Solar Energy 137 (2016) 500–515

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

Analysis of drying of melon in a solar-heat recovery assisted infrared

dryer
Mustafa Aktasß a, Seyfi Sß evik b,⇑, Ali Amini c, Ataollah Khanlari d
a
Gazi University, Technology Faculty, Energy Systems Engineering, Teknikokullar, 06500 Ankara, Turkey
b
Hitit University, Vocational School of Technical Sciences, Electrical and Energy, 19169 Çorum, Turkey
c
Atatürk University, Natural and Applied Science Institute, 25240 Erzurum, Turkey
d
Gazi University, Natural and Applied Science Institute, 06500 Ankara, Turkey

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

Article history: Infrared drying systems are popular in terms of high heat and mass transfer. By using an infrared dryer, it
Received 21 September 2015 is possible to catch fast heating and short drying time in comparison to the other drying methods. But it
Received in revised form 8 March 2016 consumes a high amount of energy. Therefore, a new type solar air collector (SAC) and air to air heat
Accepted 23 August 2016
recovery unit were added to the infrared dryer to reduce specific energy consumption. The general
Available online 6 September 2016
aim of this study is to analyze heat and mass transfer characteristics of the dryer and three-
dimensional (3-D) computational fluid dynamic (CFD) simulation and to investigate drying kinetics of
Keywords:
melon slices. Experiments were performed at 50 °C and 60 °C melon’s surface temperature and 0.5 m/s
Solar energy
Infrared dryer
air velocity. Melon slices were dried from 9 g water/g dry matter to 0.044 g water/g dry matter moisture
Heat and mass transfer content. The effective moisture diffusivity (De) values varied from 8.25
1010 to 1.24
109 m2/s. The
Melon drying average mass transfer coefficient (hm) values increase from 8.53
108 m/s at 50 °C to 1.47
107 m/s at
60 °C. Heat recovery unit has a key role in this system and it provides 23–28% of total input energy.
Average solar air collector efficiency was calculated as 50.6%. Obtained theoretical and experimental
results are in line with each other. This study shows the successful and efficient combination of solar
energy, infrared energy and heat recovery in food processing.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction
Drying is a part of the post-harvest process and it encompasses
a sequence of activities and operations that can be divided into two
main categories as natural and technical. Solar energy as an energy
source is used in both of these methods. In drying systems, energy
needed for drying is key point for evaporating the amount of mois-
ture from the product. Infrared drying as one of radiation type
methods is preferred since heat and mass transfer rate are high.
However, it has high energy consumption. Therefore, researchers
are turning to the combined dryer. Many researchers have studied
on the performance of single infrared dryer or combined one. They
used different techniques such as infrared drying (Nowak and
Lewicki, 2004; Toğrul, 2006; Nasıroğlu and Kocabıyık, 2009), infra-
red and convective drying (Kumar et al., 2006; Jaturonglumlert and
Kiatsiriroat, 2010; Supmoon and Noomhorm, 2013; Yinqiang et al.,
2014). They reported that infrared dryer has important advantages
to maintain product quality and to increase the drying rate. Recent
⇑ Corresponding author.
E-mail addresses: mustafaaktas@gazi.edu.tr (M. Aktasß), seyfisvk@hotmail.com
(S. Sßevik), ali82amini@gmail.com (A. Amini), ata_khanlari@yahoo.com (A. Khanlari).
studies show that infrared technology has priority over hot air dry-
ing (Raksakantong et al., 2011; Yinqiang et al., 2014). Combining
convective and infrared drying enhance the drying rate (Yang
et al., 2010; Supmoon and Noomhorm, 2013).
Solar air collector (SAC) can be used in drying applications
because it is feasible and economical in drying applications. They
can be used directly, indirectly or in a combination mode (S ß evik,
2014). Also, some researchers have studied different type of solar
collectors (Kavak Akpinar, 2010; Sßevik, 2014; Ramani et al.,
2010). Dissa et al. (2016) designed a new solar air collector with
a composite absorber. Its composite absorber composed of cou-
pling a non-porous absorber made of a corrugated iron sheet and
a porous absorber made of a mesh of aluminum. This air collector
is appropriate to use in drying applications because it can rise the
air temperature between 50 °C and 75 °C. Croitoru et al. (2016)
analyzed and redesigned an unglazed transpired solar collector
preheat the fresh air. They increased thermal efficiency by using
a new geometry with innovative perforation. This collector used
as dryer in some countries.
Melon (Cucumis melo L.) belongs to the cucurbitaceae family.
100 g of melon contain 8 g carbohydrate with glucose and fructose
being the most predominant sugars and also contain 1 g fiber,

http://dx.doi.org/10.1016/j.solener.2016.08.036
0038-092X/Ó 2016 Elsevier Ltd. All rights reserved.
 M. Aktasß et al. / Solar Energy 137 (2016) 500–515 501

Nomenclature

A area, m2 V reduced volume, m3

aw water activity V0 initial volume, m3
Bi Biot number V_ volumetric flow rate of air, m3/s
c specific heat capacity, kJ/kg K v kinematic viscosity, m2/s
C1 water concentration of air, kg/m3 q density, kg/m3
Cw water concentration near melon, kg/m3 g efficiency, %
D mass diffusivity, m2/s x specific humidity, kg water/kg dry air
Di Dincer number r Stefan–Boltzmann constant, 56.7
1012 kW/m2 K4
De effective diffusivity coefficient, m2/s
emission coefficient
E energy, kJ u any conserved property
e(t) instantaneous process error at time cu exchange coefficient of the entity


h enthalpy, kJ/kg z direction of diffusion process, m
h convection heat transfer coefficient, W/m2 K
hm mass transfer coefficient, m/s Abbreviations
hw wind convective coefficient, W/m2 K DR drying rate, g water/g dry matter per minute
hfg latent heat, kJ/kg MR moisture ratio, %
I instantaneous radiation, W/m2 PIR performance increase ratio, %
KP proportional gain SEC specific energy consumption, kW h/kg
k thermal conductivity of air, W/m K
kc drying constant, 1/min Subscripts
L product thickness, m a air
Ls tray length, m
aa ambient air
Lw latent heat of vaporization for water, kJ/kg abs absorbed
Md final dry weight, g ac accumulated
Mi initial wet weight, g C chamber
Mr weight after rehydration, g water
conv convection
m mass, kg db dry basis
m_ mass flow rate, kg/s d dry
mw vaporized water weight, kg
de drying efficiency
Nu Nusselt number dp drying process
Qp energy for the product heating, kJ el electricity
Qw energy for the moisture evaporation, kJ em emitted
PIR infrared lamp power, W
0 reference
Pf fan power, W ex exhaust
Pout output of the proportional controller eva evaporation
p0 controller output with zero error f fan
p vapor pressure of water, kPa
HR heat recovery
p0 vapor pressure of pure water, kPa IR infrared
Pr Prandtl number i initial
R function uncertainty ia inlet air
Ra ideal gas constant, kJ/kg K
inp inputs
Re Reynolds number mel melon
S shrinkage mp moisture production
Sc Schmidt number
oa outlet air
Sh Sherwood number o outlet
Su the source rate of u p product
X moisture content, g water/g dry matter pl plate
WR total uncertainty, %
sac solar air collector
w1, w2, wn uncertainties in the independent variables si initial surface
T temperature, °C su upside skin
T0 reference temperature, 0 °C TOT total
t time, min
w water
U heat loss coefficient, W/m2 K eq equilibrium
u velocity, m/s x independent variables

12 mg calcium, 17 mg phosphorous and 31 kcal energy (Amiri
et al., 2014). The melon has high water content, vitamins A and
C, beta-carotene and potassium. However it has a low caloric value
(Gil et al., 2006; Solval et al., 2012). World melon production is
about 25 million tons in about one million hectares and Turkey
is a major melon producer (FAO, 2010).
Melon is a short-lived product and typically is consumed raw. Its
supply to supermarkets is limited to harvest season; however, its
shelf life is limited to 15 days (Ayhan et al., 1998). Many studies
have been done to prolong shelf life. For example; Mahmoud et al.
(2008) extended the shelf life of cantaloupe with the treatment
5.0 mg1 ClO2 by 6 days compared to the untreated storage. But,
drying method can be better solution than chemical additives for
shelf life prolongation. Dried melon is a healthy snack that is rich
in antioxidants, vitamins, and minerals (Berdıyev et al., 2009). Dry-
ing of melon has been studied by few researchers (Solval et al., 2012;
Rodrigues and Fernandes, 2007; Chayjan et al., 2012; Darvishi et al.,
2015). Solval et al. (2012) reported that the juice powder produced
502 M. Aktasß et al. / Solar Energy 137 (2016) 500–515

at low temperature had higher moisture content and higher vitamin through the heat exchanger and it is subsequently exhausted
C and b-carotene than those produced at high temperature. outside.
Rodrigues and Fernandes (2007) obtained that the best operating
condition for the osmotic dehydration of melon was 42.5 °C. 2.2. Experimental procedure
Solar and infrared drying systems have been applied separately
to the different dryers in literature, but there is not any study Fresh melons were bought from the local market during the
related to solar heat recovery assisted infrared dryer (SHRAIRD). summer season of 2015. Melon’s dry weight was determined by
Also, studies on drying characteristics of melon slices are limited. drying melon samples in a drying oven at 70 ± 1 °C and atmo-
Therefore, the main objective of this study is to discuss on the spheric pressure. At the end of two consecutive measurements,
applicability of the SHRAIRD system, and the effects of the system samples were considered totally dry on condition that the weight
on the drying kinetics of melon under various conditions. For this change is below 1% (AOAC, 1990).
purpose, heat and mass transfer analysis, and computational fluid For each drying experiment, 2.5 kg of melon slices (20 mm
dynamics (CFD) analysis was done to observe flow characteristics width, 100 mm length, 10 mm thickness) were selected for drying.
in the system, also thermophysical properties during drying of pro- The experiments were conducted at 50 °C, and 60 °C set tempera-
duct were observed. tures and 0.5 m/s air velocity. When moisture content of the sam-
ples decreased to 0.044 g water/g dry matter the drying process
was stopped.
2. Material and methods
2.3. Analysis of drying system
2.1. Experimental setup
2.3.1. Theoretical calculations
SHRAIRD consists of three main parts including; solar air collec- The following equations were used to analyze the energy and
tor (SAC), a heat recovery unit and drying cabinet. Drying air pre- mass balance of the drying process. The mass and energy balance
heated in the SAC, then air temperature increases by passing during a drying process is symbolized in Fig. 2.
through a heat recovery unit and enters to the drying cabinet. In The mass and energy conservation equations can be expressed
the drying cabinet six rod types near infrared lamps (6
500 W) by the following equations (Çengel and Boles, 1994):
were mounted in three rows (with two lamps in each row), that The mass conservation of the drying air is
each of them emits radiation with a peak wavelength of
2400 nm. The single pass forced convection type SAC consists of _ ia ¼ Rm
Rm _ oa ð1Þ
a finned absorber plate, collector box, insulation, a blower and The mass conservation of moisture is
temperature sensors. Solar-heat recovery assisted infrared dryer
(SHRAIRD) is shown in Fig. 1. _ ia  xia þ m
Rðm _ oa  xoa
_ mp Þ ¼ Rm ð2Þ
K type thermocouples (Range 200 °C to +850 °C, supply 24 V- The energy conservation in the drying process is
DC, output 4–20 mA, accuracy ±0.4 °C) and thermo hygrometers !
2 2
(0–100% relative humidity, 40 to +70 °C temperature range, accu- V  V ia
Q_ dp  W
_ ¼ Rm
_ ia 
hoa 
hia þ oa ð3Þ
racy ±2% RH and ±1 °C) probes were placed in different point to mea- 2
sure temperature and relative humidity. To obtain uniform drying
Heat content of air is
conditions in drying chamber infrared lamps were controlled pro-
portionally according to dried product surface temperature.
ha ¼ cs  ðT a  T 0 Þ þ xhfg ð4aÞ
Drying air temperature was increased by passing through a
cs ¼ 1:005 þ 1:88x ð4bÞ
solar air collector and a heat exchanger. Then heat and moisture
transfer in drying chamber is performed. Moist air is passed cross Dryer energy balance is

Fig. 1. Solar-heat recovery assisted infrared dryer (SHRAIRD).

 M. Aktasß et al. / Solar Energy 137 (2016) 500–515 503

Fig. 2. Mass and energy balance.

 0:5
Energy of inlet air þ Gained energy to drying air by SACðQ_ sac Þ hfg ¼ 7:33
1012  1:6
107 T 2 for 65:57 6 Tð CÞ 6 260
þ Recovered heat by HRðQ_ HR Þ
ð13bÞ
þ Energy delivered by IR lampsðQ_ IR Þ
h The concentration of water in the vicinity of melon surface can
¼ Consumed energyðQ_ CE Þ þ Total heat lossesðQ_ loss Þ be calculated by applying ideal gas law for water vapor at satura-
i tion pressure that can be seen in Eq. (14) (Çengel and Ghajar, 2011):
þExhausted energyðQ_ e Þ ð5Þ
psu
Cw ¼ ð14Þ
Ra T su
2.3.1.1. Energy analysis. The system energy inputs consist of electric The energy that is emitted by melon (Qem) can be calculated as
energy, solar energy and recovered energy from air to air heat (Çengel and Ghajar, 2011):
exchanger.  
Einp ¼ Eel þ Esac þ EHR ð6Þ Q em ¼
rA ðT su þ 273:15Þ4  ðT aa þ 273:15Þ4 ð15Þ

Input electric energy (Eel) in drying system is divided into two
main thermal and mechanical parts including infrared lamps and
fans respectively. Energy consumed by IR lamps and fans was cal-
culated using Eq. (7):
Eel ¼ EIR þ Ef ¼ ðPIR tIR þ P f tf Þ  60
Thermal energy gained by product via infrared lamp is divided
into five parts including; heating of melon (Qmel), evaporating
melon moisture (Qeva), convective heat transfer from melon to air
flow (Qconv), emitted heat by melon (Qem) and heat loss (Qloss). Eq.
(8) explains this theory by mathematical expression:
EIR ¼ Q mel þ Q conv þ Q ev a þ Q em þ Q loss
There are some heat losses in drying chamber that is mainly
caused by convection heat transfer between the chamber walls
and the ambient air. The heat loss (Qloss) is calculated by subtract-
ing the other four energy terms (Qmel + Qconv + Qeva + Qem) from
thermal energy in Eq. (8).
ð7Þ
2.3.1.2. Solar energy. The gained energy by drying air passing
ß evik,
through SAC can be found by using the following equation (S
2014):
Accumulated energyðQ_ ac Þ þ Energy gainðQ_ sac Þ
ð8Þ ¼ Absorbed energyðQ_ abs Þ  Lost energyðQ_ loss Þ ð16aÞ

Thermal energy needed for heating of melon from initial tem-

_ pl  c  ðdT pl =dtÞ þ V_ sc  q  c  ðT sac;oa  T aai Þ
m
perature to steady state is calculated in Eq. (9) by considering tem-
perature differences between initial and steady conditions of ¼ g  I  Asac  Asac  U cv  ðT pl  T aa Þ ð16bÞ
melon.
The amount of gained energy in SAC can be found by the follow-
Q mel ¼ m0 C p;mel ðT su  T si Þ ð9Þ ing equation;

Q conv ¼ h AðT su  T i Þ ð10Þ Q_ sac ¼ V_ sac  q  c  ðT o  T i Þ ð16cÞ

Heat content of product; Thermal losses occur due to the wind. To calculate the thermal
  losses, the wind convective coefficient can be calculated by using

hp ¼ cp;p  ðT p  T 0 Þ þ X  cp;w  T p  T 0 ð11Þ the following equation (Duffie and Beckman, 2006):
The thermal energy is required to vaporize product moisture. hw ¼ 5:7 þ 3:8 u ð17Þ
This energy is related to water concentration of melon and air,
latent heat and mass transfer. The thermal energy used for evapo- The instantaneous thermal efficiency of the solar air collector is
rating moisture from product is calculated using Eq. (12) (Çengel defined as follows (Duffie and Beckman, 1991);
and Ghajar, 2011): R t2 _
Q sac dt
Q ev a ¼ hm ðC w  C 1 ÞA hfg ð12Þ g¼ t1
R t2 ð18Þ
Asac t1 ITOT dt
Heat of evaporation as a function of drying air temperature is
given by Eq. (12) (Aghbashlo et al., 2012):
2.3.1.3. Recovered heat. The energy (Qa) that is carried out by air
hfg ¼ 2:503
106  2:386
103 T for 0 6 Tð CÞ 6 65:57 ð13aÞ flow is a function of air flow and enthalpy differences between dry-
ing chamber inlet and outlet. A part of this energy (QHR) can be
504 M. Aktasß et al. / Solar Energy 137 (2016) 500–515

recovered by heat recovery unit and remaining energy (Qex) is Schmidt number that is the ratio of momentum diffusivity and
exhausted to atmosphere that can be expressed by Eq. (19). mass diffusivity (Çengel and Ghajar, 2011):

Q ex ¼ Q a  Q HR ð19aÞ
t
Sc ¼ ð27Þ
D
_ a ðh
Qa ¼ m
o 
hi Þt  60 ð19bÞ Transportation of moisture inside a melon slice can be
described by a diffusion model according to Fick’s second law
_ a C p;a ðT o  T ex Þt  60
Q HR ¼ m ð19cÞ (Thuwapanichayanan et al., 2011).
 
_ a C p;a ðT o  T i Þt  60 @X @ @X
Q HRmax ¼ m ð19dÞ ¼ De ð28Þ
@t @x @x
The specific heat capacity of the inlet air can be expressed by a
The moisture content (Xdb);
polynomial function of temperature that has been shown in Eq.
 
(20) (Zare et al., 2006): Mi  Md
X db ¼ 100 ð29Þ
Md
C p;a ¼ 1009:26  0:0040403
T þ 0:00061759
T 2
 0:0000004097T 3 ð20Þ Initial and boundary conditions of Eq. (28) are described as fol-
lows (Thuwapanichayanan et al., 2011):
The performance increase ratio (PIR) of the system that uses  
L
heat recovery is defined as follows: X ; 0 ¼ X0 ð30aÞ
2
g  g1
PIRHR ¼ 2 ð21Þ
g1 @Xð0; tÞ
¼0 ð30bÞ
where g1 is the system efficiency without heat recovery and g2 is @z
the system efficiency with heat recovery. PIR of the system that  
uses solar collector-heat recovery is defined as follows:
X 2L ; t
De ¼ hm ðX  X eq Þ ð30cÞ
@z
g  g3
PIRsacHR ¼ 4 ð22Þ
g3 Solving Eq. (28) according to the conditions in Eq. (30) for the
slab bodies moisture diffusion can be expressed as follows
where g3 is the system efficiency without solar collector-heat (Crank, 1975);
recovery and g4 is the system efficiency with solar collector-heat  
X  X eq 8 X
1
1 2 2 De
recovery. MR ¼ ¼ 2 exp ð2n þ 1Þ p t ð31Þ
The mass transfer process is controlled by moisture diffusivity X 0  X eq p n¼0 ð2n þ 1Þ2 L2
of melon in contact with air flow. The procedure for determination
For long drying periods, the first term of the series can be used
of this parameter is described as follows. Convection heat transfer
for simplifying diffusion description,
is related to flow regime type, so analyzing the flow type is done by
   
calculating the Reynolds number as follows (Çengel and Ghajar, 8 p2 De
LnðMRÞ ¼ Ln  t ð32Þ
2011): p2 L2
uLs
Re ¼ ð23Þ
t 2.3.1.4. Mass transfer coefficient. Convective mass transfer coeffi-
u is the mean velocity of air with 1 cm distance from melon’s top cient can be calculated from the Biot number (Bi) description in
surface that was 0.25 m/s in this study. By substituting L and m val- Eq. (33) (Çengel and Ghajar, 2011).
ues in Eq. (23), it is shown that external laminar flow is happened in hm L
this case. Eq. (24) is used for calculating the Nusselt number in lam- Bi ¼ ð33Þ
De
inar flow (Çengel and Ghajar, 2011):
The Biot number can be calculated according to Dincer number
Nu ¼ 0:453Re0:5 Pr 1=3 ð24Þ (Di). Relationship between two dimensionless numbers are given
in Eq. (34) (Dincer and Hussain, 2002).
Convection heat transfer coefficient of melon in contact with air
flow can be calculated by Eq. (25). 24:85
Bi ¼ 0:375
ð34aÞ
Nu k Di
h¼ ð25Þ
Ls u
Di ¼ ð34bÞ
The Sherwood number (Sh) is the ratio of mass transfer resis- kc L
tance to the diffusivity. Mass transfer coefficient (hm) is determined
where k is drying constant that can be defined by the following
based on Sh and it is defined based on characteristic length. The
expressions (Aboltins, 2013)
relationship between the hm, Sh and Schmidt number (Sc) are given
as follows; dX
kc ðX  X eq Þ ¼ ð35aÞ
hm L dt
Sh ¼ ð26aÞ Drying rate (DR);
D
dX X tþdt  X t
Sh D ¼ ð35bÞ
hm ¼ ð26bÞ dt dt
L

Sh ¼ 0:453Re0:5 Sc1=3 ð26cÞ 2.3.1.5. Specific energy consumption. Specific energy consumption
(SEC) in drying of melon is calculated by using Eq. (36) (Kassem
et al., 2011).
 M. Aktasß et al. / Solar Energy 137 (2016) 500–515 505

EIR þ Ef þ Esac chamber has been simulated in the DESIGN-MODELER subprogram

SEC ¼ ð36Þ
mw of the ANSYS software. Chamber frame has 114 ⁄ 64 ⁄ 37 cm
dimensions with 10 ⁄ 40 cm canal cross section area for input
and output port on both sides. Position of 6 infrared lamps with
2.3.1.6. Energy-drying efficiency. Energy efficiency is defined as the
5 cm distance from the top wall of the chamber and melon slices
ratio of the energy used for evaporation of moisture from the sam-
on a 40 ⁄ 80 cm area, that have been located 20 cm under the
ple to the total energy consumption.
lamps xy plane. Fig. 3b shows the melon geometry in detail as a
Qw meniscus slice with 8 cm inner diameter and 10 cm outer diame-
gee ¼ ð37Þ
RE ter, 130 degrees of arc angle and 1 cm thickness. Tray for melon
The energy used to evaporate water; spreading has not been modeled in this study to simplify the model
and avoid complicated mesh treatment procedures. As the whole
Q w ¼ hfg mw ð38Þ geometry of the chamber is symmetric by yz plane, the half geom-
The drying efficiency of the systems can be formulized as etry of drying chamber is recommended in Fig. 3 to reduce the ana-
below: lyze time and avoiding iterative operations.

Qw þ Qp 2.3.2.2. Meshing treatment. The meshing step of the drying cham-

gde ¼
RE
The demanded energy to heat the wet product was calculated
using Eq. (40).
Q p ¼ mp cp;p ðT p;o  T p;i Þ
The output of a proportional temperature controller used in the
IRD can mathematically be expressed as:
Pout ¼ K P eðtÞ þ p0
In experimental studies, uncertainty values are important to
evaluate obtained data. Total uncertainty values are estimated
through the following equation:
"
2  2  2 #1=2
@R @R @R
WR ¼ w1 þ w2 þ  þ wn
@x1 @x2 @xn
2.3.1.7. Product quality. Shrinkage occurs during melon drying. The
amount of shrinkage for dried melon is calculated as follows:
S ¼ V=V 0
Water activity (aw) is a critical parameter that determines the
quality value and shelf life of dried foods. The aw is calculated as
follows:
aw ¼ p=p0
The rehydration capacity of the dried samples is important as a
quality parameter in drying studies. The rehydration capacity (RC)
is calculated as follows (Doymaz, 2008):
RC ¼ Mr =M d
2.3.2. CFD simulation
The CFD model was developed using FLUENT. The CFD simula-
tion is used extensively in drying analysis to obtain uniform drying
and predict airflow distribution in the drying chamber due to uni-
form airflow and temperature distribution in drying chamber
affects the efficiency. Recently, it was used for solar collectors
and dryers, too. In order to determine the suitability of the design
of drying chamber, data analysis was carried out using the CFD.
2.3.2.1. Geometry design. SAC geometry specifications are based on
measurements of device dimensions in the energy system labora-
tory. Fig. 3 shows the sketch that has been simulated in DESIGN-
MODELER subprogram of the ANSYS software. Collector frame
has 80 ⁄ 50 ⁄ 11 cm dimensions with 3 cm EPS insulation material
on the walls. Position of 35 fins with 6 cm width, 10 cm height and
0.2 cm thickness has been shown in Fig. 3a.
Drying chamber geometry specifications are based on measure-
ments of prototype dimensions in the laboratory. Geometry of the
ð39Þ
ber has been done in the MESH subprogram of the ANSYS software.
Different sizing parameters have been checked to achieve a high
orthogonally factor and low aspect ratio. Mesh domain has two
parts including solid and fluid with the contact interface between
ð40Þ them. By applying the parameters mentioned above, it is pre-
dictable that mesh density nearside the IR lamps will be higher
than other vacant positions in drying chamber.
Different mesh sizes were used to test grid independence in
ð41Þ both drying solar collector and drying chamber. It was started with
coarse mesh (A) to finer mesh (C) in the solar collector and also
coarse mesh (D) to finer mesh (F) in drying chamber domains.
The average value of temperature along the measure line at solar
collector and the average value of velocity at drying chamber along
the measure line were monitored for three mesh sizes that have
ð42Þ been shown in Table 1.
For the 1.263 (m/s) uniform inlet air temperature in solar col-
lector, measured data showed a smooth increase in average tem-
perature magnitude. But for mesh sizes finer than mesh (C), the
temperate rising did not detect. Therefore, the mesh (C) with
2451000 cells was selected for the simulation analysis.
ð43Þ
For the 0.01701 (kg/s) inlet air mass flow and 4.663e5 (kg/s)
water evaporation rate from melon surfaces in drying chamber,
measured data showed a fluctuation in average velocity magni-
tude. But for mesh sizes finer than mesh (F), the fluctuations did
not detect. Therefore, the mesh (F) with 1668000 cells was selected
ð44Þ
for the simulation analysis. A detail view of the mesh domain for
solar collector and drying chamber has been presented in Fig. 4.
Major assumptions made to formulate the model were listed as
below.
ð45Þ
 Ideal gas mixture
 Steady state flow
Due to the fact that the main flow evolves in all three directions,
the use of three-dimensional grid is necessary. As a consequence,
the dependent variables of the problem are the three velocity com-
ponents u, v and win the x, y and z directions. Because the flow is
forced, the air density and viscosity can be assumed to be constant.
The physical problem can be described by a set of differential equa-
tions having the general form (Mathioulakis et al., 1998):
@ðquÞ  
þ div quu  r cu rðuÞ ¼ Su ð46Þ
@t
In the current work, the numerical computation is considered to
have converged when the scaled residual was 103 for the mass and
velocity components and about 106 for the energy component.
2.3.2.3. Flow, heat and mass transfer analysis. Flow analysis inside
the collector and drying chamber has been done in the FLUENT
subprogram of the ANSYS software. Energy equation with k-e stan-
506 M. Aktasß et al. / Solar Energy 137 (2016) 500–515

Fig. 3. Geometry of solar collector (a) and drying chamber (b).

Table 1
Grid-independence results for control volumes.

Solar collector Drying chamber

Mesh name A B C D E F
Mesh size (elements) 1615000 2015000 2451000 1640000 1644000 1668000
Mesh size (nodes) 499000 615000 741000 347000 348200 355800
Average value on measured line 320.5 (°C) 321.7 (°C) 322.9 (°C) 0.347 (m/s) 0.352 (m/s) 0.35 (m/s)

dard wall function as one of the turbulent models has been
selected to solve the coupled equations. Solid materials of collector
are copper for fins and air for chamber inlet channel that their
properties are based on the FLUENT database. The analysis process
composes of one step including air mass flow at inlet port by 310
(K) temperature and 100 (m3/h) flow rate. Heat flux to the horizon-
tal part of the collector (the average value of radiation rate) was
taken 500 (W/m2) and convection heat transfer of 20 (W/m2 K) in
the contact region of ambient air and insulated walls are intro-
duced as boundary conditions.
Table 2 shows the model specification in this study. Fluid mate-
rials for drying chamber are water vapor for melon slices evapora-
tion process and air for the chamber inlet channel that their
properties are based on the FLUENT database. Average air mass
flow, vapor mass flow and temperature values are introduced to
the appropriate boundaries to get flow characteristics inside the
drying chamber. The average values of mentioned parameters have
been calculated based on experimental data.
2.3.3. Determination of the drying parameters
Before experiments, it is necessary to choose a procedure to
determine the main parameters of drying. In this study the param-
eters necessary for drying were determined. Firstly, initial mois-
ture content of melon was calculated, that was obtained as 9 g
water/g dry matter (dry basis). Many studies on air-drying have
been showed that drying air temperature 60 °C can be considered
a satisfactory condition (Demirel and Turhan, 2003; Karim and
Hawlader, 2005; Rodrigues and Fernandes, 2007). Also, in our pre-
vious work (Aktasß et al., 2012; Sßevik, 2014), we have seen that dry-
ing air temperature of 50 °C is ideal and entails great risks of high
 M. Aktasß et al. / Solar Energy 137 (2016) 500–515 507

the dryer consisting of the infrared lamps (in three rows, with
two lamps in each row) and blower (100 m3/h, 50 Hz, 230 V) are
measured by an electricity meter (220 V, 50 Hz, one-phase, LCD
display, accuracy ±1% and uncertainty ±1.41%).

3. Results and discussion

This section describes the main results of CFD and experimental

analysis, including drying behavior and product quality. It also dis-
cusses obtained results and some of their possible implications.

3.1. CFD simulation results

Calculating for collector geometry was done and convergence

occurred after 3453 iterations. Fig. 5a shows the velocity contour
(a) solar collector

(b) drying chamber

Fig. 4. A detail view of mesh domain for solar collector and drying chamber. (a) Solar air collector

Table 2
CFD model specification.

Model Description
Multiphase Mixture
Energy On
Viscous k-e wall function-RN

temperature. High temperature drying of melon containing high

sugar content requires less drying time however higher tempera-
ture likely resulting in damaged flavor, color and nutrients of dried
products (Amiri et al., 2014). Also, Lima (2003) stated that melons
start to cook above 65 °C (Rodrigues and Fernandes, 2007). There- (b) yz plane view of dryer chamber
fore, 50–60 °C product surface temperatures were established to
apply an effective heat processing and to carry out the drying of
the products. Consumed energy in infrared lamps is controlled
more stable with proportional control. The products’ surface tem-
peratures are set at 50–60 °C by a proportional control system with
accuracy of ±3 °C and at a 0.5 m/s drying air velocity. According to
our previous work (Aktasß et al., 2013) and a similar study (Nowak
and Lewicki, 2004), the distance between the lamps and the pro-
duct, and the optimum drying air velocity were determined as
200 mm and 0.5 m/s for sliced melon drying, respectively. Thick-
ness of melon slices was selected as 10 mm due to physical appear-
ance and thermal properties of sliced melon.
The product weight loss was recorded continuously during dry- (c) xz plane view of dryer chamber
ing by using a load cell (output mV/V, 2.0, 5–12 V, 5 kg capacity,
accuracy ±1 g and uncertainty ±1.5). The energy consumption of Fig. 5. Mean air velocity contours in CFD model.
508 M. Aktasß et al. / Solar Energy 137 (2016) 500–515

on a plane parallel to xz plane in the middle section of the collector.
The flow speed value is going down by passing through the fins and
also some vortexes are composed in narrow angle regions. Fig. 5b
shows the velocity contour on a plane parallel to yz plane in the mid-
dle section of the drying chamber. Calculation processes for drying
chamber were done and convergences occurred after 2000 itera-
tions. The flow speed is going down by passing through the melon
parts and some vortexes are composed around drying product.
Fig. 6 shows the temperature distribution in the middle section
of the collector. Fins absorb a portion of radiated energy and warm
up as a result of radiation heat transfer. Air flow passing over the
fins gets some part of stored energy by convection heat transfer,
as the air warms at second half part of collector is more sensible.
Fig. 6b is showing the temperature contour in metal domain. Heat
flux to collector warm up the temperature and air flow over the
fins remove some part of receiving energy. Inlet and middle part
of collector absorber are cooled better than outlet side because
air flow is not enough to cool down all area of absorber surface.
Fig. 6c shows the temperature distribution in the middle section
of the drying chamber. Melons absorb a portion of radiated energy
and warm up as a result of the radiation heat transfer. Air passing
over the melon slices gets some part of stored energy by the con-
vection heat transfer. Fig. 6d is showing the same temperature con-
tour just in xz plane.

(a) Air domain (b) Collector domain

(c) yz plane view

(d) xz plane view

Fig. 6. Air temperature contours in CFD model.
 M. Aktasß et al. / Solar Energy 137 (2016) 500–515 509

(a) Solar air collector

(b) yz plane view

(c) xz plane view

Fig. 7. Mean air velocity vectors in CFD model.

Fig. 7a shows the main velocity vectors in the air domain inside
the air collector, by emphasizing the vortex role in contact with
fins that gives turbulent characteristics more visible. Fig. 7b and c
shows the main velocity vectors inside drying chamber.
Coordinate of two lines on the collector domain has been shown
in Fig. 6 to verify simulation results with experimental data. Line 1
is the position of thermocouples for fin temperature measurement
while line 2 is demonstrating the location of thermocouples for col-
lector down surface temperature measurement. Fig. 8 shows exper-
imental and computational temperatures of fin and absorber.
Fig. 8a shows the temperature distribution on line 1 placed on
one of fins allocated in the entrance of SAC. The temperature distri-
510 M. Aktasß et al. / Solar Energy 137 (2016) 500–515

Line 1 Comparing the experimental and the CFD analysis data revealed
320 a good correlation. Hot air temperature and velocity distribution
CFD EXP are acceptable to be used for device design.
318
Fin temperature (K)

316 3.2. Experimental results

314
High set temperature for drying in order to achieve high heat
312 transfer rate has been employed in this proposed system. By apply-
ing a high temperature to melon slices, product becomes gel-like.
310 Therefore, low temperature is more appropriate for melon drying.
So, experiments were carried out at product surface temperature of
308
0 0.02 0.04 0.06 0.08 0.1 50–60 °C. Comparing different drying conditions showed that the
Fine length (m) best choice can be varied according to the priority of energy con-
sumption or drying time criteria. ‘‘50 °C” can be the best choice
(a) fin from the energy consumption point of view, while the ‘‘60 °C”
can be the best choice from the drying time point of view. The
Line 2 energy consumption during time intervals is shown in Fig. 10.
328
CFD EXP Fig. 11 shows evaluation of input energy and consumed energy
326
Absorber temperature (K)

in drying system. As it can be seen in Fig. 11 electricity energy

324
322
320
318
316
314
312
0 0.2 0.4
Absorber length (m)
(b) absorber
Fig. 8. Experimental and computational temperatures of fin and absorber.
bution on line 1 shows a falling manner of temperature as distance
from fin basement increases, it means that air flow has cooling
effect on the fin of collector. Collected data from three thermocou-
ples shows the consistency of the CFD results with experiments.
Fig. 8b shows the temperature distribution on line 2 placed on
the absorber. The temperature distribution on line 2 shows that
air flow has cooling effect on collector absorber except exit side.
It means that air flow is not enough to cool down all surface area
of the SAC.
Fig. 9 shows the velocity magnitude z direction on a line with
1.5 cm distance from melon surface. By averaging the velocity
along this line, it is possible to say that the mean velocity of air
passing through the melon slices was 0.35 m/s. Air velocity
measurements by anemometer on the same position verifies the
simulation process.
demand is high. However, it can be generated by using photo-
voltaic panels. Also, the role of solar energy can be increased by
enlarging surface area of solar air collector. As it can be seen a large
amount of energy exhausted by air but there is a chance to regain
this energy by using heat recovery unit.
Solution results of solar collector simulation have not been used
as an input data for drying chamber directly, because the heat
recovery system has deviated data that have not been considered
0.6 0.8
in the simulation, so inlet air experimental data has been used to
set a boundary condition based on average values. Measured air
velocity in the heat recovery inlet is 0.4 m/s. Fig. 12 shows the heat
recovery unit inlet and outlet air temperatures experimentally. As
it can be seen in Fig. 12 energy saving obtained by increasing air
temperature by heat recovery unit.
The maximum efficiency of the heat recovery unit is obtained
when the temperature of the exhaust air is cooled down to ambi-
ent air temperature by passing through the heat exchanger. But the
exhaust air temperature is always above the ambient temperature
in real case. Therefore the efficiency of the heat recovery unit can
be described by dividing the recovered heat to a maximum quan-
tity in ideal case that has been shown in Fig. 13. Average efficiency
of the SAC was obtained 46.6% and 54.6% for 50 °C and 60 °C,
respectively. Fig. 14 presents the variations in the collector effi-
ciency as a function of the time. The experiments were performed
under same outside conditions. But, there is a sudden drop in the
solar radiation in Fig. 14 at 300 min for the 60 °C due to the fact
that the weather was cloudy since 300th minute to the end of
experiment. PIRHR values were obtained 0.323 at 50 °C and 0.252
at 60 °C, while PIRsac-HR values are obtained as 0.405 at 50 °C and
0.316 at 60 °C.

0.6
Total, 60°C Total, 50°C
Without HR+SAC (50 °C) Without HR+SAC (60 °C)
0.7
CFD EXP 0.5
Energy consumption (kWh)

0.6
0.4
Air velocity (m/s)

0.5

0.4 0.3

0.3 0.2
0.2
0.1
0.1

0 0
0.17 0.27 0.37 0.47 0.57 0.67 0.77 0.87 0.97 0 100 200 300 400 500
Measured distance from chamber entrance (m) Time (min)

Fig. 9. Velocity magnitude on a line with 1.5 cm distance from melon slices. Fig. 10. Energy consumption of dryer during the experiments.
 M. Aktasß et al. / Solar Energy 137 (2016) 500–515 511

Input energy (50 °C) Input energy (60 °C)

11% recovered heat 9% recovered heat

energy energy
29% 23%
electricity electricity
energy energy
solar energy solar energy
60% 68%

Output energy (50 °C) heating of product Output energy (60 °C) heating of product
1% 1% 1% 1%
exhausted heat 4% exhausted heat
7% energy energy
23% evaporation heat evaporation heat
energy 36% energy
37%
36% total heat loss total heat loss

32% energy consumption

energy consumption
of fans 21% of fans

emitted heat energy emitted heat energy

by radiation by radiation

Fig. 11. Evaluation of the energy performance of the dryer.

50 - HR inlet air temperature 50 - HR outlet air temperature 0.9

60°C 50°C
60 - HR inlet air temperature 60 - HR outlet air temperature 0.88
50
0.86
QHR / QHR-max

45 0.84
0.82
Temperature ( )

40
0.8
35 0.78
0.76
30
0.74
25 0.72
0.7
20 0 100 200 300 400 500 600
0 100 200 300 400 500
Time (min)
Time (min)
Fig. 13. Efficiency of heat recovery unit during the experiments.
Fig. 12. Heat recovery unit inlet and outlet air temperature.

Coll. eff. (60°C) Coll. Eff. (50°C) Solar rad. (60°C) Solar rad. (50°C)
SEC values are calculated by using Eq. (36). Average SEC values 80
were obtained as 2.07 kW h/kg at 50 °C and 3.12 kW h/kg at 60 °C. 70
1030
Collector efficiency (%)

Solar radiation (W/m²)

Energy and drying efficiency were calculated by using Eqs. (37)– 60 830
(40). The drying efficiency of the system with and without the HR 50
device at 50 °C were calculated average 33.4% and 22.6%, respec- 40
630

tively. The drying efficiency of the system with and without the 30 430
HR device at 60 °C were calculated average 22.4% and 16.7%, respec- 20
230
tively. Similarly, the energy efficiency of drying system with and 10
without the HR device at 50 °C was calculated average 32.2% and 0 30
0 100 200 300 400 500
21.8%, respectively. With the HR device and without the HR device
Time (min)
at 60 °C, they were calculated average 21.1% and 15.8%, respectively.
The experiments showed that the infrared radiation directly Fig. 14. Efficiency of solar air collector during the experiments.
influences the moisture diffusivity in the drying process. The pro-
portional control of infrared lamps allows increasing the moisture 60 °C, respectively. These values show the effect of the product sur-
diffusivity in the air-drying process up to a certain maximum level. face temperature on the drying time. Accordingly, the drying time
Drying times of sliced melon are 510 min and 400 min at 50 °C and was decreased at higher product surface temperature.
512 M. Aktasß et al. / Solar Energy 137 (2016) 500–515

In this paper, some thermophysical characteristics of sliced similar for both temperatures, with the values of 8.89 for 50 °C
melon have been estimated by calculation of surface heat and mass and 8.98 for 60 °C. Bi and Di values were calculated by using Eqs.
transfer coefficients and their relation to other values based on the (33) and (34). The Di values were calculated as 16586.5 and
experimental data for two different product surface temperatures. 5088.5 at product surface temperature of 50 °C and 60 °C, respec-
Dimensionless numbers were calculated by using Eqs. (23)–(27). tively. The Bim values was obtained as 14.74 and 3.98 at product
Average Nusselt number (Nu) value was calculated as 66.17 by surface temperature of 50 °C and 60 °C, respectively. The drying
using experimental data. Average convection heat transfer coeffi- process of solids can be classified as external, both internal and
cient (h) was found as 2.3 W/m2 K. Average Schmit number (Sc) external, and purely internal controlled Bim < 0.1, 0.1 < Bim < 100
was found as 1.72 and 1.77. The Sherwood numbers (Sh) are very and Bim P 100, respectively (Haghi and Amanifard, 2008). Accord-
ingly, this drying process can be classified in both internally and
externally controlled period for this study, due to the Bim number
1.61E-09 were obtained in the range of 0 < Bim < 100. Bim number was
1.41E-09 increased significantly with the increase in surface temperature.
1.21E-09 Also, the Bim value during the drying process is on the contrary
to the moisture content of product.
De (m2/s)

1.01E-09
8.1E-10 Effective moisture diffusivity (De) and mass transfer coefficient
6.1E-10 (hm) of foods are internal factors related to properties of food. The
4.1E-10 experimental MR, De and hm values calculated using Eqs. (30) and
2.1E-10
(31) for different drying conditions are presented in Figs. 15–17.
1E-11
The De covers all possible mechanisms of moisture movement
0 2 4 6 8 within the foods (Darvishi et al., 2015). The effects of drying air
X (g water/g dry matter) temperature and air velocity on the De values were observed. They
have a significant effect on the De value of melon slices. The De
Fig. 15. Variation of the De with Xdb during the experiments.
value increased with higher temperature and velocity. Also it
increased with increase in temperature and with a decrease in
moisture content (Fig. 15). De, hm and X of the product are interre-
4.5E-07 60°C lated. The De values significantly depend on infrared power and
4.0E-07
50°C temperature. The average values of De varied from 8.25
1010
3.5E-07
to 1.24
109 m2/s. The maximum De was observed at 60 °C.
3.0E-07
hm (m/s)

2.5E-07
These values are in a good agreement with the literature. Sim-
2.0E-07 ilar results were reported as 1.81
1011 and 34.26
1011 m2/s
1.5E-07 by Darvishi et al. (2015) for the convective drying of melon at
1.0E-07 40–70 °C in different slice thicknesses. The average De values of
5.1E-08 melon slices increased with an increase in surface temperature
1.0E-09
0 2 4 6 8 during the experiments.
X (g water/ g dry matter) Two kinds of equation sets have been applied to calculate mass
transfer coefficient. Eqs. (24)–(27) are expressing average descrip-
Fig. 16. Variation of the hm with Xdb during the experiments. tion of the hm during the whole drying process. Nusselt and

1
10 60°C 50°C
60°C 50°C
X (g water/ g dry matter)

9 0.8
8
7
0.6
6
MR

5
4 0.4
3
2 0.2
1
0 0
0 100 200 300 400 500 0 100 200 300 400 500
Time (min) Time (min)
(a) Variation of the moisture content as a (b) Variation of the moisture ratio
function of drying time

0.09
(g water/g dry matter min)

60°C 50°C
(g water/g dry matter min)

0.08 0.14
60°C
0.07 0.12
Drying rate

0.06 50°C
0.1
0.05
0.08
DR

0.04
0.06
0.03
0.02 0.04
0.01 0.02
0 0
0 200 400 0 1 2 3 4 5 6 7 8 9
Time (min) X (g water/ g dry matter)
(c) Variation in the drying rate (d) Variation of the DR with the Xdb

Fig. 17. Variations of Xdb, MR and DR for sliced melon during the experiments.
 M. Aktasß et al. / Solar Energy 137 (2016) 500–515 513

Sherwood number correlations are for laminar flow over a flat
plate. Eqs. (28)–(35) are expressing time dependent description
of the hm during the whole drying process. The hm values were
calculated for different product surface temperatures at 0.5 m/s
drying air velocity. The average values of the hm increase from
8.53
108 m/s at 50 °C to 1.47
107 m/s at 60 °C. For the hm,
similar results (1.07
109–10.68
109 m/s for melon and
16.1
109 for okra) reported by Darvishi et al. (2015) and
Dincer and Hussain (2002). The variations of the hm with Xdb values
for different product surface temperatures at 0.5 m/s drying air
velocity are shown in Fig. 16. At the beginning of drying, the hm
value is low but as expected it was increased with decrease
moisture content in the last phase of drying. Also, the hm value
was increased with an increase in product surface temperature.
Consequently, the moisture content and surface temperature affect
the drying time, Bim, De and hm values.
Initial moisture content of melon slices is 9 g water/g dry mat-
ter and final moisture content of dried melon is 0.044 g water/g dry
matter. Xdb, MR and DR values were calculated by Eqs. (29), (31)
and (35). Variations of Xdb, MR and DR for sliced melon were shown
in Fig. 17; (a) variation in the Xdb as a function of drying time, (b)
variation of the MR versus drying time, (c) variation in the DR
according to the drying time for different conditions and (d) varia-
tion of the DR as a function of the Xdb.
Fig. 18 shows the melon temperature at right axis and chamber
air temperature to vaporize water at left axis during experiment.
(To  Ti)/mvw is the ratio of the temperature difference between
the cabin outlet and inlet air to the vaporized water from the
melon. (To  Ti)/mvw can be accepted as a new expression for infra-
red drying process. As this value decreases, drying efficiency
increases. As it can be seen in Fig. 18, drying efficiency is low at
the end of the drying process. The melon temperature has been
measured on two points including top surface and bottom surface.
As expected, in the first period of drying, vaporized water mass is
quite high, however melon surface temperature is low. Product
temperature increases gradually by passing time until approaching
steady temperature that is close to controller set temperature.
Infrared radiation was found to be the main factor influenced the
drying kinetics of melon. It is obvious that the drying process was
shortened substantially with the increment of the infrared radiation.
Shrinkage (S), water activity (aw), rehydration capacity (RC) and
color of the product are determinant parameters for the quality of
the dried product. Therefore, they can be used as an indicator of the
quality of the dried melon samples. The amount of shrinkage of dried

T_air/m_vaporized water (50°C) T_air/m_vaporized water (60°C)

Top surface temp. (50°C) Top surface temp. (60°C)
Bottom surface temp. (50°C) Bottom surface temp. (60°C)
1 65

0.9
60
0.8

Melon temperature (°C)

0.7 55
(To-Ti)/mvw ( °C/g )

0.6
50
0.5
45
0.4

0.3 40
0.2
35
0.1

0 30
0 100 200 300 400 500
Time (min)

Fig. 18. Obtained temperatures during the experiments.

Fig. 19. View of fresh and dried melon slices.

514 M. Aktasß et al. / Solar Energy 137 (2016) 500–515
melon is calculated by using Eq. (43). At the end of the drying process,
the mass shrinkage ratio for melon was varied between 0.0909 and
0.1126. Low temperature-dried foods tend to have less shrinkage.
The S value of sliced melon was decreased with an increase in surface
temperature. The S value was varied according to slice thickness and
drying air temperature. The aw value was calculated by using Eq. (44).
The aw value was measured between the 0.52 and 0.55 (at 25 °C
ambient temperature with water activity measuring device; value
of the aw between 0 and 1, accuracy ±0.001, uncertainty ±0.002).
The growth of most harmful microorganisms can be limited by reduc-
ing the aw below 0.60 or moisture content below 10% (Jangam et al.,
2010). Accordingly, measured aw values are ideal for storage safely.
Fresh and dried melon slices is shown in Fig. 19.
The RC was used as an indicator of the quality of the dried
melon samples. It is calculated by using Eq. (45). The RC values
were measured as 5.8–6.1. The RC values of the dried samples
changed with the drying temperature on a small scale. The RC
value of the dried samples increased as rehydration time extended.
At the beginning of the process, absorption of water had rapid but
then, little moisture movement occurred. Another quality indicator
is color. There was a noticeable increase in color saturation of the
product with temperature increase. The best results (less shrink-
age, high rehydration ratio and minimum color change) were
achieved when samples were dried at low temperature.
4. Conclusions
Solar-heat recovery assisted infrared dryer was designed and
tested. Also, drying characteristics of melon slices were investi-
gated. In this study, theoretical and experimental analysis of a
new solar air collector with finned absorber were done determin-
ing the temperature distribution and efficiency of the solar air col-
lector. Solar-heat recovery assisted infrared dryer is a new
prototype. Energy demand can be reduced by adding solar collector
and heat recovery unit to the infrared dryer. Taking into consider-
ation the results of all analyses, the following remarks may be con-
cluded: The results revealed that solar-infrared dryer has been
shown good performance with the contribution of the solar energy
and heat recovery. The drying efficiency of the system with and
without the HR device at 50 °C was calculated average 33.4% and
22.6%, respectively. CFD analyses of the SAC and drying chamber
have contributed to designation of system. Also, CFD simulations
and experimental results are in good agreement. Experiment at
50 °C product surface temperature showed better qualities based
on less shrinkage and color change. Proportional control of product
surface temperature provided homogeneous and efficient drying.
As a result, the SAC and the HR unit were contributed to system
efficiently. They have significantly reduced the energy demand
without any negative effects on the quality of product. The IR
+ convective drying method can be used instead of using the IR
method alone. The results presented in this study not only show
the good performance of the SHRAIRD, but also provide important
guidance to designers. This drying system can be applied success-
fully in drying applications.
References
Aboltins, A., 2013. Theoretical study of material drying coefficient. In: 12th
Conference: Engineering for Rural Development, Jelgava, Latvia, pp. 153–158.
Aghbashlo, M., Mobli, H., Rafiee, S., Madadlou, A., 2012. Energy and exergy analyses
of the spray drying process of fish oil microencapsulation. Biosyst. Eng. 111,
229–241.
_ ß, M., Yalçın, A., Sßahin, M., 2013. Experimental investigation of drying
Aktasß, M., Ilbas
behaviours in an infrared radiation dryer. J. Faculty Eng. Archit. Gazi Univ. 28
(4), 767–775.
Aktasß, M., S
ß evik, S., Doğan, H., Öztürk, M., 2012. Drying of tomato in a photovoltaic
and thermal solar-powered continuous dryer. J. Agric. Sci. 18, 287–298.
AOAC, 1990. Official Method of Analysis, Association of Official Analytical Chemist.
IAC, Arlington, Virginia.
Amiri, M., Tavakolipour, H., Gharehyakheh, S., 2014. Modeling of melon drying by
application of microwave using mamdani fuzzy inference system. Eur. J. Exp.
Biol. 4 (1), 44–52.
Ayhan, Z., Chism, G.W., Richter, E.R., 1998. The shelf-life of minimally processed
fresh cut melons. J. Food Qual. 21 (1), 29–40.
Berdıyev, M., Arslan, D., Özcan, M.M., 2009. Nutritional composition,
microbiological and sensory properties of dried melon: a traditional Turkmen
product. Int. J. Food Sci. Nutr. 60 (1), 60–68.
Chayjan, R.A., Agha-Alizade, H.H., Barikloo, H., Soleymani, B., 2012. Modeling some
drying characteristics of cantaloupe slices. Cercetări Agronomice Moldova XLV
2 (150), 5–14.
Crank, J., 1975. The Mathematics of Diffusion, second ed. Clarendon Press, Oxford,
UK.
Croitoru, C.V., Nastase, I., Bode, F.I., Meslem, A., 2016. Thermodynamic investigation
on an innovative unglazed transpired solar collector. Sol. Energy 131, 21–29.
Çengel, Y.A., Boles, M.A., 1994. Thermodynamics: An Engineering Approach.
McGraw-Hill, NY.
Çengel, Y.A., Ghajar, A.J., 2011. Heat and Mass Transfer Fundamentals and
Applications, fourth ed. McGraw-Hill, NY, pp. 717–784.
Darvishi, H., Khodaei, J., Azadbakht, M., 2015. The parameters of mass transfer of
convective drying in sliced melon. Philipp. Agric. Sci. 98 (1), 360–372.
Demirel, D., Turhan, M., 2003. Air-drying behavior of Dwarf Cavendish and Gros
Michel banana slices. J. Food Eng. 59, 1–11.
Dincer, I., Hussain, M.M., 2002. Development of a new Bi-Di correlation for solids
drying. Heat Mass Trans. 45, 3065–3069.
Dissa, A.O., Ouoba, S., Bathiebo, D., Koulidiati, J., 2016. A study of a solar air collector
with a mixed ‘‘porous” and ‘‘non-porous” composite absorber. Sol. Energy 129,
156–174.
Doymaz, I., 2008. Convective drying kinetics of strawberry. Chem. Eng. Process. 47
(5), 914–919.
Duffie, J.A., Beckman, W.A., 1991. Solar Engineering of Thermal Process. Wiley, NY.
Duffie, J., Beckman, W., 2006. Solar Engineering of Thermal Processes, third ed. John
Wiley and Sons, Madison.
FAO, 2010. FAO Statistical Yearbook 2010 (FAOSTAT Data 2010). Available at:
<http://faostat.fao.org>.
Gil, M.I., Aguayo, E., Kader, A.A., 2006. Quality changes and nutrient retention in
fresh-cut versus whole fruits during storage. J. Agric. Food Chem. 54, 4284–
4296.
Haghi, A.K., Amanifard, N., 2008. Analysis of heat and mass transfer during
microwave drying of food products. Braz. J. Chem. Eng. 25 (3), 495–501.
Jangam, S.V., Law, C.L., Mujumdar, A.S., 2010. Drying of Foods, Vegetables and Fruits,
vol. 1. Singapore.
Jaturonglumlert, S., Kiatsiriroat, T., 2010. Heat and mass transfer in combined
convective and far-infrared drying of fruit leather. J. Food Eng. 100, 254–260.
Karim, M.A., Hawlader, M.N.A., 2005. Drying characteristics of banana: theoretical
modelling and experimental validation. J. Food Eng. 70, 35–45.
Kassem, A.S., Shokr, A.Z., El-Mahdy, A.R., Aboukarima, A.M., Hamed, E.Y., 2011.
Comparison of drying characteristics of Thompson seedless grapes using
combined microwave oven and hot air drying. J. Saudi Soc. Agric. Sci. 10, 33–40.
Kavak Akpinar, E., 2010. Drying of mint leaves in a solar dryer and under open sun:
modeling, performance analyses. Energy Convers. Manage. 51, 2407–2418.
Kumar, D.G.P., Umesh Hebbar, H., Ramesh, M.N., 2006. Suitability of thin layer
models for infrared–hot air-drying of onion slices. LWT-Food Sci. Technol. 39,
700–705.
Lima, A.S., 2003. Desidratacßa~o osmó tica de mela~o cantaloupe sob pressa~o
atmosférica e vácuo seguida de uma etapa complementar de secagem em estufa
Master Degree Thesis. Universidade Federal do Ceará, Fortaleza, Brazil.
Mahmoud, B.S.M., Vaidya, N.A., Corvalan, C.M., Linton, R.H., 2008. Inactivation
kinetics of inoculated Escherichia coli O157:H7, Listeria monocytogenes and
Salmonella Poona on whole cantaloupe by chlorine dioxide gas. Food Microbiol.
25, 857–865.
Mathioulakis, E., Karathanos, V.T., Belessiotis, V.G., 1998. Simulation of air
movement in a dryer by computational fluid dynamics: application for the
drying of fruits. J. Food Eng. 36, 183–200.
Nasıroğlu, S., Kocabıyık, H., 2009. Thin-layer infrared radiation drying of red pepper
slices. J. Food Process Eng. 32, 1–16.
Nowak, D., Lewicki, P.P., 2004. Infrared drying of apple slices. Innovative Food Sci.
Emerging Technol. 5, 353–360.
Raksakantong, P., Siriamornpun, S., Ratseewo, J., Meeso, N., 2011. Optimized drying
of kaprow leaves for industrial production of holy basil spice powder. Drying
Technol. 29 (8), 974–983.
Ramani, B.M., Gupta, A., Kumar, R., 2010. Performance of a double pass solar air
collector. Sol. Energy 84, 1929–1937.
Rodrigues, S., Fernandes, F.A.N., 2007. Dehydration of melons in a ternary system
followed by air-drying. J. Food Eng. 80, 678–687.
Solval, K.M., Sundararajan, S., Alfaro, L., Sathivel, S., 2012. Development of
cantaloupe (Cucumis melo) juice powders using spray drying technology.
LWT-Food Sci. Technol. 46, 287–293.
Supmoon, N., Noomhorm, A., 2013. Influence of combined hot air impingement and
infrared drying on drying kinetics and physical properties of potato chips.
Drying Technol. 31 (1), 24–31.
ß evik, S., 2014. Experimental investigation of a new design solar-heat pump dryer
S
under the different climatic conditions and drying behaviour of selected
products. Sol. Energy 105, 190–205.
 M. Aktasß et al. / Solar Energy 137 (2016) 500–515 515

Thuwapanichayanan, P.S., Kunwisawa, J., Soponronnarit, S., 2011. Determination of
effective moisture diffusivity and assessment of quality attributes of banana
slices during drying. LWT-Food Sci. Technol. 44, 1502–1510.
Toğrul, H., 2006. Suitable drying model for infrared drying of carrot. J. Food Eng. 77,
610–619.
Yang, J., Bingol, G., Pan, Z., Brandl, M.T., McHugh, T.H., Wang, H., 2010. Infrared
heating for dry-roasting and pasteurization of almonds. J. Food Eng. 101 (3),
273–280.
Yinqiang, S., Yang, J., Qiuhong, Y., Hua, L., Wang, H., 2014. Infrared, convective, and
sequential infrared and convective drying of wine grape pomace. Drying
Technol. 32 (6), 686–694.
Zare, D., Minaei, S., Mohamad Zadeh, M., Khoshtaghaza, M.H., 2006. Computer
simulation of rough rice drying in a batch dryer. Energy Convers. Manage. 47,
3241–3254.