ACTA PHYSICA POLONICA A

Vol. 127 (2015)

No. 4

Proceedings of the 4th International Congress APMAS2014, April 24-27, 2014, Fethiye, Turkey

Theoretical and Experimental Analysis of Drying Various

Geometrical Forms of Red Pepper
A. Onat
a
b

a, *

, A. K. Binark

Marmara University, Technical Education Faculty, Energy Department, Istanbul, Turkey

Marmara University, Technology Faculty, Department of Mechanical Engineering, Istanbul, Turkey

In this study, a counter ow convection-type dryer was designed and manufactured. During experiments, four
dierent geometrical forms of pepper specimens were dried in a complete (unsliced), perforated, crosscut, and
longitudinally sliced forms. For each type of red pepper specimens, the experiments have been conducted at air
velocity level of 0.5 m/s, relative humidity of 1015%, and temperature of 5560 C. The theoretical mathematical
model of drying process was developed, considering the pepper's bottom surface to be isolated. The drying curves
of experimental results are compared to ones obtained from the theoretical analyses. The comparison showed that
experimental results are consistent with the theoretical model. The best results, considering the drying duration,
are collected from the specimens sliced longitudinally, which were followed by the crosscut specimens, perforated
and unsliced-complete specimens, respectively. 82% water (humidity) content in 500 g sample was reduced to 4%
in longitudinally cut samples, 6% in crosscut samples, 7.5% in perforated samples and 8% in unsliced-complete
samples, after 14 hours of drying. It is suggested that regarding to easiness of processing, the crosscut red peppers
are more suitable, compared to the other geometrical forms.
DOI: 10.12693/APhysPolA.127.1388
PACS: 47.54.Bd, 47.54.De, 44.27.+g, 44.30.+v
propriate transportation limit conditions were obtained

1. Introduction

Being an industrial process, drying is commonly used

for drying of products in chemistry, textile, ceramic, construction, timber, pepper and food sectors. The energy
used in these sectors for drying has a signicant share
in total energy consumption.

This rate reaches 6% in

chemistry, 5% in textile, 11% in ceramic and construction, 11% in timber, 12% in food and 33% in paper sector [1].

Drying is one of the most signicant methods

used in preserving food products. Dried products can be

sent to be further processed or they can be put on market. To preserve food products for a long time without
decomposition, dried products can be sterilized. Drying
can be used to obtain some products out of solutions
and to reduce the mass of carried materials [2]. Turhan
and others, in their study, performed drying in regular
and irregular pepper stacks at constant temperatures of
50, 60, 70 and 80

speed of 2 m/s [3].

C, in a tunnel-type dryer with air

Queriox and Nebra dealt in their

study with drying kinetics of banana theoretically and

experimentally in dierent drying air conditions.

Ac-

cording to constant diusion coecient assumption, they

determined the humidity content experimentally, during
the dying process of banana.

The results obtained by

taking into consideration the mass ow occurring during banana's shrinkage and transportation on its surface,
showed features highly similar with other models.

Dif-

fusion and transportation coecients were obtained by

implementing experimental data. Numerical results and
experimental results were compared and the most ap-

using a constant diusion coecient [4]. Diner and Dost

researched humidity and heat conduction performances
and humidity transfer parameters of properly cut timber,
which is put to drying, Biot coecient in the ranges of

0 < Bi < 100

and

Bi > 100

[5] were obtained. In their

research, Ertekin and Sultanolu determined humidity

sorption isotherms of green and red pepper at temperatures of 30, 45 and 60

C and relative humidity values

in the range of 10% to 90%. With nonlinear regression

analysis, the acquired experiment results were compared
with other solution models [6]. Diner theoretically calculated humidity diusion and humidity conduction coecients for plane and cylindrical structures for reference
range

0 < Bi < 100

of Biot coecient, which is the one

that is used most in the practice, in his study entitled

modeling of humidity conduction in drying of solid materials. By comparing this theoretical study with experimental studies made by Akiyama for cylindrical structures and by Lewicki for plane structures (at cut onion
wedges), he found the correlation number (r

> 0.99)

between drying time and non-dimensional humidity content.

Besides, he found Biot number 3.73 for at cut

onion and 0.035 for cylindrical structured starch powder.

5.451010 m2 /s and
2
m /s respectively [7]. Simal et al. developed

He calculated diusion coecients

3.04 10

mathematical model for heat distribution during drying

in cylindrical structures.
at

0.0070 0.0014

Broccoli was cut cylindrically

m calibre and 0.020 m long.

Then

samples were kept in NaOH solution for 15 seconds at

temperature of 100

C. They performed drying process

in a hot air dryer, with a drying temperature between 50

and 90

* corresponding

author; e-mail:

ayhanonat@marmara.edu.tr

C, and with 3 kg/m s air mass ow rate. Drying

temperature was attempted to stabilize with an automatic thermostat.

(1388)

Experiments were conducted at air

Theoretical and Experimental Analysis of Drying. . .

velocity of as 2.7 m/s and drying temperature of 50, 60,

1389

air is directed onto the conveyor belt by a blowing fan.

C . By taking shrinkage into consideration

After passing over the drying peppers, having opposite

during drying, mathematical model was developed ac-

ow direction, drying air is discharged at heat trans-

70, 80 and 90

nd

cording to Fick's 2

law. Runge-Kutta-Merson method

former, in which heat storage is performed in a PCM

Eective diu-

(Phase Change Material) with the help of air suction

sion coecient was determined as a function of humidity

fan. In this way, heat recovery is achieved by discharg-

and temperature [8].

conducted ex-

ing the the used hot air over PCM and over the intake

periments with Thompson seedless raisin in a xed-type

air. Device's band speed can be changed between 0 and

commercial dryer with air velocity of 0.25, 0.50, 0.75 and

0.05 m/s with the inverter (speed control device), con-

was used for solution of the problem.

Pangavhane et al.

C, and rel-

nected to motor-reductor group which is connected with

ative humidity of 13%, 18% and 23%. They performed

band using a belt-pulley mechanism. Thus, the time be-

weighting measurement at every half an hour until the

tween the entry and the exit of a spot on the band can

humidity content in raisin was 17%. As a result of ex-

be reduced down to 1.27 minutes. In this way, product

perimental and theoretical studies, drying kinetics of thin

drying time can be changed with the change of speeds

shelled seedless Thompson raisin was solved with Page

at a stable tunnel length, thus the opportunity to dry

equality.

dierent products is created.

1.0 m/s, temperature of 50, 60, 70 and 80

Experimental and theoretical results showed

Supply and exhaust fans

similar values. It was observed that abrupt changes oc-

were employed in the experimental setup. Supply fan is

cur in the structure of raisin above drying temperature

of a radial type. Since it has a rotational speed control

C and its shell has shriveled and cracked. They

facility it is possible to send the air into the collector with

stated that the most appropriate drying temperature is

varying speeds. Temperature control in the collector was

of 70

between 5070

C [9]. Doymaz and Pala conducted a the-

performed by a PC connected PLC.

oretical and experimental study of the drying kinetics of

Drying dierent products with the aid of computer pro-

red peppers under dierent pretreatment and air drying

gram is achieved with additional heater, at dierent tem-

conditions [10]. Scala and Crapiste simulated the process

peratures, according to entered drying values.

of drying of individual pieces of red pepper under con-

signed and manufactured solar supported counter ow-

stant external conditions and, to predict changes in some

convection type dryer is presented in Fig. 1. Front view

nutritional and organoleptic attributes of the product.

of counter ow-convection type dryer is shown in Fig 1a.

Water sorption isotherms of red pepper were determined

Side view of counter ow-convection type dryer is shown

in the range 2050

C and represented by two dierent

sorption equations. Drying kinetics was described by a

The de-

in Fig 1b. Product and air ow diagram of counter owconvection type dryer is shown in Fig. 1c.

diusive model, the eective moisture diusivity ranging

from 5.01 to

8.32 1010

m /s with an activation en-

ergy of 23.35 kJ/mol were obtained. Degradation kinetics for ascorbic acid and total carotenoids were measured
in the range 5070

C and modeled as the rst-order re-

actions [11]. Faustino at al. studied the drying of green

peppers, in terms of drying kinetics evaluated at 30
40

C, 50

C, 60

C and 70

C,

C, and tted experimen-

tal data to dierent empirical kinetic models from literature. This kinetic study was then complemented with
the modeling in terms of Fick's diusion equation [12].
Vega et al. studied and modeled the drying kinetics of
the red bell pepper at dierent temperatures (50, 60, 70
and 80

C), using air velocity of 2.5 m/s. They observed

only the falling rate period during the experiment. Effective moisture diusivity was estimated to be between

3.2 109

and

11.2 109

2 1

m s

within the studied

temperature range [13].

2. Materials and methods

2.1 Solar supported counter-ow

convection dryer
Thermal input, required for the drying, was provided

Fig. 1. Designed and manufactured solar-supported

counter ow-convection type dryer, (a) front view, (b)
side view, (c) product and air ow diagram.

2.2. Determination of the amount of dry and wet

material in peppers
Drying oven method is used in this study to determine
the amount of moisture in peppers. In order to determine
the amount of dry material and of moisture, samples were

from two heat sources, a reective sun-light collector and

weighted every two hours during processing at 110

an additional sensitive resistance heater. When the ex-

in the drying oven.

ternal air is brought into the system via a duct, its tem-

change between two measurements was less then 0.04%.

perature is being increased by the resistive heater. Hot

This period lasted approximately 70 hours. Pepper with

Measurements continued until the

A. Onat, A. K. Binark

1390

total mass of 100 g was cut and turned into samples with
a mass of roughly 3 g. They were weighted on scales with
250 g capacity and 1/1000 sensitivity. Dierences of the
change rate of amount of dry material and moisture appeared, depending on the type of pepper, dierent ways
of growing, measurement errors, and seed loss in pep-

Fig. 3. Diagrammatic presentation of drying process.

per, which was cut in dierent shapes and depending on

the air speed during the measurement.

The amount of

dry material was found to be 18%, and that of moisture

was 82%. In order to determine the humidity of the air,
temperatures of a wet and a dry thermometer were mea-

According to drying system presented in Fig. 3, the

initial and the boundary conditions for system are:

=X
b = Xb X
t = 0, 0 z b, X

sured. Samples of dierent geometrical shapes prepared

for drying are presented in Fig. 2.
red pepper is shown in Fig. 2a.
is shown in Fig. 2b.
Fig. 2c.

Unsliced-complete

Perforated red pepper

t > 0, z = b D

Crosscut red pepper is shown in

Longitudinally red pepper sliced lengthwise is

shown in Fig. 2d.

t > 0, z = 0
where

0
X
|z=b
|z=b = h X
z

X
|z=0 = 0,
z

(m) is thickness of pepper meat and

h0

(m/s)

is mass transfer coecient. Solution of (2) is considered

as a product of two functions, one of which depends on
coordinate

t time.
F om = Dt
b2 ,

and the other depends on time

Using denition of Fourier number

the

time-dependent solution is obtained as

sin(n b)
X
=
2

b
+
sin(
Xb
n b) cos(n b)
n=0 n
2

cos (n z) e(n b)

F om

(3)

As a result of the numerical calculations, Biot number

0

h b
D ) is found to be 112309 for all uncut peppers. For other shapes of red pepper samples, since these

(Bi

values are practically

Fig. 2. Red pepper of dierent geometrical forms prepared for drying, (a) unsliced-complete red pepper, (b)
perforated red pepper, (c) crosscut red pepper (d) longitudinally sliced red pepper.

nd

it can be taken as

number, coecients of Eq. 3 are found.

these coecients into general solution, if
and

F om

By inserting

= Xf X
X

transformations are written in their place in

equation, if a solution is conducted for 1

st

term along (z )

(Xf ),

occurred in red pepper

stacks is acquired.

Cartesian coordinate system, humidity loss that occurs

z -axis

Bim > 100,

By using relative equations with Biot

shows the humidity loss

Law, for single dimension of

in pepper stacks of thickness along

[15].

axis in pepper stacks of (b) thickness, equation (4) that

2.3. Mathematical model

According to Fick's 2

B im

can be written




h i
f X

X
2 Dt
.
b X
= cos 2b z exp 4 b2
X

(4)

as follows.

Xf
2 Xf
,
=D
t
z 2
2
where D (m/s ) is the diusion coecient, t (s) is
and z (m) is the change of pepper stack height.

3. Results and discussion

(1)
time
For

boundary conditions it is assumed that the innite plate

of thickness

2L,

and initial temperature

T0

is suddenly

brought into contact with a uid with temperature

T .

With such boundary conditions, this equation can be

solved by thinking of it as a mass transfer equivalent
of heat transmission's process in a transient state [14]. If

is dened and written as

= Xf X ,
X

based on

the humidity of the air, we obtain

X
2X
=D 2 .
t
z

The weight loss in pepper samples of dierent geometrical forms is calculated with measurements performed
every 2 hours at a scales with 1/1000 sensitivity.

Re-

sults of experimental studies: air speed 0.5 m/s, relevant

humidity rate 1015% and drying temperature are calculated as 5560

C. Weight loss of red peppers in dierent

geometrical shapes, obtained as a result of experiment

is presented in Table I and the humidity loss (%) is presented in Table II and Fig. 4. 82% of water (humidity) in
500 g sample were reduced to 4% in longitudinally sliced
samples, to 6% in crosscut samples, to 7.5% in perforated

(2)

samples and to 8% in unsliced-complete samples after 14

hours of drying.

Theoretical and Experimental Analysis of Drying. . .

Measurement of mass (g) of red peppers with
dierent geometry, as function of time.
Drying time, [h]

Unsliced

500

375

Perforated

500

360

Crosscut

500

300

Longit. sliced

TABLE I

10

12

14

KMM

264.7 152.5 118.4 109.5 101.1 97.8

225

142.9 118.4

107

1391

90

98.9 97.2

90

176.5 121.6 111.1 104.5 97.8 95.7

90

500 272.7 157.9 115.4 107.1 102.2 95.7 93.7

90

Humidity (%) of red peppers of dierent geometry as a function of time.

TABLE II

Drying time, [h]

10

12

14

Unsliced-compl.

0.82

0.76

0.66

0.41

0.24

0.16

0.11

0.08

Perforated

0.82

0.75

0.60

0.37

0.24

0.14

0.09

0.075

Crosscut

0.82

0.70

0.49

0.26

0.19

0.12

0.08

0.06

Longitud. sliced

0.82

0.67

0.43

0.22

0.16

0.10

0.06

0.04

Fig. 4. Humidity (water) loss of red peppers in dierent geometrical shapes.

By using equation (4), obtained as a result of theoret-

(Xf ) occurred along z

pepper stacks of width b

ical analysis, humidity losses

axis,

as functions of time, in

were

calculated.

Experimental results obtained by measure-

ments performed in opposite ow transport type dryer

and theoretical results are compared in the same graphic
for peppers with dierent geometrical shapes. Figure 5a
shows comparison of drying curves for unsliced-complete,
Fig. 5b for perforated, Fig. 5c for crosscut, and Fig. 5d
for longitudinally sliced red peppers. During the rst two
hours, the humidity loss in unsliced-complete and perforated red peppers was too small to be noticeable and
supplied energy helped the product to heat and to prop-

Fig. 5. Comparison of experimental and theoretical

time-depended humidity (water) loss (%) of unslicedcomplete red peppers (a), of perforated red peppers (b),
of crosscut red peppers (c) and of longitudinally sliced
red peppers (d).
Colour determination.

TABLE IV

Samples
Colour value ASTA colour value
Unsliced-complete
91.3
70-150 ASTA
Perforated
94.1
70-150 ASTA
Crosscut
116.3
70-150 ASTA
Longitudinally sliced
119.1
70-150 TA

agate heat from the surface to inner layers. The membrane shell between red peppers delayed the transmission

is decreasing in following order: in longitudinally sliced,

of heat to products inside.

crosscut, perforated and unsliced-complete samples. The

Ash content (weight percentage).

ash contents and colours of samples obtained after drying

TABLE III

Amount Most wanted in TS 2419

[%]
I. Class
II. Class
Unsliced-complete
9.3
8.5
10.0
7.9
8.5
10.0
Perforated
Crosscut
7.7
8.5
10.0
Longitudinally sliced
7.7
8.5
10.0
Samples

are presented in Table III and Table IV respectively.

4. Conclusions

In the opposite ow type dryer device which was de-

nd

signed and produced, by taking advantage of Fick's 2

law,

mathematical model of time-depended humidity

loss, which occures along

of width

axis in a red pepper stack

is prepared. A comparison of the results, ob-

tained in experiments performed by a weight loss method,

However, since the membrane shell was torned in the

and the theoretical results obtained as a result of mathe-

crosscut and longitudinally sliced red peppers, heating

matical modeling was carried out. It is determined that

occurred in a shorter period of time and humidity loss

with the exception of the uncut red peppers, the theoret-

started immediately. Thus the humidity loss in samples

ical and experimental results are highly similar and that

A. Onat, A. K. Binark

1392

are in accordance with each other. It is found that the

change dierence between theoretical drying curve and
experimental drying curve in all uncut red peppers is a
little higher than in other samples, and that this is the
result of the air gaps between the uncut pepper stacks.
Comparison of the drying periods of red pepper samples, which are dried in dierent geometrical shapes,
shows that the shortest drying time was obtained in longitudinally cut samples, followed by the cross cut, perforated and whole red pepper samples. 82% water (humidity) ratio in 500 g sample was reduced to 4% in longitudinally cut samples, 6% in crosscut samples, 7.5%
in perforated samples and 8% in uncut whole samples
after 14 hours of drying. However, if the processing facilitation is considered, based on automation possibilities,
it is suggested to crosscut red peppers in pepper drying
processes.
References

[1] A.S. Mujumdar, Handbook of Industrial Drying, 28,

New York, 1987.
[2] N. zbalta, A. Gngr, Kurutma Sistemlerinde I
Pompa Kullam Potansiyeli, Ege University, Solar
Energy Institute, 5, zmir, 1998.
[3] M. Turhan, K.N. Turhan, F. Sahbaz, Journal of Food
Processing and Preservation 21, 223 (1997).

[4] M.R. Querioz, S.A. Nebra, Journal of Food Engineering 47, 127 (2001).
[5] I. Diner, S. Dost, Wood Science and Technology 30,
245 (1996).
[6] F.G. Ertekin, M. Sultanolu, Journal of Food Engineering 47, 225 (2001).
[7] I. Diner, Moisture Transfer Models for Solid Drying,
12th International Drying Symposium (2002).
[8] S. Simal, C. Rossello, A. Berna, A. Mulet, Journal
of Food Engineering 37, 423 (1998).
[9] D.R. Pangavhane, R.L. Sawhney, P.N. Sarsavaa,
Journal of Food Processing and Preservation 24, 335
(2000).
[10] I. Doymaz, M. Pala, Journal of Food Engineering 55,
331 (2002).
[11] K.D. Scala, G. Crapiste, Food Science and Technology 41, 789 (2008).
[12] J.M.F. Faustino, M.J. Barroca, R.P.F. Guine, Food
and Bioproducts Processing 85, 163 (2007).
[13] A. Vega, P. Fito, A. Andres, R. Lemus, Journal of
Food Engineering 79, 1460 (2007).
[14] O.F. Genceli, zml Is letim Problemleri, Birsen
Pub., Istanbul, p.320-323, 2000.
[15] O.F. Genceli, Momentum-Heat Transfer Lecture
Notes, Istanbul Technical University, stanbu, 2000.