3. Results and discussion
3.1. Drying mode 1
The recorded temperatures in apples drying at constant temperatures
are shown in Fig. 2a and the corresponding drying curves
are shown in Fig. 2b. It is clear that drying apples at lower temperatures
required longer time and more energy (Table 1). High temperature
drying was preferred in terms of time and energy saving.
However, quality evaluation showed that middle and low temperature
drying resulted in better quality (Tables 2 and 3), and high
temperature drying resulted in worse quality, probably because
of the charring effects. Although the medium drying temperature
resulted in acceptable time, energy, and quality, it could still be
improved.
In our previous study it was found that during the beginning
and middle drying stages (Li et al., 2010a, 2010b), the emanation
of the aroma from food samples was fast, as long as the moisture
removal rate. The reason might be the fast loss of moisture in these
stages created greater pores in the apples, where more and bigger
aroma molecules were able to escape from the apples; or the fast
moving moisture brought more aroma molecules out of the samples.
A large microwave power requirement at these stages also
caused product charring. If the drying rate in this stage could be
slowed down, the pores might be smaller and less aroma would
be lost, hence better product quality might be achieved.
3.2. Drying mode 2
From Tables 1–3 it can be observed that drying apple at 40 C
resulted in a good product quality, and the drying time and consumed
energy were also acceptable. To mimic and improve this
drying curve in the beginning and middle drying stages, a linear
line was developed where moisture content is decreased from
7.0 to 1.0 (d.b.) in exactly 120 min, after that the drying temperature
was maintained at 50 C until the end to reduce drying time
and energy consumption.
By applying the linear line in drying process, apple drying rate
was controlled through temperature adjustment (Fig. 3b). It can
be noticed that although the actual drying curve didn’t follow
the designed drying line exactly, the difference was small (less
than ± 0.5). This difference was hard to reduce because the drying
temperature was limited to 25–75 C and this range cannot be expanded
for the quality requirements and ambient temperature
limits. The adjusted temperature profile is shown in Fig. 3a. It
can be observed that the temperature almost increased from 30
to 50 C in 120 min, and then kept at 50 C until the end. The high
temperature at the very beginning was because the designed initial
moisture content was 7.0 (d.b.) and the actual moisture content is
different from this value. However, this would not affect the drying
effects much because it did decrease fast and did not last long. The
temperature fluctuations became larger in the final drying stage
when the mass became smaller and power density became greater,
as discussed in our previous study (Li et al., 2010d). This was also
one of the reasons why the final temperature was chosen at 50 C,as the highest temperature almost reached 75 C at certain
moments.
3.3. Drying mode 3
Although the previous drying mode did achieve a desired drying
curve, the system requirement was complex. It needed the data of
the initial moisture content, an accurate mass measurement, an
online calculation of the moisture content, and a calculation and
control of the real time drying temperature. This might be difficult
for some industrial applications.
To simplify the system, a linear temperature control method
was developed. In this method, drying temperature was increased
from 25 to 50 C in exactly 120 min and then kept at 50 C until the
end of drying. The moisture content was not controlled but recorded
for analysis purpose. The temperature control is shown in
Fig. 4a and the drying curve is shown in Fig. 4b. This method
avoided the acquirement of the initial moisture content of the sample,
and avoided the temperature fluctuation at the beginning. Also
the calculation of the online moisture content was not necessary.
The resulted drying curve is almost a linear line in the first
120 min, very similar to the drying curve in Fig. 3b. Hence, a linear
temperature profile resulted in a near linear drying curve, and the
system was greatly simplified.
In a previous study of carrot drying (Li et al., 2010a), a similar
linear temperature line at the beginning and middle stages was
achieved, although it was through a fuzzy logic control based on
aroma emanation. These results indicate that a gradually increasing
temperature at the drying beginning can achieve good control
effects and product quality, as analyzed later in this study and the
results reported in our previous paper.
3.4. Drying mode 4
As a continuous temperature change still need complex software
control and may not be applicable in some industrial practices,
it is desirable to develop a drying strategy where
temperature control can be further simplified. To address this
problem, a 3-step temperature drying was developed, i.e., drying
apples at 30 C for 60 min, at 40 C for another 60 min, and then
kept at 50 C until the end of drying. This method is a simulation
of the linear temperature controls in drying mode two and three,
but the temperature adjustment could be implemented manually
(Fig. 5a). Moisture content was also recorded but not controlled
(Fig. 5b). It can be observed that the drying curve is still a nearly
linear line, although not so smooth as in the previous modes.
Two turning points in the drying curve were corresponding to
the two temperature changes. In general, the drying curve is similar
to that in drying mode two and three with little changes.
3.5. Quality assessment
The overall results of color measurements are shown in Table 2.
In drying mode 1, 40 C drying resulted in the highest L⁄ value,
indicating a very white product. In other modes, real time and 3-
step all got the high L⁄ value. The linear control did not achieve a
high L⁄ value, but had low (absolute) a⁄ and b⁄ values, indicating
a none-burned good color.
Sensory evaluation is presented in Table 3. In drying mode one,
middle drying temperature resulted in good taste, but low temperature
resulted in good visual appearance. Among all, 40 C drying
has the best drying effects. Real time, linear, and 3-step temperature
control all resulted in good visual appearance. However, only
real time control achieved the best taste and textural consistency.
Hence, a linear drying curve is recommended in microwave drying
where drying rate can be controlled and less aroma is lost, by
which the final product quality can be improved.
Similar effects can also be found from our previous study (Li et
al., 2010a, 2010b), where real-time control always achieved the
best product quality. The reason is that the real-time control can
best adapt to the actual product’s characteristic and achieve thebest control effects, while the other fixed methods are only universal
control modes and cannot vary according to individual samples.
The novel control methods found in this study have the potential
to be used in industry. The principle is just to reduce the drying
rate in the middle drying stage. With this treatment, more preferred
aroma can be retained, charring can be avoided, drying time
can be reduced and energy can be saved. In an industrial microwave
drying system where the products are moving, infrared sensor
can be used to replace the optic fiber sensor. All other
parameters can also be optimized by specifically designed experiments
conducted in advance.