The performance of natural convection solardryers for copra production

C.K. SANKAT and R.A. ROLLE

Faculty ofEngineering, University of the West Indies, St. Augustine, Trinidad. Received 20 February 1990;accepted 25 October1990.

Sankat, C.K. and Rolle, R.A. 1991. The performance of naturalconvection solar dryers for copra production. Can. Agric. Eng.33:085-091. Three natural convection solar crop dryers of simpledesign and suitable for use by small farmers in the tropics wereevaluated with respect to their ability to dry split, dehusked coconuts.Dryer A was an indirect dryer consisting of an air pre-heater anddrying chamber. Dryers B and C were direct dryers of the cabinet type,with dryer C being simpler in design and lowest in cost Air temperatures in the dryers increased above ambient (29°C) by maximumvalues of 22°C, 11°C, and 9°C for dryers B, C and A respectively.Dryer B showed the fastest rate of coconut moisture content reductionwhile there was little difference in this parameter between dryers A andC. On the basis of the effective collector areas, estimated moisture

2 1removal rates were 0.66,0.64 and 0.58 kg H2O m" -day" for dryers C,A and B respectively. In the direct dryers, kernels which were positioned to face upwards and directly exposed to the solar radiation driedfaster than those oriented in a downward position. Dryers B and Cproduced copra of good quality, except for slight browning of theproduct in dryer B. In dryer A, 67% of the kernels were of unacceptable quality due to fungal growth.

On a evalue le rendement de trois secheurs solaires de recolte parconvection naturelle. De conception simple, ces secheurs conviennentparfaitement aux petits exploitants agricoles des tropiques. On aevalue la capacite de ces appareils a secher des noix de coco fendueset decortiquees. Le secheur A comprenait un prechauffeur d'air et unechambre de s^chage. Les secheurs B et C etaient des secheurs directsdu type armoire. C etait moins cher et de conception plus simple. Lestemperatures de l'air, dans les secheurs B, C et A, depasserentrespectivement la temperature ambiante (29°C) par desvaleurs max-imales de 22 °C, 11 °C et 9 °C. Le secheurB indiqua le taux le plusrapide de reduction d'humidite des noix de coco, tandis que lessecheurs A et C presentaient peu de differences. A partir d'aires decollecteur effectives, les taux de deshumidification ont ete estimes a0,66, 0,64 et 0,58 kg de H2O m" •jour" , pour les secheurs C, A et Brespectivement. En ce qui a trait aux secheurs directs, les noix, dontla face etait exposee vers le haut directement aux rayons solaires, ontseche plus rapidement que celles orientees vers le bas. Les secheurs Bet C produisirent du copra de bonne qualite, a 1'exception d*un legerbrunissage du produit dans le secheur B. Dans le secheur A, 67 %desnoix etaient de qualite inacceptable a cause de moisissure.

INTRODUCTION

The coconut palm (Cocos nucifera) grows widely in the Caribbean islands, either in pure stands on the large estates orintercropped on small farms. It is a palm of considerableeconomic importance, with the principal product of the coconut being the dried endosperm or kernel, commonly called"copra". This is used for the production ofedible oil. The freshcoconut kernel is usually dried from an initial moisture contentof 40-50% (wet basis) to a level below 10%, to ensure safe

CANADIAN AGRICULTURAL ENGINEERING

storage. Whole, dehusked coconuts are usually split intohalves (split nuts) and dried with the kernel in the shell. Thedried kernel or copra is more readily removed from the shellafter the drying process.

In the Caribbean islands, copra dryers commonly used byfarmers can be put into one of three categories, this beingdetermined principally by the size of the farm. The dryers usedare:

(i) Forced convection, flat bed, batch type dryers which utilisefuel oil as the energy source. The drying air temperature isusually in the range of 60-70°C. These dryers, limited innumbers and usage, are used on large (150-1300 ha) ormedium sized (5-150 ha) estates.

(ii)Natural convection, batch type copra dryers which directlyutilise the heat of combustion of coconut shells as the sole

source of energy. These are widely utilised on large andmedium sized estates.

(iii)Open, natural drying of copra using direct sunshine (sundrying). This is the simplest and cheapest method availablefor copra manufacture, usually with no initial investmentrequired, except in some cases where a concrete floor isused for spreading and drying the nuts. A drying time of6-7 days ofgood sunshine is required (ambient temperatureaveraging 30°C) for split coconutsdried in a single layer.This method is used extensively by small scale farmers,many having farm sizes less that 1 ha. However the productis generally of variable quality, often mouldy and contaminated by dust and dirt.

OBJECTIVE

The objective of this study was to develop and evaluate a solardryer of simple design, construction and operation, as well asof low initial cost for the use of small-scale Caribbean farmersand copra producers. A dryer with a batch capacity of 100-150coconuts/week of operation appeared adequate to meet theneeds of such farmers. In evaluating the dryer, its performancewas to be compared with that of two natural convection solarcrop dryers of conventional design. The drying of split nuts(kernels in shells) was to be evaluated, as this is the preferredmethod of many copra producers.

SOLAR DRYER DESIGNS

Three solar dryers were evaluated in this study and are shownschematically in Fig. 1. On these designs, two were existingdryers (dryers A and B) of previously reported design while

85

drying chamber^

sokjro/r heater*.

air inlet

air inlet

absorber^,plate

air inlet

DRYER B

transparencaver^^^

mid.

air exit

dryingtrays

air exit

drying trays

absorber plate

air exit

DSYER C

Fig. 1. Schematic of three natural convection solar cropdryers.

the third, dryer C, was designed and built specifically for thisstudy. It is noted that Trinidad's location is 10.5°N latitude,and that the optimum angle (0) that a fixed, solar collector istilted from the horizontal is a function of the latitude ($) andthe declination angle (8) of the sun, that is

0= $ + 5 (1)

The drying studies were programmed for the month ofMarch, usually the hottest and driest month of the year. Thechosen design slope (a) to the horizontal for the direct, cabinetdryers Band C was therefore 10°, asa declination angle of0°was used. For indirect dryer A consisting ofa separate solar airheater anddrying cabinet, a wassetat 33°soas toenhance theair flow through the dryer.

Table I shows the specifications of the three dryers used inthis study.

Dryer A

Kalra and Bhardwaj (1981) described a dryer of this type forfruit and vegetable products. The solar air heater was of theparallel pass type, with air moving by natural convection oneither side of a blackened, corrugated metal surface which waspositioned between a glass cover and an insulated (25 mm

Table I. Specifications of the three natural convection solar dryers used for copra drying

Specifications

Type

Collector

Size (m)

Material

Slope, a

CosPArea (m2) - AiEffective area (m ) - A2**

Tray

Size (m)

Number

Total area (m2) - A3No. split nuts - N

A2/A3

N/A2 (split nuts/m2)N/A3 (split nuts/m2)

Airflow

Port size (m)

Area (m )

Cost($US)

Dryer A

Indirect

1.84x0.93

glass

33°

23°

0.92

1.71

1.57

0.89 x 0.38

4

1.35

126

1.16

80

93

1.05x0.075

0.079

275

* P=noon anglebetweeninsolation andvectorperpendicular to collector= (ct-«|> + 8))

** Effective area= Ai cosP

86

Dryer B Dryer C

Direct Direct

1.86x0.92 2.40x1.20

glass plastic10° 10°

0° 0°

1.00 1.00

1.71 2.88

1.71 2.88

0.89x0.38 2.40x1.20

4 1

1.35 2.88

126 256

1.27 1.00

74 89

94 89

0.86x0.05 1.20x0.10

0.043 0.120

218 80

SAKATandROLLE

styrofoam) plywood bottom. The woodendrying chamberwas1.05 m highx 1.05m widex 0.38 m deep and wasdesigned tohold four trays stacked vertically. Hot air from the air heaterrose through the stack of trays and left the dryer through arectangular opening located at the top of the rear panel of thedrying chamber.

Dryer B

This dryer is a cabinet dryer of the multi-rack design, previously described by Sandhu et al. (1979). Four drying trayscould be removed or inserted into the drying cabinet throughrectangular slots (0.38 m x 0.10 m) provided for on the side ofthe cabinet. Rectangular ports provided for air movement atthe lower front, and top end of the dryer. The horizontal dryerbase (absorber plate) was made of galvanized sheeting andbetween this and the dryer support, 25 mm thick styrofoaminsulation was used.

Dryer C

This dryer of original design is shown in Figs. 2 and 3 and wasdesigned specifically for the drying of split coconuts by smallfarmers. Simplicity in design, low cost and ease of construction were the essential elements in the design considerations.It consisted of a wooden-sided cabinet with a corrugated galvanized base sheet located 0.20 m from the top. Screened airinlet and exit ports were provided for at the front and rear endsof the dryer. A transparent plastic cover was made to fit snuglyover the sides of the cabinet. The cover was made with a

wooden frame to which clear plastic sheeting was attached.When operating, the air exit side was raised and supportedabove the air inlet side, so as to provide a cover with a slopeof 10° to the horizontal.

The interior surfaces of all the drying cabinets and the airheater of Dryer A were painted with flat, black paint and thesloped solar collectors were all positioned to face south.

corrugated base sheet

clear p/ast/c cover(.removable)

air inlet(screened)

Fig. 2. Schematic of direct, cabinet copra dryer (Dryer C).

CANADIAN AGRICULTURAL ENGINEERING

MATERIALS AND METHODS

Split, dehusked coconuts were dried in the three solar dryerssimultaneously with these trials conducted over a 5-day periodin the month of March, this being in the dry season of the year.

Coconut loading

Dryers A, B, and C were loaded (single layer) with 126,126and 256 split nuts respectively giving a near uniform loadingdensity ranging from 89-94 split nuts/m2 oftray surface area(Table I). Uniformity in loading was necessary to comparedryer performances. Dehusked coconuts used in this studywere randomly selected, with the average weight per split nutbeing 0.197,0.176 and 0.182 kg for dryers A, B and C respectively. For the purposes of obtaining drying data in dryers Aand B, three pairs of split nuts were labelled, individuallyweighed and located at the centre and the two ends of eachtray. For each pair of split nuts, one was positioned with thekernel facing down, while the other was positioned with thekernel facing up and therefore directly exposed to the sun.This procedure was adopted to evaluate the effect of kernelorientation on the rate of drying. In each of dryers A and Btherefore, the weight losses with time of 24 split nut samples(4 x 6) were determined.

For dryer C, consisting of a single drying tray, fourteen splitnut samples were labelled, individually weighed and distributed across the inclined dryer surface, four on the top section,six on the middle section and four on the bottom section. Of

these fourteen samples, nine were oriented with the kernelfacing up, and directly into the sun. In all three dryers, theremaining split nuts which were being dried but not individually monitored, were generally positioned with the kernelfacing upwards.

Instrumentation

In dryers A and B thermometers were supported in the centreof each of the four trays to measure the drying air temperatures, with their bulbs well into the air stream and shielded

03/7?

•inclined iom tothe horizon talwhen operating2 Am

wooden sides

87

Fig. 3. Direct cabinet dryer containing split, dehuskedcoconuts (Dryer C).

from the sun by the crop. Three thermometers were used fordryer C, located at the top, middle and the bottom, whileambientair temperature was measured by an additional thermometer. Insolation measurements were made with ahand-held Dodge Solar Meter (Model 776) while split nutweights were measured with twoSartorius (Model 1216 MP)electronic, top-loading balances. A forced convection oven(Blue M-Model OV490A-2) was used for moisture contentdeterminations.

Procedure

Drying beganat 1600hon the first day,andcontinued to 1600hon the fifth day. Dryers were left unmodified at nights andduring any period of rain. The masses of labelled samples ofsplit nuts were measured initiallyand at 900h and 1600h eachday,whilethermometer readingsandinsolation measurementswere taken at hourly intervals. At theendof thedrying run, thekernel was readily separated from the shell of the labelled,split nuts by knocking the samples, face down, on a hardsurface. Thedryweights of thecorresponding kernel and shellsamples were determined after oven drying at 75°C for 48hours. From this data, themoisture contents of the split nutsduring drying were determined, as well as the percent kernelcontent by weight (Md) of each splitnutwhich was sampled.

Estimation of kernel weights

In thedrying process, theweights of splitnuts(kernel andshelltogether) were recorded, and therefore the average moisturecontent of such samples could be readily determined. Ofparticular interest in this study however, was the variation in thekernel moisture contentwithdrying time. Thiscouldbedetermined only from a knowledge ofthe changes inkernel weightswhich occurred with drying. Kernel weights in splitnuts weretherefore estimated as drying progressed. From weighing ofsplit coconuts, it was determined that there was a small decrease in the percentage (byweight) of kernel in the split nut,as thesplitnutlost moisture. It wasfound thatin thefresh, wetstate, split nuts (kernel and shell together) had an averagemoisture content of 33% (dry basis), with a mean kernel

content of 65.1% [(weight of kernel/weight of kernel andshell) x 100]. On complete oven drying of split nuts (0%moisture content) the mean kernel content of the samples was59.5%. From these observations, the % kernel content byweight in the split nuts was estimated as:

Y=0MX + Md (2)

where:

Y = % kernel content by weight in a split nut(kernel + shell),

X = % moisture content (dry basis) of split nuts, andMd = %kernel content by weight in sample split nuts oven

dried to 0% moisture content.

RESULTS AND DISCUSSION

In assessing the performanceof the solar dryers, the followingparameters were evaluated:

(i) air temperature(ii) kernel drying rate(iii) product quality(iv) drying cost

Drying air temperature

The weatherthroughout the dryingperiod was fine? exceptforday 3 when there was an unusual, intense rainstorm whichlasted for a couple of hours, and on day 4 in which there wasintermittent sunshine. Theair temperatures in thethree dryers,together with the ambient temperatures, averaged at hourly

55 _

50

45

40

55

30 _

25

10

A INSOLATION

• DRYER A

* T DRYER B

/ \ o DRYER C

/Vv* '

12

HOUR OF DAY

14

AMBIENT

\t

800

600

400

200

18

Fig. 4. Average daily drying air temperatures in thesolar dryers and insolation levels.

SAKATandROLLE

intervals between 900h and 1600h and for the four, full daysof drying are shown in Fig. 4. Ambient temperatures variedbetween 28 and 30°C. Drying temperatures were consistentlyhighest in dryer B, averaging 22*C above ambient between1200h - 1300h, although a peak of 35°C above ambient wasrecorded on the 3rd day ofdrying. Temperatures in dryer Cwere usually 1-2°C higher than in dryer A, with the differencein temperature between dryer C and ambient peaking at 11°Cbetween 1200h - 1300h.

The temperature spatial variability within the dryers isshown in Fig. 5, where air temperatures between 900h and1600h at various points in the dryers and averaged over thefour full days ofdrying areshown. Dryer Bshowed the smallest variability inair temperature with position indicating thatthere was efficient aircirculation within the dryer. Heat leakage from this dryer was also reduced duetogood constructionand the useof a glasscoverand insulating material. Reducedair inlet andexitareas in dryerB compared to dryers A andC(Table I) limited the air flow rate in this dryer and consequently increased the air temperature. While being desirable,it was notpossible to measure theactual air velocities through

DRYER60

55

50

~ 45

40

35

30 Jk I L

12 16 12

DRYER C

— tap— middle'" bottom

-^-'-^

-L\*^

12 H

HOUR OF DAY

Fig. 5. Average daily temperature variation in the threesolar dryers.

the dryers, given their very low values. In dryer A, an indirectsolar dryer, the highest air temperature was achieved on thelowest tray (nearest to the air heater) and the temperatureprofiles (Fig. 5) indicate an expected behaviour in that as theair rose through the trays itlost enthalpy to evaporation and theair temperature fell.

Dryer C showed the widest variation in air temperature,with the path for airmovement being the longest here i.e. 2.40m. Air temperatures in this dryer suffered from the lack ofinsulation, particularly theinclined absorber plate which wasdirectly exposed to the ambient air. In both dryers B and.C,highest temperatures were obtained in the middle rather thanat the top, probably due to the infiltration of the prevailingwinds (approaching from a north easterly direction) throughthe air exits at the top of these dryers.

Figure4 alsoshowsinsolation valuesaveraged overthefourfull daysof drying. Duringthis time, insolation averaged 21.6MJ» m" • day" with a range of 15.6 MJ« m"2- day"1 on day 4to apeak of26.6 MJ« m~2» day"1 on day 2.Kernel drying rate

Kernel weights were estimated by Eq. 2, and moisture contentsthencalculated. Thekernel moisturecontent (drybasis)for thethreedryersat various locationsand for the 5 dryingdaysaregiven in Tables II, III and IV. The moisture contents reportedare mean values obtained from the individual samples in aparticular tray or location i.e. six samples per tray in dryers Aand B, four, six and four samples from the top, middle andbottom sections respectively for dryer C. These tables alsoshow the overall, average kernel moisture content with time(mean of all samples) for particular dryers. The kernel moisture content when the kernel is facing up or down, for aparticular dryer is also shown. From these tables, Fig. 6 wasconstructed and it shows the changes in overall kernel moisture content ratio [variable moisture content (M)/initialmoisture content (Mo)] with drying time. Figure 6 shows thatin the initial period of drying the rates of copra moisturecontent reduction for the three dryers are very similar. Thisoccurs as there is considerable free surface moisture in thekernels i.e. remaining after water normally present in thecoconut kernel has been drained on splitting the shell. However, as the drying progresses, and moisture movement from

Table II. Estimated kernel moisture content (dry basis)versus drying time for Dryer A

Moisture content (%) Avera^;e moisture content (%)Day Time(hr)

Tray 1 Tray 2 Tray 3 Tray 4 Kernels up Kernels down Overall

1 16.00 51.5 59.4 47.7 52.1 51.8 53.2 52.52 9.00 41.7 47.6 39.6 42.0 41.4 43.0 42.22 16.00 34.9 34.3 32.3 31.5 32.5 32.9 32.73 9.00 31.4 34.8 29.2 26.8 29.3 30.8 30.03 16.00 25.7 28.2 24.1 23.0 24.8 24.9 24.84 9.00 23.4 25.0 22.0 20.9 22.6 22.4 22.54 16.00 19.0 21.0 18.8 17.9 19.2 19.1 19.15 9.00 19.6 20.1 17.9 17.1 18.4 18.4 18.45 16.00 15.6 16.1 13.8 13.8 14.9 14.4 14.6

CANADIAN AGRICULTURAL ENGINEERING 89

within the kernel is made increasingly difficult, dryer B showsthe highest moisture content reduction with little differencebeing exhibited between dryers AandC.This behaviour isdueto the increased air temperature in dryerB compared to dryersAandC, as previously noted. Theworkof Rajasekharan et al.(1961) clearly demonstratedthe strong influenceof dryingairtemperature on the rate of drying of copra.

Using the data of Tables II to IV, together with the initialcrop weights, moisture removal rates from the kernel averagedover the four full days ofdrying are estimated as 1.92,1.00 and0.99 kg H20/day for dryers C, A and B respectively. Whenthese rates are calculated on an effective collector area basishowever (Table I), they are 0.66,0.64 and 0.58 kg H2O m"2«day"1 for dryers C, Aand Brespectively. These results showthat while the absolute moisture removal rates are not verydifferent for the three dryers, dryer C is marginally mosteffective.

1.0

5 0.8 _

5 0.6•—

oC_>

LU

2 0.4

I

0.2 -

JL -L

DRYER A

V V DRYER B

O O DRYER C

J_ J.0 12 3 4 5

TIME OF DRYING (OAY)

Fig. 6. Drying curves of coconut kernels in the threesolar dryers.

The horizontal segments in Fig. 6 indicate no drying atnights, inthe latter halfofthe drying cycle. An examination ofthe data in Tables II and III show that in certain situationsre-absorption of moisture by thekernel occurs in thenight.

An analysis of the drying data for dryer A showed nodifference in the kernel moisture content reduction rate whenin the up or down positions. However, this factor clearlyaffected the kernel drying rate in dryer B, as shown in Fig. 7and was also observed for dryer C, though to a lesser extent.These results show that for direct solar dryers such as dryers Band C, having the kernels directly exposed to the sun willincrease the drying rate.

Product quality

Dryer C produced good, white copra. Some of the productfrom dryer B exhibited slight browning, indicating the onset ofcharring due to high temperatures and prolonged drying. It isnoted that temperatures were highest in dryer B, with peak

1.0

o2:

3 0.8o

«r

cc

2 0.6i—

zoo

'Jj

or

^0.48

0.2

JL

O KERNEL FACING UP

A KERNEL FACING DOWN

-L JL

TIME OF DRYING (DAY)

Fig. 7. Drying curves For coconut kernels orientedeither up or down in Dryer B.

Table HI. Estimated kernel moisture content (dry basis) versusdrying time for Dryer B

Moisture content (%) Average moisture content (%)Day Time(hr)

Tray 1 Tray 2 Tray 3 Tray 4 Kernels up Kernels down Overall

1 16.00 51.1 51.3 48.4 52.2 51.6 50.9 5.2

2 9.00 42.4 42.8 40.8 43.9 41.2 43.7 42.4

2 16.00 29.4 29.5 27.5 31.8 27.0 31.6 29.2

3 9.00 27.1 27.2 26.0 29.8 24.9 29.6 27.2

3 16.00 16.2 20.1 19.1 21.9 18.3 21.0 19.6

4 9.00 18.8 19.1 18.8 21.5 17.9 20.9 19.4

4 16.00 14.0 14.3 14.1 16.4 13.8 15.5 14.7

5 9.00 13.2 14.2 14.5 16.4 14.1 15.3 14.7

5 16.00 9.2 9.3 8.9 11.0 9.4 9.9 9.6

90 SAKATandROLLE

TableIV.Estimatedkernelmoisturecontent(drybasis)versusdryingtimeforDryerC

DayTime(hr)Moisturecontent(%)Averagemoisturecontent(%)

1

2

2

3

3

4

4

5

5

16.00

9.00

16.00

9.00

16.00

9.00

16.00

9.00

16.00

Top

57.9

49.2

37.5

33.6

27.6

22.1

21.8

21.3

16.2

Middle

50.0

40.4

27.7

24.4

19.8

18.4

15.2

14.7

11.2

valuesof65°Crecordedonthethirddayofdrying.IndryerA,only33%oftheproductwasofgoodquality,astheremainderexhibitedmoderatetoexcessivefungalgrowth,thisbeingmostapparentonthelowesttrays,despitetheairtemperatureanddryingratesinthesetraysbeingrelativelyhigher(Fig.5andTableII).Theresultsindicatethattheremayhavebeenzeroair-flowatnightthroughdryerAmaintainingthelowertraysinastagnant,humidenvironmentthatisfavourableforfungalgrowth,whiletheuppertrayshadsomeairexchangewiththewind.Moisturecondensationonthekernel'ssurfaceastemperaturesfallatnightwouldalsohavecontributedtothisunfavourablebehaviour.

Dryingcost

DryersA,BandCwillcurrentlycostUS$275,US$128andUS$80respectively.ThesecostsreflecttheconsiderablereductioninmaterialandlabourrequiredforconstructingdryerC.Neglectingthedifferencesinthedryingtimerequiredforabatchofsplitcoconuts,itisthereforeexpectedthattheannualthroughputofdryerCwillbeapproximatelytwiceasmuchasthatofdryersAorB.Onthebasisofinitialcost,anexpecteddryerlifeof3yearsandanannualthroughputwhichassumesadryingseasonof30weeks/yearwithdryingoccurringinweeklybatches,itisestimatedthatthedryingcostforcoprawillbe49,230and250$USpertonnefordryersC,BandArespectively.FordryerC,thecostincludestheexpectedreplacementoftheplasticcoverat15weekintervalsduringthedryingseason.

CONCLUSION

Fromtheresultspresentedabove,itisevidentthatanaturalconvectiondryerofthedirect,cabinettypeasexemplifiedbydryerCinthisstudymaybeusedbysmallfarmersintheCaribbeanislandsforthedryingofcopra.Dryingwillbe

CANADIANAGRICULTUitALENGINEERING

Bottom

52.1

42.1

30.9

27.9

25.1

23.2

20.0

19.7

15.8

KernelsupKernelsdownOverall

52.8

40.5

30.3

27.0

24.6

21.1

18.8

18.2

14.4

53.9

47.3

33.7

30.3

23.8

21.4

19.2

18.9

14.3

53.3

43.9

32.0

28.6

24.1

21.2

19.0

18.5

14.4

completedinfourdaysofsunshine,withclean,whitecopraproduced.Thedryerissimpletoconstruct,possiblybythefarmerhimself,usingmaterialswhicharereadilyavailable.Thecostofdryingisalsolowandisestimatedat$US49pertonneofcopra.Itisrecommendedthatinoperatingsuchadryer,thekernelsbepositionedtofacedirectlyintothesunshine.Indirectdryers,operatingbynaturalconvectionandconsistingofcojipledsolarairheatersanddryingchambersarenotrecommendedforcopradryingduetothepotentialforcropspoilageresultingfromstagnantconditionsandmoisturecondensation,particularlyatnights.Forsuchdryerstobeeffective,airflowshouldbemaintainedatnight,possiblythroughthermalstorageusingforexamplearockbedorsealedwatertank.

ACKNOWLEDGEMENTS

TheauthorsthankMr.O.LawrenceandMr.D.Padarath,techniciansoftheMechanicalEngineeringLaboratoriesfortheirgenerousassistanceinthisstudy.WealsothankProf.D.McGawandDr.O.Headleyfortheirsuggestions.

REFERENCES

KALRA,S.K.andK.C.BHARDWAJ.1981.UseofsimplesolardehydratorfordryingfruitandvegetableproductsJFoodSci.Technol.18:23-26.

RAJASEKHARAN,N.,D.S.BHATIAandK.M.PANDALAI.1961.Somepreliminarystudiesonmechanicaldryingofcoconuts.IndianCoconutJ.14:71-80.SANDHU,B.S.,K.D.MANNAN,G.S.DHILLON,andL.S.CHEEMA.1979.Design,developmentandperformanceofmulti-racknaturalconvectiondryers.In:SunII.Proc.ISESSilverJubileeCongress[K.N.BoerandB.H.Glenn(Eds.)],PergamonPress,NewYork,NY.

91