Improving The Operation of A Commercial Mango Dryer: Donald G. Mercer PHD Peng
Improving The Operation of A Commercial Mango Dryer: Donald G. Mercer PHD Peng
Improving The Operation of A Commercial Mango Dryer: Donald G. Mercer PHD Peng
Robertson, G.L. & Lupien, J.R. (Eds), © International Union of Food Science & Technology (2008)
Mangoes are an important commercial crop in many tropical countries. The drying of mangoes is an
ideal value-added opportunity for processors in many developing countries because the processing
requirements are relatively non-capital intensive. In addition, there is a ready market for the product
both domestically and in the export trade. The regular export trade, dominated by a few countries in
South East Asia, consists of mangoes dried with sulphites to stabilize color and with sugar added to
improve product texture. Processors in Burkina Faso, West Africa have decided to focus their
attentions on the organic market, since mangoes produced in this fashion can be sold at premium
prices. The mangoes are usually dried near where they are grown in forced-air cabinet-type dryers
fuelled by bottled gas or heated by solar radiation. Sensory quality is particularly difficult to control in
this product since, without chemical stabilization or added sugar, colour changes (due to both enzymic
and non-enzymic browning) and/or texture defects can occur. Careful control of temperature and
humidity parameters as well as drying chamber design is critical to achieving optimal product quality.
Dried mango can be used as an illustrative example of the impact of technology transfer on the
improvement of small-scale food processors in developing countries. Operational problems that are
typically encountered can prevent expansion and reduce profitability. Through technology transfer
such problems can be overcome and economic viability achieved.
INTRODUCTION
In this case study, a representative food drying operation similar to many found in developing
countries is considered. The fruit of the mango tree (Mangifera sp.) is a major commercial tropical
tree fruit, grown in many countries of the world. Although the total production of mangoes in Africa is
small in comparison to production in other areas of the world, the sale of fresh and dried mango
represents an important domestic and export value-added opportunity.
Due to the perishable nature of fresh mangoes, their export to Europe and America from Africa
represents a formidable challenge. Capital-intensive cold storage and refrigerated transportation
requirements can easily be beyond the technical capabilities of processors in poorer African nations.
Dried mangoes, however, are a less capital intensive value-added product that can be easily processed,
stored and shipped.
World-wide, dried mango represents a multi-million dollar market. Aimed primarily at the American
and Western European countries these dried mangoes, stabilized though the use of sulphites and
added sugar, can be used as an ingredient in many products such as breakfast cereals and granola-type
bars. Much of the production of dried mango for the above mentioned markets originates in Asian
countries such as the Philippines and Thailand. African processors find it difficult to compete in this
market with those countries.
Many African countries, however, have identified organically-grown and processed dried mangoes as a
market in which they can compete, especially Europe. In Burkina Faso, the mango season begins in
June, when especially certified growers (internationally certified by entities such as Eco-cert©) begin
shipping their product, primarily from the South-West, to processors throughout the country.
Although most mango dryers are centered around Bobo-Dioulasso, many processors also operate
outside that area (e.g., Ouagadougou).
In Burkina Faso, processors can operate their businesses in cooperation with other similar processors
as a kind of commodity-based cooperative or club (Fr. approche filière), or as an independent entity.
The advantage of the filière approach is enhanced access to technical support from governments and
Non-Governmental Organizations (NGOs). Mangoes are transported to the processing establishment
where they are stored (some for final ripening), sorted and washed.
Washed mangoes enter the processing areas, where they peeled and sliced by hand into about 1cm
thick slices. The slices are then arranged upon a drying frame that consists of a fine nylon mesh,
placed over a large metal screen, which in turn is supported by a wooden frame (Figure 1). Frame size
is about 0.75m by 1.0 m. The processing facility is certified organic by an internationally recognized
certifier (Eco-cert©), and meets all of the sanitation and food safety requirements of that organization.
2
Figure 2 shows the small production-scale cabinet dryers that use heated air as the drying medium.
Each dryer chamber is about 0.75m wide and about 2.5 m tall. In a typical installation there are eight
such chambers. The drying frames, containing the mango slices, are slid into an open chamber, each
chamber holding about 20 frames. Once fully loaded, the doors of the dryer are closed and drying
commences. After an appropriate time, the dried mango slices are removed from the dryer, cooled,
inspected and placed into low density polyethylene plastic bags which are then heat-sealed and
labelled.
The mangoes produced by these processors must meet organic standards. This means that sulphites
or other types of preservatives cannot be used to control enzymic browning. Rather control of
enzymic browning must be attained through careful control of product temperature and moisture
content. In addition, since no sugar can be added, the texture of the product must be controlled
through careful control of water activity (aw). In an attempt to control enzymic browning, the
processors start the drying process at very high temperatures (often as hot as 80°C). Theoretically,
thermal denaturation of the intrinsic polyphenol oxidase should occur at this temperature. Drying
temperatures are slowly reduced over an 8-12 hour period to about 65°C. There is no monitoring or
control of product moisture content, or humidity of the drying air.
Even though the dryer is being supplied with high temperature air, drying of the mango slices is not
uniform. When the drying chambers are opened after drying, some slices appear too dry and have
inconsistent colour (too dark). The texture of the dry slices also appear to be inconsistent, with some
too firm. Paradoxically, a portion of the mango slices appears to be too moist at the end of the run,
with voids, or pockets of un-dried mango contained within a pocket of dried, firm mango. If dried
mango slices are not quickly placed into heat-sealed plastic bags, they begin to noticeably darken
within 2 days.
It would seem obvious that process control has not been achieved using the present regimen.
Although denaturation of browning enzymes should occur at temperatures approximating 80°C, the
actual temperature of the mango slices will be substantially lower. In this non-adiabatic process,
water evaporation from the surfaces of the mango slices reduces internal temperatures. Consequently
little control of enzymic browning can occur, resulting in the browning phenomenon seen when dried
mangoes are exposed to the air. The texture of the mango slices appears be too dry and firm. Texture
control of the product must be achieved through careful control of aw, allowing the natural sugars
present in the fruit to act as humectants. Additional textural defects appear from the effect of case-
hardening, where too rapid evaporation of moisture from slice surfaces, hinders or blocks the
migration of moisture from the slice interior to the surface. Indeed, some slices must then be placed
back into the dryer and dried for several additional hours. Running the dryer with only a partial load
of product creates an additional cost and reduces the number of batches that can be dried over the
course of the mango processing season.
At these high drying temperatures and high air flow-rates, the problem of browning and non-uniform
drying persists. Until the design of the existing dryers and the drying regimens change, very little can
be done to significantly improve the quality and value-added opportunity of the product.
In addition to these problems with the mango drying, owner are faced with the seasonal nature of
their business. Once the crop of mangoes is harvested and dried, the equipment sits idle until the next
processing season begins. In order to maximize the use of their capital assets and spread fixed costs
over a longer time period, owners would like to diversify the drying operation by drying other
materials during the ‘off-season’. However, due to the difficulties being experienced with the mangoes,
this is not considered to be a viable option until the dryer problems are resolved.
3
DESCRIPTION OF THE CABINET DRYER
A schematic diagram of the cabinet dryer is shown in Figure 3. Dimensions of the dryer are not stated
since this is meant to represent a generic unit rather than that of a specific operation. Air is drawn
into the bottom of the cabinet dryer by fan, is heated by bottled gas, and forced into the drying
chamber. Heated air is directed across a number of drying frames containing sliced mangoes in the
first dryer section (section “A” in Figure 3). For clarity, only three frames are shown in each section of
the dryer, although in actual dryers there are often more than this number. After reaching the
downstream end of section “A”, the air is directed upwards into section “B” which is separated from
section “A” by strategically placed partial ply-wood partitions.
The air that has left section ‘A’ then reverses direction and passes across the frames of sliced mangoes
located in section ‘B’. At the end of section ‘B’, the air is once again directed upwards into the third
section of the dryer, section ‘C’, where air flow is once again reversed. Air continues up in this
serpentine way through sections ‘D’ and ‘E’, after which the air is exhausted into a plenum
arrangement, combining the exhaust of two chambers at the top of the dryers and expelled through a
common flue.
Temperature of the drying air can be monitored by three or four dial thermometers inserted through
the walls of each section. There are no automated process controls on the drying units. Airflow is
controlled by adjusting the speed of a motor on a fan that introduces the air into the first zone of the
dryer. As with most dryers of this type, the fan is usually run at full speed to maintain maximum air
flow to the dryer.
The fruit of the Mango tree (Mangifera sp.) grown in Burkina Faso are about 125 mm long and about
75 mm in width. The fruit consists of an inedible central flat seed surrounded by yellow flesh and an
outer inedible skin. The edible flesh contains about 82% moisture. The main chemical constituents on
a % dry weight basis (% d.b.) are shown in Table 1 (3).
The flesh contains about 70% sugar, mostly composed of sucrose, and a relatively large amount of
fibre, reflecting the fibrous nature of the fruit. Mango is unlike fruits such as raisin or dates which
contain large amounts of sugar, but relatively small amounts of fibre. The fibrous nature of the fruit
4
explains why dry mango tends to be somewhat firm in texture. Careful control of a w is essential if the
processor is to avoid an overly tough consistency. Figure 4 shows generic water sorption isotherms of
mango fruit (3).
Fruit Fru Glu Suc NSP NSP NSP NSP NSP NSP Protei Ash
Ara Xyl Man Gal Glu Uroni n
c acid
Raisin 51.16 43.14 0.00 0.20 0.10 0.00 0.00 0.90 1.10 3.05 2.32
Date 42.25 43.55 0.53 0.17 0.54 0.16 0.34 1.68 1.55 2.39 1.94
Pineappl 21.36 18.33 44.40 1.20 2.10 0.30 0.70 4.00 0.70 2.72 2.72
e
Mango 17.68 6.36 46.11 3.70 1.10 0.20 1.70 5.40 5.90 3.83 2.19
Fru = Fructose Glu = Glucose Suc = Sucrose Ara = Arabinose Xyl = Xylose
NSP = Non-starch Polysaccharides Man = Mannose Gal = Galactose
In general, dried fruit are considered shelf stable if the aw is at 0.6 or below. For dried mangoes, the
equilibrium moisture content (% d.b.) equivalent to an aw of 0.6 is about 15% moisture. Careful
monitoring of the mango slices as they are drying is essential to avoid over drying. Such over drying
could lead to texture defects such as tough consistency.
Through experience, the dryer operators know approximately how long it takes to process an average
load of mangoes. After drying for this amount of time, the dryer is shut down and opened to remove
the racks of dried product. In general, the mango slices in section ‘A’ at the bottom of the dryer (Figure
3) are over-dried (i.e., their moisture content is below that considered acceptable) and they are darker
in colour (probably due to Maillard reaction products) than those in section ‘E’ which is the last section
of this particular dryer. The mango slices in section ‘A’ also have a less pliable texture (due to the
extremely low aw) than the softer slices in the final sections of the dryer. Not only is there variation in
moisture content from the bottom to the top of the dryer, but moisture content variation also occurs
within each section of the dryer. There are often regions along the centre of each rack where the
mango slices are drier than those along the edges of the rack. Often, the mango slices on the bottom
rack in sections ‘B’, ‘C’, ‘D’ and ‘E’ are not as dry as those on the upper racks in these sections.
5
Faced with the problems outlined above, processors may experience excessive rejection of product
that fails to meet specifications for moisture, colour, and texture attributes. Mango slices with
excessive moisture levels (especially those in section ‘E’) must receive additional drying which reduces
the number of full loads of product that can be scheduled. Overly dry dark product cannot be sold at a
premium price, and when possible is sold at a loss to recover some of the production costs. Hand
sorting of mango slices from each dried batch to remove high moisture product for reprocessing and
overly dried product adds to the labour costs.
In order to improve the operational efficiencies of the mango drying operation, the processor should
conduct a thorough ‘audit’ of the drying process. This audit should consist of two tasks. The first task
is to evaluate the drying protocols regarding drying times and temperatures. A flow-chart diagram of
the process should be constructed which can outline where potential problems can be identified. Once
the locations of the problems have been identified, protocols can be changed to address those
problems. These protocols have a profound and direct effect upon the quality attributes of the
product. The second task should be to assess the drying equipment itself, to better measure drying
parameters and enhance drying efficiencies. Since airflow is one of the primary contributors to drying,
along with time and temperature, it seems reasonable to begin by examining airflow patterns within
the dryer. While pitot tubes or anemometers could be used to determine actual air velocities and
pressure taps could be installed in the dryer to determine uneven pressure distribution, these devices
are not really necessary for a basic examination of how the dryer is functioning.
In a developing world situation, it is important that this audit be carried out in as simple a fashion as
possible, due to limitations in access to capital and technical expertise. The audit should not be an
overly arduous task involving sophisticated analytical equipment.
Before making any modifications to the dryer or changing operating conditions, it is necessary to
understand what happens to the mangoes during the drying process. In this way, conditions can be
better matched to the needs of the product which is something often neglected in many drying
processes.
By following a volume element of air through the dryer (as shown in Figure 3), a true appreciation of
the process can be obtained. As a volume of air enters section ‘A’ of the dryer, it is at a high
temperature (e.g., 80°C) and at low absolute moisture content. Its ability to remove moisture is high,
and it readily takes up moisture from the surface of the mango slices located in section ‘A’. The air will
also be distributed in a relatively uniform manner due to the inherent design of the air inlet into this
section. The relatively dry air absorbs moisture from the product, and its (the air) temperature
decreases. As the moisture content of the mangoes decreases, their temperature comes into
equilibrium with that of the air. This process continues in section ‘A’ as the air gains more moisture
and loses more heat. It is important to realize that although the main purpose of the very high air
temperature (besides removing moisture) is to denature the polyphenol oxidase, no such denaturation
can occur, since its (the air) temperature is reduced well below that necessary for enzyme inactivation.
As a result, enzymic browning can occur (2). Of perhaps more serious consequence, the high rate of
moisture loss from the surface of the mango slices creates a surface layer of very dry mango. This
surface layer, once formed, has a very limited ability for rehydration, and can prevent moisture from
migrating from the interior of the mango slice to its surface. This phenomenon, known as case-
hardening, can block further moisture loss and can lead to the development of interior pockets of
moisture which are very hard to eliminate.
As the air continues on its journey, the combination of heat loss and increased level of saturation
reduces the water removal capacity of the air as it enters section ‘B’. In section ‘B’, the air will still
have sufficient ability to remove water from the mangoes, but it will gain additional moisture and lose
more heat as it does so. As the air enters section ‘C’, its water removal capacity will be further
diminished and it will once more gain moisture and lose heat as it passes across the mango slices in
this dryer section. By the time the air reaches sections ‘D’ and ‘E’ in the dryer, it will no longer be able
6
to remove moisture from the product at the same rate that it did when it entered section ‘A’ of the
dryer.
If the dryer is operated to produce a satisfactory product moisture in the final section, product in the
four previous sections will most likely be overly dried and be suffering other quality attribute
problems such as off-colour development and textural defects. From this, it is evident that measures
must be taken to ensure that product in each section of the dryer receives a uniformly distributed
supply of air at a suitable temperature and initial moisture content to promote optimal quality
attributes and efficient drying.
Psychrometric charts coupled with wet and dry bulb temperatures, or dry bulb temperatures and
relative humidity readings, taken throughout the dryer can provide a more exact indication of what is
occurring to the air as it passes through the various dryer sections. However, simply envisioning that
the air is becoming more heavily loaded with moisture and is losing heat as it does so, points out the
need for an alternate approach to the drying protocol and how air flow is managed in the dryer.
Figure 5 shows the effects of air velocity (at 50°C) on the water removal from mango slices in a
laboratory-scale tray dryer (1). By plotting the moisture ratio (i.e., the dry basis moisture of the
mango slices at time “t” divided by their initial dry basis moisture), drying test runs can be compared
directly without the distraction of having different starting moistures. In these tests, observations
were taken every fifteen minutes over the course of the dryer runs. By comparing the moisture ratios
for the mango drying at 50°C with air velocities of 0.2 m/s and 0.5 m/s as shown in Figure 5, it can be
seen that increasing the linear air velocity across the surface of the mango slices has a pronounced
effect on the rate at which moisture is removed. This can be attributed, in part, to disruption of the
stagnant boundary layer of air that impedes the removal of water from the saturated surface of the
mango slices. With continuous removal of moisture from the product surface, drying efficiencies are
enhanced.
Figure 6 shows how increasing the temperature with an air velocity of 0.5 m/s can add to the
efficiency of water removal for the mangoes (1). Water removal rates during the first 2.5 hours for the
drying runs reported in Figures 5 and 6 are presented in Table 2. These values were obtained from
plots representing the initial constant rate drying period for mangoes (d.b. moisture versus time),
which are not shown in this paper.
7
Table 2: Water Removal Rates of Mango Slices under Various Drying Conditions
Figure 6: Moisture Ratios versus Drying Time for Mangoes in a Tray Dryer at Various Temperatures
Photographs of the mango slices at the start and end of a drying run appear as Figures 7 and 8,
respectively. Figure 8a shows the dried mango slices on the metal rack inside the laboratory-scale tray
dryer, while Figure 8b shows them after they have been removed from the dryer.
8
Figure 8a: Slices of Mango after Drying on Rack in Tray Dryer
Figure 8b: Slices of Mango after Drying in Tray Dryer (removed from dryer)
Using this information, it becomes quite obvious that drying efficiency increases with an increase in air
velocity and temperature (for a given moisture content). In order to produce optimal drying
conditions to achieve improved product quality, the drying protocol needs to be carefully monitored.
The moisture content and temperature of the drying air need to be as constant as possible as it passes
across the complete drying frame. The uniformity of the air distribution patterns must also be
increased in order to maintain optimal air velocity.
With enhanced air distribution, the drying protocol should be altered to take advantage of these
improvements. Initial drying air temperature should be reduced from the current 80°C to a more
moderate 50°C. At the lower water removal rates, moisture from the mango slice interior will have
time to migrate to the surface, where it can in turn evaporate. This lower initial water removal rate
will avoid problems associated with case-hardening. As drying proceeds, the temperature of the
drying air should be slowing increased, still allowing for the movement of water within the mango
slices, but also compensating for the lower water moisture removal rates, the result of lowering aw and
increased water binding. Drying air (and mango slice) temperatures as high as 80°C could be tolerated
in order to denature browning enzymes; as the product nears the monolayer moisture point. As the
product approaches the monolayer moisture point, however, drying air temperature should again be
reduced to more moderate temperatures (50°C) to reduce the production of Maillard reaction
products (2).
9
TASK TWO - ASSESSMENT OF DRYING EQUIPMENT
Looking at Figure 3, it can be seen that air entering section ‘A’ of the dryer strikes a vertical wall at the
end of the section and is then forced upwards into section ‘B’ where it strikes the bottom surface of the
horizontal divider between the top of section ‘B’ and the bottom of section ‘C. This flow pattern is
repeated each time the air leaves one section of the dryer and enters the next section. The
combination of two 90° turns for the air leaving one section and entering the next creates non-uniform
air distribution patterns.
One potential way to create a more even air distribution pattern would be to install louvers at the
downstream end of each section mounted at a 45° angle to deflect the air upwards into the next
section. Similarly, louvers could be mounted at a 45° angle at the upstream end of the following
sections to deflect horizontally across the racks of mango slices. In this way, the lower racks of
product would receive more airflow than they did previously. Figure 9 shows how modifications could
be made by installing louvers and deflectors to direct the air as it travels from one section of the dryer
to the next.
By improving the air distribution pattern in the cabinet dryer used for the mangoes, higher linear
velocities could be achieved to enhance water removal and drying uniformity. Increased turbulence
across the surface of the racks can reduce the ‘wall effects’ noted in many dryers with poor airflow
distribution. It may also be advisable to install several deflectors at the upstream end of each dryer
section to deflect air towards the walls of the dryer and minimize side-to-side moisture variations.
There is one significant feature of the dryer design in Figure 3 and the improved design with louvers
shown in Figure 9 that still needs to be addressed. Air passing over the mango slices in each dryer
section travels to the next section where it is expected to remove water from the mango slices located
there. ‘Water removal capacity’ is an effective manner in which to envision the ability of air to remove
moisture from a product. To determine the water removal capacity of air, it is necessary to know how
much water the air contains compared to how much water the air could hold if it was fully saturated at
that temperature. Absolute moisture contents can be obtained from psychrometric charts, when the
dry bulb temperature and wet bulb temperature or relative humidity of the air, are known.
The difference between the amount of water the air is holding and the amount of water it can hold
when saturated (i.e., 100% relative humidity) can be described as its maximum water removal
capacity.
10
Figure 10 shows a proposed re-designed cabinet dryer which addresses the problems experienced
with the original dryer shown in Figures 3 and 9. Heated air is blown into a plenum that runs the full
height of the cabinet dryer. Several entry points, instead of one single supply point, could enhance the
uniformity of air distribution into the heated air plenum. Perforated panels form a wall between the
heated air plenum and the five sections containing product to be dried. These panels consist of sheet
metal or other appropriate matter with small holes (e.g., 2 cm diameter) drilled at regular intervals to
provide an open area of approximately 30% to 40% of the total panel area. They act to create a back-
pressure in the hot air supply plenum which then promotes uniform distribution of the air into the
drying sections. In this way, all five sections of the dryer will receive a relatively uniform supply of
heated air with enough water removal capacity to remove moisture from the mango slices. The
velocity of the air will also be sufficient to sweep away the stagnant boundary layers that inhibit
moisture removal. Once the air leaves each dryer section, it passes through a second set of perforated
plates designed to maintain uniform air distribution inside the drying sections. The air then enters an
exhaust plenum where it is directed out of the dryer. The exhaust air could then pass through a
counter-current heat exchanger where its residual heat could be used to warm fresh incoming air.
Optionally, a portion of the exhaust air could be mixed with fresh incoming air to recover some of the
heat, providing the exhaust air’s water content is not too high for this purpose.
For smaller dryer loads, sections of the dryer could be blocked off and only as many sections as
required would need to be used. This further enhances the flexibility and efficiency of the dryer’s
operation.
ECONOMIC OPPORTUNITIES
By reassessing drying protocols and re-designing dryers, processors can reduce product variability,
thereby lowering losses due to excessive rejection of product that fails to meet specifications for
moisture, colour, and texture attributes. This would lead to improved efficiency and enhanced
profitability. More runs could be conducted during the harvest season since there would be less need
to re-dry mango slices from runs which previously contained overly moist product.
Increased flexibility and uniformity in dryer operation would allow processors to seek out alternate
products to dry when mangoes are out of season. Other fruits and vegetables requiring different
11
drying conditions could be dried by altering drying protocols and dryer configurations to meet the
different drying needs. Having the dryer in operation throughout the year would maximize the
utilization of the business’s capital assets and take advantage of a variety of opportunities.
When dryers are operated year-round, rural employment can become less seasonal and more
sustainable. These benefits could then spread throughout the local community. With a larger
industrial processing base, capacity expansion or additional diversification could be possible. This
would further increase rural employment, providing economic benefits for all. By understanding the
drying kinetics of each product being dried, process operating protocols could be modified to match
air temperature to the drying needs of the product. During the initial and later stages of drying, lower
temperatures may be required by the product. This would allow the dryer to be operated at lower
temperatures for as much as half of the drying cycle, resulting in fuel savings and lowering production
cost. Not only would this result in fuel cost savings, but over-drying and under-drying could be
reduced thereby improving product quality, resulting in more saleable product per dryer run with
subsequent improved profitability.
SUMMARY
Small and medium-scale businesses that produce value-added products provide opportunities for
economic development in many developing countries. These businesses create jobs and provide much
needed incomes for the urban and rural poor. These value-added products, produced from locally
sourced raw materials, are not capital intensive, and take advantage of local labour markets, providing
decent incomes for those that otherwise would continue to exist in impoverished conditions.
It is important for donor nations, in their official developmental assistance (ODA) programs, to
support private sector entrepreneurs in the developing world. With such support, the entrepreneurs
will be in a better position to identify markets and develop suitable processing technologies.
Continued improvements in production processes can result in new and expanded markets, which in
turn can enhance efficiencies and profitability.
Although technology transfer from developed nations is an essential part of any donor nation’s ODA, it
is important for them to understand the economic and infrastructural constraints found in the
developing world. In environments which lack capital with poor infrastructure, optimization of
relatively simple, straight-forward processes should be favoured. By examining the mango drying
business from a fundamental perspective, basic changes can be recommended that have the potential
to overcome most of the difficulties outlined in this case study. It is important, however, for developed
countries to promote appropriate technologies: to do otherwise would be to chance disappointing
results.
REFERENCES
2. Devereau, A.D.; Myhara, R.; Anderson, C. Physical Factors in Post-Harvest Quality. In:
Crop Post-Harvest: Science and Technology, Vol 1. Principles and Practice. Golob, P.,
Farrell, G., Orchard, J.E., Eds; Blackwell Science Ltd: London, United Kingdom, 2002; pp. 69-92.
3. Myhara, R.M.; Sablani, S.S. Unification of fruit moisture sorption isotherms using artificial
neural networks. Drying Technology 2001, 19 (8), 1543-1554.
ACKNOWLEDGEMENT
The financial support of the Canadian International Development Agency (CIDA) for this project is
gratefully acknowledged.
12