DESIGN PARAMETERS OF FLUIDIZATION PR§NC§PLES
FOR FORAGE HARVESTlNG AND PROCESSING

Thesis for the Degree of Ph. D.
MICHIGAN STATE UNIVERSITY

LESTER FRANK WHETNEY
I 9 6 4

 

7 1“"I‘FI'I

 

This is to certify that the 1

thesis entitled '

 

Design Parameters of Fluidization Principles
For Forage Harvesting and Processing

presented by

Lester Frank Whitney

has been accepted towards fulfillment ‘
of the requirements for l

Major professor

Date-W !

 

 

 

 

 

______._——--"W~v--*-' '-

 

 

ABSTRACT

DESIGN PARAMETERS OF FLUIDIZATION PRINCIPLES
FOR PORAGE HARVESTING AND PROCESSING

By Lester Frank Whitney

The value of the annual hay crop in the United States
exceeds $2.3 billion, with conservative estimates of 28%
loss using conventional harvest and storage practices. To
reduce this loss, a different system approach is considered.

The concept involves stripping the leaves from growing
alfalfa plants and alloving the stems to grow new leaves for
future harvest. The leaves and minor stems would then be
rapidly dried utilizing fluidization drying. The design
parameters for continuous fluidized drying are the main con-
cern of this research.

Leaf stripping of alfalfa plants was attempted on a
limited experimental basis and was found to be feasible.

The desirability of leaf stripping as a cultural practice
has yet to be established. The possibility of accOmplishing
leaf stripping by mechanical means has yet to be developed.
Such a procedure is necessary because of the incompatibility
of fluidizing both stems and leaves in the same operation.
Furthermore, seventy percent of the nutritive value of the
plant is in the leaves and can be recovered by removal of
50% of the plant water. The drying is enhanced by the more

desirable heat and mass transfer prepertiea of the leaf.

Susceptibility of leaves to damage at the high temper—
atures associated with this dehydration was investigated
by single leaf studies and the mathematical description of
the time-temperature~damage point was determined. Leaves
were dried to equilibrium with air at 500'F in 15 seconds
without damage to the product. The drying rates of alfalfa
leaves at high temperatures up to 800°! were established
as a family of exponential functions.

Flake particle behavior relating the Sphericity of the
particle to the change in moisture content in a fluidized
bed was investigated. This parameter is of vital concern in
predicting the fluidizing velocities and voidage of the bed.
Single leaf sphericity values were determined using a mercury
displacement method developed for this study. The sphericity
data were statistically analyzed and this parameter was
found to be nearly constant for the total moisture content
range. A discernible difference was found between small
and large alfalfa leaves.

lass behavior of leaves in a small diameter drying tower
was investigated to relate the mass velocity of the drying
medium to a drying index of the leaves under study. This
parameter had been previously investigated using pilot
process, continuous flow equipment at inlet temperatures
above 720°. However, the macroscopic scale of the laboratory

apparatus developed for this study yielded similar relation-

ships at inlet temperatures of 200 to 300°F. The small
diameter drying tower required the use of leaves smaller
than alfalfa. Birdsfoot trefoil, another trifoliate plant,
was used in this phase of the study.

Fluidization drying principles were found to be well
adapted to forage leading to the conclusion that the system
concept was feasible, warranting further research activity.
Application of this system might appear as a farmstead pro~

cess perhaps utilizing unused noncombustible silos or a field

alfalfa leaf combine.

Approved WM W
Aka? /9(V’

DESIGN PARAMETERS OF PLUIDIZATION PRINCIPLES
FOR FORAGE HARVESTING AND PROCESSING

By
Lester Frank Whitney

A THESIS

Submitted to the School for Advanced Graduate
Studies of Iichigan State University
in partial fulfillment of the
requirements for the degree of

DOCTOR OF PHILOSOPHY

Department of Agricultural Engineering
1964

PLEASE NOTE:

Figure pages are not original

copy. They tend to curl. Filmed

in the best possible way.
University Microfilms, Inc.

Lrsr or rrcunss . . . . . . . . . . . . . . . .

LIST OF TABLES . . . . . . . . . . . . . . . . .
ACKNO'LEDGEIENTS . . . . . . . .
'1 TA 0 O O O O O O 0 O I 0 0 O O O O O O O 0 O 0

INTRODUCTION’ . . . . . . . . . . . . . . . . . .
REVIH' 0? LITERATURE . . . . . . . . . . . . . .

Pluidized Bed Drying of Alfalfao I . . . . .
Behavior of Pluidized Biological laterials .
Divergence of velocity Field . . . . . . . . .

THEORETICAL oossxnsasrtons . . . . . . . . . . .
srarsnssrlor rs: paosLsa . . . . . . . . . . . .
PnerlrnAar INVESTIGATIONS . . . . . . . . . . .

Forage Plant Selection . . . . . . . . . . . .
Plant Part Dimensions . . . . . . . . . . . .
Stokes Law Application to Alfalfa Plant Parts
Single Particle Behavior in Suspension . . .
Stripping Leaves As A Cultural Practice . . .

Ernsararxran TECHNIQUES . . . . . . . . . . .

Time-Temperature-Damage Point Relationships of
Alfalfa Leaves ; . . . . . . . . . . . . . .

Damage Point Tests . . . . . . . . . . . . .

Sphericity-loisture content Relationships .

Sphericity Determination Tests At Various loisture

Cbntehts . . .

lass Yelocity-Temperature-Ioisture Removal
Relationships . . . . . . . . . . . . . . .

Instrumentation of Drying Tower . . . . . . .

Drying Test Procedures . . . . . . . . . . . .

Page
RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . 86

Damage Point Relationships . . . . . . . . . . . . . 88
Drying at Ultra-High Temperatures . . . . . . . . . . 91
Sphericity-loisture Content Relationships . . . . . . 98
lass Rate of Plow, Temperature, and lbisture Removal
Relationships . . . . . . . . . . . . . . . . . . . 102

sumarmconcwsxoxs ..............'..109
FUTUREINVRSTIGATIONS.................112
“maxim

AMI“: O O O O O O O O O O O O C O 0 O O O O O O O O 119

A. Summary of Damage Peint and Drying Data for
Single Alfalfa Leaves . . . . . . . . . . . . . . 120

B. Summary of Sphericity-loisture Content Data for
Alfalfa and Birdsfoot Trefoil Leaves . . . . . . 130

C. Terminal velocity Determinations

a - Smoke Powder . . . . . . . . . . . . . . . . 136
b - Alfalfa Stems and Leaves . . . . . . . . . . 139

D. Summary of lass Rate of Flow and loisture
Buoyal D‘t‘ O O O O O 0 O I O O O O O O O O 0 e 141

woodman»

10
11
12
13
14
15
16

17

18
19
20

21

111

LIST OF FIGURES

Process Plow Diagram . . . . . .g. . . . . . .

Impression Of Alfalfa Upper Epidermis- 1200x

Stomata Fully Open; Growing Plant . . . . . .
Empression Of Alfalfa Upper Rpidermis- lZOOX
Stomata Closed After Cut rive Iinutes . . . .
Alfalfa And Birdsfoot Trefoil Leaves . . . .
Dried Birdsfoot Trefoil Leaves . . . . . . . .
Audio-Time Recording Apparatus . . . . . . . .
Temperature Indicating Apparatus . . . . . . .
Velocity Recording Apparatus . . . . . . . . .
Single Leaf Drying Apparatus . . . . . . . . .
Analytical Balance . . . . . . . . . . . . . .
Leaf Velume leasuring Apparatus-Disassembled .
Leaf volume leasuring Apparatus-Assembled . .
Velume leasuring Apparatus . . . . . . . . . .
Pluidized Bed Drying Tower . . . . . . . . .
Drying Tower And neat Source . . . . . . . . .
Drying Tower And Seat Source Iith Sample Port
Disassembled . . . . . . . . . . . . . . . . .
Assembled Drying Tower And neat Source Showing
Smoke Powder Injection . . . . . . . . . . . .
Photo Cell Circuit & Output . . . . . . . . .
Damage Point Relationships . . . . . . . . . .

Drying Curves For Alfalfa Leaves At Ultra-
High Temperatures . . . . . . . . . . . . .

Relation Of loisture Centent Ratio To Time
For Drying Alfalfa Leaves At Ultra-nigh
TemperatureS...............

8

54
54
55
55
56
56
57
58
59

61

61
82

92

22

23
24

25

26

27
28

iv

LIST OF FIGURES (Canto)

Relationship Of Drying Constants To Air

Temperature For Alfalfa Leaves .

Sphericity Versus % leisture Content (D.B.).

Relationship Between rlov Of Hot Air And
Initial And Final Leaf Ibisture Content

Regression Line "Y" Intercept vs Inlet
Tenperature (For lass Velocity-Drying Data).

Regression Line Slope vs Inlet Tenperature

(For lass velocity-Drying Data)
Velocity Variation-Natural Drafts
Nemograph For Evaluation of Particle Terminal

Velocity, Leva (1959)

100

104

106

106
107

138

LIST OF TABLES

 

Page

1 Physical Properties Of Fresh Cut Trifoliate

P1 ant Par t8 0 O O O o O I O O I O 0 O O I O 42
2 laterial For Drying Tower And Heat Source . . . 60
3 Drying Curves for Alfalfa Leaves At Ultra~Righ

Temperatures . o o . . . . . . . . . . . . 91
4 Moisture Content Ratio, Time, And Temperature

Relationships . . . . . . . . . . . . . . . . 94
5 Sphericity Relationships For Leaves At Variable

loisture Contents . . . . . . . . . . . . . . . 99
6 Error Analysis Of Sphericity Results For Fresh

Cut And Dried Leaves . . . . . . . . . . 101
7 lass Velocity, Temperature, And loisture Removal

Relationships . . . . . . . . . . . . . . . . . 103
A-l Summary Of Damage Point And Drying Data For

Single Alfalfa Leaves - Temperature 281°? . . . 120

A—2 Summary Of Damage Point And Drying Data For
Single Alfalfa Leaves - Temperature 296°! . . . 121

A—3 Summary Of Damage Point And Drying Data For
Single Alfalfa Leaves - Temperature 311'? . . . 122
A-4 Summary Of Damage Point And Drying Data For
Single Alfalfa Leaves - Temperature 379°! . . . 123

A-5 Summary Of Damage Point And Drying Data For

Single Alfalfa Leaves - Temperature 455’? . . . 124

A~6 Summary Of Damage Point And Drying Data For
Single Alfalfa Leaves - Temperature 515’? . . . 125

A-7 Summary Of Damage Point And Drying Data For

Single Alfalfa Leaves - Temperature 524‘? . . . 126
A-8 Summary Of Damage Point And Drying Data For
Single Alfalfa Leaves - Temperature 570°? . . . 127

A-9 Summary Of Damage Point And Drying Data For
Single Alfalfa Leaves - Temperature 640°! . . . 127

A~10 Summary Of Damage Point And Drying Data For

Single Alfalfa Leaves - Temperature 661‘F . . . 128

 

sxm'r

 

vi

LIST OF TABLES (Cont.)

Page

A-ll Summary Of Damage Point And Drying Data For
Single Alfalfa Leaves — Temperature 761°! . . . 128

A-IZ Summary Of Damage Point And Drying Data For
Single Alfalfa Leaves ~ Temperature 819‘? . . . 129

A-lS Summary Of Sphericity-loisture Centent Data,
Large Alfalfa Leaves . . . . . . . . . . . . . 130

A-ls Summary Of Sphericity-loisture Content Data,
ledium Alfalfa Leaves . . . . . . . . . . . . . 131

A~l5 Summary Of Spherdcity-lbisture Content Data,
Small Alfalfa Leaves . . . . . . . . . . . . . 132

A-lfi Summary Of Sphericity-Ioisture Content Data,
BirdBIOOt “.1011 O O O O O O C O 0 O O O O O 134

A-17 Summary Of lass Rate Of Plov And loisture
. Removal Data, Temperature - 200‘! . . . . . . . 143

A-18 Summary Of lass Rate Of Plow And loisture
Removal Data, Temperature - 250'! . . . . . . . 144

A-19 Summary Of lass Rate Of Plov And Ioisture
Removal Data, Temperature - 300°? . . . . . . . 145

 

 

vii

ACKNOWLEDGEIENTS

The author wishes to express his sincere appreciation
and gratitude to:

Dr. Carl W. Hall, Professor, Agricultural Engineering
Department, for his constant inspiration, encouragement,
interest and supervision under whom this investigation has
been conducted.

Dr. Robert I. Kleis, Read, Agricultural Engineering
Department, University of Massachusetts, for his interest,
encouragement and assistance which permitted the undertaking
of this program.

Dr. A. I. Parrall, Chairman, Agricultural Engineering
Department, for his administrative activities in arranging
for an assistantship in the early months of the program
and for funds in support of this project.

Dr. Frederick R. Buelow, Associate Professor, Agri-
cultural Rngineering Department, for his assistance in
guiding this program as well as his active interest in the
instrumentation of the apparatus.

Dr. Rolland T. Rinkle, Professor, lechanical Engineering
Department and Dr. George E. lase, Associate Professor,
letallurgy, lechanics and laterials Science Department,

for their interest and assistance in guiding this program.

viii

His wife, Phyllis, and seven children for their

encouragement, understanding, sacrifices, and assistance
which have made this program possible and necessary.

The Charles R. Rood Foundation and the lassachusetts

Society for Promotion of Agriculture for fellowship grants
The

which supported the early portion of the program.
National Science Foundation for a Science Faculty Fellowship

which supported the completion of the program.

ix

VITA

Lester Frank Whitney
Candidate for the degree of
Doctor of Philosophy

Final Examination: January 20, 1964, 1:00 P.Mo Room 218,
Agricultural Engineering Building.
Dissertation: Design Parameters of Fluidization Prin-

ciples for Forage Harvesting and
Processing.

Outline of Studies:
Major Subject: Agricultural Engineering

Minor Subjects: Applied Mechanics

Mechanical Engineering
Biographical Items:
Born: March 21, 1928, New Bedford, Massachusetts
Family Status: Married to Phyllis M. Burrill, 1950;
one girl and six boys.

High School: Rampden Academy, mepden, Maine

Undergraduate Studies: University of Maine, BSAE, 1949

Graduate Studies: Michigan State College, MSAR, 1951
University of Massachusetts, 1959—62

Michigan State University, 1962-64

Experience:

Graduate Assistant, Michigan State College,
1949-50
Consultant Agricultural Engineer, Jack R.
Kelly Onion Farms, Parma, Michigan
1950-51
Design and Development Engineer, Ariens
Company, Brillion, Wisconsin
1951-53
Consultant Agricultural Engineer, Maine
.Potato Growers, Presque Isle, Maine
1953-54
Plant Engineer, Assistant Chief Engineer,
Assistant Plant Superintendent, Virthmore
Feed Division, Corn Products Company,
Ialtham, Mass. 1954-59
Assistant Professor in Agricultural Engi—
neering, University of Massachusetts
1959-82
Graduate Assistant, Fellow in Agricultural
Engineering, Michigan State University
1962-83
Associate Professor in Agricultural Engi-

neering, University of Massachusetts

1964

xi

Professional Affiliations: Member, American Society
of Agricultural Engineers
Member, New England Retail
Farm and Power Equipment
Association

INTRODUCTION

The value of the annual hay crop in the United States
exceeds $2.3 billion and in Michigan is $60 million. Con-
servative estimates indicate losses of 28% caused by
present harvest, handling and storage practices. Because
of these losses and the uncertainties imposed by weather-
related factors, the utilization of present systems which
result in dried forage as a crop form is being seriously
questioned. It appears that radically different system
approaches to this problem are needed to minimize crop
losses, harvest costs and weather dependence, thus sub-
stantially increasing the desirability and efficiency of
dried forage systems.

Ideally, such a system should have the following
characteristics and requirements:

1. Virtual independence of the weather insofar as
drying is concerned.

2. Minimal number of machine components and process
operations.

3. Simplicity of operation.

4. Forage product should evolve as one which lends
itself to completely automatic bulk handling methods.

5. Forage quality should be high.

6. Adaptable to all sized units-~the family farm
as well as the commercial enterprise.

7. Initial investment should be economically sound.

8. Costs of operation should be partially offset by
a decrease in yield losses now incurred.

In brief, a concept is presented herein along with
results of initial investigations which support the feasi-
bility of such an idealized system. It involves the
stripping of leaves from the growing alfalfa plant, leaving
the standing stems to regenerate new leaves for future
harvest. The stripped leaves and minor stems at 75% m.c.,
w.b., would then be dried immediately to 20% m.c. utilizing
fluidisation drying principles at high temperatures. The
dried leaves could in turn be pelletized and handled in
bulk, much as the current practice in handling grains. A
process flow diagram illustrates the possibilities as
shown in Figure 1 for either a field system, such as an
alfalfa combine or a farmstead system, as might be con-
structed from a noncombustible silo.

The key to success of these systems is the continuous
drying process by the fluidization of alfalfa leaves.
Application of known physical laws indicates the incom-
Patibility of the simultaneous suspension of stems and
leaves. Experimentation has verified that the theoretical

terminal velocity required for a dried stem section of

m0<zOPm OF

      

>mo I o._.<m

 

e!

 

D<meEz<m
OF Hmoamza‘zb

 

 

 

 

 

 

 

 

 

 

on. M
I, E

T.

: oz<m._<m_ m
S

.v G

N

- m

:33 cm>z__ B
Y

A

.’ H

h 92¢ # M
N

O

.’. n

- N
, w
ToEozoowgoi m
- C

 

weak—Ohm 0...
0230241 QUH<EOHD<

.’

MN_.__.w.._.._mQ

 

 

02;mo owm ONE
1054... mDODZCZOU

+

Q<ubmzz<w
O... onaszC.

*

mw><m.._ Emhm
mo\oz< aOIU

 

 

 

 

 

 

 

 

STEAD DRYI N G SYST EM

IDEALIZE D FARM

 

wo<mOHm Oh
$23024: QwH<EOHD<

o<mhm2z<m

O... kzocmzs:

a.

MN_._.w4.._wn.

 

 

 

 

:5 oz< “=sz
.mzazou “an:

 

 

 

LIZED FIELD DRYING SYSTEM

 

QEA

PROCESS F LOW DIAGRAM

FIGURE l.

 

arbitrary length exceeds that which will vertically trans-
port a fresh-cut leaf. Thus, stems would need to be
altered from their natural configuration to be considered
in the same drying stage as leaves.

Since approximately 70% of the food value in the
alfalfa plant is in the leaves and minor stems, the advisa—
bility of drying stems is questioned. Schrenk reports
that leaves make up 51.1% of the plant weight at the one
tenth bloom stage of maturity. Thus, by removing 50% of
the plant water, 70% of the food value is recovered. lore-
over, the water in the leaf is far easier to remove, since,
among other factors, the smallest dimension is of the
order of 0.01 inches while that of the stem is approximately
0.1 inches. Since the time required to dry a particle to
a given moisture content increases approximately as the
square of the least dimension, according to Van Arsdel
(1963), theoretically it should take about one hundred times
as long to dry the stem to the same stage as the leaf.

The rate of leaf drying will be higher than for the plant
as a whole.

The work reported in this thesis is an analysis of
this overall concept with research activity specifically
directed to the determination of the design parameters of

fluidization principles for alfalfa dehydration.

Fragmentary initial findings relative to the cultural
practices and plant physiology which relate to such a
system are included.

Since this concept is far from traditional methods
of forage harvesting, supporting basic information appli-
cable to the system is lacking. Further, it is anticipated
that eventual application of such a drying process might

encompass other agricultural products as well.

REVIEW 0? LITERATURE

The general concept of stripping alfalfa leaves with
immediate dehydration in the field or farmstead appears
not to have been reported as a specific research activity.
However, considerable research has been reported which is
applicable to alfalfa dehydration. As a result, many
large commercial installations have been built and suc-
cessfully Operated over the past thirty years. However,
very little basic information is available on this process
for alfalfa which would be adaptable to systems involving
leaves only. Further, the cost of dehydration equipment
which has been developed for commercial use is prohibitive
for smaller farm-sized units.

Such a system might be found by using fluidized bed
drying principles. Fluidization has provoked widespread
interest in the petro-chemical industry and has resulted
in tremendous research efforts in the past twenty years.
These research efforts have dealt almost exclusively with
non-biological, granular materials. Paralleling this
interest, food researchers have developed fluidized bed
drying and spray drying processes for a wide variety of

agricultural products. Consequently, the process is well

defined both mathematically and experimentally insofar as
mass and heat transfer relationships are concerned.

Thus, the literature review will be directed to the
establishment of the design parameters for fluidized bed

drying of alfalfa leaves.

Desirability of Drying Leaves Only

Stripping leaves from the growing alfalfa plant as a
possible cultural practice, thereby allowing the plant stems
to produce more leaves, is a relatively new idea submitted,
by Hall, McColly, and Buchele (1959). waever, although
not supported by research results, there appears to be sound
basis for considering this since leaves contain twice as
much protein as stems as reported by Schrenk, et a1 (1959);
carotene content and vitamin A are similarly associated.
Considering an alfalfa leaf product and basing calculations
on information by Schrenk (1959), it can be shown that a
relatively constant product of 25% protein might be expected
for stages of maturity before full bloom. This is con—
sistent with the requirements of commercial alfalfa dehy-
drators who prefer to cut at early stages of maturity to
get high quality alfalfa with a high percentage of leaves.
Present practice results in a product with continuing
decrease in protein content as the season progresses,

necessitating the blending of low and high protein content

material to meet the guarantee specifications. According
to wale (1962), storage for various grades and subsequent
blending constitute additional costs and quality control
problems, (which might be reduced by considering leaves
only). Grinding in preparation for pelleting appears to
be unnecessary, which might result in further savings when
considering leaves alone.

It has also been reported by Schrenk (1959) that
leaves constitute approximately 50d of the weight of the
plant. Since leaves and stems have about the same moisture
content (with seasonal variations), it can be concluded
that by removing 50% of the water that 70% of the protein
can be retained in a dry product with 25% protein. This
process may make dehydration more Justifiable for farm-
sized units as well as for the large commercial ones.

The present practice of handling and curing hay
results in huge losses. Field losses of 28%, mentioned
by Hall (1957), 39%, by Schoenleber (1949), and 40%, by
Bohnstedt (1944), are Just a few of the references
suggesting the magnitudes which add to the undesirability
of a dried forage form. Thus, without elaborating on the
inherent problems associated with the unevenness of drying
of the various plant parts, it becomes apparent that a
process which permits the harvesting of fresh cut material

by a single pass should result in less losses.

From a physical standpoint, leaves dry much quicker
than unaltered stems as reported by Pedersen (1960). This
is because of the relative magnitudes of the least
dimensions as well as the relatively large surface area
exposure, in addition to other physiological factors.
Leaf thickness dimensions are of the order of 0.01 in.
while those of the stems are of the order of 0.1 in.
Altering the natural stem configuration changes this
relationship. Bhan (1959) demonstrated that the drying
rate of stems cut § in. long approached that of leaves.
he further showed that by increasing the exposed area by
splitting the stem, the drying rate is significantly
increased. Pederson (1960) demonstrated similar behavior
by various other treatments of hay and found that "hard
crushed" stems dried faster than leaves.

Other factors which affect the drying rate of leaves
depend on diffusion and certain internal mechanisms which
are not completely understood. Whitney and Veeks (1963)
observed that for high temperature - short time vacuum
drying of pre—frozen samples where the cells were killed
and ruptured, the leaf literally exploded. Deducing that
the stomata behavior might be involved, observations were
made using a well known method whereby a silicon cast is
made of the leaf epidermis. The cast is then coated with
clear fingernail polish. After drying, the resulting

10

peel is observed directly under a microscope revealing
the stomata behavior. Alfalfa leaves on the growing plant
were "observed" at a time when stomata were thought to be
open. After the initial impressions were made, the plant
was cut and five minutes later impressions were taken of
adjacent leaves when it was found that the stomata were
completely closed. This is clearly shown in Figures 2 and
3. It is well known through studies of tranSpiration by
Loftfield (1921) and many others, that the transfer of
water from the growing plant is controlled by the stomata
cells. From the available information, it is still diffi-
cult to predict behavior of stomata because of the inter-
relation of uncontrolled factors such as light, temperature,
relative humidity, wind velocity, and soil moisture. When
the stomata are closed, the epidermis is very effectively
sealed. From these considerations it appears that the
condition of the stomata may influence the transfer of plant
held water. This points to the desirability of immediate
field drying when the stomata are open, effecting an
accelerated mass transfer of water. No information is
available relating high temperature drying rates with the

condition of stomata cells.

11

 

Figure 2. Impression of Alfalfa Upper Epidermis-IQOOX
Stomata Fully Open; Growing Plant

 

Figure 3. Impression Of Alfalfa Upper Epidermis-l200x
Stomats Closed After Cut Five Minutes

12

High Temperature-Short Time Drying of Heat
Sensitive Materials

Drying rates of alfalfa as well as other plant parts
are well established for many temperatures with most
research conducted at temperatures below 200°F. Research
on drying alfalfa is virtually nonexistent for temperatures
at which damage is known to occur although it has been well
established that prolonged heating at temperatures above
160°F. results in dry matter loss and/or damage. However,
heat sensitive materials, such as alfalfa, are in fact
dried at much higher temperatures. Schoenleber (1950)
reports inlet temperatures for dehydration of alfalfa at
lOOO‘P. with the outlet at 250-325°F. Schrenk, et al (1959)
report inlet temperatures as high as 1500’1. Sewever,
the susceptibility to damage of alfalfa at high temperatures
and resulting damage point relationship are not reported.
Headley and Hall (1963) report a method for determining.
damage points and drying rates of corn kernels at various
high temperatures up to 800°F. Their method of presenting
these relationships are particularly applicable to alfalfa
drying.

Corder, et al (1956) report on the drying of sawdust
at 1200°r. with little tendency to char or burn. The short

time exposure of this material to high temperatures

13

permits this process. Particles at high moisture content
upon entering a high temperature medium remain at the
adiabatic saturation temperature of the gas. For air,

this depends on the wet bulb temperature and is usually
below 212°F. until the moisture is evaporated on the sur—
face during the constant rate period of drying. Heat
evaporates the water after it moves to the surface of the
particle by vapor diffusion during the falling rate period
of drying and subsequently cools the air. For these
processes, by the time the moisture has been removed from
the particle moving along with the gas medium, the gas

has cooled to such a low temperature that there is negligible
damage to the particle before it passes out of the process
stage. Subsequent cooling of the particle may be necessary.
Contact times vary from a few seconds for the suspension
drying of sawdust to three minutes for alfalfa dehydration.
The larger particles tend to move more slowly through a
given stage while the smaller particles have a higher
velocity. The larger particles, which require a longer
time to be heated, remain in the heat transfer medium
longer. As a result, the various sized particles attain
a fairly uniform moisture content upon leaving a process

stage.

14

Fluidized Bed Drying of Alfalfa

High temperature - short time drying of forage is
associated with continuous processes usually utilizing
direct heating, for which several such systems have been
developed. The most predominant system in alfalfa process-
ing according to Schoenleber (1949) is the rotary drum
dehydrator which permits a wide variation of particle size
for a chopped alfalfa plant. In an effort to increase the
drying area, the material is maintained in a quasi-fluidized
state by the mechanical rotation of the drum with subsequent
tumbling of the material resulting in parallel flow drying.
Another continuous process which might be adapted to
alfalfa is the fluidized bed dryer. Schrenk, et al (1959)
described this process in some detail to which they refer
as a pneumatic type alfalfa dryer. Several foreign manu-
facturers of such a system for which claims are made of
successful drying of chopped grasses, alfalfa, sugar beet
taps, etc. are cited. A pertinent remark is of interest:
"Iaterial having the shape of a flake is ideally suited'.
for drying to low—moisture values in a pneumatic-type
dryer because of short distances moisture migrates." This
thought has been expressed by several researchers. As a
result of this recognition, experimental direct-fired
Pneumatic dryers, for which many of the operating design

parameters have been established concerning the mass

 

15

velocities, drying tower dimensions, feed rates and temper—
ature requirements, have been constructed and tested at
Kinsas State University. A relationship between the mass
flow rates of the drying medium and the initial and final
moisture content of alfalfa has been established for inlet
temperatures above 720'!. These results show that as the
mass velocity of the drying medium increases, the residence
time of the particle decreases and consequently, the amount
of moisture removed decreases. This was previously reported
by Gregg (1955) in his work on the suspension drying of
sawdust.

Schrenk concludes that the pneumatic type dryer shows
promise in the field of artificial drying of alfalfa, but
that further work is required to determine the proper degree
of field chopping and control features to yield a dried
product of uniform quality. The apparent need for evaluat-
ing the behavior of alfalfa leaves is clearly implied.

Theoretical considerations of fluidized bed drying
have been similarly well established both mathematically
and experimentally. Love (1959) has written one of the
most recent texts which summarizes the significant works

in fluidization. lass and heat transfer experimental
results are presented in semi-empirical form. Over one

hundred papers are presented each year in the general area

16

of fluidization. Sbwever, non-biological materials which
can be idealized as spheres are usually considered.
Applicability of these findings to biological materials

is not obvious.

Behavior of Fluidized Biological laterials

When considering particles in a fluidized state, (which
might be idealized as spheres, i.e. sand, catalysts, granules,
and coal dust) it can be assumed that volumes do not change
except for such mechanical processes as attrition.

A biological material such as sawdust might also be
idealized, since it does not change substantially in shape;
however, it does shrink somewhat. Rice (1960) explored the
behavior of various shapes and sizes of particles in vertical
air streams for which he was able to determine terminal
velocities and relative stabilities. Pettyjohn and
Christiansen (1948) conducted extensive studies on isometric
particle terminal velocities.

The behavior of a particle which changes density,
shape, and size, all simultaneously, is not readily idealized.
An alfalfa leaf is such a particle. Initially, a leaf
might be considered as a thin wafer or disk. when a disk

is oriented such that the large cross-sectional area is
normal to the stream of fluid, it experiences a drag force

for which a particular drag coefficient is obtainable.

 

17

Later, as the particle orients itself such that the normal
of the large area becomes perpendicular to the stream,
the drag coefficient decreases. Thus, based on data by
Binder (1943), the drag force can be calculated and shown
to be continually changing while the gravitational force
remains relatively constant. A resulting oscillating motion
can be expected. In addition, the shape has been shown to
change quite radically because of wilting, curling, and
shriveling-—each stage with indeterminate drag coefficients
resulting in a completely randomized motion much as might
be visualized for molecular movement of a gas. Othmer (1956)
reports that particle movement is not entirely random.
Particles will rise at the center of the bed where the gas
velocity is highest and fall at the wall of the vessel
where the velocity of gas is at a minimum. The value of
individual particle observation is questioned by Quinn (1963).
He suggests a statistical approach to this problem for
fluidized bed studies. Persson (1957) reports a similar
approach for air separation of grain combine materials.
Bailey, Fan and Stewart (1961) predict the instability of
a fluidized bed by statistical analysis.

Kennet (1950) followed by Anderson (1951) conducted
research on the simultaneous grinding and drying of alfalfa
to reduce the particle size before the drying process, to

obtain a more rapid drying rate. The principle was

18

demonstrated but resulted in extremely high power costs

and instability of the grinding machinery, virtually

rendering the process unworkable. Bbwever, the mass

transfer coefficients were determined.

Kennet (1950) expressed the mass transfer of water

from small particles of alfalfa to the carrying medium

for a pilot direct-fired dryer as follows:

 

 

w - .00014 c A up: (92 (51'2“ F32! ”/3
P31 in FL ( FL)
Where:
I - rate of mass transfer, No. moles/hr.

G - mass velocity, lb/ftz/hr.

a - effective area of mass transfer/unit volume

of bed, sq. ft./cu. ft.

In - mean molecular wt. of the gas stream, 153/

mole

v - volume of mass transfer space, cu. ft.

AP, - log mean partial pressure difference at the

terminals, atmospheres

log mean partial pressure of the non-trans-

ferred gases in the gas film, ’atmospheres

Dp - diameter of particle, ft.

[1,. dynamic viscosity of gas, lb/hr. ft.

P .. 4.11.1" of gas stream, lb/cu. ft.

Dv - diffusivity of gas in the film, sq. ft./hr.

19

D G - Reynold's number, dimensionless

._E__
p.

__lf_'- Schmidt's number, dimensionless
PM

The rapid rate of drying in a fluidized bed was due to

the great drying surfaces, and not particularly to the

high mass transfer coefficient. This is shown by a state-

ment by Corder (1955) in which he states "...the drying

rate of wood is governed by heat transfer rather than

diffusion when the wood particle is less than i in. thick.

Heat transfer, therefore, is the controlling factor in sus-

pension drying."

Divergence of Velocity Field
The effects of the size and shape of particles in a

fluidized bed are well established. The velocity require-

ments for such a particle as it loses moisture are not

readily understood from the literature. As a particle of

a given moisture content is suspended in a vertical stream

of hot air, the particle progressively loses moisture, thus

becoming lighter. What was once the terminal velocity,

later becomes the vertical transport velocity and the

particle moves upward until such time as the velocity of the

air decreases to a value where the particle will again

remain suspended. The particle drying in a tower requires

I negatively diverging velocity field according to the

20

definition by Davis (1952). This idea has been implied by

Gregg (1955) and Schrenk (1959) noted the sensitivity of

velocity to temperature, but did not describe the divergence

of the velocity field. For a finite stage of a tower

process, a particular residence time would be required as

d by the mass and heat transfer and damage
residence time,

governs point

relationship. Since the velocity controls

it becomes necessary to determine the divergence for preper

design. Velocity divergence can be obtained structurally

with a tower which is smaller at the bottom. Such a device,

referred to as a diffuser, is described by Noel, et al

(1954) for a pilot plant model of an air-lift dryer used

for potato granules. Schrenk also describes a similar

device in their experimental dryer, which is apparently used

to separate large stems from the field chopped material

before drying further up in the tower.

Negative velocity divergence, as a result of the

negative temperature divergence, is desirable from a drying

time standpoint and may be of considerable consequence.

The temperature decreases owing to cooling of the heat

transfer medium, as a result of evaporating water from the

leaf surface and by heat loss from the hot tower to the

surrounding atmosphere at a lower temperature. This

phenomenon is exemplified by stack design in which the

carbon particles in flue gases must be expelled out of the

stack. Here, a constant velocity, or possibly a positively

21

diverging velocity is required. Smoke stacks are usually

smaller at the top to structurally negate the effects of

the negative temperature divergence. The diverging velocity

is undesirable in this particular case.
The positive divergence due to addition of water vapor
as it is evaporated from a moist particle progressing up

the tower must be added to the negative divergence produced
by the cooling effect on the heat transfer medium. It

can be shown that this partially negates the desirable
velocity divergence and must be accounted for in a differ-

ential equation which considers the laws of continuity,

motion and thermodynamics. The treatment is based largely

on the approach presented by Bird, Stewart, and Lightfoot

(1962).

22

THEORETICAL CONSIDERATIONS

In order to obtain the differential equation leading
to the relationships of momentum, heat, and mass transfer,
the law of conservation of mass is applied to a small

volume element within a flowing fluid, in which a particle

having "zero" volume is suspended. The particle is assumed

to have an arbitrary moisture content which is releasing
aw .
water vapor at the rate 3? ’

z ’ T 2:4:I'Az

 

 

 

 

 

 

 

 

2
! p—u—(x+Axl
‘ Y‘I‘Ay’
| 2&2)
: 9x 2
Particle
z - - 7t— -— 4 -- —- m o
I / \ A mass - w
t; 1 ' Velocity - Vi - 0
Oman-'3 l pd;
Y

 

 

 

113-2. . .
x... ,

23

The volume AxAyAz is fixed in space through which a fluid
of density fDand velocity V is flowing and suspending a
particle of "zero" volume.

The equation of conservation of mass leads directly
to the equation of continuity as follows:

rate of mass - rate of mass - rate of mass
accumulation in out

AWAZ 3’? + AWAZ 331:: ' Ayaz (pVx)Ix - (P‘ixllfiox
+Amz PVy) Iy - (,OVy) Iymy
+Axoy pvz)Iz —- (PVz)Iz+Az
. . Jr;

Divide right hand side (rhs) and left hand side (lhs)
by was and take limits as Ax, Ay, Az approach zero:

g€+3¥=_[g_ipvx+g7pvy+gzpvz]+o (1)
= __fqg!§ ._ Vx饧‘_ . . .

Rearranging terms:
- - V aV' + avz
a + V a + V + V2 a£>+ 6w 3 x +
aé) xng y y as 315 p3! 3y 2
This is the most general case of the equation of con-
tinuity with the additional term accounting for the mass
release of water vapor by a particle in suspension. Since

a compressible gas with changing density and viscosity is

considered, the general form must be used throughout.

24

Similarly, the equation of motion can be written by
summing forces on the system:
rate of momentum} rate of rate of um of
(acceleration) - momenta -moment +{gorces
syst
This can be expanded to indicate the nature of the
forces as follows:

rate of increase rate of increase
of momentum per of momentum per -
r

+
nit volume - flui nit volume - vapo
ate of momentum pressure force

gain by convectio + on element per +
or unit volume unit volume

4.

rate of momentum
gain by viscous gravitational
ransfer per u.v force on +

lement per u.v
gravitational
force on

particle mas

+

Drag force on}
particle

By Newton's Law of Viscosity, the viscous forces acting

on the faces of the volume element for laminar flow are:

'u

>I'I
*4I<

That is, the force per unit area is pr0portional to the
velocity decrease in the distance from the location of

force F; the constant of proportionality,fL, is the
viscosity of the fluid. The force per unit volume is the

stress on the surface. The shear stress exerted in the

x-direction on a fluid surface of constant y by the fluid

25

in the region of lesser y is designated as'Txy, and the x-
component of the fluid velocity vector is Vx. Thus,

Newton's law can be written:
Y
Y8 ALE?!

Six expressions for the stresses can be obtained in
terms of the velocities:
V

V an + 8 + aVz
TXX -- -2I~lg_x_x + 2/3’J‘ X 3’7! 3;]

plus two similar expressions for‘Tyy and‘Tzz. Expressions

for 7&2 - Thy and 758 - ng are similar to:

v V

The equation of motion can be written for the components:

gfpvx+gf 'Vx" '[§vavx+3;PVY'x*§-5PV'V’J

-[:% 'rxx + SFTYX + 327”]
.. g . pgx + (try/341'“),

plus two more equations for y and z components.

The equation of motion for the system can be rearranged
With the help of the equation of continuity and substitution

for the stress terms; considering the x component as follows:

26

.. _ v _
pgeg. "* {first M" 51,3133]

9 aV’ 6V ~a 3V
+§i[‘3‘5+&1] + 5&3}! + 379 +ng

plus two more equations for y and 2 components, where 2%:

is the substantial time derivative. These equations along
with the equations of state, the dependency of viscosity

and the boundary and initial conditions completely determine
the pressure, density and velocity components in a flowing
isothermal fluid with a suspended particle. The equations
can be expressed in terms of cylindrical coordinates (r, 9,
z). Since the z—coordinate is the only one of interest in
the simplest case, the equation can be written as follows:

v -
3:! 4' Yr avz ‘0’ v9 avz 4* V2 8 z + _a__ ' VZ sz+

3F— i-“ST 3‘? at
- 7' l a79z + 37h
5: B‘s—(r rz)+?ae—‘ z
. - v av ,
Where. ::f%{2fi a: -2 l &S%. (r Vi) + %_g%g + Z2
3V9 + 1 8V: ,
§E_ F 59—
- ~iL a: z +- éfigj
as

The equation of mechanical energy can now be written by

forming the scalar product of the local velocity V with the

equation of motion:

27

rate of increase

in kinetic energy -
per unit volume

of fluid and vapor

et rate of input of rate of work done
kinetic energy by - by pressure of +
virtue of bulk flow surroundings on

volume element

rate of reversible rate of work done
- conversion to - by viscous forces +
nternal energy n volume element
rate of irrever- rate of work done
- sible conversion - {by gravity force
to internal energy on volume element
The increase of internal energy due to the simultaneous heat

transfer of the system must be added. The energy balance

expressing the first law of thermodynamics is as follows:

of internal

{fate of accumulation}
kinetic energy

rate of internal rate of internal

and kinetic energ _ and kinetic energy +
in, out,

by convection by convection

not rate of heat not rate of work

added by — done by system

conduction on surroundings

The rate of accumulation of internal kinetic energy

within AWAZ is:

in v2)
Axayaz :33? (p0 + isz) + AXAYAZ 3-;

where U is the internal energy per unit mass of the fluid

in the element.

28

The rate of convection of internal and kinetic energy

into the element is:

AYAZ{Vx (PU + ipvz) Ix - Vx (PU + észflxflSx

+ . . . (two more terms)

The not rate of energy input by conduction is:

£9112 {081x " (”In-Ax + ' ° '
where qx, qy, and qz are the x, y, and z-components of the
heat flux vector 3.

Work done by the fluid element against its surroundings
is accountable in two parts: the work against the volume
forces (e.g. gravity) and the work against the surface
forces (e.g. bouyant or drag). The rate of doing work

against the gravitational force E per unit mass is:

o
‘Pammz (szx + Vysy + VzSz) - yup/'03:: - sy - sz)
0
plus the drag force on the particle +‘yyv’KF dx - de - Fdz)

Since the particle is assumed to be stationary, there is no

rate of work.

The rate of doing work against the static pressure p

at the six faces of,AxayAz is:

NAz {(pVx)|x+Ax " (PVX)IX} + .

Similarly, the rate of doing work against the viscous

forces is:

29

MAI {(Txxvx " xyvy " szvz) X+A* " (7'qu - Txyvy - Tszzjx}

+ .

Summing the above expressions for energy and equating
according to the first law of thermodynamics, then treating
as in the derivation for the equation of continuity, the
equation of energy can be obtained in terms of the energy
and momentum fluxes. The following equation of energy is in
terms of CV, the heat capacity at constant volume and T,
the temperature as it changes with respect to time 3% -
Appropriate consideration is made of the potential energy

change. The expression in cylindrical coordinates is:

pcv[g+vr3¥+¥gg§+fig_§]+SEC§wV2)'

l a (r q ) + l 309 + 39%] +
{35; " ear 3—

1 v) 16's 6% +
TB¥]p[FS'F ” ” +r§r+sr]

.. T av T 1 ava + v + T avz} +
avr +

_ V' 7' 3V ..
Whats) +-:-g-.=J+ "(5.! 5.—
mega?»

whore qr --kg_! ’ q9-—k%3-g' and qzu-kgg
r

k - thermal conductivity

30

For the most general case,[.LandPare functions of T
and are not constant which does not permit the usual
simplifications. Thus, the general form of the equation
of energy has been written which must be used with the added
term for the vapor producing effect of the suspended particle.
These equations have been based on laminar flow. If
turbulent flow is considered, the values in the equation
are for "time-smoothed" quantities to which terms must be
added to account for the fluctuations about the "time-
smoothed" curves.

Applying the energy equation to obtain the differential
equation for flow in a round vertical pipe, the z-components
only are considered, making simplifying assumptions as
follows, neglecting turbulent fluctuations:

VD - Vi - 0

V2 not a function of 9 .1 3N2 - 0

V2 is a function of r, but if r is large,

Viagconstant and glg - 0

6r 2
Similarly for r and/l, 3&1 -%'_11: - 3;; - 0. Thus, by

omitting the inappropriate terms, the energy equation

reduces to:

2 _ V;
cv(%¥+vzgg)-g€(§wvzz)-kg_;§ T(g¥)P (32—)

‘ gaggéf

whereP-p(T) andfl- F”)

31

This second order differential equation would need to
be solved for V2 and T2 between the limits of the boundary
conditions. However, with the complexities of changing f)
andfl, the equation is virtually unsolvable; especially
with the vapor producing effect of the particle in suspension.
However, the "unsolvability" of the equation is generally
implied by reason of the usual requirements for assuming
that the density, viscosity, and fluid mass of the system
remain constant. As a result, semi-empirical and dimension-
al analysis approaches have been taken, e.g. Chilton and
Colburn (1934), Kettenring, et al (1950) and Sieder and
Tate (1936), to mention but a few researchers in this area.

By examination of this equation, several predictions

and generalizations can be made:

1) The equation is of second order for both temperature
and pressure.

2) The divergence of velocity and temperature play
prominent roles in the mathematical description of the
process.

3) The addition of the vapor producing effect from
a "zero" volume particle is evident and directly influences
the effect of the divergence.

4) Turbulent flow, necessitating the addition of

terms to account for the fluctuations of velocity and

temperature would further complicate the differential

equation,

32

The solution of this equation would probably be best
carried out using numerical methods or finite difference
methods, but will not be pursued further herein.

It has been assumed that the particle is in a suspended
state with zero velocity. This suggests a terminal
velocity, the calculation of which is applicable to non-
isometric particles. One such empirical formula for non-
spherical, isometric particles results from extensive work
by PettyJohn and Christiansen (1948). Stoke's law was

modified with an application accuracy of i 2%, as follows:

Vt - K (FL - Pk) 8c dp2 (laminar flow)
WW

where x - 0.843 log $8 , Stoke's law shape factor

Application is for isometric particles only with
sphericityqbg of not less than 0.60. Sphericity has been
found to be a satisfactory criterion of the effect of
particle shape on the resistance to motion of particles
moving in a fluid. The extent to which it can be applied
to non-isometric particles is not known.

Under the effect of highly turbulent flows, the
following equation (PettyJohn and Christiansen, 1948) can

be used with an accuracy of i.‘%°

'Vt -;/Qf‘ do (st"F?f) 8c__
3 Cr p f c 3000 (Re (200 .000)

 

 

33

where Cr - 5.31 - 4.88qbs, the coefficient of resistance
dB - spherical diameter (diameter of sphere having the
same volume as the particle), cm
dp - projected diameter of particle, cm
gc - acceleration of gravity, cm/sec2
F} - density of the fluid, g/cc
F; - density of the particle, g/cc
IL - viscosity of fluid, g/cm sec
Vt - terminal velocity of particle (terminal velocity),
cm/sec
Sphericity is a parameter of considerable interest.
It is defined by Love (1959) as follows:

qbs -'%¥ (dimensionless)

where A - the surface area of an arbitrarily shaped particle
Ap - the surface area of a spherical particle having
the same volume as the particle of arbitrary shape
Since a spherical particle is a body that will provide
a given mass with the least surface area, values ofqbg will

always be less than unity. By applying the basic geometrical

relationships, a more convenient form results:

- /3
45' $26511

where V - volume of particle, in3; A - area of particle, 152

34

While the nature of granular particles does not permit
direct calculation of sphericity because of dimensions and
uneveness of surface, sphericity of flakes, e.g. alfalfa
leaves, is obtainable directly from measurements. Pressure
drop tests and displacement determinations are used to
obtainCDg for granules. No information is available for
alfalfa leaves, especially as sphericity may be affected by
a progressive change in moisture content.

linimum fluid voidage,é m1, is of particular interest
because of the relationship of the overall space requirements
for a process. The leastE m1 expected can be calculated

from the relation as follows:

Gmf - 0-005 up“ Be P: (9. — P1) <l>32 6.13
(1 - 6M)

 

where on, - fluid mass velocity for minimum fluidization,
lb/hr. 1t2

However, this expression is derived for particles with
qh. far in excess for that which might be expected for a
particle such as a leaf. This seriously limits the value
of the above correlation. A generalization is offered by
the important observation that “(13. decreases, elf
increases. Thus, the population of particles per unit
volume is much less dense for a flake than for a sphere.
Further, according to Leva, the less spherical the particle,

the more interstitial space will be required to permit

35

motion. The diameter of a particle such as a leaf is far
beyond that which is referred to in the literature for
which the voidage emf can be predicted.

Fluidization in the classical sense begins with a
fixed bed and gradual increase in the flow rate of fluid
until the particles become fluidized. It is questionable
whether a fluidized state can be induced from a fixed bed
of particles of the characteristics of alfalfa leaves with~
out some means of agitation. Thus, it appears that the
particles must enter the fluidizing vessel in a suspended
state.

Prediction of the point of incipient fluidization is
based on the relationships as presented; but for flake
particles these relations have been found to be unreliable

and beyond the range of the published research results to

date.

 

36

STATEIENT OF THE PROBLEI

The general objective of this research is to determine
the design parameters and supporting information for a
continuous drying process for alfalfa leaves which is ful-
filled by fluidization drying principles. Since these prin-
ciples are well established insofar as the behavior of
particles and the mechanics of heat and mass transfer for
non-biologicals, the specific objectives will be oriented
to biological applications. Several aspects will be investi-
gated for which initial determinations will be made as
follows:

1. To determine the tolerance and susceptibility to
damage of alfalfa plant parts exposed to air flows at high
temperatures.

2. To determine the drying rates of alfalfa leaves
at high temperatures incidental to the damage point deter-
mination.

3. To determine the behavior of alfalfa plant parts
in vertical hot air streams as the plant part loses water.

4. To determine the air velocity divergence require-
ments in a model drying tower which might also provide the
required drying time for an arbitrary drying stage.

37

5. To indicate possible cultural practices and plant
physiological factors which relate to the mass transfer of

plant held water, and how these may affect the drying process.

38

PRELIIINARY INVESTIGATIONS

An evaluation of the fluidized bed drying process for
any arbitrary material must be preceded by a knowledge of
its physical characteristics. Specifically, the dimensions
and physical phenomena associated with forage must be

considered as a prelude to such a study.

Forage Plant Selection

Two types of plants are in common usage for dried
forage former-grasses and legumes. Neither plant type in
a basally severed, unaltered form can be considered as
capable of being fluidized in the normal sense. To approach
some semblance of becoming fluidized, grasses would need to
be chopped finely, still with some degree of non-homogeneity
between the chapped stem and leaf blade. Similarly, even
greater non-homogeneity is encountered when considering
the various parts of a trifoliated plant because the stems
constitute one half of the weight. After consideration of
the relative merits and value of the various plants and
their parts, the broad leaf has been chosen to most nearly
approach acceptability for this drying process study. Any
plant part, were it chopped or ground finely enough,
might be fluidized provided that the adverse affects of

39

juicing and subsequent agglomeration of particles could
be overcome. Kennet (1950) was seriously handicapped
because of his inability to overcome this difficulty. Be
partially negated this effect by feeding back dried
material.

There are several legumes which are commonly grown
for dried forage of which a few are: alfalfa, clovers
(white, red, ladino), lespedeza and birdsfoot trefoil. Of
these, alfalfa is of the most economic importance, and was
selected as the main object of study.

Leaf sizes vary considerably on any given alfalfa
plant with a leaf weight ratio (maximum: minimum sized
leaves) usually greater than 3:1. An arbitrary classi-
fication of leaf size was made, based largely on visual
inspection as to population on the plant. Only plants in
the pro-bloom to 10% bloom stage were considered to reduce
the effects of culture and plant oriented variables.
DuPuits variety of alfalfa was selected because of the
seedling vigor and quick recovery after cutting. Also,
birdsfoot trefoil, which has smaller leaves with a maximum
leaf weight ratio of about 1.5:1, was also used because of
the miniaturization required for small bore pilot studies.

Figure 4A illustrates an alfalfa plant with widely
varying sizes of leaves in which the saw tooth edge of the

leaf can be seen easily; Figure 4B shows 1 in. diameter

Figure 4.

4O

 

Alfalfa And Birdsfoot Trefoil Leaves

 

Figure 5. Dried Birdsfoot Trefoil Leaves

41

leaf disks made with a conventional paper punch. This
size of flake particle was found to be most compatible
with small bore column drying, however, because of the
difficulties in producing large quantities of these small
disks, it was decided to use birdsfoot trefoil leaves.
Figure 4C illustrates typical trefoil showing smooth edges
on individual leaves and the uniformity in size of leaves
growing on the plant stem. Leaves as shown were plucked
individually from the plant to exclude the minor stem
attachment. This eliminated hair like appendages which were
found to interfere with the freedom of particle movement

in a small bore column.

Plant Part Dimensions

The dimensions of various trifoliate plant parts are
listed in Table 1 to provide some index of the relative
sizes which were considered. The data obtained were for
fresh cut, typical plant types; pro-bloom - 10% bloom stage
of maturity. Leaf sizes were dependent on a visual select-
ion based on the most frequently occurring size range.
Tabulated values are the numerical average of ten typical
plant parts in each category. The one inch stem section

was taken at the midheight of the plant.

42

 

 

 

 

 

 

 

 

 

 

«so.o aooo.o ao>o~o ooaeos
oo~.o woo«.o ao>omo com
ooo.o oooo.o success ooouuoaam
oo.om on.oo uvo~.o wmooo.o ooo.o Homo.o «assume
box «a ou\os as «as .aoa< on naval emu”
somd .mmflmsea .essno> seduces .ueeosesn .asw«03 seem sens!
ooemssm Immoso
o.vo «.mv nooo.o ov~.o mo.o oooo.o ao>oso cusses
o.vo o.se nooo.o men.o so.” soon.o ao>omo one
m.sv s.me vuso.o Hoao.o nu~.o voso.o ”soothe oooooeaam
v.~o e.ge oooo.o oo¢.H~no.o vo.HoH.o moo.uo~o.o Hoaaououmaume
v.uo o.~v mooo.o «Hoauuvo.o us.uwm.o «Ho.uovo.o assooauauoous<
o.mo e.gv oooo.o ooo.unoo.o oo.uoo.o ooo.uooo.o omaasnoumaos<
omxnou «exam as as on m «saw
.aoa< .amauooo .uaooaoaoe .on:~o> .moa< .oomeom «one
see new oeaoonosm

 

mbmdm b2<dm HF<HAOhHmF PDQ mmm¢h ho mflnH¢MEOEA A<Unmhlh

.H HAmsh

43

Measurements were obtained by various means, and
the dimensions presented are in the raw units commensurate
with the particular instrument used.

Weights were obtained using a Iettler analytical
balance, accurate to 0.00005 g, shown in Figure 10. The
accuracy used in this study was 0.0001 g because the natural
moisture loss from a leaf during the time required to
make the measurement was usually discernable.

Projected leaf areas were obtained with a planimeter.
The leaf was placed under a transparent plastic film and
the perimeter was traced three times to obtain an average
area, in2. Because of the saw tooth edge of the alfalfa
leaf, an average perimeter was measured along what might be
roughly compared to the pitch circle of a gear. Accuracies
obtainable were i 0.005 inz.

One other way of measuring leaf areas is to punch out
a disk of known area from an object leaf, weigh this disk,
weigh the remainder of the object leaf and then relate
the weight of the known area to the weight of the unknown.
The obvious disadvantage of this is the destructiveness
of the measuring technique. It is necessary to keep the
leaf intact as volume measurements are made at different
moisture contents. One other disadvantage is that if the
disk is cut out of a leaf area with veins, the area or

volume of the leaf may be distorted. The punch method was

not used because of these reasons.

44

Leaf thicknesses and stem diameters were obtained
using a one-inch micrometer with accuracies of 1 0.0001
in. Hewever, because of the lack of hardness of a leaf or
stem, the necessary feel for micrometer measurements was
difficult to obtain, and an accuracy of i 0.001 was
estimated. Extremely delicate handling was required to
reduce the possibility of crushing of the leaf cells.
Leaves were measured in the non—veinous zones of constant
thickness. Stems were idealized as being of round cross
section, and measurements were made accordingly.

Vblumes at this stage of investigation were determined
from the thickness and area measurements. Another method

will be discussed later.

Stokes Law Application to Alfalfa Plant Parts
When considering leaves and stems in a vertical stream
of air, it appears that leaves might be supported by less
velocity than the stem section. Theoretical terminal

velocities can be determined from Stoke's law (Binder,

1947).

 

Cd f3 V2 A
D' 2

where D - drag force, lbs.

Cd - drag coefficient, dimensionless

P - density of fluid, lb/ft3

A - projected area, ft2

45

V - velocity, ft/sec.; Vt - terminal velocity
L - inside diameter of tube, ft.
}L- dynamic viscosity, lb/ft sec2
l - weight of particle, lb.
At the terminal velocity of a particle, the drag force
must equal the weight of the particle:
D - W

Therefore, the equation can be rewritten:

2
CdPV-t
w-———2————
2 2 I
Vt "W
- W
Vt f3Cd A

Calculation of the terminal velocities for a typical
stem and leaf, both fresh cut and dried were made using
this formula. The leaf shape was idealized as a circular
disk of constant cross section (the stem as a perfect
cylinder), for purposes of calculation and selection of
the drag coefficient. The temperature of the air was
selected at 300°! in the calculations as shown in the

appendix.

A summary of the terminal velocities is as follows:

46

 

. Theoretical
Terminal Velocity
Alfalfa Plant Part Stage Ft/Iin
Leaf Fresh 71.0
Leaf Dried 47.0
stem (1 in. long) Fresh 174.0
Stem (l in. long) Dried 110.0

This demonstrates theoretically that the wet leaf would
be vertically transported out of a column at the terminal
velocity of the dried stem. Thus, tentatively it can be
concluded that the leaf and unaltered stem are incompatible
insofar as a given fluidization stage is concerned. Two
alternatives might be: 1. to alter the stem by crushing
(thus rendering it more nearly the same area-weight ratio
as the leaf) or 2. separate the stems from the leaves.

For reasons of simplification, the second alternative
was selected for this study, although additional argument
has been advanced previously to consider discounting the
stems completely.

Preliminary flotation tests in a small bore drying
column completely substantiated this hypothesis. The
resulting separation of stems from the leaves suggested
applications which are in widespread use for other materials
and processes. If in a column of vertically rising air

(with high temperature heat source at the bottom), the

47

stem was found to drop, subsequent drying and dry matter
loss occurred such that the completely charred particle

eventually attained an area-weight ratio which allowed it
to be transported vertically out of the column. However,
the basic objective of the drying process was not met and

further consideration was pointless.

Single Particle Behavior in Suspension

The behavior of a flake particle in a vertically rising
hot air stream is characterized by a continually changing
orientation with respect to the direction of flow. In
addition, a decrease in moisture content creates a con-
tinually changing area-weight ratio. Both occurrences
result in a continual change in terminal velocity require-
ments.

Qualitative movement of the flake particle could be
predicted from the change in drag coefficient for the
changes in particle orientation. While a study of individual
particle movement is of questionable value, an awareness
of this behavior appeared to be worthwhile. In brief, a
wildly oscillating movement could be expected with an
average velocity of the particle in the upward direction.

To observe this behavior, a small bore glass drying
column was used. The location of the particle versus time

was recorded using audio-time recording apparatus as shown

48

in Figure 6. The method required the use of a dictaphone
as a recorder and an electronic metronome for time indi-
cations. The metronome clicks as well as the observed
location of the particle were audibly recorded. This record
was later transcribed as written data.

Results of these preliminary observations indicate
that particle behavior in the center of a column are as
expected-—i.e., oscillations characterized by large amplitudes
and variable periods with eventual movement out of the column.
The amplitudes are not constant, varying from a fraction of
one inch to ten feet, apparently limited only by the size of
apparatus used. Particle behavior in or near the boundary
layer is considerably different. The velocity of fluid
flowing in a pipe is zero at the inside surface, rapidly
increasing in the boundary layer toward the center. For
turbulent flow, the velocity rapidly approaches the maximum,
such that vAve - 0.8 Vlax (Bird et al, 1961). When a
particle comes in contact with the reduced velocity zone,
it might be expected to drOp, since the velocity is below
that which will support the particle. This was not observed,
however. Not only does the particle change in moisture
content and orientation, but it changes in size,and shape,
shrivelling and curling at random. Thus, when a curled

flake comes in contact with the inside surface, part of the

49

leaf is subjected to relatively high velocities while part
is touching the surface. The result can be predicted by
Bernoulli's principle for an air-foil section. The result-
ing lateral force created a frictional force which negated
the gravitational force on the particle. In short, the
particle may adhere to the inside surface. This was the
behavior observed.

For single particles, this adherence is an undesirable
phenomenon. Eventually dry matter loss results in product
damage even though the temperature at the surface is less
than at the center of the stream, and a longer residence
time can be tolerated. However, it was found that with
sizeable quantities of leaves, a rapid, random, molecular-
like motion results. The effects as observed for single
particles are considerably reduced and a smaller amplitude,
higher frequency (nearly constant) oscillation is evident
in all directions in a highly agitated mass. Particles
which adhere to the surface are dislodged into the stream
and replaced by the other moving particles. A relatively
constant velocity of the agitated mass up the column was
observed, dependent on temperature differentials and
velocities superimposed on the natural draft.

Clusters of leaves joined by minor stems as found in
trifoliate plants were observed to behave well in a
fluidized state, limited in these studies by the small bore

apparatus.

In all cases, the particles were dropped into the air
stream from the three foot level. It was found to be
extremely difficult to produce a fluidized bed from a fixed
bed of flaked particles. For all practical purposes, this
method of fluidization would be considered impossible and

undesirable because of product damage.

Stripping Leaves As A Cultural Practice

An evaluation of fluidizing (terminal) velocities for
forage indicated that leaves would require different process
stages than stems. Other information from the literature
(Schrenk, 1959) led to the realization that stems could not
be dried as economically as leaves. These conclusions
supported the suggestion that leaves might be stripped from
the stems, possibly as the plant continued to grow in the
field.

As a preliminary investigation, this idea was explored
on an informal basis. A small plot of alfalfa, variety-
DuPuits, was divided into three parts. The experiment was
conducted on second crop growth at the pre-bloom stage, the
chronology of which is recorded to best illustrate the
phenomenon.

On 7/3/62 a third of the plants was manually stripped
of leaves by a harsh pulling and squeezing action, leaving

the stripped stems standing. The leaves were not plucked

51

singly. Another portion of the plot was cut off leaving
conventional two inch stubble; the remainder of the plot
was left standing. Unseasonably dry weather prevailed. On
7/6/62, the stripped plants showed signs of leaf recovery
with tiny leaves beginning to appear. No growth was evident
in the cut section, and the control was in the early stages
of bloom.

Continued recovery of the stripped plants was evident
on 7/13/62 with remarkable recovery and thick growth of
leaves noted. The cut section had sprouted the third crap
and was about 2 in. high. The control was well into full
bloom with some blossoms turning brown. Continued dry weather
approaches drought conditions.

Sixteen days after stripping, the plants had recovered
to the point where the first blooms began to appear, with
notations made of intent to restrip leaves. The third crop
from the cut plot had reached a growth of 4 in. at this
time, and the control had gone to seed and was lodging.

Recovery appeared arrested on 7/23 and 7/27 with much
smaller leaves noted than control or new growth. Appearance
of leaf happers was evident with plant damage noted.

Bloom stage was not progressing. New growth on cut section

had reached height of 8 in.

52

On July 20, the plants were re-stripped with no
apparent growth benefits after July 23. Re-stripping might
well have been accomplished a week sooner. The cut section
remained at 8 inches, and the control still showed green
pigment, but was no longer accounted for.

Upon observation on August 3 (after a heavy rain the
previous evening, the first in over a month), the stripped
plants showed signs of recovery. new growth in the cut
section was up to about 9 in.

The stripped plants continued to show signs of recovery
on 8/9 and 8/17 again with small leaves in dense clusters,
similar in appearance to the first stripping. The control
and cut section had stopped showing signs of growth activity
because of the dry weather.

On September 4, the stripped plants exhibited signs
of the pre-bloom stage. On September 11 the plant did not
show healthy growth signs with continued dry weather pre-
vailing. Leaf hopper damage was evident.

Stripping for the third time was accomplished on
September 20, but plants had not progressed much beyond the
pro-bloom stage. The cut portion was beginning to show
growth activity after some rainfall.

No further growth activity of the stripped plants was
noted and the condition of the plants for weathering the
winter was extremely poor. An experiment was conducted

during the following season with similar results.

53

It was not the intent of this preliminary study to
establish the desirability of leaf stripping as a cultural
practice. Quantitatively, leaf production by stripping
outproduced the conventionally cut method, and the feasi-
bility of stripping leaves from the growing plant was
demonstrated. Obviously, much more research by crOp
scientists is required to establish the desirability-—the
details of the cultural practice, and the conditions under
which this practice may or may not be justified. In
addition, methods for the removal of leaves mechanically
would need to be considered and then developed.

Stripping leaves as a prelude to the fluidized bed
drying process appears to be highly desirable and should
be further explored in a more saphisticated and detailed

manner, possibly as a separate research activity.

54

 

Figure 6. Audio-Time Recording Apparatus

 

Figure 7. Temperature Indicating Apparatus

 

\

P~ ure b. Velm‘jty R4 (‘Hl‘dlnfl Apparatm;

 

 

. . [ll'l.l.l
II|I|II||II||t

55

 

Figure 9. Single Leaf Drying Apparatus

 

Yi Hrv 10. Analytical Balance

56

 

Figure 11. Leaf Volume Measuring Apparatus-Disassembled

 

 

 

Figure 12. Leaf Volume Measuring Apparatus-Assembled

57

   

VALVED ASPIRATOR BULB
(PROPIPETTE " TM.)

 

 

 

 

 

   
  
 

 

 

 

 

 

 

 

 

W40
_.09
_.08 \
_.O7
_.06
_.05
h—‘RUBBER ST “04
: E13:57:»: OFFER ..____ O I ml PIPETTE
" _.03 or equivelont small
bore Iube
._.OZ
-.--0'
.~ 4" DRYING TUBE :“00 m"
RESERVOIR I
I
I E GROUND GLASS
STAINLESS JOWT
WIRE CLIP
25 ml FLASK WITH
TAKE- OFF AT
LEAF BOTTOM

 

O—INDEX LINE

ON SMALL
BORE TUBING

MERCURY FILLED

 

 

 

 

Q

FIGURE l3. VOLUME MEASURING APPARATUS

58

 

59

 

 

 

 

 

.13

/
«a, «fi— [9

 

 

 

 

 

flag/[ml ‘Cu c T.C. (4) 6

g' ...d. \\
I . ,. .\
I t
.‘ . a s‘
I e
} o q \ ‘.
I x . .
I H .
v t. ’
I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.1...

 

 

AIR- LOCK SAMPLE
INJECTOR

 

Ill Ill/[l

__‘

 

 

 

 

_ l/[I/f/[l/llf //

($-

 

 

FIGURE I5. DRYING TOWER AND HEAT SOURCE

TABLE 2.

tomslcicnnmwtow

HHHHHHHH
NGU‘AQNHO

MATERIAL FOR DRYING TOWER AND HEAT SOURCE

Removable Slide

Air Lock Sidewall

1}" I.D. Pyrex Glass and Fittings
Stainless Steel Spacer Sleeve
Thermocouple Imbedded in Sauereisen Cement
#930 RCA Photocell

flailing Tube Shield w/ l/8" Hole
3-7/8" I.D. Acrylic Plastic Tubing
10 Position Rotary Sample Collector
Removable Sample Container

Glass WOol Insulation

Electric Heating Coils

Perforated Plate

Asbestos Gasket

Aluminum Flange

Sample Port
1/8" flesh Capper Screened Air Outlet

 

Figure 16. Drying Tower And Heat Source
With Sample Port Disassembled

 

Figure 17. Assembled Drying Tower And Heat Source
Shouing Smoke Powder Injection

62

EXPERIMENTAL TECHN IQUES

The broad scope of the system concept for which this
research was conducted required a wide range of experi-
mental techniques. While the basic principles of fluidized
bed drying have been well defined both mathematically and
experimentally, parameters related to forage materials have
not been established. The process for flake particles which
are heat sensitive and which are subjected to changing
moisture content, size and shape has not been reported.

As a result, information has been required in several defined
sub-phases with apparatus and experimental techniques
develOped for each as follows:

1. Time-temperature-damage point relationships.

2. lSphericity determinations related to moisture loss.

3. Mass velocity-temperature-moisture removal

relationships

Time-Temperature-Damage Point Relationships of
Alfalfa Leaves
From the results of the preliminary work, only the
leaves of alfalfa were considered in the drying tests.
Single leaves were subjected to high temperatures for which

the drying rates were determined and the damage points

63

were fixed. The technique and apparatus developed and
reported by Beadley and Hall (1963) were modified to meet
the requirements of alfalfa leaf drying. This apparatus

is illustrated in Figure 9 showing the leaf, (B), held by
the tip in a small spring clasp (D). This assembly was
rotated with the leaf hanging down into a vertical hot air
stream. The speed of the universal motor, a i in. electric
drill, (A), was varied by means of a variable voltage auto-
transformer (C).

Heated air was supplied at various temperatures and
velocities by varying the speed of a small fan with a
variable autotransformer. Two heating coils from a con-
ventional hot plate were placed on edge so that they lay in
parallel vertical planes 3 inches apart. In order to aid
in converting the radiant energy to convective heating, a
perforated steel disk was mounted between the two coils to
absorb the radiant energy. The entire assembly was enclosed
in a cylindrical sheet metal container and insulated with
aluminum foil and glass wool. The electric conductors
were external to this assembly for reasons of overheating.
Forced air at room temperature was supplied at various
velocities through an inlet orifice in the bottom quadrant
of the curved surface centering on the burner planes. The
air outlet was through an orifice in the top quadrant,
then through an attached 90° elbow discharge conduit such

that the object being heated could not "see" the intense

64

J

radiant energy source. Heat was transferred from the
radiant energy source to the metal plate; air forced through
this assembly was heated primarily by convection heat
transfer. Variable temperatures were attained by controll-
ing the voltage applied to the heating coils and by

changing the combination of the various segments of the
coils by means of the switching controls (6).

Temperatures of the heated air at the outlet were
sensed by an iron-constantan thermocouple installed in the
center of the hot air discharge such that it too could not
"see" the intense radiant energy source. A 3-in. lead wire
to the thermocouple formed a probe parallel to the air
stream and partially compensated for error from conduction
to the outside of the duct. The millivolt output was
determined using a Leeds a Northrup manual potentiometer
shown in Figure 7C. The reference junction was immersed
in an ice bath (70). Temperature determinations were
accurate to): Q'F. Calibration of thermocouples was made
at 32°F and 212°F.

Air velocities were determined using a vane anemometer;
however, the instrument could not be subjected to high
temperatures but for a short period. The moisture content
of the atmospheric air at the entrance was determined using
a sling psychrometer. Time of exposure was determined

using a stop watch (93).

65

lanipulation of the leaf specimens in the hot air
stream required the use of tweezers to handle the leaf (9F),
and pliers to cpen the spring clasp. Time of exposure was
taken from the instant the leaf was placed in the center
of the air stream. Hanipulations were rather cumbersome,
requiring as long as three seconds although they did not
adversely affect the air flow or exposure of the leaf to
the air flow. At ultra high temperatures, where the allow-
able time of exposure was less than ten seconds, the leaf
was fixed to a needle holder by piercing in a loose weave.
The holder was then turned by the fingers to provide the
desired rotation. The leaf could then be inserted into and

out of the air stream in approximately Q second.

Damage Point Tests
Each test was run at constant temperature and constant
air flow. This required a steady state condition for the
heating apparatus. To achieve this, an equilibration period
of approximately two hours was required. Since the
apparatus was dependent on an electrical source, any slight
fluctuation in voltage would result in a slight change in

temperature.

Discharge temperatures depended on the rate of air
flow through the heater for any given power input. To

maintain a constant discharge velocity at all temperatures,

66

this required a variable mass velocity at the inlet. All
tests were run with an air flow of approximately 60 fpm,
the theoretical fluidizing velocity for fresh cut leaves.

The leaf was rotated in the heated air stream at about
45 rpm. This simulated the action that a leaf might under-
go in a freely suspended fluidized state. Stationary leaf
drying tests did not prove to be satisfactory for two
reasons. 1. The downstream side of the leaf was shielded
from the air blast, and was not exposed to the same heat
transfer condition; and 2. The change in Young's modulus of
elasticity (as the leaf progressed from the fresh cut- to
the wilted - to the dried) allowed the leaf to be affected
by the vertical air flow and to curl up similar to the action
of a bimetal thermostat.

Leaf samples were plucked individually from the plant
at the time of the test. Field plants were cut at random
and kept fresh by immersing the stems in water to reduce
the effect of wilting in the laboratory. All minor stems
were excluded from the leaf proper.

After procuring the sample, it was weighed on the
precision analytical balance (Figure 10) to obtain the wet
weight and immediately subjected to the test. After the
drying test, it was weighed again to obtain the dried weight.
Subsequent drying of the single leaf for a prolonged period

at the test temperature was accomplished until no further

67

moisture loss was noted. Final weighing established the
bone dry weight.

Tune of exposure to cause visible damage was the basic
objective of the test. To determine this, an initial period
was selected which was estimated to not damage the leaf.

The tests were run for a given leaf to a particular moisture
content. Subsequent leaves were dried during periods which
were gradually increased by 1 second intervals. When the
time of exposure was established such that all moisture was
removed, further exposure to the drying medium would result
in dry matter loss and eventual browning over the entire
leaf. This point was one criterion of the damage point
definition.

At extremely high temperatures-above 500°F, the leaf
was observed to char at the tips and edges. Until the
damage points were approached, replicated drying tests were
not conducted. In the region of the damage point, as many
as ten replicates were made to fix this point, allowing
for the variation in moisture content of the leaves, effect
of leaf position on the plant, size of leaf and individual
plant variations. At extremely high temperatures, the
charring point was quite obvious, requiring as few as five
replicates.

Four hundred and twenty-five tests at 13 temperatures
ranging from 281° to 819° were conducted to provide the

data for this phase of the research.

68

Sphericity-loisture Content Relationships

Particle sphericity has been designated as a parameter
of considerable interest in fluidized bed drying. Sewever,
for flake particles the usual methods for determining this
criteria are ineffectual. The problem increased in com-
plexity when considering a flake particle with changing
moisture content.

The necessary information to determine the sphericity

is revealed by the relationship:

qbs - v 2/3
0.205A

Ieasurement of the projected area of a leaf by a
planimeter has been described in an earlier section. The
total area,A, for the particle would be twice the projected
area added to the area of the edge around the perimeter.
The perimeter area is of second order magnitude in compari-
son to the total surface area and was neglected.

volume indications would be possible using area and
thickness data if they were reliable. Because of the
veinous structure of the leaf, an average thickness is
difficult to obtain; particularly so for the subtle changes
when the moisture content is gradually reduced.

As a consequence, another technique was considered
using Archimedes principle of displacement. The requirements
for the displaced liquid were such that a leaf would be

virtually unaffected by absorption, adsorption, or chemical

69

reaction. Kerosene had been used by Kennet (1950) for
alfalfa density determinations. Single determinations were
required and then the material was discharged with little
need for concern with post measurement adhesion. However,
with the alternated volume determination and drying tests,
this liquid would not lend itself to this procedure. As
a consequence, mercury was selected.

Criteria for the volume measuring apparatus were: 1.
easy access with an Opening large enough to accommodate a
leaf without crushing or curling; 2. a range of measurements
estimated to be from .005 to .10 ml; 3. simplicity of Opera-
tion to effect relatively quick reading. No commercial
sources for this specialized equipment were found.

. Two displacement techniques were considered. The first
attempt made use of an open vessel filled to the extreme
limits of the meniscus such that the slightest addition
resulted in a break in surface tension. The volume spilled
into a receiving receptacle was supposedly equal to the
volume of the object being measured. For relatively large
volumes, this method was reasonably successful. But for
extremely small volumes, such as a single leaf with minute
changes in volume from a change in moisture content, the

method was entirely unsatisfactory.

 

70

The method arrived at consisted of immersing the leaf
in a vessel sealed with a tightly fitting stopper out of
which the displaced liquid could be volumetrically measured
to the accuracy required. The apparatus was designed and
developed with the use of standard chemical glassware, altered
and joined by glass blowing techniques. The finished
‘apparatus is shown in Figures 11 and 12. A schematic diagram,
Figure 13, shows additional details of construction.

A 25 ml flask with a standard tapered ground glass open—
ing at the top provided for a receiving vessel with the
required ease of access to accommodate a leaf. A take-off
at the bottom was added, then fused to 3/16" glass tubing
bent at two right angles forming a "0", into a short length
of small bore precision glass tubing (0.001 mm dia.). At
the top of this leg of the "U", a reservoir (4" drying tube)
was fused to form a monolithic glass assembly. Early models
were joined using "Tygon" plastic tubing but proved unsatis-
factory.

Into the ground glass mouth, a mating hollow stopper
was fitted, out of which another length of 0.001 mm tubing
was fused so that this stem provided the means for measuring
the volume. Actually, a 0.1 ml pipette was used, which
provided the built-in calibration required. Further sub-
division was made by dividing the 0.01 ml division into ten
parts so that readings of 0.001 ml could be discerned.

71

A small stainless wire clip was bent to fit snugly
into the hollow stapper so that when the stem was removed,
the clip came out also. This assembly to which the leaf
has been fixed can be seen in Figure 11 lying alongside the
apparatus. The ground glass seat provided for a precision
placement to an accuracy of i .001 ml.

The scheme of operation required that the mercury filled
apparatus be pumped out of the reservoir under pressure
until one leg was indexed to a line on the small bore tubing.
The quantity of mercury in the system was adjusted such that
the volume would show on the leg of the measuring stem out
of the receiving vessel. A precise register was found to be
unimportant, providing that the initial level was noted.

The system would then be allowed to seek its own level,
about half way down into the flask and the stem assembly
removed. The leaf whose volume was to be measured was
clipped into position, held at the bottom of the leaf so
that natural buoyancy tended to keep the leaf in a distended
position. The stem was refitted into the joint, and the
level of the mercury was again indexed as before. Because
of the added displacement of the leaf, the level of the
mercury in the measuring stem rose appropriately so that

the difference in levels was a direct measure of the volume.

 

72

The liquid levels were manipulated by a hand squeezed
aspirator bulb fitted into the mouth of the reservoir.
First models employed the use of a stop cock in conjunction
with a conventional aspirator bulb to hold the liquid
levels while readings were taken. However, a much simpli-
fied maneuver was possible with the use of a valved aspi—
rator bulb ("Propipette" T.l.). By pressing the various
valves marked S, R and A with the fingers while the hand
squeezed the bulb, an effective control of the system was
possible after much practice.

The instrument was surprisingly effective and capable
of detecting extremely small changes in volumes. Two
difficulties developed in the measurement of leaves. Re-
peated measurements inevitably contaminated the mercury such
that a coating of dirt appeared on the glass parts. This
was particularly troublesome in the small bore tube sections,
culminating in a separation of the mercury. This required
the periodic removal of all the mercury and complete
cleansing of the glassware and the mercury. A gold ring
mercury cleaner was found to be particularly helpful in
this periodic purge, since the dirt was non-metallic and
could be adequately removed by this method.

A second, more serious difficulty, arose from the
effect of the relatively high hydrostatic pressure developed

by the head of mercury in the measuring stem. This was

73

variable in height, averaging about six inches Hg. This
is approximately 7 ft. HQO and unquestionably affected the
leaf cells by crushing them to a small yet underdetermined
degree. However, the measured volumes very nearly checked
with the calculated volumes from the area-thickness data.
The method required considerable delicacy, especially for

leaves at low moisture content.

Sphericity Determination Tests At Various
loisture Contents

The objective of these tests was to determine the
sphericity of the leaf, and how the sphericity was affected
by a continual reduction in moisture content. A series of
tests were run on various sizes of alfalfa and birdsfoot
trefoil leaves. Samples were procured as before, but with
ten replicates examined simultaneously. The procedure follows.

A leaf was plucked from the plant, weighed on the
analytical balance; its area was determined by means of three
planimeter tracings; its volume was determined as outlined
above. The leaf was removed using tweezers taking care to
remove any adhering mercury droplets. The leaf was re—
weighed to assure that no mercury had clung to it and was
then placed in a drying receptacle with a screened bottom.
This receptacle was subjected to a convective heat source
at approximately 160°F. Slight pressure was applied by a

small wooden block placed on the leaf while drying to prevent

74

it from curling as it dried. Thus, upon removal after a
predetermined period, the procedure could be repeated.
The flat configuration was particularly critical for area
determinations, especially at the lower moisture contents
when the leaf became extremely brittle.

A second leaf was similarly treated and so on until
at the end of the 10th leaf, by which time the first had
reached a lower moisture content level at which the whole
procedure was repeated. The moisture level was brought
down to the equilibrium moisture content at 160°F. This
level was attained in six steps, with measurements made at
each step.

It is interesting to note that volume measurements
were made even with the extremely brittle dried leaves with-
out fracturing the leaf. Extreme care in handling was
exercised, but the submersion of the leaf in mercury appar-
ently was so gentle that no adverse affects were noted.

Two hundred and seventy—three volume determinations on
60 leaves were made for moisture contents ranging from 300%
d.b. to nearly zero moisture content to provide the data

for this phase of the research.

75

lass Velocity-Temperature—loisture Removal
Relationships

Certain operating parameters of fluidized bed drying
of alfalfa, referred to by Schrenk (1959) as a pneumatic
type alfalfa dryer, have been established from the pilot
process standpoint. The parameters have been established
for the chopped whole plant substantiating that stems must
be removed at some stage in the process. The behavior of
leaves only and their fluidizing velocity requirements as
they lose moisture remained to be established.

To observe this behavior, a transparent drying tower
was considered to be ideally suited for these studies.
Since the temperatures of the drying process were to be
considerably higher than the melting point of Acrylic plastic,
Pyrex glass pipe was selected as the experimental tower.
This equipment is readily available with expanded ends to
accommodate standard metal flanged joints; glass fittings
such as elbows, Tees and "U" bends are standard equipment.
The 1% in. I.D. tubing was selected for reasons of compati-
bility with the heat source apparatus available from another
phase of the research. The apparatus was designed to comply
with laboratory techniques as compared to pilot plant pro-
cedures. Available ceiling heights in the drying labo-

ratory limited the effective height of the drying column to
13 ft.

76

The completely assembled apparatus is shown in Figure
14 with clarifying details shown in Figures 15, 16 and 17,
and Table 2.

The heat source previously described was positioned so
that the discharge had a vertical orientation. The inlet
was fitted with a variable rectangular area orifice. A
sliding valve in the inlet duct was provided with a vernier
screw adjustment (Figure 16D). The outlet was fitted with a
i in. dis. copper tube bent 90° so that smoke could be
injected through an orifice in the side of the duct in the
direction of flow (170). Copper mesh screening, l/16 in.,
was fitted at the discharge to prevent leaves from accidently
dropping into the heat source chamber. Conventional flui-
dized bed bottoms are of porous vitreous clay, but this is
for purposes of small diameter particles and was not
considered for this application.

Aluminum glass-pipe flanges permitted tight fitting
joints. The assembly was altered slightly so that asbestos
gaskets were substituted for rubber ones normally used.

At the three foot level joint, an air lock sample injector
was installed. The sample injector is shown disassembled
at the bottom of the drying tower (Figure 16A). There are
three essential parts to this assembly: 1. the aluminum
sample port, 2. the sheet metal sliding air valve, and 3.

the air lock side wall. These parts are shown as items

77

No. 16, l, and 2 in Figure 15. The assembly was bolted
between two flanges. Operation of the air lock required
the hand assembly of Nb. 1 and 2 forming a scoop into which
the sample was placed. This loose assembly was inserted
into the sample port and No. l was withdrawn leaving No. 2.
in place to form the inner surface of the tube. Thus, the
sample became fluidized immediately without blowing back
out through the sample port. The assembled sample port is
shown in Figure 17E.

The drying tower was completed with the addition of a
ten foot section of pipe. At the top, a 6 in. U bend was
installed to retrieve the dried sample. A return pipe was
installed so that the sample after drying was guided to the
sample collection containers below. An Opening to provide
for the exhaust of the moisture laden air was installed at
the base of the U bend on the down leg. Exhausting air
in the up leg was impossible because the reduction in
velocity would be such that the dried product could not be
carried up over and into the return line. The exhaust
Opening was made from 1/16 in. screening fixed in a
cylindrical frame such that the free area was twice that
of the cross sectional area of the pipe. A machined
stainless pipe spacer on the up leg balanced the dimensional

requirements of the system.

78

Since the apparatus was constructed as a semi-permanent
fixture in the drying laboratory to accommodate future
research activity, it was decided to partially insulate
the down leg by encasing the glass pipe in a 3-7/8 in. I.D.
acrylic plastic jacket 1/8 in. thick (Figure 14A). This
would alter the heat transfer characteristics of the drying
tower and provide for further variation in air velocities
when the roles of the two legs were interchanged by merely
interchanging the heat source and U bend fixtures.

Dried samples were collected in 10 paper containers
fixed on a rotary disk holder. By this method, samples could

be collected at various times to determine the drying rates.

Instrumentation of Drying Tower

Instrumentation of the tower provided for temperature
and velocity determinations. Copper-constantan thermo-
couples (No. 24 wire) were imbedded in the glass walls at
the 6 in. level and at four foot intervals thereafter.

Ho1es were drilled into the glass walls allowing for the
T.C. junctions to be mounted just at the inside surface.
Permanent mounting was effected with the use of Saureisen
cement, a liquid paste which dried to a heat resistant
ceramic. The thermocouple leads were connected to a rotary

switch leading to the potentiometer (Figure 14H).

79

Velocity measurements were made extremely difficult
by reason of the low fluidizing velocity requirements, and
the high temperatures of the flowing fluid. After exploring
many possibilities, smoke injection was selected as the
device by which velocity measurements could be most quickly
and accurately made. The original scheme involved two #930
RCA photo-electric cells mounted at the top and bottom of
the drying tower so that a light beam from a D.C. lamp would
shine through the glass onto the sensing element. Any
interception in the light beam would cause a signal to be
transmitted through an amplifier and appear on a constant
chart speed oscillograph. This apparatus group is shown on
Figure 8 (E and F). An acid vapor smoke gun was used in
early trials, but upon the addition of heat, the vapors were
found to not condense sufficiently to be detected by the
photocells. As a result, a black powder, "Norit A"-—an
activated charcoal product, provided the most effective
"smoke".

Theoretical considerations predicted the prominent role
of a velocity divergence, and this in fact was observed by
the noticeable deceleration of the smoke plug in the drying
tower for heated air flows. Yet, when the velocity for an
isothermal-unheated air flow was observed, there was no
Perceptible deceleration. As a result, it was decided to

increase the number of photocells to four and finally to

80

six to provide five separate velocities along the 13 ft.
drying tower. From this information, the velocity divergence
3; could be obtained from a plot of the velocity versus

axial distance.

Several troublesome problems arose when the number of
photocells was increased from two to six, which are worthy
of note for future users of this system. The electronic
circuitry best suited for this type of system is a parallel
connection of the tubes, as shown in Figure 18. However, as
the number of cells was increased, the total resistance of
the circuit was decreased and the signal became less and
less pronounced requiring greater amplifications. Signal
'output was improved by shielding the photocell with a small
piece of mailing tube, covered on the top,with a 1/8 in. hole
in the side. The shield could be rotated to increase or
decrease the light intensity so that the system could be
balanced. Even with these measures, the total amplification
required was in the order of 106:1. Pick—up of stray
signals was particularly troublesome at this level and even
with the most elaborate precautions taken, e.g. metallic
shielding of all wire leads, 60 cycle hum was difficult to
obliterate. Success was finally achieved after a tenacious

trial and error effort which yielded the correct combination

Of capacitors in parallel and in series with the primary

circuit.

81

Further difficulties in the behavior of the powder smoke
signal became apparent with the addition of more photocells.
To more clearly understand these problems, consideration of
the nature of the signal must be made.

The volume of the powder injection was approximately
0.025 cu. in. mixed with 4 cu. in. of air as the propelling
medium supplied by an aspirator bulb squeezed by hand (Figure
17H). The injection velocity was calculated at 13 ft/sec.
This was designed to approximate the fluidizing velocity.

The plug of smoke emerged as a fairly well defined cylinder
about 4 in. long. A particle size analysis of the powder

- was made and it was found to contain 81% by weight of particles
measuring less than 0.05 mm. Terminal velocities of the
powder were calculated by Newton's equation, verified by

means of a nomograph (see appendix) and observed in a re-
fracted light source to be approximately 0.1 ft/sec. with
close agreement.

As the smoke intercepted the first photocell, a clear
sharp signal appeared on the oscillograph; paper speed was
constant at 25 cm/sec. However, as the smoke progressed
uP the tower, particles in the boundary layer decelerated
sharply, some falling downward, as the bulk continued upward.
The effect was that the signal trailed off, becoming weaker
as the front continued upward, and finally towards the t0P

of the tower, the signal to two photocells was affected.

82

 

I‘ 6— T 930 RCAphotoceIg

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

IL
71
‘7’rL
W} IZMfl
/ y lit—I'm»:-
// 90 VOLT k
“J4 /, n BATTERY 5W
,-._../ . l
H ‘ A .OH/‘f %F—
M/
/ o_
1 / In
_-/ AMPLIFIER
I06: I
9 9
. I g
_ /" n ‘ ‘

 

 

————"’/‘/
,1 71%

 

OUTPUT

 

 

 

 

 

 

 

a ”5”” TEST#7b
N
“we/T V9/l7/63
_ ;/
‘r—N—v—fi/ N
“LIAM—r” e
\.K\ jk-__,
L's—M

 

OSCILLOGRAPH

 

24.9
M UNITS/SEC

PAPER STRIP
OUT PUT

O

 

 

FIGURE l8 PHOTO CELL CIRCUIT 8I OUTPUT

83

The net result was very slight signals for the 4th, 5th and
6th photocells. This was partially compensated for by
balancing the light signals to decrease or increase the
light signal as needed. A sample signal output for which
the difficulty of interpretation is apparent is shown in
Figure 18. The tangent points of the signal trace was
finally considered the most nearly representative of the time
required for travel of the smoke plug from one cell to the
next. Various appurtenances prevented the equal spacing of
the photo cells and this was taken into account in the
velocity calculations. For each velocity determination,
three replicates were taken as part of the data. Apart from
the drying tests, the natural draft velocity characteristics
were determined for several temperatures at the tower inlet.
In addition to the previously mentioned temperature and
velocity determinations, the temperature profiles at the
2-3/4 ft. and 13 ft. levels were taken for each test condition
as well as at the center of the pipe at the "0" level of
the drying tower. A thermocouple probe with 3 in. of lead

wire parallel to the stream was used for this purpose.

Drying Test Procedures
As already suggested, this small bore drying tower was
incompatible with the space requirements for fluidized
alfalfa leaves. It was found that i-in. diameter disks

84

punched from alfalfa leaves were much more desirable from
a fluidized bed behavior standpoint. This was correlated
with the experiences of Leva (1957) and Gregg.(1956), al-
though no information was available to accurately predict
the d/D requirements. As a compromise, birdsfoot trefoil
was selected to most nearly approach the miniaturization
desired. Batch drying tests were considered because of the
control made possible by the use of small quantities as
well as the opportunity to observe the drying patterns at
various fluidizing velocities.

The test batch sample consisted of 100 trefoil leaves,
weighed as a mass on the analytical balance. The sample
was injected into the hot air stream through the sample port
as previously described. At this instant, time was started.
The sample either sank immediately forming a fixed bed or
became fluidized, depending on the combination of velocities
from the natural draft of a particular source temperature
and the superimposed forced air velocity. Where the leaves
sank immediately, fluidization was not attained until after
sufficient moisture removal had been effected such that
the partially dried flakes eventually fluidized spontaneously.
Or, if the velocity was sufficient to fluidize immediately,
the leaves gradually progressed up the tower as an agitating
mass, with some dropouts. The initial behavior of the

sample was noted as well as the time when the first leaves

85

began to drop down the recovery leg. From this, the average
velocity of the sample as it progressed up the tube could
be calculated, if required.

After the first dried leaves appeared in the first
sample recovery container, ten seconds elapsed before the
rotary holder was advanced to the second container and so on.
The test was ended at the end of a 4-min. period with a
total of twelve samples taken. This meant that after two
minutes of sample recovery at lO-sec. intervals, all leaves
that would come out had done so. Leaves in each of the
containers of the segmented dried sample were then indi-
vidually counted and weighed. Oven drying followed from
which the dry weights were determined. Thus, moisture con-
tents could be determined for purposes of examining the
drying patterns.

Three replicates for each temperature-velocity com-
bination were run. Twenty-four combinations were selected
from which twelve produced immediate fluidizing velocities.
A total of 72 drying tests were conducted for a range of
inlet temperatures from 200 to 300°F. The temperature
range was limited by the apparatus available. However, the
data obtained provided information which had not been

reported heretofore.

86

RESULTS AND DISCUSSION

The sensitivity of alfalfa leaves to damage at high
temperatures is one of the basic limiting criteria for
the design of a high—temperature, short-time process such
as fluidized bed drying. Determination of these relation-
ships on a limited scale has been accomplished as a minor
but important phase of the research. The results are
presented in a manner similar to those of Headley and Hall
(1963) with the necessary modifications to comply with the
needs of the product under investigation.

Certain visual observations have necessarily become
of qualitative importance and should be mentioned as part
of the results. Damage to alfalfa leaves manifests itself
as either: 1. a visible charring of the perimeter while
the inner areas remain in a green, wilted, partially dried
state or, 2. a relatively uniform moisture removal over
the entire leaf area to as nearly a bone dry state as
possible, after which continued exposure to high temper-
atures results in dry matter loss with subsequent break-
down of nutritive compounds in the product. The various
types of damage can be seen from close inspection of Figure
5. Figure 5A illustrates leaves which have been dried at

a temperature of 500°F for 15 seconds, which were removed

87

from the drying medium at the 25% m.c.d.b. level. Below the
sample are freshly plucked leaves for color comparison,

This important result clearly shows desirable color fixation,
indicating a preservation of the nutrients in the product.
This is typical of all drying results for leaves at temper-
atures below 500°F. It was possible to remove moisture to or
below the 70°F equilibrium moisture content approaching the
bone dry state without visual damage.

For further comparison, Figure SB shows leaves dried from
the fresh cut stage down to the same equilibrium moisture
content at 140°F over a conventional long time drying cycle
of approximately 8 hours. The dull olive drab green clearly
indicates a loss of pigmentation indicative of carotene and
other vitamin destruction. The severe curling of these leaves
is contrasted with the comparatively unaffected leaves of 5A.

Various types of damage are shown in the remaining
samples. The sampling in Figure 5C illustrates the effect of
continued exposure of leaves to temperatures of 300 to 500°F
with a definite browning over the entire surface of the leaf.
Figure 5E illustrates leaves dried at higher temperatures
with blackening of the perimeters clearly evident. At the
time of the tests, while the edges were brittle and charred,
the inner area was still at a wilted, partially dried

stage. Finally, Figure 5D shows leaves which can best

88

be described as being "pOpped". The ultra-high temperatures
at which these leaves were exposed for a relatively long
time (e.g. 10 seconds, after damage first occurred at 3
seconds) caused the vapor pressure to build up within the
leaf so fast that the lower epidermis actually parted from
the leaf cell structure. It might be added that later
research reported by Whitney and Weeks (1963) disclosed

that in all probability, the stomata of the leaf surfaces
were closed for these tests effectively sealing the epidermis
so that normal mass transfer of the plant held water was

not possible with the fast buildup of vapor pressure. All
these occurrences are to be expected from basic consider-
ations of the heat and mass transfer phenomena, and could
probably be predicted from mathematical analysis. This

was not attempted at this stage of develOpment of the

research.

Damage Point Relationships
Graphical presentation of the data is presented in

Figure 19, in which the quantitive relationships are clearly
shown. As stated in an earlier section, damage points at
thirteen temperatures were investigated, and each point

was established with from 5 to 10 replicates. The data
plotted on semi-logarithmic graph paper as two straight
lines with a change in slope at 500°F. The need for the

statistical determination of a regression line was

89

IO O 0F’/ /,x
//

900%;

8 O O of;
700~ \

  

\
. ~\\_\

\
. \ - \
opt—s-..- -_;¥‘b‘_¥:—l

SINGLE ALFALFA LEAVES FROM MIDDLE OF PLANT.
DAMAGE AREA— lOO°/oDRY MATTER OR VISIBLE
CHARRING, (WHICHEVER OCCURS FIRST)

GOOI- ‘2‘

     
 

-.0069 O
500 I"

\\
4OOL- / \\\

300 —
25% MC. DB LINE

FROM EXTRACTION OF \ \
DRYING RATE DATA '

ZOO

 

a
\T

I—
b
P
I'-
b-
I—
P
I...
I.-

I00

90
80
7O
60
50
4O
30
20"
IO -

 

 

1 l l 1 l J 1 #L l i 1

IO 20 30 4O 50 60 7O 80 90 lOO IIO
DRYING TIME- e- SECONDS

FIGURE I9 DAMAGE POINT RELATIONSHIPS

MOISTURE REMOVED,PERCENT DRYING TEMPERATURE —- T- F°

 

 

9O

obviously not necessary in this rare case. Results were

surprisingly comparable to those of Headley and Hall (1963)

in their work with single kernels of corn, including a

similarity in the critical temperature at the break in the

curve. The equation of the damage point graph relates

the temperature and drying time in the following manner:

-.0069
T - 580 e

Where T - temperature, °F 5 500

6 - time, sec.
e - base (2.718) of natural system of logarithms

Directly below, and parallel to this line, another
relationship can be seen which has been obtained from the
data from which the time to attain a 25% moisture content

(d.b.) level can be predicted for the range of temperatures

and times explored.
The graph at the bottom half of Figure 19 shows a plot

of the percentage of moisture removed before damage versus

time, clearly illustrating the effects of drying at temper-

atures greater than 500°F. From this presentation, it is

possible to determine the effects of any drying stage.
Starting on the Y-axis at any design temperature and moving

horizontally, the damage point on the curve establishes
the allowable exposure time. Tracing vertically to the

second graph, this point determines the percentage of

91

moisture removal on the ordinate of the lower graph. The

value of this presentation in process stage design consider-

ations is self evident.

Drying at Ultra High Temperatures
As a by-product of the data obtained to determine the

damage point relationship, certain other drying data were
recorded as related to the time of exposure of the single

leaf to air at the various temperatures as the damage point

was approached from the fresh cut stage. This portion of

the experiment was not statistically designed but several

other relationships are revealed: 1) drying curves for

alfalfa leaves, 2) moisture content ratio-time relationships
and 3) a plot of drying constants versus temperature.
A well defined family of drying curves was obtained

as shown in Figure 20 for which the percent moisture content
(dry basis) was plotted versus time, 9. A table of these

results follows:

DRYING CURVES FOR ALFALFA LEAVES AT ULTRA-

 

 

 

 

TABLE 3.
HIGH rsspsaarunss
[Tamperature % Moisture Content, d.b.,
°F ' _ Ae-kO
_ . . 73 __ _
281 2516-0276
311 234e--0339
379 3309-.0589
427 488e‘°0999
455 2450e‘°2789‘
524 1375e'.2679

 

 

92

30E
.5: k4 mm><mm 534.144 10.... mm>mno oz_>mo ON mm

a o u:—#
mozonmm o ow 0n ON 0. 0

mm mahdmmac‘w ._. 10.1I4m

       

 
 
 

oo. om om 0.1. cm
.5 R...
.I.
.m w
m
S
l.
m
23
3
m
8...

 

O
2
0
IO
N
Iq‘nTw ' M 3

 

 

 

 

 

93

Theoretically, at time "zero" the percent moisture

content at all temperatures should be consistent with that

of fresh cut material of approximately 300% d.b. HOwever,

as can be seen from the equations of the curves, discrepan-
cies are exhibited, possibly due to the effect of the constant ‘
drying rate periods at the higher temperatures.

Further analysis of these data in a more conventional

way related the logarithm of the moisture content ratio,

I — , to the drying time, 0. The regression lines at the

- Ie

various high temperatures were determined from the data.
The equilibrium moisture content, He, was found to rapidly
approach zero and was neglected in the calculations of the
moisture content ratio.

This was permissible as will be shown using data from

Bakker-Arkema (1962) to evaluate the constants c and n in

equation (2-1) from Hall (1957):
l-RH - e'c'l'len

From Bakker-Arkema
RH -0.9, T - 120°F (580°A), He - 22% d.b.
RH -0.8, T - 120°F (580°A), lg - 17% d.b.
two equations can be written:
1 - .9 - e-c(580) (22)n
—c(530) (17)n

the unknown constants can be

Thus,

1 - .8 - e
Solving simultaneously,

solved as follows:

94

c - 0.0000547
n - 1.386
Finally: 1-an - 6'0-0000547Tl61-336
From the psychrometric chart at high temperatures, the
relative humidity is seen to rapidly approach zero above
212°F. For example, at a temperature of 250°F, the relative
humidity was extrapolated at .001. Solving for Me in the
final expression:
1_.001 _ .999 _ e-o.0000547(710) 261-386
Figure 21 shows the family of regression lines for the
various drying temperatures. As before, the curves fail to

intercept the Y axis at the expected common initial moisture

content ratio M - Mg - 1.
lo - Me

A tabulation of the equation of the curves is as follows:

TABLE 4. MOISTURE CONTENT RATIO, TIME, AND TEMPERATURE

 

 

RELATIONSHIPS

Moisture Content St d d
Drying Ratio _ an ar

Temperatures M - !§ - Be k9 Deviation

’1 lo - lo izcr

1.469e-050335’ 0.468
236 1.8476'0-0439 0.694
311 4.284e'0-0659 0.911
379 11.3599-0-1199 0.776
455 4.7126'0-1359 0.455
515 22.8740-0’4246 0.943
524 36.966e‘0-3909 0.579
570 8.2486’0-3109 0.382
640 11.5886‘0-4fgg 0.321
661 3.289e'0’3 0.336
761 1.6006‘0-3099 0.821
819 1.5156‘0-3366 0.448

 

 

 

 

 

95

 

I. ‘
.80.
.70.

.401

 

 

 

M-M

 

.06-
.05' (‘
28I°

”03. 52491 “155° 379° 3II°/‘ 295.

 

MOISTURE CONTENT RATIO,

 

 

 

0 I0 2'0 3'0 40 5'0 6'0 70 86 90 I00
TIME,9.SEc0NDs
FIGURE 2| RELATION OF MOISTURE CONTENT RATIO TO
TIME FOR DRYING ALFALFA LEAVES AT

ULTRA-HIGH TEMPERATURES

96

A semi-log plot of the drying constant, k, reveals
the relationship shown in Figure 22. The logarithm of the
drying constant, k, was plotted versus the temperature, T,
as a straight line up to 524°F. This corresponds to the
temperature at which drying occurred without visible charring
shown in Figure 19. At higher temperatures, charring
occurred at the perimeter of the leaf while the center area
was at a much higher mOisture content. It can be seen that
k remains at a nearly constant level, although much more
experimentation is required to verify the various implications
of these preliminary results.

The equation of the curve from 281°F to 524°F has been
determined as follows:
k - 0.002660-0097T
This relationship can be substituted into a general

expression relating the three variables:
a _ u _ Be-O.0026900-0097T
uo-ue
Further analysis is not attempted here because of the
limitations of the data available, and the deviation from
the basic objective of the experiment design. However, the
results are interesting and worthy of note at this point

in the dissertation.

DRYING CONSTANT , k

97

.401

.30.

CHAR DAMAGE OCCURS BEFORE
DRYING IS COMPLETE AT TEMPS.

ABOVE 500° F

 

.I04
.09‘

=0.0026 e0.0097 T

.07‘
.09

ed

.04‘

 

 

300 400 500 600 700 800
TEMPERATURE ,T, °F

FIGURE 22 RELATIONSHIP OF DRYING CONSTANTS TO AIR
TEMPERATURE FOR ALFALFA LEAVES

98

Sphericity-Moisture Content Relationships

The basic relationships involving the mass rate of flow
for fluidization, Gm, voidage,E mf, and particle diameters, .
Dp are dependent on the sphericity,qbs, of the particle.

This parameter is usually unavailable for flake particles,
seriously limiting use of the equations which allow deter-
mination of the mass velocity at the onset of fluidization.
Equally important is the variation in sphericity as the parti-
cle decreases in moisture content, possibly with reduced
physical characteristics. In the design of the process

stages this becomes of primary concern. This phase of the
research has been devoted to the investigation of the sphe-
ricity parameter.

Classification of leaves was as stated in the preliminary
investigations. Values of sphericity for large, medium, and
small alfalfa as well as birdsfoot trefoil, were determined
according to the procedure previously described. These were
plotted on cartesian coordinates with moisture content as
the abscissa shown in Figure 23. Considerable scatter of
the data was prevalent, but statistical analysis permitted
determination of the regression lines. The equations of

the curves are tabulated as follows:

99

TABLE 5. SPHERICITY RELATIONSHIPS FOR LEAVES AT VARIABLE
'IOISTURE CONTENTS

 

 

, f§tandard

Leaf Classification Sphericity, Deviation,
and type (tbs - mQ.C.) + b :0”
I

Large alfalfa 0.0000753(M.C.) +.1025 0.0081
Medium alfalfa 0.0000631(M.C.) +.1055 0.0109
Small alfalfa 0.0001305(M.C.) +.1637 0.0314
Birdsfoot trefoil 0.00001 (M.C.) +.1799 0.0215

 

 

 

 

 

 

When considering the magnitudes of the coefficient m,
it can be shown that the influence of the moisture content
is minor (less than 20%). This influence of itself would
have little effect on the process, and the sphericity might
be considered nearly constant over the entire moisture con-
tent range.

An error analysis of the experimental techniques has
been made pointing out certain limitations of the experi-
mental technique. The accuracies of the apparatus were such
that the volume of the leaf could be determined to 10.001 ml
and the area to iO-005 in2. However, because of the natural
shrinkage of the leaf, the volumes and areas decreased
substantially such that the percentage of error increased
sharply as the leaf lost moisture content. Also, when con—
sidering the large differential between the sizes of small
and large alfalfa leaves, the accuracies were influenced in-

favor of the larger objects.

 

100

am“: FZUPZOU MEDPQOZ o\o mama“; PGEMImm MN szoC

3.3. 2;. 23.200 Enema: .\.

OON 00.
a 1 A

00m
_ _

soon 30. .+ 8.3 3880.. me $343 <55: woman
mo.o.ammo..+l.u.zv fiooooo. . we 83m... <anan .2285
38.38.... 1.0.5: 3280. "me 323 453.3 .335

98.“ 8:162:88. "we gob”: Ecumosm

sees

 

 

 

'||I| L

Ill.dllslb.sll.lll.mll_sII.sII_mII_III III III_III III III

C?

SI

‘52

2

m_.

 

 

30 ‘AiIDIHBHdS

I:

0..

ON.

Since 0 - V’Z/3 ,
“07255

A0. 59;: (AV) +55:- (AA)

Ad’s .- 2/3 :1/3
0.205A

101

(AV) -

the maximum erroraOs, is:

QQA)

0.205A2

The analysis for the results is tabulated as follows:

 

 

 

TABLE 6. ERROR ANALYSIS OF SPHERICITY RESULTS FOR FRESH CUT
AND DRIED LEAVES

Lei?‘

Classi- Shrinkagel Maximum % Error Statistical
fication va/vw TSUfiirm.cdeb. TfiiTm.c.dTb. Error
Large
Alfalfa .28 .i 2.4 .i 6.8 i 7.8
Medium
Alfalfa .33 i 3.8 110.2 :10.3
Small
Alfalfa .35 i 7.6 :17.5 ‘119.0
Birdsfoot
Trefoil .42 i 5.57 i15-2 ‘112.0

 

 

 

 

 

 

From this analysis,

it is concluded that the error of

the experimental technique developed increases as the area

and/or volume of the object decreases.

Further, error must

be expected because of the influence of the hydrostatic

pressure in crushing the leaf cells.

This can only be

analyzed by microscopic inspection by which comparisons can

be made of cells before and after stress.

However, it was

found that the immersion technique was gentle enough not to

 

102

affect even the most brittle dried leaf. MicroscOpic veri—
fication was not made at this time.

The sphericity determinations indicate that the smaller
leaves are more easily fluidized. For all leaves, the
voidage émf can be expected to approach unity. Positive
evaluations of voidage could not be determined because of the
small scale of the apparatus, but as a result of the tests,

the process can be identified as a dilute fluidized bed.

Mass Rate of Flow, Temperature, and Moisture
Removal Relationships

Investigation of the fluidized bed drying process has
been conducted using laboratory scaled apparatus and tech-
niques described earlier. This has differed considerably
from the experimental techniques reported by Schrenk, whereby
a continuous pilot process was required to yield meaningful
design parameters. The velocity requirements of the drying
particle by Schrenk in terms of the mass velocity and the
results are presented as a regression line with a dimension-
less parameter, !Q - l as the ordinate and the mass velocity
per unit area aswihe abcissa. The data available were for
inlet temperatures greater than 720°F, considering freshly
chopped alfalfa plants as the feed. Provisions were pro-
vided to separate out the stems.

It seemed desirable to obtain information which would

establish some relationships with the inlet temperature and

which would be more appropriate for leaves only. With these

 

III‘
ll

 

 

103

objectives, the tests were conducted at variable inlet
temperatures commensurate with the limitations of the small
bore laboratory apparatus. The results are superimposed on
Schrenk's information to further expand the limits of the
available parameters. Regression lines were obtained by
statistical analysis and are presented in Figure 24. The

tabulation of the relationships follows:

TABLE 7. MASS VELOCITY, TEMPERATURE, AND MOISTURE
REMOVAL RELATIONSHIPS

 

 

 

Inlet Siandard
Temperature 19 - 1 Deviation,
°F .11 ,Sicr
200° 1018.78(vmasa)-5.5242 °1535
250° 1014-1°(vmass)'4-3269 .1244
300° 1011-85(Vmass)‘3-775° .1199

720° 1019»5 (Vmass)‘5°5
(Schrenk

1959
estimated)

 

 

 

 

 

Analysis reveals a family of curves for which a re-
lationship between the inlet temperatures and the coefficients
(or exponent ) might be expected. These relationships
might be deduced from a plot of the regression line coef-
ficients and Y-intercepts in Figures 25 and 26. Continuity
with the established data by Schrenk is not evident, how-

ever, the material being considered is substantially

104

 

IO
s—I
a-I
7..
6..

54
4..

 
   
  
 
   

dNLET TEMPERATURE 720°F

DATA FRW SCHRENK I I959)

I6 -5. 614-2
IO (Vmoss) n H s 5.5
IO (Vmoss)

(esIImoIed)

     

IO 4 o(Vmos ST 32 ., G

-.- x

0.8"

0.64
0.5—4

0.4w

0.3-I

O
W 0' M” I’L/

Wo- — MOISTURE IN FEED — '33 bonedry IreIoII
O. [-1 Ib

” =MOISTURE IN PRODUCT - Tg-bonedry IreIoII

QZITITI TIIT

I
. I50 00 2000 2500 3300
MASS VELOCITY THROUGH DRYING COLUMN. Ib/I‘Ir -sq.II, Vmoss

FIGURE 24, RELATIONSHIP BETWEEN FLOW OF
HOT AIR AND INITIAL AND FINAL
LEAF MOISTURE CONTENT

   
 
 

 

 

 

105

different and could well account for some of the discrepancy.
If the relationship as indicated by three points were
acceptable as a straight line, then the expression relating
the regression line slope and Y-intercept could be written
easily. However, a sufficient range of temperatures was not
possible on which to base conclusive statements as to what
this relationship is at this stage of the development with
the scale of apparatus available.

Sensitivity of the fluidizing velocity to changes in
temperature along the length of the tower has been shown.
This is well founded upon examination of the differential
equation which clearly reveals the existence of a divergent
velocity field. This has also been observed and measured.
Typical divergence can be seen from an examination of the
graphical presentation, Figure 27, where the divergences
resulting from natural drafts at various inlet temperatures
are obvious. A divergence was detected for all combinations
of inlet temperatures and superimposed velocities. The
apparatus adequately served as a simulator of the continu-
ous process. A larger scale, with continuous flow of ma-
terials would result in heat transfer and evaporation of
water vapor with subsequent cooling of the drying gas
along the length of the tower, hence a divergence of the
velocity field. The divergence would be affected by the

mass transfer of plant held water released to the drying

O

106

 

‘55

$7)

NQE

r5;

 

REGRESSION LINE “v“ 'INTERCEPT
G

 

 

’§\"o
/’ '
I
’I
,/ CHOPPED
, ALFALFA
I PLANTS
BIRDSFOOT TREFOIL ,4
LEAVES7 ,1”
I”’
0'”
' T fi‘hT'
200 250 300 700

I
FIGURE 25. REGRESSION LINE “Y" INTERCEPT vs INLET
TEMPERATURE IFOR MASS VELOCITY-DRYING DATA)

 

 

 

 

 

-35,
LU
CL
3 °~4
“L4. “~4~ CHOPPED
I; ‘~ ‘ ALFALFA
3 BIRDSFOOT ‘x PLANTS
Z TREFOIL ~°\
9 LEAVES \
«‘2- - ‘
U 5 ‘\\
U
I

'64#y sf

500 700

INLET TEMPERATURE , mm", 'F

I T
200 250

FIGURE 26. ‘REGRESSION LINE SLOPE vs INLET
TEMPERATURE (FOR MASS VELocITv-Davmc DATA)

VELOCITY, vz , II/sec.

107

 

II-I

IO.

9-I

6-I

5..
J.
L

0

 

   
  
 
   
   
  

FLUIDIZING VELOCITY

for BIRDSFOOT TREFOIL LEAVES
(300% m.c.d.b.)

FLUIDIZING VELOCITY
-—- -—-—-_-
.OhTII

for ALFALFA LEAVES (300%mc

AV2
2 : VELOCITY
DIVERGENCE

(Uninsu ated Tex
A 2 quss pIpe Iipayn ID.

Room temps-80¢)

”I”
N<

 

T I 1 l ' '

I1
.2 3 4 5 6 T 8 9 I0

HEIGHT ABOVE HEAT SOURCE.Z . ft.
FIGURE 27. VELOCITY VARIATION-NATURAL DRAFTS

108

gases as water vapor. In order to fully explore this
further, a pilot scale observation would be required to
further verify the theory.

The significance of this phenomenon is that as the
leaf dries, becoming lighter, it does in fact require less
velocity to support it. To accomplish this, it might be
assumed that a tower needs to be designed smaller at the
bottom, so that for a given mass velocity at which fluidization
occurs, the absolute velocity would decrease further up the
column to accommodate the drying particle. However, the
divergence phenomenon in a tower of constant crossection
appears to accomplish the same objective, and considerable

economics of construction can be obtained for the process

under investigation.

109

SUMMARY AND CONCLUSIONS

The scope of the investigation of an entirely new
systems concept of forage harvesting and handling has been
such that a detailed study of all the facets uncovered is
virtually impossible as a single effort. No pretense is
made to imply that this accomplishment has been effected,
desirable as it may be. However, the investigations have
demonstrated that the system is feasible; that much more
research depth is warranted in all facets, and is essential
for the complete description of the process.

Results to date can be summarized as follows:

1. Leaf stripping has been demonstrated, and the
necessity of such a maneuver has been established as a pre-
lude to the fluidized bed drying process. Whether or not
this may be a desirable cultural practice has yet to be
established. Even if the plants are cut in a conventional
manner and then stripped of leaves, discharging the stems
back on to the harvested field area, no increase in present
field loss would be experienced and a considerable increase
in drying rate and efficiency would be accomplished, making
the process desirable from several considerations.

2. The damage point relationships have been definitely
established for a specific set of conditions of alfalfa

leaves and the results correlate well with those of other

110

products. Further information is needed to explore the
effects of a wider range of plant and culture oriented
variables. The damage point relationships can be exprSSSSd
mathematically.

3. The effects of ultra high temperature on the drying
of alfalfa leaves have been explored, largely as a by-
product of data taken for another phase. Verification of
the mathematical relationships must follow based on sta-
tistically designed experiments.

4. Particle behavior can be predicted based on studies
of the sphericityof the leaf. Variations in moisture content
have been explored and the particles, in spite of substantial
shrinkage, exhibit nearly constant characteristics over the
entire range of moisture contents. Process design of the
mass rate of flow can be made as a result of these studies.

5. The effect of mass velocities on fluidized bed
drying have been established showing a definite relationship
to work that has been previously reported. Forther data
is necessary at inlet temperatures higher than those attain—
able with laboratory scaled apparatus to firmly establish
the relationship with the inlet temperature. . I

6. Divergence of the velocity field has been demon-
strated and measured to relate this phenomenon with the
needs of a flaked particle drying in free suspension. Re—

finement of the measuring technique is necessary to provide

111

for clear cut velocity determinations. Design of the velocity

divergence must be a part of the considerations of the
process.

In conclusion, fluidization principles have been demon-
strated to be well suited to forage harvesting and handling.
The limitations relating to the incompatibility of various
plant parts in a fluidizing medium have been resolved by the
realization that effective separation of the major categories
must be accomplished. This may turn out to be the major
hurdle in the successful adoption of this system. However,
if a practical leaf stripping or stem altering operation can be
developed, the system will be competitive with existing ones.
The principles could be applied to a process which might
conceivably evolve as an alfalfa leaf combine whereby the
fresh cut leaf is dried immediately to a moisture content
commensurate with safe storage. A forage product could evolve

which lends itself to completely automated handling, which

has yet to be achieved with forage.

112

FUTURE INVESTIGATIONS
1. Expansion of the information on time-temperature
damage point relationships to correlate the effects of such
variables as stage of plant maturity, seasonal variation,

plant varieties, location of leaves on plant, etc.

2. Investigation of the drying rate relations for ultra-

high drying temperatures by statistically designed experi-

mentation.

3. Evaluation of the sphericities of other forage
plants and agricultural products as a prelude to the appli-
cation of fluidized bed drying principles.

4. Investigation of the change in Young's modulus of
plant parts related to the change in moisture content as
revealed by observations of the behavior of individual leaves
at accelerated drying rates of ultra-high temperatures.

The affect on other hay harvesting and handling pro—
cesses might be the objective of such a study.

5. Investigation of the feasibility and desirability

of stripping leaves from the growing alfalfa plant as a

cultural practice. Evaluation should be with the cooperation

of crop scientists.

6. Determination of methods to strip leaves mechanically

from the alfalfa plant, either standing in the field and/or

after mowing.

 

113

7. Evaluation of the parameters for fluidized bed
drying which will ultimately lead to the design and develop-
ment of this continuous drying process for alfalfa and

other agricultural products as either:
a) farmstead stationary installations, and/or

b) field combines to harvest alfalfa leaves and flash
dry them from the fresh cut stage to the safe storage

moisture content.

8. Investigation on the effect of stomata behavior on

the drying rates of alfalfa leaves.

10.

11.

114

REFERENCES

Anderson, Clyde G. (1951). An investigation<>f a con-
tinuous process for the flash drying and grinding of
alfalfa. Thesis for degree of M.S., Michigan State Univ.,

East Lansing (Unpublished).

Bailie, R. C., L. T. Fan and J. J. Stewart (1961). Insta-
bility of fluidized beds. Jour. of Chem. and Eng. Data 6

(3):469-473.

Bakker-Arkema, F. W., C. W. Hall and E. J. Benne (1962).
Equilibrium moisture content of alfalfa. Quarterly Bulletin,
Mich. Agr. Expt. Sta., Michigan State Univ., East Lansing

44 (3):492-496.

Barger, E. L. (1945). Mechanization of sumac leaf harvest-
ing and processing. Agr. Eng. 26 (6):243—245.

Barington, R. D., P. W. Davis and H. S. Wilgus (1952).
Conservation of sugar beet tops by dehydration. Colorado
Agr. Exp. Sta. Toch. 801. 47, 19 pp.

Bennett, C. A. and F. L. Gerdes (1941). The vertical drier
for seed cotton. Misc. Pub. No. 239, USDA, Washington,

D.C. 32 pp.

Bhan, A. K. (1959). Some engineering investigations on
the effect of exposed area on the drying of alfalfa. Thesis
for degree of M.S., Rutgers, The State University, New

Brunswick, N.J. (Unpublished).

Binder, R. C. (1943). Fluid Mechanics. Prentice—Hall,
Inc., New Yerk. 307 pp.

Bird, R. D., W. E. Stewart, and E. N. Lightfoot (1962).
Transport Phenomena. John Wiley and Sons, Inc., New York.

780 pp.

Bohnstedt, G. (1944). Nutritional values of hay and silage
as affected by harvesting, processing and storage. Agr.
Eng. 25 (9):337-341.

Chilton, T. H. and A. P. Colburn (1934). Mass transfer
coefficients. Ind. and Eng. Chem. 26:1183.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

115

Corder, S. E., C. 0. Morris, and G. H. Atherton (1956).
Suspension drying of sawdust. Mech. Eng. 78 (7):627-629.

Cummings, E. W. and T. T. Sherwood (1934). Drying of
solids. Ind. Eng. Chem. 26 (12):1046.

Davis, H. F. (1962). Introduction to Vector Analysis.
Allyn and Bacon, Inc., Boston. 359pp.

 

DeLeon, Paul (1945). Economics of processing animal feeds
by mechanical dehydration. Agr. Eng. 26 (3):103.

Franks, G. N. and C. S. Shaw (1962). Multipath drying at
cotton gins for controlling moisture in cotton. USDA, ARS
Rep‘t. No. ARS 42-69. 12 pp.

Gallaher, George L. (1949). Fundamentals of drying Agri-
cultural crops. Thesis for degree of M.S., Michigan State
Univ., East Lansing. (Unpublished).

Gerdes, F. L. and W. J. Martin (1949). Drying seed cotton.
USDA Leaflet No. 181, Washington, D.C. 8 pp.

Gilbert, M. et a1. (1955). Velocity lag of particles in
linearly accelerated combustion gases. Jet PrOpulsion 25:
26-30.

Gordon, E. D. and W. M. Hurst (1953). Artificial drying
of forage crops. USDA Circular 443, Washington,D.C. 28 pp.

Graumann, H. O., J. E. Webster, E. L. Canode, and H. F.
Murphy (1954). Harvest practices on the performance of
alfalfa. Oklahoma Agr. Expt. Sta. B-433. 23 pp.

Gregg, Gale R. (1955). Suspension drying of sawdust.
Thesis for degree of M.S., Oregon State Univ. (Unpublished).

Hall, C. W. (1957). Drying*Farm Crops. Edwards Brothers,
Ann Arbor, Mich. 336 pp.

 

Hall, C. W., H. F. McColly and W. F. Buchele (1959).
Developments in harvesting, storing and handling silage--
new concepts and possible applications. Paper presented
M.S.U. Silage Conference. (Unpublished).

Headley, V. A. and C. W. Hall (1963). Drying rates of
shelled corn with air temperatures up to 800°F. ASAE
Paper 63-24 presented at Michigan Section, Plymouth, Mich.
8 pp. (Unpublished).

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

116

Hedlin, C. P. and J. W. Garland (1959). An experimental
continuous flow hay drier. Report on Agr. Eng. ProJ. No.
29, Ontario Agr. College. June.

Henderson, 8. M. and R. L. Perry (1955). Agricultural
Process Engineerigg. John Wiley and Sons, Inc., New York.
402 pp.

Hoyt, Paul and Richard Bradfield (1962). Effect of varying
leaf area by partial defoliation and plant density on dry
matter production in corn. Agronomy Jour. 54:523.

Johnstone, H. F. and R. L. Pigford (1941). Heat transfer
to clouds of falling particles. Trans. Amer. Inst. Chem.
Eng. Board 37: 96-133.

Kennet, Wilbur William (1950). An investigation of the
simultaneous drying and grinding of alfalfa. Thesis for
degree of M.S., Michigan State Univ., East Lansing. (Un-
published).

Kettenring, K. W., E. L. Manderfield and J. M. Smith (1950).
Heat and mass transfer in fluidized systems. Chem. Eng.
Prog. 46:139.

Koegle, John S. (1949). An investigation of the drying
rates of alfalfa in a rotary kiln. Thesis for degree of
M.S., Kansas State College, Manhattan. (Unpublished).

Kroll, K. (1959). Trochner und Trocknungsverfahren. Verlag,
Berlin. 588 pp.

 

Lamouria, Lloyd H. (1950). Design and performance of a
pilot plant dehydration for agricultural products. Thesis
for degree of Ph.D., Iowa State College, Ames. (Unpublished).

Leva, M. (1959). Fluidization. McGraw Hill Book Co.,
New York. 327 pp.

Liang, S. C. (1951). On the calculation of surface area.
Jour. of Physics and Colloidal Chem. 55:1410-1412.

Loftfield, J. V. G. (1921), The behavior of stomata.
Carnegie Inst. Pub. 314, Washington, D.C.

Longhouse, A. D. (1950). The application of fluidization
to conveying grain. Prog. Rep. Agr. Eng. 31 (7):349,352.

39.

40.

41.

42.

43.

44.

45.

46.

47.

48.

49.

50.

 

117

Lorenzen, Robert T. (1958). The effect of moisture on
weight--volume relationships of small grains. Paper
presented at annual meeting ASAE, Santa Barbara, Calif.
June, 1958.

Moore, Walter J. (1955). Physical Chemistry. Prentice-
Hall, Inc., New Jersey. 633 pp.

 

Neel, G. H., G. S. Smith, M. W. Cole, R. L. Olson, W. D.
Harrington and W. R. Mullins (1953). Drying problems in
the add back process for production of potato granules.
Food Technology 8 (5):230, 234.

Othmer, D. F. (1956). Fluidization. Reinhold Publishing
Corp., New York. 220 pp.

 

Pabis, Jan (1960). Application of fluidization process

to drying of grain on industrial scale. Paper presented
International Conference of Crop Drying, Paznan, Poland,
March, 1960.

Pederson, T. T. (1962). The effect of conditioning and
physical environment on the field drying of alfalfa.
Thesis for degree of M.S., Michigan State Univ., East
Lansing. (Unpublished).

Persson, S. (1957). Eigenschaften des Veinigungsgutes in
Mahdveschern, Landtechnicsh Forschung 7 (2):41—45.

Pettyjohn, E. S. and E. B. Christiansen (1948). Effect
of particle shape on free-settling rates of isometric
particles. Chem. Eng. Prog. 44:157-172.

Quinn, Martin F. (1963). Fluidized bed dryers. Ind. and
Eng. Chem. 55:18-24.

Rice, Charles Erskine (1960). The effect of particle
size, shape and density on minimum pneumatic suspension
and vertical transport velocities. Thesis for degree
of Ph.D., Michigan State Univ., East Lansing. (Un-
published).

Richey, Lacey T. (1956). C00perative alfalfa dehydration.
USDA FCS Circular 12, June.

Rush, R. M. andB. B. Reilly (1943). Direct-fired air
heaters designed for dehydration and chemical processes.
Mech. Eng. 65 (6):511.

51.

52.

53.

54.

55.

56.

57.

58.

59.

61.

62.

63.

118

Schneider, Adorf (1954). Investigations on the drying
characteristics of alfalfa and sugar beet leaf in single
layers and by passing the air through bulk materials.
Technical College of Munich.

Schoenleber, L. G. (1949). Engineering problems of
harvesting, conditioning and preserving hay. Paper pre-
sented at 11th annual meeting of Agronomists, W. Spring-
field. Mass.

Schoenleber, L. G. (1950). Operation and performance of
alfalfa dehydrators in central U.S. USDA Inf. Series No.
101, April.

Schrenk, W. D., H. L. Mitchell, R. E. Silker, E. L.
Sorensen, W. H. Hbrnstead, R. G. Taecker and C. C.
Burkhardt (1959). Dehydrated alfalfa. Agr. Exp. Sta.
Bul. 409, Kansas State College, Manhattan. 64 pp.

Thompson, C. Ray (1952). Method of stabilizing carotene
at high temperatures. Agr. Eng. 33 (l):19, 20.

Towle, H. E. (1962). Interview with manager of commercial
dehydration plant. Consolidated Mills, Inc., Blissfield,
Michigan. (Unpublished).

von Loesecke, H. W. (1955). Drying and Dehydration of

 

Foods. 2nd ed. Reinhold Publishing Corp., Newfi73Fk. 300

PP-

Wang, Jaw Kai (1958). Theory of drying. Thesis for degree
of Ph.D., Michigan State Univ., East Lansing. (Unpublished).

WanJura, Donald F. (1962). Drying variable relationship
for seed cotton dried by its conveying air. Thesis for
degree of M.S., Clemson College, Clemson, S.C. (Unpublished).

Whitney, L. F. and S. A. Weeks (1963). The application of
vacuum drying principles to forage. Annual report to
Regional Committee NE-13, Hatch 190. (Unpublished).

Wodicka, Virgil and Lamar Kishlar (1939). Some carotene,
protein and fiber values of dehydrated alfalfa meals.
Agr. Eng. 20 (3):109-110, 114.

Ziemba, John V. (1962). Now - drying without heat. Food
Eng. 34 (7):S4-85.

Zink, Frank J. (1946). Moisture content at which alfalfa
leaves shatter. Agr. Eng. 17 (8):329-330.

 

119

APPENDICES

120

APPENDIX A

TABLE A-l. SUMMARY OF DAMAGE POINT AND DRYING DATA FOR
SINGLE ALFALFA LEAVES - TEMPERATURE 281°F.

 

 

Leaf Initial M.C. Final M.C. M -Me % H20 Time,
No. % d.b. % d.b. H3:i3 removed seconds
1 213.2 171.9 0.8063 19.4 10
2 194.2 147.1 0.7575 24.3 15
3 133.7 97.0 0.7255 27.5 17
4 215.8 131.6 0.6098 39.0 20
5 191.3 136.4 0.7130 28.7 22
6 277.5 170.6 0.6148 38.5 25
7 243.6 299.7 0.9429 46.3 27
8 310.3 217.2 0.6999 30.0 30
9 192.5 95.5 0.4961 50.4 32
10 310.1 195.3 0.6298 37.0 35
11 192.8 88.8 0.4606 53.9 37
12 114.9 37.4 0.3255 67.4 40
13 208.6 86.7 0.4156 58.4 42
14 211.5 77.0 0.3641 63.6 45
15 263.1 113.9 0.4329 56.7 47
16 174.5 44.1 0.2527 74.7 50
17 158.7 28.9 0.1821 81.7 55
18 203.0 36.0 0.1773 82.3 60
19 250.8 64.2 0.2560 73.2 65
20 325.8 100.0 0.3069 69.3 70
21 312.3 84.4 0.2703 72.9 75
22 176.4 31.9 0.1808 87.9 80
23 185.0 55.0 0.2973 70.3 85
24 166.7 12.8 0.0768 92.3 90
25 199.4 19.8 0.0993 90.0 95
26 183.0 4.1 0.0224 ,97.7 100
27 238.2 3.5 0.0147 98.5 105
28 247.2 0.0 0.0 100.0 110
Date: 7/28/62 Stage of Maturity: Pro-bloom,
Air Velocity: 58 fpm 0 80°F 2nd cutting
Rotation; 40 rpm Equilibrium m.c.: He 9 0

No discoloration

121

 

 

 

TABLE A-2. SUMMARY OF DAMAGE POINT AND DRYING DATA FOR
SINGLE ALFALFA LEAVES - TEMPERATURE 296°F.

Leaf Initial M.C. Final M.C. M - % H20 Time,
No. % d.b. % d.b. 33:53 removed seconds

1 218.7 128.0 0.5853 41.5 20

2 255.7 166.0 0.6492 35.0 25

3 227.1 113.1 0.4980 50.2 30

4 263.9 98.8 0.3739 62.6 35

5 234.4 90.6 0.3865 61.3 40

6 179.3 76.7 0.4278 57.2 45

7 274.4 40.7 0.1483 85.2 50

8 259.2 54.1 0.2087 79.1 55

9 167.5 11.4 0.0681 93.2 60

10 238.6 16.9 0.0708 92.9 65

11 223.4 47.8 0.2140 78.6 70

12 236.5 36.5 0.1543 84.6 75

13 188.3 4.7 0.0250 97.5 80

14 311.8 22.6 0.0725 92.8 85

15 230.0 40.6 0.1765 82.4 85

16 226.3 0 0.0000 100.0 90

17 287.8 2.4 0.0083 99.2 95

18 247.3 7.6 0.0307 96.9 92

19 318.1 0 0.0000 100.0 92
Date: 7/31/62 Stage of Maturity: Dre-bloom,

Air velocity:

Rotation:

58 fps 080°F
40 rpm

2nd cutting
Equilibrium m.c., Me an 0
No discoloration

122

 

 

 

TABLE A-3. SUMMARY OF DAMAGE POINT AND DRYING DATA FOR
SINGLE ALFALFA LEAVES - TEMPERATURE 311’F.

Leaf Initial M.C. Final M.C. M - % BIO Time

No. % d.b. % d.b. i332: remoged seconds

1 235.9 159.8 0.6774 32.2 20

2 135.5 74.5 0.5498 45.0 25

3 188.3 67.0 0.3558 64.4 30

4 179.5 89.4 0.4981 50.2 35

5 175.2 50.4 0.2877 71.2 40

6 169.9 58.8 0.3461 65.4 45

7 164.6 22.9 0.1391 86.1 50

8 180.7 27.1 0.1500 85.0 60

9 184.8 63.6 0.3442 65.6 55
10 244.4 82.5 0.3580 66.2 65
11 206.3 7.8 0.0378 96.2 70
12 219.8 1.0 0.0045 99.5 90
13 248.6 13.1 0.0527 94.7 75
14 186.7 0.9 0.0048 99.8 80
15 165.8 2.0 0.0121 98.8 82
16 243.6 10.7 0.0439 95.6 85
17 342.2 13.2 0.0386 96.1 90
18 347.5 0.0 0.0000 100.0 95
19 219.5 0.0 0.0000 100.0 100
20 188.9 0.7 0.0037 99.7 105
21 252.1 3.1 0.0123 98.8 90
22 234.0 2.1 0.0090 99.1 95
23 210.9 0.0 0.0000 100.0 100

Date: 7/31/62 Stage of Maturity: Pro-bloom,

2nd cutting
Equilibrium m.c., Mesa 0
No discoloration

Air Velocity: 58 fpm @ 80°F
Rotation: 40 rpm

123

 

 

 

TABLE A-4. SUIIARY OR DAMAGE POINT AND DRYING DATA FOR
SINGLE ALFALFA LEAVES - TEMPERATURE 379?.
Leaf Initial M.C. Final M.C. M - % 820 Time
No. % d.b. % d.b. M;:%: removed seconds
1 219.5 125.8 0.5731 42.7 20
2 188.9 87.7 0.4643 53.9 25
3 204.1 86.3 0.4228 57.7 30
4 157.1 23.3 0.1483 85.2 35
5 198.3 52.3 0.2637 73.6 40
6 174.8 11.7 0.0669 93.3 45
7 267.9 5.1 0.0190 98.1 50
8 318.2 5.7 0.0179 98.2 55
9 278.7 0 0.0000 100.0 ' 60
10 311.5 0 0.0000 100.0 65
11 268.6 5.9 0.0220 97.8 55
12 168.1 1.1 0.0065 99.4 57
13 237.3 5.9 0.0249 97.5 60
14 139.9 4.7 0.0336 97.7 60
15 130.5 1.7 0.0130 98.7 60
16 135.7 0.8 0.0057 99.4 60
17 169.3 0.0 0.0000 100.0 65
18 141.2 0.7 0.0050 99.5 62
19 215.9 108.4 0.5020 49.8 20
20 196.0 96.8 0.3893 50.6 25
21 168.2 75.4 0.4483 55.1 30
22 237.6 75.2 0.3165 68.3 35
23 233.3 57.1 0.2447 75.5 40
24 218.5 18.5 0.0847 91.5 45
25 216.7 3.1 0.0143 98.6 50
26 215.2 4.3 0.0200 98.0 55
27 211.4 1.9 0.0090 99.1 60
28 243.9 0.4 0.0016 99.6 62
Date: 8/1/62 Stage of Maturity: 10% bloom,
Air Velocity: 45 fpm @ 80°F 2nd cut
Rotation: 40 rpm Equilibrium m.c. , IIe ~ 0

No discoloration

TABLE A-5.

SUMMARY OR DAMAGE POINT AND DRYING DATA FOR
SINGLE ALIALFA LEAVES - TEMPERATURE 455°F.

 

 

 

Leaf Initial M.C. Final M.C. M - % 820 Time
No. % d.b. % d.b. i3:%§ removed seconds
1 280.4 164.9 0.5881 41.2 10
2 310.6 156.5 0.5042 49.6 15
3 245.9 69.7 0.2834 71.6 20
4 246.2 77.3 0.3140 68.6 22
5 259.8 14.5 0.0751 92.5 25
6 222.7 13.6 0.0611 93.9 27*

7 225.6 11.1 0.0492 95.1 27*
8 268.2 9.4 0.0350 96.5 27
9 277.7 8.5 0.0306 96.9 27
10 285.2 0 0.0000 100.0 30*
11 305.6 0 0.0000 100.0 30*
12 312.1 173.6 0.5562 44.4 10
13 271.4 122.4 0.4510 54.9 15
14 265.5 44.0 0.1657 83.4 20
15 196.3 16.0 0.0815 91.8 25
16 282.7 10.7 0.0385 96.2 27*
17 251.4 0 0.0000 100.0 30*
Date: 8/4/62 Stage of Maturity: 10% bloom,

Air Velocity:
Rotation: 40

45 fpm @ 85°F 2nd cut
rpm Equilibrium m.c. , Me is 0
*Discoloration-brown edge

TABLE A-6.

SUMMARY OP DAMAGE POINT AND DRYING DATA FOR

SINGLE ALFALFA LEAVES - TEMPERATURE 515°P.

 

 

Leaf Initial M.C. Final M.C. M -!g % H20 Time,
No. % d.b. % d.b. Mb-Me removed seconds
1 276.6 64.9 0.2346 76.5 10
2 324.1 10.3 0.0318 96.8 15
3 311.3 1.6 0.0051 99.4 20
4 271.0 104.3 0.3849 61.5 10
5 255.8 34.9 0.1364 86.4 15
6 342.8 0 0.0000 100.0 20*
7 302.2 4.5‘ 0.0149 98.5 17
8 275.0 3.1 0.0113 98.9 18
9 264.2 1.5 0.0057 99.4 20*
10 337.2 2.3 0.0068 99.3 19
11 292.0 2.0 0.0068 99.3 19*
12 264.5 12.9 0.0488 95.1 16
13 283.9 33.3 0.1173 88.3 16

 

Stage of Maturity: Pre-bloom,
2nd cut

Equilibrium m.c. , Men: 0

*Discoloration« brown edges

Date: 7/27/62
Air Velocity: 58 fpm 0 80°F
Rotation: 40 rpm

126

TABLE A-7. SUMMARY OF DAMAGE POINT AND DRYING DATA FOR
SINGLE ALFALFA LEAVES - TEMPERATURE 524°F.

 

 

 

Leaf Initial M.C. Final M C M -!fi % H20 Time,
No. % d.b. % d.b Mo-Me removed seconds
1 340.2 146.4 0.4303 57.0 10
2 322.1 142.1 0.4412 55.9 11
3 308.3 98.6 0.3198 68.0 12
4 331.7 115.8 0.3491 65.1 13
5 328.8 57.6 0.1752 82.5 14
6 318.2 56.8 0.1785 82.1 15
7 303.4 44.8 0.1477 85.2 16*

8 320.8 16.7 0.0521 94.8 17
9 307.2 20.3 0.0661 93.4 18*
10 319.7 3.3 0.0103 98.9 18*
11 275.3 5.2 0.0189 98.1 19*
12 319.6 5.4 0.0169 98.3 19*
Date: 8/4/62 Stage of Maturity: 10% bloom,

Air Velocity: 45 fps 0 85°F 2nd cut
Rotation: 40 rpm Equilibrium m.c., MON 0

1'*Discoloration--brown tip, edge

127

 

 

 

TABLE A-8. SUMMARY OF DAMAGE POINT AND DRYING DATA FOR
SINGLE ALFALFA LEAVES - TEMPERATURE 570°F.
Leaf Initial M.C. Final M.C. M ~M§ % H20 Time,
No. % d.b. % d.b. M'o-Me removed seconds
1 289.1 79.5 0.2750 72.0 10
2 315.7 61.4 0.1945 80.5 11
3 381.8 122.7 0.3214 67.9 12
4 353.7 75.6 0.2137 78.6 13
5 349.3 54.8 0.1569 84.3 14*
6 350.8 18.6 0.0530 94.7 15*
7 312.9 28.6 0.0914 90.9 14*
Date: 8/4/62 Stage of Maturity: 10% bloom,

Air Velocity:

45 fpm 0 80°F 2nd cut

 

 

 

Rotation: 40 rpm Equilibrium m.c. , Me an 0
*Discoloration-brown edges
TABLE A-9. SUMMARY OF DAMAGE POINT AND DRYING DATA FOR
SINGLE ALFALFA LEAVES - TEMPERATURE 640°F.
Leaf Initial M.C. Final M.C. M - % 820 Time,
No. % d.b. % d.b. Mb-Eg removed seconds
1 298.1 203.8 0.6837 31.6 5
2 331.0 207.0 0.6254 37.5 6
3 284.8 153.8 0.5400 46.0 7
4 306.4 119.4 0.3897 61.6 8
5 374.6 103.2 0.2755 72.4 9
6 244.9 46.1 0.1882 81.2 10*
7 267.1 35.6 0.1333 86.7 10*
8 267.5 58.4 0.2183 78.2 10*
9 328.1 101.0 0.3078 69.2 9*
10 308.5 101.9 0.3303 67.0 9*
11 331.3 91.3 0.2756 72.5 9*
12 292.8 85.7 0.2927 70.7 9*
Date: 8/5/62 Stage of Maturity: 10% bloom,

Air Velocity:
Rotation: 40

2nd cut
Equilibrium m.c., Me‘s 0
*Discoloration-black edges

45 fpm 0 85°F
rpm

TABLE Ar10. SUMMARY OF DAMAGE POINT AND DRYING DATA

128

FOR

SINGLE ALFALFA LEAVES - TEMPERATURE 661°F.

 

 

 

 

 

 

Leaf Initial M.C. Final M.C. M - % H20 Time
NO. % d.b. % d.b. M33M§ removed seconds
1 308.0 144.0 0.4675 53.2 5
2 283.1 107.8 0.3808 62.9 6
3 321.8 111.5 0.3465 65.3 7
4 300.0 101.1 0.3370 66.3 8*

5 252.8' 46.1 0.1824 81.8 9*
6 289.9 65.2 0.2249 77.5 8*
7 289.2 66.2 0.2289 77.1 8*
8 313.7 102.7 0.3274 67.2 8*
Date: 8/5/62 Stage of Maturity: Pre-bloom,
Air Velocity: 45 fpm 0 85°F 2nd out
Rotation: 40 rpm Equilibrium m.c., M94» 0
*Discoloration-black tips
TABLE A-ll. SUMMARY OF DAMAGE POINT AND DRYING DATA FOR
SINGLE ALFALFA LEAVES - TEMPERATURE 761°F.
Leaf Initial M.C. Final M.C M ~M§ % H20 Time,
No. % d.b. % d.b. Mo-Me - removed seconds
1 335.2 282.4 0.8425 15.8 2
2 283.3 179.4 0.6333 36.7 3
3 400.0 134.2 0.3355 66.4 4:
4 350.0 215.3 0.6151 38.5 4*
5 318.6 128.6 0.4036 59.6 4*
6 339.4 157.6 0.4644 53.6 5*
7 297.7 “71.0 0.2385 76.0 5*
8 295.9 136.5 0.4613 53.9 4*
9 313.8 184.6 0.5883 41.2 4:
Date: 8/5/62 Stage of Maturity: Pro-bloom,

Air Velocity: 45 fpm 0 85°F
40 rpm

Rotation:

2nd cut
Equilibrium m.c., Me‘v 0
*Discoloration-black edge

129

 

 

 

 

TABLE A-12. SUMMARY OF DAMAGE POINT AND DRYING DATA FOR

SINGLE ALFALFA LEAVES - TEMPERATURE 819°F.

Leaf Initial M.C. Final M.C. M -Me % H20 Time,
No. % d.b. % d.b. Mo-Me removed seconds

1 296.0 205.3 0.6936 30.6 2

2 330.1 220.4 0.6677 33.2 3*

3 359.6 184.2 0.5122 48.8 3*

4 320.0 36.0 0.2688 73.1 4*

5 208.5 107.3 0.5146 48.5 4*

6 361.8 240.4 0.6645 33.5 3*

7 340.0 173.3 0.5097 49.0 3*
Date: 8/5/62 Stage of Maturity: Pre-bloom,

Air Velocity:

Rotation:

45 fpm @ 85°F
40 rpm

2nd cut
Equilibrium m.c. , Me nu 0

*Discoloration-black edge

130

APPENDIX B

TABLE A-13. SUMMARY OF SPHERICITY-MOISTURE CONTENT
DATA, LARGE ALFALFA LEAVES

 

 

Total
Leaf Vbl. surfacg area Sphzgicity Moisture
No m.1. in 8 content Density_
1a 0.067 0.934 0.133 359.8 0.906
b 0.058 0.934 0.121 280.3 0.866
c 0.051 0.820 0.127 200.8 0.778
d 0.018 0.566 0.092 0 0.733
2a 0.076 1.094 0.124 287.0 0.784
b 0.067 1.094 0.114 251.3 0.807
c 0.055 0.900 0.122 194.8 0.825
d 0.023 0.626 0.098 0 0.670
3a 0.068 0.986 0.128 346.8 0.913
b 0.062 0.934 0.127 284.2 0.861
c 0.046 0.766 0.127 192.8 0.885
, d 0.018 0.434 0.120 0 0.772
4a 0.074 1.100 0.121 341.3 0.853
b 0.063 1.040 0.115 261.5 0.821
c 0.047 0.886 0.111 169.2 0.819
d 0.022 0.606 0.098 0 0.650
5a 0.064 1.066 0.115 350.8 0.902
b 0.054 0.946 0.114 242.2 0.811
c 0.037 0.706 0.119 113.3 0.738
d 0.018 0.560 0.093 O 0.711
6a 0.071 0.980 0.132 316.0 0.879
b 0.065 0.854 0.143 281.3 0.880
c 0.052 0.906 0.116 192.0 0.842
d 0.020 0.546 0.102 0 0.750
7a 0.069 1.074 0.118 336.2 0.872
b 0.062 0.946 0.125 271.7 0.827
c 0.050 0.886 0.116 89.9 0.800
d 0.020 0.586 0.095 0 0.690
8a 0.072 1.066 0.122 313.8 0.914
b 0.062 0.934 0.127 240.2 0.873
c 0.051 0.820 0.127 157.9 0.804
d 0.022 0.580 0.102 0 0.723
9a 0.065 0.960 0.127 344.4 0.862
b 0.056 0.906 0.122 286.5 0.870
c 0.041 0.754 0.119 151.6 0.773
d 0.017 0.534 0.094 0 0.741
10a 0.055 0.874 0.126 298.4 0.884
b 0.048 0.806 0.124 245.1 0.877
c 0.030 0.654 0.112 137.7 0.967
d 0.017 0.460 0.109 0 0.718

 

131

TABLE A-13 (Cont.)

Date: 9/12/63 Ave. Volume, Fresh cut, Vw '

Stage of Maturity: Pre-bloom, 0.071 ml.
3rd cut Ave. Velume, Dried, Vd -
0.020 ml.
Shrinkage, 3g - 0.28

Vw

TABLE A-14. SUMMARY OF SPHERICITY-MOISTURE CONTENT

DATA, MEDIUM ALFALFA LEAVES

 

 

Total
Leaf Vbl. surfacg area Sphe icity Moisture
No m.1. in 95s content Density_
1a 0.038 0.78 0.1096 291.5 0.845
b 0.030 0.70 0.1046 198.7 0.817
c 0.019 0.56 0.0961 90.2 0.821
d 0.012 0.42 0.0941 3.7 0.708
o 0.012 0.42 0.0941 2.4 0.700
2a 0.040 0.70 0.1258 246.0 0.883
b 0.030 0.58 0.1257 181.4 0.1060
c 0.027 0.54 0.1257 153.1 0.1059
d 0.026 0.52 0.1274 109.7 0.912
e 0.014 0.44 0.0996 19.4 0.964
3a 0.039 0.72 0.1204 180.8 0.864
b 0.025 0.56 0.1231 110.0 1.008
c 0.018 0.50 0.1161 65.8 1.106
d 0.018 0.44 0.1328 19.2 0.794
e 0.012 0.40 0.0987 0.0 1.000
4a 0.035 0.60 0.1350 277.2 0.851
b 0.027 0.54 0.1258 172.1 0.796
c 0.022 0.48 0.1241 118.9 0.786
d 0.018 0.44 0.1328 62.0 0.711
e 0.011 0.36 0.1035 11.4 0.782
5a 0.046 0.88 0.1106 402.3 0.939
b 0.039 0.72 0.1212 296.5 0.874
c 0.031 0.64 0.1166 222.1 0.894
.d 0.023 0.58 0.1060 111.6 0.791
e 0.010 0.46 0.0770 0.0 0.860
6a 0.037 0.72 0.1114 254.1 1.043
b 0.026 0.62 0.1073 109.2 0.876
c 0.020 0.48 0.1165 55.9 0.850
d 0.017 0.46 0.1094 21.8 0.759
e 0.016 0.46 0.1049 0.0 0.681
7a 0.046 0.84 0.1158 236.0 0.907
b 0.038 0.74 0.1153 183.9 0.926
c 0.033 0.64 0.1218 147.6 0.930
d 0.025 0.62 0.1046 89.5 0.940
e 0.014 0.50 0.0870 0.0 0.885

132

TABLE A-14 (Cont.)

 

 

 

8a 0.042 0.82 0.1114 203.0 0.959
b 0.035 0.70 0.1153 157.9 0.980
c 0.030 0.66 0.1108 124.8 0.996
d 0.027 0.62 0.1099 81.9 0.896
e 0.016 0.54 0.0888 0.0 0.831
9a 0.034 0.66 0.1204 245.3 0.965
b 0.027 0.52 0.1317 133.7 0.822
c 0.019 0.44 0.1228 72.6 0.863
d 0.016 0.42 0.1147 20.0 0.712
e 0.011 0.36 0.1035 0.0 0.863
10a 0.047 0.74 0.1337 184.3 0.847
b 0.040 0.70 0.1258 159.3 0.907
c 0.032 0.68 0.1122 120.0 0.963
d 0.027 0.60 0.1135 45.7 0.888
e 0.016 0.48 0.1006 0.0 0.875
Date: 8/16/63 Ave. Vblume, Fresh cut, Vw
Stage of Maturity: Pre-bloom, 0.038 ml.
3rd cut Ave. Vblume, Dried, Vd -
0.013 ml.
Shrinkage, Ed - 0.33
Vi
TABLE A-15. SUMMARY OF SPHERICITY-MOISTURE CONTENT
DATA, SMALL ALFALFA LEAVES
ITEtal
Leaf Vol. surfacg area Sphgsicity Moisture
No m.1. in 8 ‘__content Density;
1a 0.020 0.314 0.178 272.0 0.930
b 0.017 0.294 0.171 234.0 0.982
c 0.010 0.240 0.147 104.0 1.02
d 0.007 0.186 0.150 0.0 0.714
2a 0.028 0.374 0.186 238.4 0.882
b 0.021 0.380 0.151 212.3 1.09
c 0.010 0.374 0.093 101.4 1.47
d 0.009 0.294 0.112 0.0 0.811
3a 0.023 0.320 0.189 242.6 0.909
b 0.022 0.300 0.196 209.8 0.859
c 0.016 0.300 0.159 80.3 0.688
d 0.008 0.214 0.141 0.0 0.763

133

TABLE A-15 (Cont.)

 

4a 0.020 0.260 0.215 264.7 0.930
b 0.015 0.254 0.182 225.5 1.107
c 0.009 0.260 0.127 127.5 1.289
d 0.008 0.194 0.155 0.0 0.638

5a 0.019 0.274 0.197 215.6 0.747
b 0.017 0.274 0.183 173.3 0.724
c 0.012 0.272 0.145 55.6 0.583
d 0.010 0.214 0.163 0.0 0.450

6a 0.022 0.274 0.217 314.3 0.659
b 0.020 0.266 0.208 228.6 0.575
c 0.016 0.240 0.201 42.9 0.313
d 0.007 0.174 0.158 0.0 0.500

7a 0.023 0.240 0.255 290.9 0.748
b 0.019 0.246 0.221 218.2 0.736
c 0.015 0.226 0.206 88.6 0.553
d 0.006 0.174 0.143 0.0 0.733

8s 0.027 0.340 0.199 284.4 0.911
b 0.025 0.314 0.208 228.1 0.840
c 0.025 0.312 0.208 123.4 0.572
d 0.008 0.226 0.135 0.0 0.800

9a 0.020 0.320 0.173 251.8 0.985
b 0.019 0.294 0.184 185.7 0.842
c 0.018 0.294 0.178 78.6 0.555
d 0.008 0.180 0.168 0.0 0.700

10a 0.025 0.300 0.214 290.0 0.796
b 0.022 0.280 0.213 260.0 0.918
c 0.018 0.290 0.181 126.0 0.628
d 0.009 0.174 0.187 0.0 0.555
Date: 9/13/63 Ave. Volume, Fresh cut, V. -
Stage of Maturity: Pre-bloom, 0.023 ml.
3rd cut Ave. Vblume, Dried, Va - 0.008
ml.

Shrinkage, $§ - 0.35
w

134

TABLE Ar16. SUMMARY OF SPHERICITY-MOISTURE CONTENT
DATA, BIRDSFOOT TREFOIL .

 

 

Total
Leaf Vol. surfacg area Sphericity Moisture
No. II. 1. in $8 content Densim
1a 0.026 0.320 0.207 387.5 0.900
b 0.023 0.294 0.207 322.9 0.812
c 0.020 0.254 0.205 218.8 0.765
d 0.018 0.240 0.217 100.0 0.533
e 0.012 0.186 0.213 8.3 0.433
f 0.010 0.180 0.195 0.0 0.480
2a 0.028 0.360 0.193 373.7 0.964
b 0.023 0.336 0.182 235.1 0.830
c 0.020 0.334 0.167 133.3 0.665
d 0.018 0.280 0.186 54.4 0.489
e 0.015 0.240 0.192 0.0 0.380
f 0.011 0.226 0.165 0.0 0.518
3a 0.017 0.260 0.192 377.1 0.982
b 0.014 0.260 0.168 294.3 0.986
c 0.014 0.254 0.172 228.6 0.821
d 0.013 0.186 0.224 82.9 0.492
e 0.010 0.166 0.211 2.9 0.360
f 0.009 0.160 0.205 0.0 0.389
4a 0.020 0.386 0.144 358.7 1.055
b 0.018 0.354 0.147 223.9 0.827
c 0.016 0.306 0.157 106.5 0.594
d 0.013 0.246 0.170 28.3 0.454
e 0.010 0.234 0.150 0.0 0.460
f 0.009 0.226 0.145 0.0 0.511
5a 0.024 0.294 0.213 300.0 0.800
b 0.022 0.286 0.208 245.8 0.755
c 0.020 0.286 0.195 204.2 0.730
d 0.018 0.274 0.190 143.8 0.650
e 0.012 0.220 0.180 35.4 0.542
f 0.011 0.200 0.187 0.0 0.436
6a 0.023 0.386 0.158 395.2 0.904
b 0.021 0.340 0.169 316.7 0.833
c 0.019 0.326 0.165 231.0 0.731
d 0.016 0.280 0.172 69.0 0.444
e 0.014 0.214 0.205 2.4 0.307
f 0.014 0.206 0.213, 0.0 0.300
7a 0.025 0.320 0.202 395.5 0.872
b 0.021 0.300 0.192 334.1 0.909
c 0.019 0.240 0.225 270.5 0.858
d 0.013 0.226 0.185 143.2 0.823
9 0.008 0.220 0.138 18.2 0.650
f 0.007 0.160 0.185 0.0 0.628

135

TABLE A-16 (Cont.)

 

8a 0.028 0.420 0.166 310.9 0.939
b 0.024 0.354 0.177 259.4 0.958
c 0.021 0.346 0.166 215.6 0.962
d 0.018 0.340 0.153 157.8 0.917
e 0.015 0.286 0.161 62.5 0.693
f 0.010 0.226 0.155 0.0 0.640
9a 0.021 0.326 0.176 362.7 1.124
b 0.020 0.306 0.182 311.8 1.050
c 0.016 0.274 0.176 352.9 1.125
d 0.014 0.260 0.181 170.6 0.986
e 0.012 0.242 0.164 54.9 0.658
f 0.010 0.226 0.155 0.0 0.510
10a 0.024 0.340 0.185 475.7 0.888
b 0.022 0.326 0.182 391.9 0.827
c 0.017 0.280 0.178 291.9 0.853
d 0.016 0.274 0.175 172.3 0.631
e 0.013 0.220 0.190 32.4 0.377
f 0.009 0.186 0.176 0.0 0.411
Date: 9/9/63 Ave. Velume, Fresh cut, V. -
Stage of Maturity: 10% bloom, 0.029 ml.
3rd cut Ave. VOIume, Dried, Vd -
0.01 ml.

Shrinkage, 1g - 0.42
Vw

136

APPENDIX C

SMOKE POWDER TERMINAL VELOCITY

Smoke powder: Norit "A" (activated charcoal)
Chemical compound: Carbon

Sieve Analysis:

 

 

 

Particle 511., Weight Over
Mesh Size 3911““ Screen, %
60 0.25 or greater 0.29
140 0.10 or greater 1.97
300 0.05 or greater 16.75
Thru 300 less than 0.05 81.00
Density of air at 212°F, 1% R.H.: F} - 0.06 lb. per cu. ft.

Viscosity: 1L - 0.0218 centipoises

(Density of particle: £2 - 33.0 lb. per cu. ft.
Q - Pf - 33.0 - 0.06 - 32.94 lb. per cu. ft.
Diameter of particle -0.05 mm - 0.00197 inches

From Figure 28, the terminal velocity Ut can be determined

by the following steps of procedure:

1. Start at left hand axis F} - 0.06 .......... ..pt 1
2. Connect with fL- 0.0218 ....... .............. ..pt 2
3. Intercept "a" axis at ......... ........ ... ..... pt 3
4. Also intercept "c" axis at ....................pt 3’
5. From pt 3 on "a" axis to ( F; - F2) - 32.94 ..pt 4
6. Cross "b" axis at .............................pt 5
7. From pt 5 to particle diameter,Dp - 0.00144....pt 6

137

8. Intersect "d" axis at .............. ........... pt 7
9. Follow pattern to "e" axis at ......... ........pt 8
10. Connect 3' with pt 8, intersect "f" axis at ...pt 9
11. Connect pt 6 with pt 9, intersect Ut at .......pt 10
From this analysis, the terminal velocity Ut - 0.1 fps.
This value checks with observed free fall values within 10%.
With the small magnitudes in comparison with the velocity of

the fluidizing air, this was neglected in the velocity calcu-

lations.

 

138

 

particle dia.,

e

it I T T I TI WTTTTTTTIT TOT T I I 111 YTITTIWT
O . "

o. 8

O

 

ngnflITl Irntn
—' O
O _'.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

'o ..
. '. o,
m 1. t. t '.
3 ' "
I e
3;; J 1 1 11 1 ~11111:1 1 1 124 1,J_1_1_111111uu11__-1_i_11-i. i 1 11-111111111L_ J
o .. ..e ‘e O O
o .. O s. _: —
I ..
‘ e
U ..u; @_L .e
w. .0. .0.
3 O O
.Q:- o 3,. ‘..
5.” 11111121111111 19.1111 111 111111111
lm 0- 9 °.. ‘.°. 0
Q ', .0 O
D O O. .—
._ O. to.
0 e
e .‘
e. ..O.
.0
O... .3
e. ‘e
e‘
0... g
0
e. a
O
O
a 1...- 1.1, .9. 1--,___JI
0. .g
e.‘.
0.0‘
o —__._._. W ._ # _ _#__.
.0
§ "
O
O
2 "
O
0
e- ' o
is 11111111 1111111111 1111111111 1111111131. 6.1111111 111
- a: o O , "= '. O
:3 0 o '- "' 0 ’oo‘ 0.
2 "' °. 0
D .0.
D °o
o O
0..
5 °.
c1 3.
e O
D O .0
1 O O o O —.0.
.._ ._ N _ .1 q .0
Q. 111111 11 111111111I11_1 111111111111 11 111111 .

 

FIGURE 28 NOMOGRAPH FOR EVALUATION OF PARTICLE
TERMINAL VELOCITY, LEVA (I959)

139

THEORETICAL TERMINAL VELOCITY FOR ALFALFA STEMS
AND LEAVES

Stoke' s Law, Binder (1943), is applied to particles of

the following dimensions to determine their terminal veloci-

 

 

ties:
Dried
Fresh Stem Dried Stem Fresh Leaf
Quantity (1 in.) 25% (mcdb) Leaf 25% mcdb
Diameter, d, in. 0.090 0.073 xxx xxx
Thickness, t, in. xxx xxx 0.010 0.009
Projected
Area, Ap, rt2x104 6.25 5.07 25.7 16.6
Crossection
Area, Ac, inleo3 6.36 4.20 xxx xxx
Volume, v, in3x103 6.36 420 2.9 .97
Weight, w, 1b.x105 18.3 7.52 9.02 3.15
Density, P.1b/cu.ft. 50.20 30.85 54.52 52.85
Area/Weight, ft2/1b 3.41 6.75 28.38 52.70

 

Stems are approximated as cylinders, Cd - 1.7, Re - 50
Leaves are approximated as disks, Cd - 1.2, Re - 150

Sample calculation:

- 3
Assume: air entering e 500°F, f)- 4.13 x 10 2 lb/ft
1L- 1.900 x 10"5 lb/ft sec.

—2 3
air leaving 0 300°F, fD- 5.23 x 10 lb/ft
IL- 1.615 x 10"5 lb/ft sec.

140_

For example:

Fresh cut stem, 1 inch long, air entering 0 500°F

-5
Utz .. 2 (18.3 X 1.0—2— 4 .. 8.33
1.7 (4.13 x 10 )(6.25 x 10- )

 

Ut - 2.89 ft/sec; 174 ft/min

Re - (4.13 x 10‘2)(2.89)(0.09) - 47
(1.90*x 10:57 ‘12“

Similarly, the following values of at can be determined

along with Reynolds number, Re.

 

Air Temp. Term Velocity, Ut Reynolds
0

 

 

Particle P ftZSec ft/min Number Re
Fresh Stem (l in.) 500 2.89 174 47
Dried Stem (1 in.) 300 1.83 110 36
Fresh Leaf 500 1.19 71 148

Dried Leaf 300 .78 47 117

 

141

APPENDIX D

SAMPLE CALCULATIONS FOR MASS RATE OF FLOW AND MOISTURE
REMOVAL RELATIONSHIPS

Test No.: 15c

Date: 9/22/63

Material: Fresh Birdsfoot Trefoil Leaves

Quantity: 1.0501 g (89 leaves)

Stage of Maturity: 10% Bloom, 3rd cut

Atmos. Air: Temp. - 85°F; 3.2. - 25$;f3- 0.072 1b/ft3
Air In: Mean Temp. - 242°r; R.H. - 0.8%;f3- 0.05 1b/ft3
Air Out: Mean Temp. -141°r; R.H. - 6%;f3- 0.066 lb/ft3
Fluidization Immediate

Divergence of velocity Field Data,Test 15c

 

Photocells, 1-2 2-3 3-4 4-5 5-6
Dist. Between, ft. 2.63 2.42 2.38 2.54 2.08
Oscillograph Chart

Output, Units* 4.8 5.1 5.7 6.2 6.0

Time, Secs. 0.19 0.20 0.23 0.25 0.24
Velocity, fps 13.6 11.8 10.4 10.2 8.7

 

*Oscillograph chart output speed - 25 units/sec., one unit -
i centimeter.

Time - No. units - 6.0 - 0.24 secs.
units/sec. 25—

 

Velocity - 2.08 - 8.7 ft/sec.
0‘24

lass Velocity, ib/rt2 hr. - Velocity x Density x 3600
- 8.7 x 0.066 x 3600 - 2060 111/it2 hr.

142

Moisture Removal Data, Test 15c

 

 

Time, Initial ngves Dry M.C. Leaves Sub Total

Secs. M.C. % d.b. Out ‘ % d.b. Out % Leaves out %
10 380.2 17 209.2 19.1 19.1
20 380.2 30 198.0 33.7 52.8
30 380.2 13 173.9 14.6 67.4
40 380.2 4 154.8 4.5 71.9
50 380.2 0 - - -
60 380.2 3 156.7 3.4 75.3
70 380.2 2 80.6 2.2 77.5
80 380.2 1 7.1 1.1 78.6
90 380.2 3 22.6 3.4 82.0
100 380.2 - - - -

110 380.2 - - - r

4 min 380.2 16 103.8 18.0 100.0

 

a) Weight of feed, fresh cut leaves - 1.0501 g
b) Weight of feed, bone dry - 0.2187 g
c) Weight of total H20 in feed, wb - 0.8314 g
d) Initial moisture content, % d.b. - 380.2%
e) Weight of H20 removed in process, wo-wl - 0.3729 g
f) H20 removed %, wo-wl x 100 - 44.8%
'5

Schrenk's parameter (1959) - IQ - 1
'1
wb - weight of H20 in feed 1b/lb bone dry
wl - weight of H20 in product lb/lb bone dry

wn—w] - B’O.552

'0 '0
h) _1_ - 19 - 1.815
(8) '1

8)1-(f)-1-

1) In -1 - 1.815 -1 =0.815.
'1

143

I TI"!!!

$0.0

 

.. '1 II": I.-. I! ll.

 

 

3.5 ommm n.a~ m.m~ mm
o~.o H.m omwfl o.Ha v.0H am
mo.o v.m cess o.HH v.e~ om
NH.O $.0H omen o.cH m.ca Oh
m~.o c.mH coon w.a w.mH Db
NH.o m.oH Gama m.m «.ma and
v~.o o.NH cmNN $.m «.mH nwu
mH.O m.mH ovum m.m v.mH sh
HN.m m.®H owam v.w N.¢u an
m~.o m.n~ coda o.m m.m~ on
mn.o e.NN “new v.m h.n~ an
ne.o e.mm coma N.b N.na no
me.o m.nm mama m.b n.n~ um
mv.c N.Nm mama u.v n.va 09

H- fluxes .mmusa s .8: use\ss can can .62
Hm>osem cusp-«o: ve>osem cum muss> .Blo sa< masooHe> Radoo~o> uses
no oaamuansm waxMII ocssuusm.

 

.hoOON I Hmbhcmmmlflh
.<H<D A<>Oimm mmDBmHO: 92¢ BOJh ho HB<¢ mm<l ho imdxxbm

.h~1< fldm<9

In

'w-D‘O“ .h... .

 

144

vN.c
ON.o
Hm.o
Nm.c
MH.H
nw.o
Nw.o
no.6

T Hw\0memwww
Heaosem ehsumaoa

'9"- ... '1551151 '1'. '51!’|11"1I"$!-.".i

 

 

 

 

 

 

”.md odvfl 0.0H N.nH aha
«.mfi oHnN $.0H u.vH aha
v.mN oHnN .¢.OH m.v~ aha
m.v¢ owwd w.b n.nH 00H
H.mn aflbfi ”.5 H.NA 90H
n.0v OOGH o.w Q.MH mmn
u.vv omen h.m o.nH 00H
m.mm OQON b.m u.vfi and
a ..2 new}: use .Iumm- .oz

oo>osom cm: mnes> .Blo ma< hudooHeb hedooH0> cash

no easMflMms: Waxm eosduusm

 

. .hoomu I HMDB<¢HAIMB
<B<D A<>Ofimm mmbkmnoa Q24 loam ho HP<¢ mn<fl ho Vm<xxbm

.mHI< HJm<F

145

 

vH.o

w.HN

 

 

 

omen o.ss «.sH as"

ov.c e.mm omen s.n~ «.ss new

ma.o m.nH omen p.82 «.ms can

ma.s o.nn omen H.~H n.6H con

ss.o e.nv cans H.HH m.eH new

an.~ 6.6» oven s.o~ «.sd sen

62.8 s.mm omen n.s m.ns new

as.s «.86 case 6.3 s.n~ ass
H- Hs\osnaoauan w .as sexes new new .62
12552— eusvnaom ceases—em cum mmma> .mlo s2. massed; 5:001; amen.

11:11-- 1.1: no tawnymnam wdxm soasnwmmI

 

 

 

. .hocom l Hflbfi<flflhflflk
<B<Q A<>OSM¢ HmDBmHOI 92¢ BQQh ho HF<¢ mm<l ho >¢<xlbm .QHI< mandk

.. ’1' .

 

 

 

I '

 

w .n#.. .
Ii - .
- . . 13.....1 1‘13!

 

1111
‘1. .1} I
111111.51...- 1‘! -151!

. .I
.11 a- I! ...M- .8.» $1.261nt -1
. .‘i
ii

 

"I71111111111111111“

   

..4 —~si_ _ _ — — _ _ _ - a - _ I