. Table 5 represents the maximum
longitudinal flow; hence decreases the film thickness at the brick side percentage increasing in local Nu as the bricks spacing changed from
surfaces which improves the convective heat transfer. For setting 7 in 5 to 26 mm in the presence and absence of UV. It can be noticed that the
presence of U-shape vanes, the wall column pushes the air flow away maximum percentage increasing of about 16% is obtained for pair (5 &
from the adjacent side wall of the tunnel to turn towards both sides of 6) in the transversal brick in absence of guide vanes with higher flow
longitudinal brick in the middle column. The results provided that the superficial velocity. Moreover, the guide vanes give another increasing
longitudinal bricks are subjected to a sufficiently large flow rate that in the longitudinal Nusselt number as shown in Table 5. Regarding pair
(2 & 7), the maximum percentage increasing of about 26.41% is

Fig. 7. Longitudinal Nu versus Re for the middle column of the seven settings.

271
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Table 5 loss for no vanes as a penalty. The values of average Nusselt number
Maximum percentage increasing in local Nusselt number as brick spacing increases almost linearly as Re increases for all plots as shown in Fig. 9.
changes from 5 to 26 mm for flow velocity, u, for guide vanes (UV). This can be attributed to increasing vortex length and velocity values to
Longitudinal middle (LM) Transversal middle (TM) mix the flow around bricks. Moreover, settings 7 and 2 give maximum
enhancement while setting 3 has the lowest value of average Nu.
7&2 5&6 7&2 5&6 Table 7 represents the maximum enhancement percentage in
average Nusselt number due to using UV over SV technique. The table
u % u % u % u %
shows that the maximum enhancement of Nuavg for UV with θ= 135o
No vanes u1 4.95 u2 12.00 u5 7.41 u6 16.05 increases 27.18% more than SV with no significant effect on the pres-
θ = 150o u6 24.30 u1 12.83 u6 9.02 u4 7.22 sure drop. The increasing of the percentage average Nu more than unity
θ = 135o u6 26.41 u1 8.99 u6 13.30 u1 4.65
for different attack angles due to the formation of recirculation zones
θ = 120o u3 13.39 u6 12.02 u6 7.32 u6 4.48
and unstable vortex nearest to the two side walls and the roof of tunnel
kiln. Hence, the boundary layers are destroyed in these zones so the
obtained for the longitudinal brick at higher flow superficial velocity heat transfer increases.
and attack angle 135°.
4.5. Heat exchange average rate
4.2.1.3. Columns spacing effect. Actually, settings 1 and 2 have the same
spacing between bricks in all columns. Therefore, the effect of space Productivity-dependent study is essential when considering the
between columns (S) could be observed in Table 6. As the space average Nu variations with guide vanes for same brick arrangement
between columns increases, the maximum increase in Nu for settings. This may alert the design issues related to energy cost, pro-
longitudinal and transversal bricks reaches about 5.76% and 13.07%, ductivity, and pumping power considerations during cooling in tunnel
respectively for no vanes. Furthermore, there is a maximum increase in kiln. Hence, attention was paid to help in giving practical engineering
the longitudinal Nu by about 47.26% at θ= 120o at the higher solutions to the heat transfer augmentation considerations. The kiln
superficial flow velocity. This is related to the turbulence in the flow productivity is presented in the form of (Qavg = havgAw;bΔT) that has
between columns and the highest flow resistance in the space between been presented by abo Zayian [20] that was based on the assumptions
bricks. of no heat loss and radiation effects. The variations of average heat
transfer rate (Qavg) against the mass flux (M) are shown in Fig. 10 for
4.3. Guide vane type effect different brick settings arrangement and attack angles.
It is found that as the mass flux increases, the average heat transfer
The longitudinal local Nusselt number ratio is presented to explain rate increases rapidly for all plots. For mass flux (M) values ranged from
the augmentation or reduction in heat transfer due to the different 4 to 7 kg/m2. s in the presence of guide vanes, settings 7 and 2 have the
positions of two types of guide vanes (UV and SV). The local Nusselt highest values of heat transfer rate in comparison with other settings.
number of longitudinal middle brick when using U-shape guide vanes On the other hand, by examining the heat transfer rate variations, it is
(NuLM,UV) is normalized by the corresponding one when using side observed that setting 3 provides a lowest enhancement than other cases
vanes (NuLM,SV). This ratio (NuLM,UV/NuLM,SV) of relative local Nusselt in presence or absence of guide vanes. This is attributed to that setting 3
number can show the effect of vane position on the heat transfer for has a high porosity with small solid bricks mass. The maximum en-
different brick settings arrangement. The combined effects of attack hancement in heat transfer rate of about 48% is observed for setting 7
angle, Reynolds number, and brick setting on the heat transfer which with guide vane angle θ= 135o and at a mass flux = 7 kg/m2. s com-
are presented in terms of (NuLM,UV/NuLM,SV) are shown in Fig. 8. From pared with that in the absence of guide vanes. This may be attributed to
Fig. 8, it can be noticed the increase of the relative local Nusselt number the intensive turbulence mixing due to the high void fraction as well as
more than unity for all plots and it increases rapidly to a maximum the heat transfer improvement in average Nusselt number.
value of 1.48 times of that for setting 7 with side vanes at Re = 26,000 In addition, as attack angle decreases from 150° to 120° the average
and θ= 135o . The maximum values of (NuLM,UV/NuLM,SV) differ with heat transfer rate increases for all studied settings as shown in Fig. 10. It
the angle of attack and brick setting. The minimum enhancement of the is found that for low range of mass flux, the production time is reduced
heat transfer occurs at the lowest value of Reynolds number where the for setting 2 while it has a lowest productivity. For large range of mass
relative local Nusselt number for UV reaches to 1.02, 1.04, and 1.06 flux, settings 4 and 1 have a high productivity with moderate produc-
times for that of SV and at attack angles 150°, 120°, and 135°, respec- tion time when using U-shape guide vanes with different attack angles.
tively. It is found that as the Reynolds number increases up to 22,000 Herein, Fig. 11 represents the effect of guide vane type (UV and SV)
the relative local Nu (NuLM,UV/NuLM,SV) increases and thus the en- on the average heat transfer rate for four settings (1, 2, 4, and 7). These
hancement ratio increases for settings 5 and 6 at attack angle 120°. four settings are chosen as two of them (1 & 4) have the highest range of
Then the (NuLM,UV/NuLM,SV) decreases after further increases of Rey- mass flux and the others (2 & 7) have the lowest range of mass flux and
nolds number. This is because of the decreasing of the turbulence largest heat transfer rate. The figure demonstrates that the results
mixing of the boundary layer of the fluid flow near the wall of the
longitudinal middle brick at this small attack angle, and consequently, Table 6
lowest heat transfer takes place from the LM brick to the nearest cold Maximum percentage increasing in Nusselt number as columns spacing (S)
air stream. changes from 19 to 58 mm for (UV).
Flow velocity, u No vanes θ = 150o θ = 135o θ = 120o
4.4. Average Nusselt number
Long. Trans. Long. Trans. Long. Trans. Long. Trans.

Fig. 9 shows experimentally the average Nusselt number variations u1 4.68 9.52 20.62 24.97 25.47 29.46 28.08 28.84
for different brick arrangement settings in presence of UV with three u2 5.10 12.43 29.57 29.10 35.70 34.09 34.54 33.32
different attack angles and compared with no vanes. For Re = 26,000, u3 5.52 14.04 32.07 34.88 38.96 42.42 37.55 39.53
u4 5.76 12.27 33.51 33.66 40.86 39.51 47.26 36.52
best cooling is seen for setting 7 with θ= 135o . Generally, cooling of the
u5 1.50 11.32 27.03 33.00 34.29 39.00 46.39 42.12
brick settings using UV is better than the cooling of the same setting u6 1.57 13.07 25.92 31.89 33.40 37.93 45.90 43.28
without vanes by about 37%–48% with (1.3–1.5) times the pressure

272
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Fig. 8. Effect of vane position on the longitudinal Nusselt number ratio for different settings.

values of the UV are larger than that of SV for all settings (1, 2, 4, and Table 7
7). Moreover, there is a maximum enhancement in average Nusselt Maximum enhancement percentage in average Nusselt number for using U
number of about 17% for setting 7 with UV compared to SV as shown in shape vanes over side vanes.
Table 7. This is attributed to the good mixing which is provided be- θ = 150 θ = 135 θ = 120
tween the main flow in the middle column and the fluid flow between
columns due to using UV. The U-shape vanes generate vortices which % Re % Re % Re
turn the flow field perpendicular to the main flow direction. Therefore,
Setting 1 5.31 23,344 7.89 19,878 19.06 25,479
the production time of settings which use the UV technique is smaller Setting 2 8.10 20,989 10.52 20,989 12.03 22,461
compared to the SV technique. As a result, it is recommended to use the Setting 3 15.49 27,634 12.63 25,270 11.08 21,564
U-shape guide vane over the side wall guide vane to reduce the pro- Setting 4 11.65 27,335 16.32 27,335 18.62 27,335
duction time. Consequently, the energy consumption in tunnel kilns Setting 5 7.13 25,992 14.38 25,992 16.74 22,414
Setting 6 9.78 24,380 10.04 17,966 11.31 17,966
which use the U-shape guide vanes can be reduced. Setting 7 14.64 25,360 27.18 25,360 18.01 25,360

Fig. 9. Average Nu versus Re for all studied settings with the three different vanes angles.

273
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Fig. 10. Average heat transfer rate for the different settings and vanes angles versus mass flux (M).

4.6. Thermal performance analysis pumping power ratio (Qavg/PP) than higher values. The dependence of
(Qavg/PP) on the brick arrangement setting is more significant while the
It is found that the heat transfer features depend greatly on the dependence of (Qavg) on the same settings is not significant for the
attack angle and the type of guide vanes for constant mass flow rate present mentioned range of Reynolds number. The maximum en-
constraint as shown in the previous section. On the other hand, for hancement in heat transfer rate to pumping power ratio (Qavg/PP)
practical application, an equal pumping power constraint must be taken (highest performance) is about 25 for setting 7 in the absence of vanes
into consideration and the heat transfer rate to pumping power ratio while that is obtained in the presence of vanes with angle 150° and 135°
(Qavg/PP) must be provided for comparing the heat transfer perfor- reached about 23 at Re = 14,000. Setting 7 permits air flow improve-
mance in the absence and presence of attack angles. The (Qavg/PP) ratio ment and provide the highest heat transfer (highest performance) at a
considers the pressure drop across the brick arrangement settings. given mass flow rate with a high reduction in pressure drop. The
Hence, the pressure drop across each setting is important regarding the minimum enhancement in heat transfer rate to pumping power ratio
heat transfer enhancement. Fig. 12 shows the effect of Reynolds number (Qavg/PP) was about 3 for setting 4 with and without vanes. It can be
on the heat transfer rate to pumping power ratio (Qavg/PP) for different observed that the performance ratio increases with the decrease of
attack angles and different settings. It is observed that, when the Reynolds number.
pumping power constraint is considered, the (Qavg/PP) ratio has a
lower value than that of equal mass flow rate constraint (Qavg) as shown
previously in Fig. 10. Also, it is found that the low values of Reynolds 4.7. Performance criteria
number have a more significant effect on the heat transfer rate to
To evaluate the performance criteria, an equal pumping power

Fig. 11. Variations of the average heat transfer rate against the mass flux for different vanes positions and different vane angles.

274
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Fig. 12. Variations of heat transfer rate to pumping power (PP) ratio against Re for different settings and angles of attack for UV.

constraint must be considered during the heat transfer process. Figs. 13 and 14 illustrate the combined effects of Re and θ on Nu and
Therefore, combining friction factor of settings with vanes (UV and SV) friction factor ratio at constant pumping power (PP) for the two guide
and without vanes, the following relation is applied [27,28]; vanes; UV and SV, respectively. Generally, Re has a small significant
effect on (η), but the setting and the attack angle have a great influence
Nuθ/Nu
η= on performance criteria (η), as can be observed from Figs. 13 and 14.
(f θ/f)1/3 (8)
As shown in Fig. 13 it can be noticed that the performance criteria

Fig. 13. Performance criteria versus Reynolds number for the different investigated settings for (UV).

275
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

Fig. 14. Performance criteria versus Reynolds number for the different investigated settings for (SV).

(η) , is greater than unity for all the studied settings except for setting 4 S  −0.313 ⎛ S ⎞0.255 ⎛ θ ⎞−0.639
Nuavg = 1.141Re0.427 ⎛ ⎞ ε
with θ= 120o . The figure shows that setting 7 has the best performance ⎝ a ⎠ ⎝ b⎠ ⎝ 180 ⎠ (9)
by about 30% with the UV at Re = 25,360 at θ= 135o .
Fig. 14 represents the performance criteria (η) for side wall guide For SV (side wall guide vanes)
vanes (SV). It can be noticed that, η is greater than unity for all the
studied settings except for setting 4 with all angles and setting 1 with S  −0.259 ⎛ S ⎞0.213 ⎛ θ ⎞−0.346
Nuavg = 2.362Re0.358 ⎛ ⎞ ε
θ= 120o . This can be attributed to the increase in the pressure drop in ⎝ a ⎠ ⎝ b⎠ ⎝ 180 ⎠ (10)
these two settings because they have the largest number of bricks. The
These correlations satisfy the present experimental data within ±
figure shows that nearly setting 7 has the best performance for the three
15%, ± 14 maximum deviation for UV and SV, respectively as shown
attack angles with maximum performance of about 10% at
in Fig. 15 for Reynolds number range (13,609 to ≤ Re ≤ 27,634), and
Re = 25,360.
setting characteristics ratios (0.33 ≤ (S/a ) ≤ 1.0),
(0.84 ≤ (εS/b ) ≤ 3.08) and (120o ≤ θ ≤ 180o ).
4.8. Experimental correlations

From the experimental results for both SV and UV two empirical 5. Conclusions
correlations are obtained for the average Nusselt number (Nuavg) in
terms of Reynolds number (Re), brick setting dimensionless groups ( S ), The cooling zone of brick tunnel kiln is simulated experimentally in
a
S the present work. The thermal performance of the cooling zone is ex-
(ε b ), and attack angle (θ) as follows:
perimentally investigated for seven different brick settings.
Furthermore, an augmentation technique using two different guide
For UV (U-shape guide vanes)
vanes; side wall guide vanes (SV) and U-shape guide vanes (UV) with

Fig. 15. Correlated Nusselt number versus experimental Nusselt number for the two guide vane types.

276
H.A. Refaey et al. International Journal of Thermal Sciences 130 (2018) 264–277

attack angles 120o , 135o and 150o are presented. [4] D.R. Dugwell, D.E. Oakley, A model of heat transfer in tunnel kilns used for firing
The following points could be withdrawn from the experimental refractories, Int J Heat Mass Tran 31 (11) (1988) 2381–2390.
[5] R. Riedel, The optimization of tunnel kilns by utilization of convection,
results: Ziegelindustrie Int 42 (7–8) (1989) 373–379.
[6] G.S. Almeida, J.B. Silva, C.J. Silva, R. Swarnakar, G.A. Neves, A.G. Lima, Heat and
• For the seven bricks arrangement settings, the attack angle for U- mass transport in an industrial tunnel dryer: modeling and simulation applied to
hollow bricks, Appl Therm Eng 55 (2013) 78–86.
shape guide vanes (UV) has a great influence on the average Nusselt [7] H.A. Refaey, E. Specht, Flow field visualization to simulate the burning of sani-
number rather than the side wall guide vanes (SV). taryware in tunnel Kilns, Proceedings of ICFD11: Eleventh International Conference
• As the attack angle of guide vanes decreases, the average Nusselt of Fluid Dynamics December 19-21, 2013 Alexandria, Egypt.
[8] E. Mancuhan, K. Kucukada, E. Alpman, Mathematical modeling and simulation of
number increases for both guide vanes positions.
•
the preheating zone of a tunnel kiln, J Therm Sci Technol 31 (2) (2011) 79–86.
The U-shape guide vane improves the flow characteristics around a [9] S. Kaya, K. Kucukada, E. Mancuhan, Model-based optimization of heat recovery in
longitudinal middle brick and thus the heat transfer rate increases the cooling zone of a tunnel kiln, Appl Therm Eng 28 (2008) 633–641.
[10] V. de Paulo Nicolau, A.P. Dadam, Numerical and experimental thermal analysis of a
compared to the bricks in wall column.
•
tunnel Kiln used in ceramic production, J Braz Soc Mech Sci Eng 31 (4) (2009)
The maximum enhancement in relative local Nu reached about 297–304.
68.79% in the presence of UV and it is obtained for setting 3 at [11] M.F. Naccache, M.S.P. Gomes, A.O. Nieckele, Numerical simulation of flow and
θ= 120o and Re = 23,560. heat transfer through a tunnel kiln, 18th International Congress of Mechanical

•
Engineering November 6-11, 2005 Ouro Preto, MG.
As the space between bricks and columns increases, the longitudinal [12] R.H. Essenhigh, Studies in Furnace Analysis: prediction of tunnel kiln performance
and transversal Nusselt number increase in the absence and pre- by application of the integral energy equation, Energy Fuel 15 (3) (2001) 552–558.
sence of both guide vanes (UV and SV). [13] G.M. Santos, Study of thermal behavior of a tunnel kiln used in red ceramic in-

•
dustry, M.Sc. dissertation Federal University of Santa Catarina, Santa Catarina,
The maximum enhancement of about 94.5% was found for long- Brazil, 2001 (in portuguese).
itudinal brick at middle column in setting 2 at θ= 135o , and [14] H.A. Refaey, E. Specht, M.R. Salem, Influence of fuel distribution and heat transfer
Re = 22,407. on energy consumption in tunnel Kilns, Int J Adv Eng Technol 8 (3) (2015)

• The values of average Nusselt number increase almost linearly as Re

281–293.
[15] D.R. Dugwell, D.E. Oakley, Correlation of convective heat transfer data for tunnel
increases and the rate of increasing depends upon the setting char- kilns, J Inst Energy 61 (1988) 165–171.
acteristics. [16] H. Roth, Combined heating and power systems in brickworks, Ziegelindustrie Int 46

• Setting 7 permits air flow improvement and provides the highest (9) (1993) 545–552.
[17] Nahed Soussi, Wassim Kriaa, Hatem Mhiri, Philippe Bournot, Reduction of the
heat transfer (highest performance) at a given mass flow rate with a energy consumption of a tunnel kiln by optimization of the recovered air mass flow
high reduction in pressure drop. from the cooling zone to the firing zone, Appl Therm Eng 124 (2017) 1382–1391.

• Settings 7 and 2 provide the highest values of heat transfer rate in [18] J. Durakovic, S. Delalic, Temperature field analysis of tunnel kiln for brick pro-
duction, Material Geoenviron 53 (3) (2006) 403–408.
compared to other settings while setting 3 has a lowest enhance- [19] S.A. Karaush, Y.I. Chizhik, E.G. Bober, Optimization of ceramic setting as a function
ment. of their heat absorption from the radiating walls of the furnace, Glass Ceram 54 (5)

• The maximum enhancement in heat transfer rate reached about (1997) 190–192.
[20] H.Z. Abou-Ziyan, Convective heat transfer from different brick arrangements in
48% for setting 7 compared with that in the absence of guide vanes tunnel kilns, Appl Therm Eng 24 (2004) 171–191.
for θ= 135o and M = 7 kg/m2. s [21] T. Ros-Dosdá, I. Celades, E. Monfort, P. Fullana-i-Palmer, Environmental profile of

• Generally, settings 4 and 1 provide a moderate production time with Spanish porcelain stoneware tiles, Int J Life Cycle Assess (July, 2017) 1–19.
[22] H.A. Refaey, Ali A. Abdel-Aziz, R.K. Ali, H.E. Abdelrahman, M.R. Salem,
highest productivity while setting 2 has lowest productivity with Augmentation of convective heat transfer in the cooling zone of brick tunnel kiln
small production time. using guide vanes: an experimental study, Int J Therm Sci 122 (2017) 172–185.
• As the vane attack angle decreases the average heat transfer rate [23] Tino Redemann, Eckehard Specht, Mathematical model to investigate the influence
of circulation systems on the firing of ceramics, Energy Procedia 120 (2017)
increases and thus the enhancement ratio increases. 620–627.
• Based on the experimental results, the average Nusselt number [24] M.V. Araújo1, R.S. Santos, R.M. Silva, A.G. Barbosa de Lima, Drying of industrial
hollow ceramic brick: analysis of the moisture content and temperature parameters,
correlation is obtained related Reynolds number, brick setting di-
Defect Diffusion Forum 380 (2017) 72–78.
mensionless groups, and angle of attack.
[25] Massimo Milani, Luca Montorsi, Matteo Stefani, Roberto Saponelli,
Maurizio Lizzano, Numerical analysis of an entire ceramic kiln under actual oper-
References ating conditions for the energy efficiency improvement, J Environ Manage 203
(Part 3) (2017) 1026–1037.
[26] S.J. Kline, F.A. McClintock, Describing uncertainties in single-sample experiments,
[1] H.A. Refaey, Mathematical model to analyze the heat transfer in tunnel kilns for Mech Eng 75 (1) (1953) 3–8.
burning of ceramics, Ph.D. dissertation Otto- von Guericke University, Magdeburg, [27] R.L. Webb, Performance evaluation criteria for use of enhanced heat transfer sur-
Germany, 2013. faces in heat exchanger design, Int J Heat Mass Tran 24 (1981) 715–726.
[2] Y.U. Boming, Dynamic modeling of tunnel kiln, Heat Tran Eng 15 (2) (1994) 39–53. [28] J.F. Fan, W.K. Ding, J.F. Zhang, Y.L. He, W.Q. Tao, A performance evaluation plot of
[3] A.H. Tehzeeb, M. Bhuiyan, N. Jayasuriya, Evaluation of brick Kiln performances enhanced heat transfer techniques oriented for 0 energy-saving, Int J Heat Mass
using computational fluid dynamics (CFD), Energy Environ Eng J 1 (2) (2012) Tran 52 (2) (2009) 33–34.
86–93.

277