See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/266508199

Modeling some drying characteristics of garlic sheets under semi fluidized

and fluidized bed conditions

Article in Research in Agricultural Engineering · June 2012

DOI: 10.17221/42/2011-RAE

CITATIONS READS

17 2,151

3 authors:

Reza Amiri Chayjan Kamran Salari

Bu-Ali Sina University University of Tehran
119 PUBLICATIONS 1,785 CITATIONS 14 PUBLICATIONS 197 CITATIONS

SEE PROFILE SEE PROFILE

Behdad Shadidi
Buali Sina University
13 PUBLICATIONS 202 CITATIONS

SEE PROFILE

All content following this page was uploaded by Kamran Salari on 01 September 2016.

The user has requested enhancement of the downloaded file.

Res. Agr. Eng. Vol. 58, 2012, No. 2: 73–82

Modeling some drying characteristics of garlic sheets

under semi fluidized and fluidized bed conditions

R. Amiri Chayjan, K. Salari, B. Shadidi

Department of Agricultural Machinery Engineering, Faculty of Agriculture,

Bu-Ali Sina University, Hamedan, Iran

Abstract

Amiri Chayjan R., Salari K., Shadidi B., 2012. Modeling some drying characteristics of garlic sheets under
semi fluidized and fluidized bed conditions. Res. Agr. Eng., 58: 73–82.

Thin layer drying properties of high moisture garlic sheets under semi fluidized and fluidized bed conditions with high
initial moisture content (about 154.26% d.b.) were studied. Air temperatures of 50, 60, 70 and 80°C were applied to
garlic samples. Among the applied models, Page model was the best to predict the thin layer drying behavior of garlic
sheets. Using this model, correlation coefficient (R2) was high for all drying cases. The computed values of Deff were
between 3.38 × 10–10 and 2.54 × 10–9 m2/s during the falling rate drying. Values of Deff for garlic sheets were also in-
creased with increasing in input air temperature. Activation energy values were varied between 51.32 and 60.58 kJ/mol
for 50 to 80°C, respectively. The specific energy consumption (SEC) for garlic specimens was placed in the range of
0.316 × 106 and 0.979 × 106 kJ/kg from 50 to 80°C, respectively. An increase in air temperature caused decrease in
SEC value. Application of semi fluidized bed convective drying with temperature between 50 and 60°C was suitable to
produce dried garlic.

Keywords: energy; diffusivity; garlic; moisture ratio; Page model

Garlic (Allium sativum L.) is an important crop in of the garlic will quickly appeared after dormancy.
the world. Due to its therapeutic properties it has When sprouting is started, the nutritional storage
been cultivated in many countries. Garlic is also in the garlic will be decreased and due to wither-
usually used as a flavoring agent; it may be used in ing the stem and trunk, garlic quality is obviously
the shape of powder or granule as a valuable condi- reduced. Drying is one of the best methods for pre-
ment for foods. serving the garlic quality.
When the garlic bulb is cut or split, pungency It is proved that the quality of dried food products
of flavor is diffused, while fresh garlic bulb has no are strongly affected by applied drying methods and
distinct pungency. Process of drying garlic is usu- various physical, chemical and biological changes
ally evaluated based on many indices such as: dry- may be created in the food material (Krokida et
ing time, effective moisture diffusivity, activation al. 2000). In other words, some properties of food-
energy, specific consumption of energy and color. stuffs such as color, structure, aroma compounds
Applying of proper drying method is a key to pro- and nutritional substance were changed or deterio-
ducing high quality dried garlic with a higher dif- rated. These changes may tend to reduce the prod-
fusivity, lower drying time and energy and better uct quality (Pezzutti, Crapiste 1997).
appearance as similar to fresh garlic as possible. Hot-air drying of garlic slices in a common fixed
If the environmental conditions are suitable and bed method is unfortunately not suitable due to a
the rest period passed about 60–80 days, the sprout significant decrease in the quality of dried product

73
Vol. 58, 2012, No. 2: 73–82 Res. Agr. Eng.

related to the fresh one. Applying high temperature high moisture garlic sheets and to compute the effec-
(about 60°C) in a fixed bed drying causes an in- tive moisture diffusivity, activation energy and specif-
crease in drying period, energy consumption, color ic energy consumption of high moisture garlic sheets
degradation and mass transfer. during falling rate of semi fluidized and fluidized bed
In recent years, fluidized bed drying was inves- thin layer drying method and their relation to input
tigated and utilized as a new method for obtaining parameters such as air temperature and velocity.
dried foodstuffs with high quality (Poomsa-Ad et
al. 2002; Cubillos, Reyes 2003; Amiri Chayjan
et al. 2009; Gazor 2009). Fluidized bed drying is MATERIAL AND METHODS
rapid and more uniform compared with fixed bed
drying (Soponronnarit et al. 1997). Fluidized Fluidized bed drying condition
bed drying was employed for drying of some ag-
ricultural grain products such as: broad beans In order to apply fluidized bed condition in dry-
(Hashemi et al. 2009), milky mushroom (Arumu- ing of garlic sheets, maximum pressure drop against
ganathan et al. 2009), rough rice (Amiri Chay- air flow velocity in fluidization curve was deter-
jan et al. 2009), green beans (Souraki, Mowla mined (Kunii, Levenspiel 1991). A digital appa-
2007) and corn (Soponronnarit et al. 1997). ratus (Standard ST-8897, Standard Instruments
Suspending of grain particles in air flow is known Co., Ltd., Hong Kong, China) utilized for recording
as fluidization. At the beginning of air passing both static pressure drop (kPa) and outlet air ve-
through grain bed, a fixed bed will be created. locity (m/s). About 25 g garlic sample was used in
With gradually increasing in air flow rate, a bed fluidization experiments. First experimental point
of fluffy material is obtained, namely minimum was determined at air velocity about 2.8 m/s as a
fluidized bed (semi fluidized bed); afterward with semi fluidized bed condition. Two other points
another increase in air flow rate, bubbling fluid- were in fluidized bed condition with air velocities
ized bed and transportation phenomenon would of 4.26 and 5.4 m/s, respectively. These points and
be observed. At the semi fluidized bed, maximum related fluidization curve were shown in Fig. 1. The
value of pressure drop is obtained and particles net static pressure of garlic bed for points of A, B
weight is equal to frictional force between bed and C were obtained 0.048, 0.037 and 0.026 kPa.
particles (Kunii, Levenspiel 1991; Soponron- As can be observed, net static pressure for point A
narit et al. 1997). (Fig. 1) is maximized and is defined as semi fluid-
Mathematical modeling of garlic drying using ized condition.
defined models can precisely predict the drying
kinetics in a drying system (Kaleta, Górnicki
2010). Effective moisture diffusivity is affected by Experimental setup
air temperature and velocity (Senadeera et al.
2000, 2003). All effective parameters on the mass Fresh garlic was purchased from a local farm
transfer phenomenon in drying process are repre- in September 2010. Garlic samples were kept in a
sented by effective moisture diffusivity (Hashemi
et al. 2009). Activation energy is important for es- 0.06
timation of minimum energy requirement for car- 0.05
Pressure drop (kPa)

rying out the drying process of garlic (Aghbashlo

0.04
et al. 2008). All consumed energy for transferring
of 1 kg water from fresh garlic id defined as specific 0.03
energy consumption (SEC) (Koyuncu et al. 2007). 0.02 A B C
Although many investigations were carried out 0.01
on drying indices for various crops and agricul-
0
tural products, no study reported about drying of 0 2 4 6 8
fresh garlic sheets in semi fluidized and fluidized
Air velocity (m/s)
bed dryer. Also drying properties of garlic sheets in
fluidized bed drying are not available. Fig. 1. Fluidization curve of garlic slices and selected points
The main goals of this study were to find a math- as experimental cases: (A) semi fluidized bed (2.8 m/s),
ematical model for predicting the drying behavior of (B) fluidized bed (4.26 m/s), (C) fluidized bed (5.4 m/s)

74
Res. Agr. Eng. Vol. 58, 2012, No. 2: 73–82

refrigerator at 3 ± 1°C. Ambient air temperature The slope (B) is determined by plotting time
and relative air humidity changed from 26 to 32°C against ln(MR) as follows:
and from 22 to 31%, respectively. During the dry- π 2 Deff
ing experiments, inlet and outlet temperatures of B= (4)
4 L2
the drying chamber and the ambient air relative
humidity and temperature were measured. An ex- Activation energy was obtained using an Arrhe-
perimental fluidized bed dryer utilized to perform nius type equation (Babalis, Belessiotis 2004):
the experiments. Applied air temperatures in the  E 
experiments were 50, 60, 70, and 80°C. In total, Deff = D0 exp  − a  (5)
 RT 
12 experiments were conducted. Initial and final
moisture contents determination of garlic samples For calculating Ea, Eq. (5) can be linearized as fol-
was conducted using gravimetric method at 103°C lows:
(Vázquez et al. 1999). Drying process was started
 E  1 
at initial moisture content of about 154.26% (d.b.) ln( Deff ) = ln( D0 ) −  a   (6)
and finished at final moisture content of about 5%  R  T 

(d.b.). where:
Ea – activation energy (kJ/mol)
R – universal gas constant (8.3143 kJ/mol.K)
Modeling T – absolute air temperature (K)
D0 – pre-exponential factor of the equation (m2/s)
Fick’s second law for moisture diffusion in infi-
nite slab was employed for modeling. Due to minor After plotting of 1/T against ln(Deff) using Eq. (6),
diameter of garlic sheets, that were much smaller six fitted models were obtained as straight lines
than their major diameter, we therefore assumed with the slope of B1.
that the garlic sheets are similar to infinite slab. E
After solution of Fickian equation for an infinite B1 = a
R (7)
slab, the following equation was obtained and can
describe effective moisture diffusivity changes in Simplified form of Eq. (3) for all models can be
garlic sheets: written as exponential equations as follows:
M − Me 8 ∞
1  − Deff (2n − 1) 2 π 2t  M − Me
MR = =
M 0 − M e π2
∑ (2n − 1) 2
exp
4 L2
 (1) MR =
M0 − Me
= a exp(−kt ) (8)
n =1  
where: They are mathematical models for prediction of
MR – moisture ratio (decimal) garlic drying kinetics in thin layer mode. Some of
M – moisture content at any time (kgwater/kgdry mater) these models were employed in this study (Table 1).
M0 – initial moisture content (kgwater/kgdry mater) The value of equilibrium moisture content (Me) is
n = 1, 2, 3, ... – number of terms taken into consideration very small compared to M and M0 (Hassan-Beygi
t – time of drying (s) et al. 2009). Thus (M – Me)/(M0 – Me) was changed
Deff – effective moisture diffusivity (m2/s) to M/M0. Therefore all models in Table 1 and the
L – thickness of garlic sheet (m) basic Eq. (2) can be written as follows:
M
For a long drying period, only the first term of MR = (9)
M0
Eq. (1) can be considered. This simplification has
no negative effect on the final answer (Ramesh et Specific energy consumption (SEC) was calculat-
al. 2001): ed using the following model (Zhang et al. 2002):
 M − Me  8  D π 2t 
ln (MR) = ln   = ln  2  −  eff 2  (2) SEC =
(C Pa + C Pv ha )Qt (Tin − Tam )
(10)
 M0 − Me   π   4L  mvVh
The simplified form of MR is as follows: where:
SEC – specific energy consumption (kJ/kg)
 8   π 2 Deff t  CPv,CPa – specific heat capacity of vapor and air, respec-
MR =  2  exp 2
 (3)
π   4L  tively, (1,004.16 and 1,828.8 J/kg°C)

75
Vol. 58, 2012, No. 2: 73–82 Res. Agr. Eng.

Table 1. Thin layer drying models used in thin layer modeling of high moisture garlic

Model Equation1 References

n
Demir et al. MR = a exp(–kt) + b Demir et al. (2007)
Page model MR = exp(–ktn) Arumuganathan et al.(2009)
Two-term exponential MR = a exp(–kt) + (1 – b)exp(–kct) Sharaf-Elden et al. (1980)
Henderson and Pabis MR = a exp(–kt) Yaldiz et al. (2001)
Logarithmic MR = a exp(–kt) + c Yaldiz et al. (2001)
Lewis MR = exp(–kt) Aghbashlo et al. (2009)
2
Wang and Singh MR = 1 + at + bt Wang, Singh (1978)

1
a, b, c, k, k0, k1, n – drying constants

Q – inlet air to drying chamber (m3/s) RESULTS AND DISCUSSION

t – total drying time (1/min)
ha – absolute air humidity (kgvapor/kgdry air) Mathematical modeling
Tin, Tam – inlet air to drying chamber and ambient air
temperatures, respectively (°C) The drying kinetic of garlic sheets in semi fluidized
mv – mass of removal water (kg) and fluidized bed conditions obtained at different
Vh – specific air volume (m3/kg) temperatures. Results proved that drying air tem-
perature is the most important parameter in drying
Curve Expert (version 1.4) software (Microsoft kinetic. With an increase in air temperature, drying
Corporation, Mississippi, USA) was used to fit the time was reduced.
mathematical models to experimental data. Three Modeling of drying experiments was performed
comparative indices were used as the goodness of using non-linear regression analyses of Curve Expert
fit and to select the best model such as: (1) correla- (version 1.4). Comparative indices for all drying cases
tion coefficient (R2), (2) chi-square (χ2) and (3) root are represented in Table 2. Demir et al., Page model,
mean square error (RMSE). The best model should Two term exponential, Logarithmic and Wang and
have the highest R2 value and the lowest χ2 and Singh models have the R2 value greater than 0.99. Fi-
RMSE values (Demir et al. 2004; Erenturk et al. nally, Page model was selected as the best model for
2004). These indices are as follow: prediction the drying kinetic of high moisture garlic.
N Coefficients of Page model for all drying curves are
∑ [ MRexp,i − MR pre,i ] presented in Table 3. The average R2 value of Page
R2 = 1− i =1
(11) model (0.9986) showed that the selected model is
 n 
N  ∑ MR pre,i  suitable for prediction of garlic drying process.
∑ MR pre,i − k =1 N 
k =1 
 
  Effective moisture diffusivity
N
∑ (MRexp,i − MR pre,i )2 Variations of the ln(MR) against drying time (hour)
χ2 = i =1 (12) in all bed conditions and temperature levels for thin
N−z layer drying of garlic sheets were depicted in Fig. 2.
1 With regard to low thickness of specimens (about
1 N 2 1 mm), one falling rate was obtained in drying of
RMSE =  ∑ ( MRexp,i − MR pre,i ) 2  (13) garlic sheets. Trend of drying curves proved that the
N i =1 
slope of lines in all bed conditions increased with an
where: increase in temperature values. Effect of air velocity
MRexp,i – experimental moisture ratio of ith data on the slope of Deff for different bed conditions was
MRpre,i – predicted moisture ratio of ith data negligible; hence changes in air velocity in domain of
N – number of observations fluidized bed had no effective change in Deff value, es-
z – number of drying constants pecially for low temperatures. Eqs 2 and 4 were used

76
 Table 2. Comparative indices of statistical model parameters for thin layer drying of high moisture garlic sheets

R2 χ2 RMSE
Air temperature
Model
(°C)
2.8 m/s 4.26 m/s 5.4 m/s 2.8 m/s 4.26 m/s 5.4 m/s 2.8 m/s 4.26 m/s 5.4 m/s
Res. Agr. Eng.

50 0.9990 0.9992 0.9964 0.0017 0.0011 0.0054 0.0372 0.0301 0.0648

60 0.9992 0.9996 0.9990 0.0010 0.0004 0.0012 0.0278 0.0174 0.0300
Demir et al.
70 0.9966 0.9950 0.9948 0.0032 0.0055 0.0046 0.0451 0.0605 0.0525
80 0.9994 0.9986 0.9982 0.0004 0.0008 0.0013 0.0130 0.0185 0.0254
50 0.9994 0.9980 0.9992 0.0010 0.0033 0.0010 0.0301 0.0548 0.0298
60 0.9992 0.9994 0.9992 0.0009 0.0005 0.0007 0.0282 0.0210 0.0247
Page
70 0.9994 0.9992 0.9974 0.0005 0.0007 0.0014 0.0202 0.0241 0.0334
80 0.9952 0.9938 0.9962 0.0035 0.0043 0.0007 0.0500 0.0554 0.0229
50 0.9988 0.9992 0.9950 0.0019 0.0012 0.0078 0.0415 0.0323 0.0832
Two-term 60 0.9990 0.9994 0.9986 0.0011 0.0005 0.0017 0.0312 0.0210 0.0385
exponential 70 0.9958 0.9936 0.9936 0.0041 0.0071 0.0058 0.0579 0.0769 0.0681
80 0.9992 0.9986 0.9980 0.0004 0.0009 0.0014 0.0169 0.0253 0.0324
50 0.9972 0.9964 0.9868 0.0046 0.0058 0.0206 0.0646 0.0727 0.1353
60 0.9982 0.9976 0.9986 0.0016 0.0028 0.0017 0.0377 0.0497 0.0385
Henderson and Pabis
70 0.9906 0.9894 0.9912 0.0039 0.0119 0.0080 0.0872 0.0995 0.0800
80 0.9641 0.9629 0.9946 0.0265 0.0258 0.0039 0.1375 0.1357 0.0540
50 0.9990 0.9992 0.9964 0.0017 0.0011 0.0054 0.0383 0.0309 0.0670
60 0.9992 0.9996 0.9990 0.0010 0.0004 0.0012 0.0288 0.0181 0.0312
Logarithmic
70 0.9966 0.9950 0.9948 0.0032 0.0055 0.0046 0.0482 0.0642 0.0567
80 0.9994 0.9986 0.9982 0.0004 0.0009 0.0013 0.0151 0.0226 0.0285
50 0.9960 0.9960 0.9789 0.0065 0.0068 0.0332 0.0787 0.0806 0.1770
60 0.9984 0.9964 0.9978 0.0021 0.0041 0.0027 0.0445 0.0621 0.0503
Lewis
70 0.9864 0.9846 0.9872 0.0134 0.0173 0.0117 0.1103 0.1259 0.1026
80 0.9582 0.9562 0.9934 0.0308 0.0305 0.0047 0.1624 0.1616 0.0641
50 0.9976 0.9950 0.9976 0.0039 0.0084 0.0012 0.0595 0.0875 0.0573
60 0.9942 0.9986 0.9960 0.0079 0.0014 0.0014 0.0837 0.0351 0.0661
Wang and Singh
70 0.9980 0.9976 0.9964 0.0018 0.0027 0.0084 0.0383 0.0474 0.0513
80 0.9994 0.9986 0.9968 0.0003 0.0008 0.0115 0.0146 0.0239 0.0406

77
Vol. 58, 2012, No. 2: 73–82
Vol. 58, 2012, No. 2: 73–82 Res. Agr. Eng.

Table 3. Coefficients of Page model for prediction of Effect of bed condition on Deff
kinetic drying of high moisture garlic sheets
Variations of Deff against air temperature at dif-
Bed condition Temperature (°C) k n
ferent bed conditions are plotted in Fig. 3. Three
50 3.8534 1.1019 exponential models were fitted to the obtained val-
Semi fluidized 60 4.6402 1.0508 ues of Deff. These models as well as related R2 values
bed (2.8 m/s) 70 9.7840 1.2351 are represented in Tables 5 and 6. Results showed
80 28.748 1.4898
that the maximum value of Deff was obtained at the
highest air temperature level; they also indicated
50 3.2959 1.0781 that applying fluidized bed condition caused a de-
Fluidized bed 60 4.6924 1.0972 crease in Deff at upper air temperatures. Drying air
(4.26 m/s) 70 9.7901 1.2573 contact with garlic sheets at semi fluidized bed was
80 24.009 1.4791 the most effective due to its highest values of Deff.
Also at lower air temperatures, no significant dif-
50 5.3186 1.2938
ference was observed between Deff values of bed
Fluidized bed 60 5.5939 1.0706 conditions (air velocity levels), as applying mini-
(5.4 m/s) 70 10.997 1.2182 mum fluidized bed condition with lower air veloc-
80 11.681 1.1118 ity was even more suitable. Four quadratic models
were applied to fit on the obtained moisture diffu-
for calculation of Deff values. These values for all bed sivity. Applied models and related R2 values for four
conditions and temperatures are presented in Table 4. air temperatures are presented in Table 6.
Maximum value of Deff (2.54 × 10–9 m2/s) among all
drying cases was obtained for semi fluidized condi-
tion with air velocity of 2.8 m/s and air temperature Computation of activation energy
of 80°C. This is due to the most effective contact be-
tween garlic sheets and air flow. Also minimum value Fig. 4 shows the curve of ln(Deff) against 1/T for
of Deff (3.38 × 10–10 m2/s) belonged to fluidized condi- drying of garlic sheets in falling rate period. Ac-
tion with air velocity of 5.4 m/s and air temperature of tivation energy (Ea) was calculated using Eq. (6).
50°C. This is due to the low energy rate transferred to The obtained values of Ea for all bed conditions as
the garlic sheet and also low effective contact in tur- well as R2 values are represented in Table 7. Val-
bulent flow of fluidized bed. ues of Ea for food and agricultural crops generally
Values of Deff for high moisture garlic sheets are varied between 12.7 and 130 kJ/mol (Aghbashlo
greatly affected by drying air temperature. As can et al. 2008). Minimum and maximum values of
be seen in Table 4, an increase of 10°C in input Ea for high moisture garlic sheets varied between
drying temperature, cause about twice increase in 51.32 and 60.58 kJ/mol for all bed conditions. Two
Deff values. This pattern was observed in previous modes of moisture in agricultural materials are
studies, such as: peaches (Kingsly et al. 2007) and surface and chemical. Because most of the water in
plums (Goyal et al. 2007). high moisture garlic in falling rate period is in the

Time (h)
Time (hour)
0.0
0 0.5 1 1.5 2 2.5 3 3.5
-0.5 5.4 m/s

-1.0
ln(MR)
ln(MR)

-1.5 80 °C
70 °C
60 °C
-2.0 50 °C

-2.5 Fig. 2. ln(MR) against time (s) for thin-layer

drying of high moisture corn and all bed con-
-3.0 ditions

78
Res. Agr. Eng. Vol. 58, 2012, No. 2: 73–82

Table 4. Effective moisture diffusivity and correlation coefficient for three bed conditions at different temperatures

Semi fluidized bed Fluidized bed Fluidized bed

T (°C) (V = 2.8 m/s) (V = 4.26 m/s) (V = 5.4 m/s)
Deff (m2/s) R2 Deff (m2/s) R2 Deff (m2/s) R2
50 3.72 × 10–10 0.9934 3.63 × 10–10 0.9856 3.38 × 10–10 0.9935
–10 –10 –10
60 6.12 × 10 0.9914 5.97 × 10 0.9864 5.64 × 10 0.9921
–09 –09 –10
70 1.17 × 10 0.9933 1.07 × 10 0.9908 9.75 × 10 0.9877
–09 –09 –09
80 2.54 × 10 0.9832 2.19 × 10 0.9850 1.71 × 10 0.9821

form of chemical absorption, relatively more en- was increased as air temperature was decreased.
ergy is consumed to exhaust water and undesirable Air velocity has a direct relation with SEC, as an in-
change in chemical properties is noticeable in this crease in air velocity caused an intensive increase in
period (Amiri Chayjan et al. 2009). SEC. The minimum value of SEC (0.316 × 106 kJ/kg)
Maximum value of Ea belonged to the minimum was obtained for minimum fluidized bed condition
fluidized bed condition with air velocity 2.8 m/s. and drying air temperatures 90°C. The maximum
With increasing air velocity, the activation energy value of SEC (0.979 × 106 kJ/kg) was obtained for
was decreased. A linear equation is fitted to the cal- fluidized bed condition with air velocity of 5.4 m/s
culated data of Ea against air velocity as follows: and drying air temperature 50°C. Results indicated
that an increase in drying temperature affect SEC
Ea = –3.5451v – 70.673 R2 = 0.9950 (14) inversely. With increasing air velocity, effective con-
tact between air and garlic sheets was contrarily in-
creased but an output energy loss was increased and
Specific energy consumption SEC was therefore increased. Three linear models
were applied to fit the SEC data in semi fluidized bed
Computed SEC values for drying of samples are and fluidized bed conditions as follows:
presented in Fig. 5. It can be observed that the SEC SEC = 5,298.1T + 106 R2 = 0.9941 (2.8 m/s) (15)

Table 5. Fitted exponential models to Deff values for different bed conditions

Air velocity (m/s) Model R2

Semi fluidized bed (2.8 m/s) Deff = 2 × 10–10 × exp(0.641T) 0.9904
Fluidized bed (4.26 m/s) Deff = 2 × 10–10 × exp(0.5976T) 0.9937
Fluidized bed (5.4 m/s) Deff = 2 × 10–10 × exp(0.5419T) 0.9995

Table 6. Fitted power models to Deff value for different air temperatures

Air temperature (°C) Model R2

50 Deff = –6 × 10–12 × T2 + 3 × 10–11 × T + 3 × 10–10 1
–12 2 –11 –10
60 Deff = –7 × 10 × T + 4 × 10 × T + 6 × 10 1
–12 2 –12 –09
70 Deff = –9 × 10 × T – 3 × 10 × T + 1 × 10 1
80 Deff = –6 × 10–11 × T2 + 2 × 10–10 × T + 2 × 10–09 1

Table 7. Activation energy and related correlation coefficient for different bed conditions

Bed condition Semi fluidized bed (V = 2.8 m/s) Fluidized bed (V = 4.26 m/s) Fluidized bed (V = 5.4 m/s)
7,286.3 6,795.4 6,172.1
Equation ln(Deff) = 0.76 – ln(Deff) = 0.766 – ln(Deff) = –2.7285 –
T T T
Ea (kJ/mol) 60.58 56.50 51.32
R2 0.9846 0.9888 0.9976

79
50 3.72E-10 50 3.63E-10 0.8661 3.9E-10
60 6.12E-10 60 5.97E-10 1.3433 6.05E-10
70 1.17E-09 70 1.07E-09 2.1641 9.75E-10
80 2.54E-09 80 2.19E-09 3.8081 1.71E-09

Vol. 58, 2012, No. 2: 73–82 Res. Agr. Eng.

3.00E-09 2.8 m/s Fig. 3. Influence of air temperature on effec-

tive moisture diffusivity (Deff) in thin-layer
2.50E-09 4.26 m/s
drying of high moisture garlic in different
5.4 m/s bed conditions
2.00E-09
Deff (m2/s)
Deff (m2/s)

1.50E-09

2.8 1.00E-09
2.8 4.26 4.26 5.4 5.4
FFP FFP FFP FFP FFP FFP
0.003096 -21.7112 0.003096 -21.7377 0.003096 -21.808
0.003003 5.00E-10
-21.2135 0.003003 -21.2398 0.003003 -21.296
0.002915 -20.5686 0.002915 -20.6523 0.002915 -20.749
0.002833 -19.7894 0.002833 -19.9417 0.002833 -20.1839
0.00E+00
50 60 70 80
Air temperature (°C)
1/T
1/T
-19.5
0.0028 0.00285 0.0029 0.00295 0.003 0.00305 0.0031 0.00315

-20

-20.5 2.8 m/s

eff))

4.26 m/s
ln(Deff
ln(D

5.4 m/s
-21

-21.5
Fig. 4. ln(Deff) versus 1/T at different levels
of air velocities for thin-layer drying of high
-22 moisture garlic

SEC = –6,085.7T + 106 R2 = 0.9850 (4.26 m/s) (16) as the best method applied in this study. Colour
analysis was accomplished on dried garlic samples
SEC = –6,596.6T + 106 R2 = 0.9899 (5.4 m/s) (17) from semi fluidized method. Results showed that
the garlic sheets dried at 50 and 60°C have accept-
able colour. But its color at 70 and 80°C changed to
Comparison between dried garlic sheets brown or black and transformation also occurred.
If a fluidized bed industrial dryer is designed for
With regard to calculated indices such as Deff, Ea garlic drying, then quality of final dry product
and SEC, semi fluidized bed condition was selected should be considered.

1.2E+06 2.8 m/s 4.26 m/s 5.4 m/s

1.1E+06
1.0E+06
9.0E+05
8.0E+05
SEC (kJ/kg)

7.0E+05
6.0E+05
5.0E+05
4.0E+05
3.0E+05 Fig. 5. Specific energy consumption for
2.0E+05 thin layer drying of high moisture garlic
40 50 60 70 80 90 sheets at different air temperatures of
Temperature (°C) semi fluidized and fluidized bed drying

80
Res. Agr. Eng. Vol. 58, 2012, No. 2: 73–82

CONCLUSIONS Cubillos F., Reyes A., 2003. Drying of carrots in a fluid-

ized bed. II. Design of a model based on a modular neural
(1) The Page model was the best for prediction network approach. Drying Technology, 21: 1185–1196.
of high moisture garlic sheets drying kinetics Demir V., Gunhan T., Yagcioglu A.K., Degirmencioglu
in the semi fluidized and fluidized bed condi- A., 2004. Mathematical modelling and the determination of
tions. some quality parameters of air-dried bay leaves. Biosystems
(2) Effective moisture diffu-sivity of garlic sheets Engineering, 88: 325–335.
varied between 3.38 × 10–10 and 2.54 × 10–9 m2/s Demir V., Gunhan T., Yagcioglu A.K., 2007. Mathemati-
in this study. Increase in air temperature in cal modelling of convection drying of green table olives.
each bed condition caused increase in Deff value Biosystems Engineering, 98: 47–53.
whereas increase in air velocity in an air temper- Erenturk S., Gulaboglu M.S., Gultekin S., 2004. The
ature level decreased the Deff value. thin layer drying characteristics of rosehip. Biosystems
(3) The activation energy Ea for garlic sheets in the Engineering, 89: 159–166.
drying experiments varied between 51.32 and Gazor H.R., 2009. Modeling drying kinetics of canola in
60.58 kJ/mol and these values correspond with fluidized bed dryer. Iranian Journal of Biosystems Engi-
the activation energy of other agricultural and neering, 40: 175–184. (in Farsi)
food products in a general range reported by Goyal R.K., Kingsly A.R.P., Manikantan M.R., Ilyas
many researchers. S.M., 2007. Mathematical modeling of thin layer drying
(4) Specific energy consumption calculated for gar- kinetics of plum in a tunnel dryer. Journal of Food Engi-
lic sheets thin layer drying experiments varied neering, 79: 176–180.
between 0.316 × 106 and 0.979 × 106 kJ/kg. Hashemi G., Mowla D., Kazemeini M., 2009. Moisture
(5) A comparison between garlic slices that are diffusivity and shrinkage of broad beans during bulk
dried at different air temperature levels using drying in an inert medium fluidized bed dryer assisted
semi fluidized bed condition indicates that the by dielectric heating. Journal of Food Engineering, 92:
usage of convective dryer with temperature be- 331–338.
tween 50 and 60°C is suitable to preserve natu- Hassan-Beygi S.R., Aghbashlo M., Kianmehr M.H.,
ral color and shape of dried garlic sheets. Massah J., 2009. Drying characteristics of walnut (Juglans
regia L.) during convection drying. International Agro-
References physics, 23: 129–135.
Kaleta A., Górnicki K., 2010. Some remarks on evaluation
Aghbashlo M., Kianmehr M.H., Samimi-Akhijahani of drying models of red beet particles. Energy Conversion
H., 2008. Influence of drying conditions on the effective and Management, 51: 2967–2978.
moisture diffusivity, energy of activation and energy con- Kingsly A.R.P., Goyal R.K., Manikantan M.R., Ilyas
sumption during the thin-layer drying of beriberi fruit S.M., 2007. Effects of pretreatments and drying air tem-
(Berberidaceae). Energy Conversion and Management, perature on drying behaviour of peach slice. International
49: 2865–2871. Journal of Food Science and Technology, 42: 65–69.
Aghbashlo M., Kianmehr M.H., Khani S., Ghasemi M., Koyuncu T., Pinar Y., Lule F., 2007. Convective drying
2009. Mathematical modeling of thin-layer drying of carrot. characteristics of azarole red (Crataegus monogyna Jacq.)
International Agrophysics, 23: 313–317. and yellow (Crataegus aronia Bosc.) fruits. Journal of Food
Amiri Chayjan R., Khosh Taghaza M.H., Montazer Engineering, 78: 1471–1475.
Gh., Minaee S., Alizadeh M.R., 2009. Estimation of head Krokida M.K., Kiranoudis C.T., Maroulis Z.B., Mari-
rice yield using artificial neural networks for fluidized bed nos K., 2000. Drying related properties of apple. Drying
drying of rough rice. Journal of Science and Technology Technology, 18: 1251–1267.
and Natural Resource, 13: 285–299. (in Farsi) Kunii D., Levenspiel O., 1991. Fluidisation Engineering.
Arumuganathan T., Manikantan M.R., Rai R.D., Stoneham, Butterworth-Heinemann.
Anandakumar S., Khare V., 2009. Mathematical mod- Pezzutti A., Crapiste G.H., 1997. Sorptional equilibrium
eling of drying kinetics of milky mushroom in a fluidized and drying characteristics of garlic. Journal of Food Engi-
bed dryer. International Agrophysics, 23: 1–7. neering, 31: 113–123.
Babalis S.J., Belessiotis V.G., 2004. Influence of drying Poomsa-Ad N., Soponronnarit S., Prachayawarakorn
conditions on the drying constants and moisture diffusiv- S., Terdyothin A., 2002. Effect of tempering on subse-
ity during the thin-layer drying of figs. Journal of Food quent drying of paddy using fluidization technique. Drying
Engineering, 65: 449–58. Technology, 20: 195–210.

81
 Vol. 58, 2012, No. 2: 73–82 Res. Agr. Eng.

Ramesh M.N., Wolf W., Tevini D., Jung G., 2001. Influence bed dryer with energy carrier: modeling with and without
of processing parameters on the drying of spice paprika. shrinkage. Journal of Food Engineering, 88: 9–19.
Journal of Food Engineering, 49: 63–72. Vázquez G., Chenlo F., Moreira R., Costoyas A., 1999.
Senadeera W., Bhandari B.G., Wijesinghe B., 2000. Physi- The dehydration of garlic: desorption isotherms and model-
cal properties and fluidization behavior of fresh green bean ling of drying kinetics. Drying Technology, 17: 1095–1108.
particulates during fluidized bed drying. Transactions of Wang C.Y., Singh R.P., 1978. Use of variable equilibrium
the IChemE, 78(C): 43–47. moisture content in modelling rice drying. Transactions
Senadeera W., Bhandari B.G., Wijesinghe B., 2003. of the ASAE, 11: 668–672
Influence of shapes of selected vegetables materials on Yaldiz O., Ertekin C., Uzun H.I., 2001. Mathematical
drying kinetics during fluidize bed drying. Journal of Food modelling of thin layer solar drying of sultana grapes.
Engineering, 58: 277–283. Energy, 26: 457–465.
Sharaf-Elden Y.I., Blaisdell J.L., Hamdy M.Y., 1980. A Zhang Q., Yang S.X., Mittal G.S., YI S., 2002. Prediction
model for ear corn drying. Transactions of the ASAE, 23: of performance indices and optimal parameters of rough
1261–1265. rice drying using neural network. Biosystems Engineering,
Soponronnarit S., Pongtornkulpanich A., Prachaya- 83: 281–290.
warakorn S., 1997. Drying characteristics of corn in
fluidized bed dryer. Drying Technology, 15: 1603–1615. Received for publication June 7, 2011
Souraki B.A., Mowla D., 2007. Axial and radial moisture Accepted after corrections August 23, 2011
diffusivity in cylindrical fresh green beans in a fluidized

Corresponding author:

Dr. Reza Amiri Chayjan, Bu-Ali Sina University, Faculty of Agriculture,

Department of Agricultural Machinery Engineering, 6517833131 Hamedan, Iran
phone: + 98 811 423 8193, fax: + 98 811 442 7012, e-mail: amirireza@basu.ac.ir

82

View publication stats