2011 International Conference on Alternative Energy in Developing Countries and Emerging Economies
- 169 -
An infrared dryer for this experiment was constructed
as shown in Fig. 3. The electric IR heater power was of
500
u
8 W. (220 V
AC
). The inlet drying air temperature
was controlled by a PID controller with an accuracy of
r
1
q
C. the electric IR heat sources was fixed at 15 cm
distance from rubber sheet samples.
D. Drying experiment
Drying conditions were carried on as follows: infrared
(IR) radiation drying, hot air (HA) drying, green house
(GH) drying, green house and natural convection (GH-
NC) and open sun drying (OS) (convention drying). To
confirm long life storage of dried rubber sheet, desired
final moisture content of dried rubber sheet was
approximately 1
r
0.5% dry-basis [1]. The bulk
temperatures, dry bulb, wet bulb of ambient air
temperature and temperature inside drying chamber were
measured by K-typed thermocouple, which was
continuously monitored by a Wisco data logger with an
accuracy of
r
0.1ºC. The relative humidity was
determined by dry and wet bulb temperatures. The
moisture content was determined by ASAE method [3].
E. Drying kinetic and mathematics model
Relationship between moisture ratios and drying time
normally is determined by three major mathematical
drying models as follows: (1) simulation including heat
and mass transfer, (2) the diffusion model and (3) the
empirical or semi-empirical model mostly developed
from experimental data. The first and second model also
called as theoretical model and semi theoretical model.
Both of them are solved by complicated mathematical
analysis. The rest of model is quite suitable for the
experimental data but it has limitation use because it is
in-situ condition of each experiment.
To simplify prediction of moisture transfer, the
moisture content is normally shown in terms of moisture
ratio (MR). The MR value is defined as ratio of moisture
difference between real time moisture and equilibrium
moisture content to moisture difference between initial
moisture content and equilibrium moisture content. The
following equation was written as below:
(5)
From the thin layer drying equations and geometric
shapes can write a differential equation of moisture
diffusion, which the ratio of moisture to form flat sheets.
It uses the five-term equation to the drying equation (6):
(6)
An effective diffusion coefficient (D) is namely
described by the Arrhenius type equation as follows [12]:
(7)
The constant values in these models were estimated
by the non-linear regression analysis from the
experimental data. The suitability of the equation was
evaluated and compared using the coefficient of
determination (R
2
) and root mean square error (RMSE)
which indicates the fitting ability of a model to a data set
for selecting the best equation to describe the
experimental data. Follows equations were written:
(8)
F. Quality analysis and specific energy consumption
(SEC)
(a) Quality analysis and shrinkage
Due to having no standard rubber sheet quality control
by systematic method, the market determination of
physical quality for rib smoke, air dried sheet and
unsmoked sheet rubber was only evaluated by visual
observation. Thus, this physical quality analysis of dried
unsmoked sheet (USS) rubber is novel method to
evaluate its quality. The USS rubber sheet is one of the
best raw natural rubber (NR) materials because it was
dried under the low temperature drying which
insignificant degraded quality of NR product. Thus, the
physical quality of USS was also compared to STR 5L
standard class which was examined as follows: %DIRT,
%ASH, %VM, %N
2
, %P
O
, %PRI and Color value.
The percentage of shrinkage of sample rubber was
determined by an average of that measured with vernier
calipers with an accuracy of ±0.05 mm and the
percentage of shrinkage was defined as follows:
(9)
(b) Determination of specific energy consumption
The specific energy consumption of the drying was
evaluated by watt-hour meter while drying rate was
calculated by moisture content transfer per drying time,
calculation in equation (10) and (11), respectively.
(10)
(11)
IV.
R
ESULTS AND
D
ISCUSSIONS
A. Equilibrium moisture content (EMC)
The four different EMC models were listed in Table I.
The experimental results were mathematical simulated
by non-linear regression method. The constant values in
each model were evaluated. The indices for estimating
the errors associated with the models, which are the
coefficient of determination (R
2
) and root means square
error (RMSE) value as shown in Tables II at temperature
of 40-60
q
C. The results showed that predicted values by
modified Halsey model have good relation to the
experimental values, corresponding to its higher R
2
and
lowest RMSE value compared to the others. The R
2
and
RMSE value of Modified Halsey model is 0.953 and
0.000288.
dW)fM i (M
3.6P
SEC
u
100
final
H
) final
H initial
(H
(%)
Shrinkage
u
dW
time
Drying
)fM- i (M
rate
Drying
u
»
»
»
»
»
»
¼
º
«
«
«
«
«
«
¬
ª
¸
¸
¹
·
¨
¨
©
§
¸
¸
¹
·
¨
¨
©
§
¸
¸
¹
·
¨
¨
©
§
¸
¸
¹
·
¨
¨
©
§
¸
¸
¹
·
¨
¨
©
§
¸
¸
¹
·
¨
¨
©
§
2l
Dt 2 81
π
exp
81
1
2l
Dt 2 49
π
exp
49
1
2l
Dt 2 25
π
exp
25
1
2l
Dt 29
π
exp
9
1
2l
Dt 2
π
exp
2
π
8
MR
»
»
¼
º
«
«
¬
ª
c
abs RT
aE
expDD
»
»
¼
º
«
«
¬
ª
eqMiM
eqMtM
MR
N
2) i,
experiment
Data
N
1 i
i
predicted,
(Data
RMSE
¦
