Mathematical Description of the Drying Rate of Fully Exposed Corn

J. M. Troeger and W V. Hukill

MEMBER ASAE FELLOW ASAE

ISTORICALLY, grain has been preserved by reducing its moisture content to a safe storage level. Low moisture content results in decrease of respiration by the grain and inhibition of activity by microflora in the grain, both of which are sources of deterioration in grain quality. For drying, grain is usually placed in a bulk container and air is passed through it. As moisture leaves the grain in response to a vapor pressure gradient between the air and the kernel, it is picked up by the air and moved out of the grain mass. Within this grain mass, there is a zone of drying in which each layer within the mass reacts as a single layer fully exposed to its surroundings. A straightforward analysis of a layer within a grain mass is complicated by the fact that the conditions surrounding each layer are continually changing in response to moisture loss by the grain of previous layers. Before the response of a layer to continually varying conditions can be described, it must be possible to describe the drying process in response to constant surrounding conditions. Accordingly, an experiment (Troeger, 1967) was designed in which single layers of corn were fully exposed to constant surrounding conditions, and the response of the kernels was measured. From the data, a mathematical model was developed to describe this response. REVIEW OF PREVIOUS WORK

1956) which is based on the hypothesis that the rate of transfer of a diffusing substance through a unit area of a section is proportional to the concentration gradient measured normal to the section. The theoretical drying curve has been calculated for various regularly shaped objects (references 3, 10, 11, 13) and the results compared with actual drying data for wheat (Becker, 1959) and corn (Chittenden and Hustrulid, 1966, and Hamdy and Barre, 1969). These comparisons show good agreement during the early portion of the drying curve but are poorer predicters as the drying curve approaches equilibrium. So although this hypothesis is valid for the general diffusion situation, its application to diffusion from a grain kernel must account for the non-homogeneous composition and irregular shape of the kernel. Several researchers (Henderson and Pabis, 1961 and Hustrulid and Flikke, 1959) have applied the relationship: dm/dt = -k(m - m ) where m = average moisture content (dry basis) k = diffusion coefficient t = time m = e e quilbrium moisture content This model is analogous to the diffusion law and can be readily integrated to

predict the moisture as a function of time: m - m e = ce" kt where c = constant of integration e = exponential function A plot of this model on semilogarithmic paper gives a straight line but a plot of drying data on the same coordinates shows that the model fails to fit the data. A plot of the differential form of the equation (Fig. 2) on logarithmic coordinates shows that the rate of change of moisture in the grain kernel is more nearly proportional to the moisture gradient raised to a power greater than unity: dm/dt = -k(m - m ) a where a = arbitrary exponent Babbit (1940) points out that, although the moisture concentration gradient is convenient to use, it is actually the vapor pressure gradient which controls the diffusion rate. Using this assumption, Hukill and Schmidt (1960) developed a model (using two aspects) which satisfactorily described the single layer drying curve of grain sorghum. They developed an equation using the square of the vapor pressure gradient as the driving force by defining kernel

The search for an accurate method of describing the drying rate of grain under constant environmental conditions has been reported often in the literature. Many researchers begin with a theoretical approach using Fick's law (Crank,
Paper No. 70-324 was presented at the Annual Meeting of the American Society of Agricultural Engineers at Minneapolis, Minn., July 1970, on a program arranged by the Electric Power and Processing Division. Approved as Journal Paper No. J-6570 of the Iowa Agriculture and Home Economics Experiment Station, Ames, Iowa. Project No. 1808. The authors are: J. M. TROEGER, Agricultural Engineer, Coastal Plain Experiment Station, AERD, ARS, USDA, Tifton, Ga; and W. V. HUKILL, Collaborator, AERD, ARS, USDA, Iowa State University, Ames, Iowa. 1971 TRANSACTIONS OF THE ASAE

PUMP

Oft
FIG. 1 Schematic diagram of drying system.
1153

SAMPLE

NO, 2 3 0 4 4 1 0

Temperature - I 2 6 F R e l a t i v e H u m i d i t y - 18.6 % I n i t i a l M o i s t u r e - 41.8 %(d.b.)

dm/dt=-k(m-me)

FIG. 2 Plot of drying equation: dm/dt = -k(m-me)a humidity. Kernel humidity is the relative humidity that is in equilibrium with the average moisture content of the kernel. Relative humidity is directly related to vapor pressure so the kernel humidity gradient is effectively a vapor pressure gradient. This model gives a good fit of the data throughout the drying curve. EQUIPMENT A laboratory dryer was designed and constructed for obtaining data on the drying rates of single layers of corn. The dryer system consisted of an exposure chamber for holding the samples, a water bath and packed tower in which air was saturated at the desired dewpoint temperature, and a heater section for heating the saturated air to the desired dry bulb temperature. Fig. 1 shows a schematic diagram of the system. Since saturating the air at the dewpoint temperature limited the tests to dewpoints above 34 F., the system was later modified to pass the air through beds of silica gel dessicant. This extended the lower dewpoint limit to -50 F. The exposure chamber was divided into four sections, each with separate velocity controls. Each section held four test samples, giving a total of 16 samples for each temperature-relative humidity setting. Samples weighing approximately 50 g each were placed one layer deep in the sample trays. The samples were weighed periodically on an analytical balance weighing to 0.0001 g. The air was first conditioned by passing it upward through a packed tower while water moved downward
1154

through ceramic saddle packing material. Water for the packed tower was supplied from a water bath whose temperature was maintained at the desired dewpoint temperature by a relay controlled by a mercury contact sensor submerged in the bath. Additional heat for high temperature tests was provided by fixed heaters. Cooling coils, through which a refrigerant circulated, were used for dewpoints below ambient temperature. Water in the bath was continuously agitated to keep the temperature uniform throughout the bath. The saturated air, after leaving the packed tower, passed through a bank of electric resistance heaters that raised the temperature to the desired dry bulb temperature. A wire-wound resistance type sensor placed at the entrance to the sample chamber and shielded from the heater elements, maintained the desired dry bulb temperature. The air then passed through the samples and was exhausted to the atmosphere. Temperatures throughout the system were monitored by thermocouples read with a null-type precision potentiometer. Relative humidities were computed as the ratio of the vapor pressures associated with the dewpoint temperature and the dry bulb temperature. Air velocity in each cell was measured with a thermocouple anemometer. Moisture contents were determined at the conclusion of each test by placing the samples in an oven at 217 F. for 72 hr.

DEVELOPMENT OF THE MATHEMATICAL MODEL Drying is a complex phenomenon that might be considered as a coupled flow process involving several basic mechanisms, each acting simultaneously with several others. Each of these basic mechanisms is influenced to some extent by each of the other mechanisms. Vapor diffuses out of the kernel in response to a vapor pressure gradient while heat flows into the kernel in response to a temperature gradient. At the same time, the drying process may be influenced by certain chemical and biological processes that up to now have not been described well enough to evaluate their effects. The relationship expressed by the equation: dm/dt = - k ( m - m e ) a [1]

where m = average moisture content (dry basis) t = time (min.) a = arbitrary exponent k = diffusion coefficient m e = equilibrium moisture content indicates that the rate of moisture removal is proportional to the " a " power of the difference between the average moisture content of the kernel and its equilibrium moisture content. Equilibrium moisture content is a function primarily of the temperature and relative humidity of the surrounding air. A plot of the logarithm of drying rate (-dm/dt) versus logarithm of the moisture potential difference (m-m e ) (Fig. 2) shows that the data roughly approximates a straight line. This plot also indicates that assignment of the value a = 2.2 gives a reasonable fit. The plot when a = 1 is also indicated since this value has been used by previous researchers in attempting to fit a curve over the early portion of the drying curve. Using this plot for samples dried under a wide range of conditions, it was observed that a three-aspect model would do an excellent job of describing the entire drying curve. To further facilitate describing the curve in mathematical form, the data were reduced to dimensionless form by using the coordinates of moisture ratio (m r ): (m-m e )/(m 0 -m e ) where m 0 = initial average moisture (dry basis) and rate ratio (r r ):

EXPERIMENTS Tests were conducted that included nominal levels of these variables: Dry bulb temperature: 90, 125, 160 F. Relative humidity: 0, 5, 10, 20, 40, 60, 80 percent Air velocity: 20, 40, 80, 160 ft. per min. Initial moisture: 21, 31, 36, 42 percent dry basis Not all combinations of all variables were used. A test normally consisted of a given temperature and relative humidity with the 16 samples comprising all combinations of initial moisture and air velocity levels listed. A total of 350 corn samples were dried. Tests were usually run for two weeks although some were curtailed earlier because of equipment problems and a few were continued for four weeks. When the equipment functioned satisfactorily, weight losses were observed to the end of each test.

(dm/dt)/(dm/dt)
1971

= 0.2

TRANSACTIONS OF THE ASAE

INITIAL MOISTURE = 22.4

continuous through the transition region, the moistures at transition are given by:
i/(q,
m

xl

(P1VP2I2)

V+m
[4c]
3

1/(Q
m

-Q

x2

(P2%'P3%

+m [4d]

EVALUATION OF THE MODEL A knowledge of the equilibrium moisture associated with a given set of temperature and relative humidity is essential for the use of this model. Equilibrium moisture for each sample was estimated by extrapolating the square root of the drying rate to obtain the moisture content at zero drying rate. To evaluate the coefficients and exponents in terms of the experimental data, it was necessary to arbitrarily set the transition points to bound the region over which each aspect would be fitted. After studying data covering a

0.01

0.10 t\ = ( M - Me ) / ( M0- Me)

Mr = (M-M e ) / (Mo-Me)

FIG. 3 Initial moisture effect (dimensionless plot).

FIG. 4 Relative humidity effect (dimensionless plot).

The base rate at m r = 0.2 was chosen because this point nearly always appeared in the relatively stable second aspect. Representative data showing effects of initial moisture and of relative humidity on the drying data are given in Figs. 3 and 4. The first aspect is influenced to some extent by the initial and predrying (including storage) conditions. These conditions include the initial average moisture as well as the distribution of that moisture within the kernel. Stored samples would tend to have a more uniform moisture throughout the kernel while corn taken directly from the field would have a moisture gradient already established. The second aspect appears to be relatively stable, its slope influenced only by the relative humidity. Characteristics of the third aspect are difficult to describe in terms of the test variables. Most of the variations appear to be a result of insufficient or erratic data during this portion of the drying curve. The model can be stated in the following mathematical form: dm/dt = -a dm/dt = -a (m-m e ) (m-m ) >m>

level, these initial and boundary conditions are imposed: (a) at t = 0, m = mQ (b) at the transition points, time is continuous (c) at the transition points, drying rate (dm/dt) is continuous The integrated equations are: t = Px ( m - m / t = p2(m-m2)2 t = p3 (m-mj
whe
3 1

(mo-me)Ql

m^m^m.,
o

xi
m

[3al
J

-p2(mxl-me) - p3 (mx2-me)
b)
3

'xl + tx2

xl

> m > m

x2

t3bl [3c]

mx2>m>me

-l/a(l 1-b

= 1
The times at the transition points are:
^ 1

wide range of test conditions, these values were used: m x A = 0.40 (m o -m e') + m e . . . v l
L

[5al J

xl -Pi

P 1 ( mx l

(m o -me v

[4a]

m x2 = 0.12 (m o -m e y) + m e . . . L J [5bl v 0 The exponents were first estimated by fitting the logarithmic forms of equations [ 2 a ] , [2b] and [2c] : log (-dm/dt) = log a + b log (m-m )

[2a] [2b] [2c]

[6]
to the experimental data by the method of least squares, minimizing the squared deviations of log (-dm/dt). These results were examined to determine if the exponents could be predicted as functions of the test variables. After considering a number of combinations of the test variables, these empirical equations were selected: q- = -3.98 + 2.87 m ~1 o

m x l ^ m ^ m x2 _ m x2 > m > m e 0

dm/dt = -a_ (m-m

where a,b = arbitrary parameters mQ = initial moisture (dry basis) m x l , m x 2 = transition moistures m = equilibrium moisture t = time (min.) To obtain equations for predicting the time associated with a given moisture
1971 TRANSACTIONS OF THE ASAE

x2

p 0 (m
r

-m )
e'

2 \ x2
xl

P2(mxl-me)

+ t

[4b]

Using the third boundary condition listed above, i.e., that the drying rate is

[0.019/(h e + 0.015)] + 0.016T

[7a]
1155

-s <
CD

SAMPLE NO. I 314410 Temperature - 9 2 F Relative Humidity - I 9,2% Initial Moisture - 4 3 . 0 % (d.b.) Air Velocity - 160 ft /min

>Q a;ENT ce TURE

SAMPLE NO. 1502310 Temperature - 9 0 " F Relative Humidity - 5 8 . 7 % ' Initial M o i s t u r e - 3 6.0 %(d.b.) Air Velocity - 4 0 ft./min.

SAMPLE NO. 2301310 Temperature - I26F Relative Humidity - 18.9 % Initial Moisture - 3 6 . 0 % (d.b.) Air Velocity - 20 f t . / m i n .

" slA
\
Experimental Predicted O CN/T 2

TIME, HOURS

4
(Sample No.

^--

^___
TIME , HOURS

T I M E , HOURS

F I G . 5 Prediction 1314410).

plot

F I G . 6 Prediction 1502310).

plot

(Sample

No.

F I G . 7 Prediction 2301310).

plot

(Sample

No.

q 2 = - e x p ( 0 . 8 1 0 - 3.11 h j . . q3 = -1.0 where h e = relative h u m i d i t y (decimal form) T

[7b] [7c]

T h e validity of t h e above empirical e q u a t i o n s is limited, however, t o t h e range of t h e variables used in these tests. DISCUSSION O F RESSULTS T h e m o d e l assumes t h a t t h e drying curve can be r e p r e s e n t e d as a function of t h e moisture gradient ( m - m e ) raised t o s o m e power. C o m p a r i s o n with previous models shows t h a t there is a distinct advantage w h e n t h e e x p o n e n t of ( m - m e ) is allowed t o be a value o t h e r t h a n u n i t y . Also, use of three aspects rather t h a n a single aspect allows a good fit of t h e entire drying curve. This m o d e l of t h e drying curve is relatively easy t o apply, even using h a n d m e t h o d s . Consider this e x a m p l e : What is t h e time required t o dry a single layer of fully e x p o s e d shelled

( 0 . 0 3 0 ) ( 9 0 ) - (0.002)(160)] p

= 28.8

= exp [ 2 . 8 2 + ( 7 . 4 9 ) (0.21 vO.67

( 0 . 0 1 7 9 ) (90)] = 4 6 . 5 of the air qx = - 3 . 9 8 + ( 2 . 8 7 ) ( 0 . 4 2 8 ) ( 0 . 0 1 9 ) / ( 0 . 2 1 5 ) + ( 0 . 0 1 6 ) (90) = q 2 = - e x p [ 0 . 8 1 0 - (3.11) (0.20)] = -1.207 The first transition t i m e , given b y equation [ 4 a ] , is: -1.40 t x l = (28.8) ( 0 . 1 4 4 ) " (28.8) (0.360)'- 1 . 4 0 3 1 1 min 1.40

= dry bulb t e m p e r a t u r e of the air

(deg.F)
exp = e x p o n e n t i a l function Using t h e e x p o n e n t s c o m p u t e d b y e q u a t i o n s [7a] and [ 7 b ] , e q u a t i o n s [ 3 a ] , [ 3 b ] and [3c] were fitted t o t h e e x p e r i m e n t a l data b y t h e m e t h o d of least squares, minimizing t h e squared time deviations. T h e coefficients obtained by this p r o c e d u r e were e x a m i n e d t o d e t e r m i n e their d e p e n d e n c e on the test variables. Empirical e q u a t i o n s were developed t o predict coefficients p1 a n d P2:
Pl

The desired time is predicted b y equation [ 3 b ] : t=(46.5)(0.095)"1-207 (46.5)(0.212)""1-2O7 + 311 =

= e x p (-2.45 + 6.42 m * ' 2 5 - 3.15 h e + 9.62 m Q yfiT

+ 0 . 0 3 0 T - 0.002 V) . [8a] [8b]

8 0 9 min or 13 1 / 2 hr The i n p u t c o n d i t i o n s for this example are similar t o t h o s e for the sample p l o t t e d in Fig. 5. T h e p r o c e d u r e o u t l i n e d in t h e example, while possible b y h a n d m e t h o d s , is even m o r e readily a d a p t a b l e t o c o m p u ter m e t h o d s . SUMMARY E x p e r i m e n t a l data describing t h e single layer drying curve for shelled corn were o b t a i n e d in t h e l a b o r a t o r y u n d e r c o n s t a n t controlled c o n d i t i o n s . T h e variables considered included t h e dry b u l b t e m p e r a t u r e of t h e air (90 t o 1 6 0 F . ) , (Continued on page 1162)
-o.

p 2 = e x p [ 2 . 8 2 + 7.49 (h e + 0 . 0 1 ) 0 ' 6 7 - 0 . 0 1 7 9 T ] where V = air velocity, ft per min It should be n o t e d t h a t b y defining b o t h p x a n d p 2 empirically, t h e assumption of a c o n t i n u o u s drying rate t h r o u g h the first transition p o i n t ( b o u n d a r y condition n u m b e r 3 above), c a n n o t be m a d e . A b e t t e r fit of t h e d a t a is obtained, however, w h e n b o t h are defined empirically. F u r t h e r refinement of e q u a t i o n s [8a] a n d [ 8 b ] might allow t h e a s s u m p t i o n of a c o n t i n u o u s drying rate t h r o u g h t h e first transition p o i n t . T h e t h i r d coefficient ( p 3 ) was defined in t e r m s of t h e o t h e r t w o : ( q 9 - qQ) p_ = 0.12 ( m - m j (P2q2/q/

corn from 30 t o 14 percent wet basis (42.8 t o 1 6 . 3 p e r c e n t dry basis) u n d e r the conditions: Dry b u l b t e m p e r a t u r e of t h e air = 90 F . Relative h u m i d i t y = 20 p e r c e n t Air velocity = 1 6 0 ft per min. min. F r o m an equilibrium curve (6) for shelled corn, t h e equilibrium moisture for these conditions is f o u n d t o be 6.80 p e r c e n t (dry basis). T h e moisture differences are c o m p u t e d as: m - m = 0 . 4 2 8 - 0.068 = 0 . 3 6 0 o e m x l - m e = 0.40 (0.360) = 0 . 1 4 4 (see e q u a t i o n [ 5 a ] ) m-m = 0 . 1 6 3 - 0 . 0 6 8 = 0.095

< m
QO.

. .

[8c]

SAMPLE NO. 3121410 Temp erature - 16 1 F Relative Humidity - 4 . 7 % Initial Moisture- 4 4. 5 %<di>.) Air Velocity - 2 0 f t , / m i n .

2
UJ

This did allow t h e a s s u m p t i o n of a c o n t i n u o u s drying rate t h r o u g h t h e seco n d transition p o i n t . Plots of t h e predicted drying curve along with actual data are presented in Fig. 5 t h r o u g h 8. These plots represent typical samples covering a range of t h e test variables.
1156

Experimenfal Predicted

T h e coefficients a n d e x p o n e n t s are c o m p u t e d from e q u a t i o n s [ 7 a ] , [ 7 b ] , [8a] and [ 8 b ] : p =exp [-2.45+(5.42) (0.428)1*25-

MOISTU 10

^V

10

20 30 T I M E , HOURS

40

50

60

(3.15) (0.20) + (9.62) (0.428) ( V 0 . 2 0 ) +

F I G . 8 Prediction 3121410). 1971

plot

(Sample

No.

TRANSACTIONS OF THE ASAE

2 An estimate of the peak discharge rate and the runoff volume of each storm. 3 A topographic map for determining the hypsometric integral. 4 A soils map for measuring the portion of the watershed made up of blackland soils. 5 An estimate of the portion of the watershed that will be cultivated. 6 An estimate of the portion of the watershed that will be cultivated but not terraced. Several conclusions can be made as a result of this investigation. It is feasible to develop a procedure for predicting sediment yield from small watersheds based on climatic factors, watershed characteristics, and land use and treatment factors. A combination of these variables explained 63 percent of the variation in sediment concentration from small blackland watersheds. Storm characteristics accounted for the largest variation in sediment concentration. The next largest variation was explained by a land treatment factor. Smaller variations

were accounted for by land use and watershed characteristics.

References 1 Ackermann, W. C. and R. L. Corinth. 1962. An empirical equation for reservoir sedimentation. Internat. Assoc. Sci. Hydrol., Commission of Land Erosion, Publ. No. 59:359-366. 2 Anderson, H. W. 1954. Suspended sediment discharge as related to streamflow, topography, soil and land use. Trans. Am. Geophys. Union 35(2): 268-281. 3ARS Special Report. 1961. A universal equation for predicting rainfall-erosion losses, USDA-ARS 22-66, 11 p. 4 Beer, C. E., C. W. Farnham and H. G. Heinemann. 1966. Evaluating sedimentation prediction techniques in Western Iowa. TRANSACTIONS OF THE ASAE 9(6):828-831. 5 Betson, R. P. 1963. Building a nonlinear sediment yield model. Paper No. 10, Proceedings of the Federal Inter-Agency Sedimentation Conference, Symposium 1, Land Erosion and Control, p. 65-72. 6Cooley, W. W. and P. R. Lohnes. 1962. Multivariate procedures for behavioral sciences. John Wiley and Sons, Inc. 7 Glymph, L. M. 1954. Studies of sediment yields from watersheds. Paper presented at the Tenth General Assembly of the International Union of Geology and Geophysics, Rome, Italy, September 15-19. 8Gottschalk, L. C. 1964. Handbook of applied hydrology. Section 17-1 Sedimentation, Part I, Reservoir Sedimentation, Ven Te Chow, Editor-in-Chief, McGraw-Hill Company, p. 17-3. 9 Gottschalk, L. C. and G. M. Brune. 1950. Sediment design criteria for the Missouri Basin Loess Hills. USDA, SCS-Tp-97, Milwaukee, Wise. 21 p

10Kendall, M. G. 1957. A course in multivariate analysis. Hafner Publishing Co., New York, 185 p. UManer, S. B. 1962. Factors influencing sediment delivery ratios in the Blackland Prairies Land Resource Area. USDA Soil Conservation Service, Fort Worth, Texas, 10 P. 12Musgrave, G. W. 1947. Quantitative evaluation of factors in water erosiona first approximation. Jour. Soil and Water Cons. 2(3):133-138. 13Roehl, J. W. 1962. Sediment source areas, delivery ratios and influencing morphological factors. Interna. Assoc. Sci. Hydrol., Commission of Land Erosion, Publ. No. 59:202-213. 14 Smith, R. M., R. C. Henderson and O. J. Tippit. 1954. Summary of soil and water conservation research from the Blackland E x p e r i m e n t S t a t i o n , Temple, Texas, 1942-1953. Texas Agr. Exp. Sta. Bui. No. 781, p. 42-52. 15 Soil Conservation Service National Engineering Handbook. 1964. Section 4 Hydrology, Part I, Watershed Planning, Chapter 10, p. 10.5. 16 Strahler, A. N. 1964. Handbook of applied hydrology. Section 4-II Geology, Part II, Quantative Geomorphology of Drainage Basins and Channel Networks, Ven Te Chow, Editor-in-Chief, Mc-Graw-Hill Company, p. 4-68. 17 Strahler, A. N. 1957. Quantitative analysis of watershed geomorphology. Trans. Am. Geophys. Union 38(6):913-920. 18 Wallis, J. R. and H. W. Anderson. 1965. An application of multivariate analysis to sediment network design. Proceedings of the International Association of Scientific Hydrology. Symposium of Hydrometerologic Network Design, Publ. No. 67, p. 357-378, Quebec City, Canada, June 15-22. 19 Williams, J. R. 1969. The prediction of sediment yields from small blackland watersheds. MS thesis, Texas A&M University, College Station, Texas, 60 p.

Mathematical Description of drying Rate of Corn

(Continued from page 1156) the relative humidity of the air (0 to 80 percent), the velocity of the air 20 to 160 ft per min) and the initial moisture content of the grain (21 to 42 percent dry basis). The samples, most of which remained in the dryer for two weeks, showed a continual decrease in weight, but never attained a constant weight. A mathematical model was developed for describing the data. The model contained three aspects of the form: dm/dt = a(m-m e ) b . Coefficients and exponents for this model are described in terms of the test variables so that the model gives a satisfactory fit of all the experimental data. The use of this model should be beneficial in describing deep bed drying of corn. References 1 Babbitt, J. D. 1940. Permeability of hygroscopic materials to water vapor. Canadian Journal of Research Section A, 18:105-121. 2 Becker, H. A. 1959. A Study of diffusion in solids of arbitrary shape with application to the drying of the wheat kernel. J o u r n a l of Applied Polymer Science 1:212-226. 3 Becker, H. A. and H. R. Sallans, 1955. A study of internal moisture movement in the drying of the wheat kernel. Cereal Chemistry 32:212-226. 4 Chittenden, D. H. and A. Hustrulid. 1966. Determining drying constants of shelled corn. TRANSACTIONS OF THE ASAE 9(l):52-55. 5 Crank, J. 1956. Mathematics of Diffusion. Oxford University Press, London, England. 6 Hall, C. W. and J. H. Rodriquez-Arias. 1958. Equilibrium moisture content of shelled corn. AGRICULTURAL ENGINEERING 39(8):466-470. 7 Handy, M. Y. and H. J. Barre. 1969. Evaluating film coefficient in single-kernel drying. TRANSACTIONS OF THE ASAE 12(2): 205-208. 8 Henderson, S. M., and S. Pabis. 1961. Grain drying theory. I. Temperature effect on the drying coefficient. Journal of Agricultural Engineering Research 6:169-174. 9HukilL W. V. and J. L. Schmidt. 1960. Drying rate of fully exposed grain kernels. TRANSACTIONS OF THE ASAE 3(l):71-80. 10 Hustrulid, A. and A. M. Flikke. 1959. Theoretical drying curve for shelled corn. TRANSACTIONS OF THE ASAE 2:112-114. 11 Pabis, S., and S. M. Henderson. 1961 Grain drying theory. II. A critical analysis of the drying curve for shelled maize. Journal of Agricultural Engineering Research 6:272-277. 12Troeger, J. M. 1967. Development of a mathematical model for predicting the drying rate of single layers of shelled corn. Unpublished Ph.D. Dissertation, Iowa State University of Science and Technology, Ames, Iowa. Available from University Microfilms, Ann Arbor, Mich. (Order No. 67-13006). 13 Wang, J. H. and C. W. Hall. 1961. Moisture movement of hygroscopic materials. TRANSACTIONS OF THE ASAE 4(l):33-36.

1162

1971

TRANSACTIONS OF THE ASAE