Professional Documents
Culture Documents
Thin-Layer Drying of Melon Seed: Evaluate
Thin-Layer Drying of Melon Seed: Evaluate
0.0. Ajibola
ABSTRACT
Moisture equilibrium data and thin-layer drying rates for melon seeds at
different temperatures (4O”C- 70°C) and relative humidities (ZO%-88%)
were determined using static gravimetric methods. A nonlinear least-
squares regression program was used to evaluatefive desorption isotherm
models and three thin-layer drying models. The modified Halsey model
gave the least standard error of estimate of 0.4% for equilibrium moisture
content and 4.8% for equilibrium relative humidity. None of the three
thin-layer drying models used to evaluate the thin-layer drying results was
significantly better than the others in predicting drying. The exponential
model in which the drying constant is a function of temperature and
relative humidity was found adequate for predicting thin-layer drying of
melon seed.
NOTATION
INTRODUCTION
LITERATURE REVIEW
TABLE 1
Equilibrium Moisture Content - Equilibrium Relative Humidity Models
Henderson
(Henderson, 1952)
M = ln(l-He)
e [ -aT 1““I
1 - H, = exp( - aTM,h)
Chung-Pfost
(Chung & Pfost, 1967)
Exponential model
M-M,
M. _M =expkW
1 e
This model has been used by Ross and White (1972) to study the drying
characteristics of white and yellow corn; by White et al. (1981) in their
study of the drying behaviour of fully exposed popcorn, and by Verma et
al. (1985) in describing thin-layer drying of rice. It was selected for the
current study because of its simplicity and ease of application. The
constant K is usually related to the initial moisture content of the food
and the temperature and relative humidity of the drying air.
Page’s model
M-M,
M. _M =expNW)
1 e
The constants K, and N are determined experimentally. The model has
been used by Wang and Singh (1978) to describe the drying charac-
teristics of rough rice; by White et al. (1981) in the drying of soya beans;
and by Li and Morey (1987) in the study of the thin-layer drying of
American ginseng. In these studies K, and N were related to initial
moisture content, air temperature and air relative humidity. In studies
comparing Page’s model with other commonly used thin-layer models,
the model was found to give the best fit for thin-layer drying data are in-
shell pecan (Chinnan, 1984) and sunflower seed (Syarief et al., 1984).
Diffusion model
The diffusion model (using Fick’s second law) has been used to describe
drying in the falling-rate period. It assumes that the resistance to
Thin-layer drying of melon seed 309
The model has been used by Graham et al. (1983) to describe the drying
of wheat and corn and by Verma et al. (1985) to describe the drying of
rice.
EXPERIMENTAL PROCEDURE
The melon seeds used in this study were obtained from the local market.
The seeds had an initial moisture content of 21.6% (dry basis) and were
of the variety Bara.
Thin-layer drying
Analysis of results
The equilibrium moisture data were analysed using the five sets of
models shown in Table 1. The values of constants were estimated for the
models first by taking M, as the dependent variable and secondly by
taking He as the dependent variable. The models were linearized by
logarithmic transformation and the linear least-squares estimates of the
constants obtained using the experimental data. The correlation coef-
ficients were found to be very high in all cases (r* > 0.90). However, with
this kind of transformation it is unlikely that the errors in the
transformed dependent variable will be normally distributed over the
range of the transformed independent variables. The constant estimates
were, therefore, used as initial estimates in a nonlinear regression
procedure, GENSTAT OPTIMIZATION (NAG, 198 1), to provide
accurate least-squares estimates of the constants of the models from the
Thin-layer drying of melon seed 311
TABLE 2
Relative Humidities of the Saturated Salt Solution at the Temperatures Used (Young,
1967)
40 50 60 70
10 20 30 40
Relative humdlty P/J
Fig. 1. Experimental equilibrium moisture content results for melon seed and
predicted desorption isotherms using modified Halsey model. Temperature: X ,4o"C; 0,
50°C; A, 60°C; 0,7o”C.
Thin-layer drying of melon seed 313
TABLE 4
Estimated Constants and Standard Errors for Equilibrium Relative Humidity Models
model (1.7%) and the Chung model (l-9%), and is similar to the standard
error obtained for winged bean (0*9%) using the modified Halsey model
(Ajibola, 1986b). The smallest standard error of estimate of He (4.8%)
was obtained by using the modified Halsey model (Table 3). This is
similar to the value of 4.5% obtained when the modified Halsey model
was fitted to winged bean data (Ajibola, 1986 b ).
Figure 1 shows a comparison of the equilibrium moisture data
obtained and the isotherm using the modified Halsey (M,) model. The
deviations of the data from the model were large at relative humidity
values of 20% and 90%. At other relative humidity values the deviations
were not more than 1%.
Thin-layer drying
08
2 0.6
e
fO5
3
;;
504
Iz
03
o-2
01
I 1 I I I ,
Fig. 3. Effect of relative humidity of drying air on drying curve: x , 50%; l, 40%; 0,
30%; A, 20%; 0, 10%. ‘Moisture ratio’=(M- M,)/(M, -MC).
50 10 1.862 0 4.82 x 1O-J 1.191 5 0,776 9 4.40 x 10-5 0.427 9 2.734 7 0.174 4 6.40 x 1O-5
50 20 1.094 4 1.38 x 10-j 1.102 4 0.626 6 1.73 x 1o-q 0.738 2 I.870 8 0.134 3 1.05 x 10-4
50 30 1.012 2 4.27 x lo-” 1.014 6 0.982 3 4.50x lo-” 0.934 3 1.071 0 0.052 9 2.30 x lo-”
50 40 0.891 5 3.58 x 10m3 1.057 4 0.581 4 1.69 x 1O-1 0.404 7 15.861 0 0,032 5 1.40 x 10-j
50 50 0.936 2 5.00 x 10-5 0.946 5 0.960 9 4.00 x lo-’ 0.968 2 0.979 2 0.081 7 6.00 x lO-h
60 10 1.572 7 5.31 x lo-” 1.563 8 0.981 4 5.82 x lo- 0.928 0 1.743 0 0.128 8 4.61 x lo-”
60 20 1.517 9 6.14 x 1O-J 1.443 6 0,808 7 4.45 x 10-J 0.847 1 1.923 6 0.182 1 2.56 x 1O-1
60 30 1.328 6 5.52 x lo- 1.323 2 0.970 6 5.88 x lo-” 0.933 5 1.482 9 0.106 4 4.18 x 10-j
60 40 1.285 0 1.46 x 10-j 1.238 8 0.588 6 3.26 x lo-” 0.793 3 1.999 0 0.107 6 2.40 x lo-’
60 50 1.142 2 3.64 x 1O-1 1.144 2 0.924 3 3.53 x lo-” 0.914 3 1.314 8 0.135 3 1.58 x 10-j
1.5-
14.
13.
1.2
7
5 11
- i
* /
2 lo-
m
;;
Fj o-9-
”
~0s
>,
b
07r
10 20 30 40 50
Relative humidity W
Fig. 4. Change in value of drying constant with temperature and relative humidity.
Temperature: x ,4o”C; 0, 50°C; 0,6O”C.
Fig. 5. Comparison of observed and predicted moisture ratio for samples dried at
60°C. Relative humidity: X , 10%; Fi, ‘_9$, 0, 50%. ‘Moisture ratio’ = (M - Me)/
I e.
for in-shell pecan (Chinnan, 1984). The effect of relative humidity on the
drying constant has, however, not been clearly established.
The drying constant K obtained was related to the drying temperature
for each relative humidity setting using the Arrhenius-type relationship
Thin-layerdrying of melon seed 319
in eqn (7). An attempt was made to relate the constants K, and C to rela-
tive humidity:
K = 2 X lo4 exp[ -(C, + C,H)/T,] (8)
The standard error of estimate of K is O-05. A comparison of values of K
predicted by the model with those calculated from drying results is
presented in Fig. 4. The figure shows good agreement between experi-
mental data and the model. Experimental values of moisture ratio were
compared with values predicted by the use of eqns (3) and (8) for some
experimental conditions as in Fig. 5. The results show good prediction of
moisture ratio throughout the duration of drying.
REFERENCES