Iowa State University Capstones, Theses and

Retrospective Theses and Dissertations

Dissertations

1969

Development of a compression type corn threshing

cylinder
Robert Eugene Fox
Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd

Part of the Agriculture Commons, and the Bioresource and Agricultural Engineering Commons

Recommended Citation
Fox, Robert Eugene, "Development of a compression type corn threshing cylinder" (1969). Retrospective Theses and Dissertations.
14503.
https://lib.dr.iastate.edu/rtd/14503

This Thesis is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital
Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University Digital
Repository. For more information, please contact digirep@iastate.edu.
 DEVELOPMENT OF A COMPRESSION TYPE

CORN THRESHING CYLINDER

by

Robert Eugene Fox

A Thesis Submitted to the

Graduate Faculty in Partial Fulfillment of

The Requirements for the Degree of

MASTER OF SCIENCE

Major Subject: Agricultural Engineering

Approved:

In Charge of Major Work

Head of Major Departm-ant

Dean of Graduate College

Iowa State University

Of Science and Technology
Ames, Iowa

1969
 11

TABLE OF CONTENTS

page

INTRODUCTION 1

REVIEW OF LITERATURE 5

Results of Tests on Existing Machines 5

New Methods to Reduce Damage 6

OBJECTIVES 10

KhlRNEL DETACHMENT TESTS 11

Test Procedure 11

Results 13

DESIGN AND PRELIMINARY TESTS OF EXPERIMENTAL

THRESHER ig

Ancestor of the Design 18

Principal of Operation 18

Description of Design 20

Theoretical Considerations 26

Preliminary Tests 30

DAMAGE STUDIES 34

Damage Determination Trials 35

Damage Tests 40

Statistical model 40
Equipment 40

Combine cylinder test stand 40

Rubber roller cylinder 44
Sample dryer 44

Procedure 44
Kernel damage 46
 Ill

page

Results 48
Problems of the rubber roller sheller 56

TORQUE AND FORCE STUDIES 58

Review of Literature 58

Equipment 53

Procedure 60

Results 65

HIGH SPEED PHOTOOIAPHY STUDY 71

Review of Literature 71

Equipment 72

Procedure 72

Results 73

SUMMARY 78

CONCLUSIONS 81

RECOMMENDATIONS FOR FURTHER RESEARCH 82

fe-
BIBLIOGRAPHY 83

ACKNOWLEDGMENTS 87

APPENDIX A: DERIVATIONS OF THEORETICAL

EQUATIONS 88

APPENDIX Bs DATA RECORDED FOR DAMAGE STUDIES

FOR THE RUBBER ROLLER SHELLER 92

APPENDIX C; DATA RECORDED FOR DAMAGE STUDIES

FOR THE COMBINE CYLINDER 97
 INTRODUCTION

Corn is the predominate crop produced in the United

States. The production has more than doubled in the

last 30 years from 2.1 to 4.5 billion bushels. In the

last 15 years there has been a marked change from

harvesting corn on the ear to field shelling. In 1967

39.2 percent of the corn produced in Iowa was field

shelled and the percent is even higher in the eastern

corn belt—64 percent in Illinois and 70.9 percent in

Indiana.

Beldin and Skromme (4) report that the first

experimental corn shelling attachment for harvesting

was designed between 1940 and 1946. The first experiment

at the University of Illinois on harvesting corn by a

combine was made in 1950. A few ears of corn were tossed

into a combine and the cob came out whole and completely

stripped of corn. Encouraged by this trial the first

laboratory studies of cylinder shelling were set up

and continued through 1952. The first commercial corn

head for the combine was introduced in 1954. This switch

from picking ear corn to shelling the ear in the field

has been beneficial to the farmer in many ways;

1. The corn could be stored in smaller bins.

2, The corn could be harvested earlier and dried

 with the newly developed drying equipment.
3. The shelled corn is much easier to handle

than ear corn.

Unfortunately, however, field shelling has brought with

it one major problem--damage to the grain caused by the
impact of the combine cylinder on kernels at high moisture
content. This injury to the grain has many adverse effects,
the most important being the economic loss to the farmer.
The USDA grain standards used by most farm elevators
allow some crackage in corn without a price penalty but
they do not reflect the loss that occurs during storage.
Corn that is damaged at a high moisture content is sus

ceptible to mold growth and resulting spoilage. Mold

growth in the damaged kernels is directly related to
the amount of damage, thus drying time must be reduced
as damage is increased. Two factors affecting the quality
of stored grain are:

1- The continuing growth of the seed

2. Deterioration of the seed from the growth of
molds and fungi.

Saul and Steele (29) measured CO2 production in high

moisture grain and found that corn at a 5 percent kernel
damage level can be stored at a 28 percent moisture
content and 70° temperature for approximately 320 hours
with only a 1 percent loss in dry matter. At the same
 temperature and moisture content corn at a 30 percent
damage level can withstand only 100 hours of storage
before a 1 percent loss in dry matter occurs. Damaged
corn is also much more susceptible to attack by insects
since the pericarp or seed coat has been punctured.
A less apparent problem with high moisture field
shelling is the loss that occurs in the field. Aside
from the kernels left on the cob and broken chips
that are lost out the back of the combine or picker
sheller. Miles (17) noted that imperfect shelling occurred
when kernels were removed from the cob at high moisture
levels. The tips of the kernels were broken off during
shelling and remained intact with the cob. Johnson (12)
studied the problem further by using three treatments;
field shelling, picking and hand harvesting at several
stages of harvest. The shelled corn and ear corn were
both dried and weighed for dry matter. The results showed
that the dry matter harvested decreased in corn above 25
percent moisture content. He gives two reasons for this:
1. The tips lost due to imperfect shelling and
left in the field on the cob

2. The kernels not being physiologically mature.

The tips remaining on the cob plus chips that are lost
in the field are called invisible loss, but are definitely
caused by the cylinder.
 This study was concerned with developing a shelling

mechanism that would reduce the a-nount of damage in

shelling corn. The design, development and operating

characteristics of a low impact type sheller using rubber

rollers instead of conventional rasp-bars and concaves

are discussed.

To supplement this study, force measurements from

the concave of a threshing cylinder were studied and com

pared to the forces of the newly designed machine. High

speed photography was used to visually measure velocities

of the grain and coefficients of restitution of kernels

randomly striking the concave. The films were also

used to study the shelling process of the rubber roller

sheller.
 REVIEW OF LITERATURE

Results of Tests on Existing Machines

Much work has been done by various research workers

in testing different threshing devices and evaluating

the damage suffered by the threshed grain. Johnson (12)
observed the influence of kernel moisture content on

crackage from field shelling with a combine. As a

measure of damage, he used the percent of the grain that

would pass through a 12/64 round hole sieve. He found

that crackage increased exponentially from 0.5 percent

at 20 percent moisture content to 3.5 percent crackage
at 35 percent moisture content. Pickard (27) found the
same results using a 10/64 round hole sieve. Other

investigators visually separated all broken kernels and

chips from the whole kernels. Lamp (16), Morrison (25),
Barkstrom (3), Burrough and Garbage (5) and Waelti (32),
using visual inspection, drew similar shaped exponential
curves but found iiuch higher levels of damage than
Johnson and Pickard. They found kernel damage ranged
as high as 40 percent damage at 35 percent moisture content
Arnold and Lake (2) used acid germination for a damage
criterion and found that for a cylinder speed from 3500
to 6500 FPM the germination was reduced from 90 percent
to 70 percent, respectively. These investigations seem
to sho-w that moisture content is the most important plant

parameter with respect to seed damage.

The most significant parameter for the threshing

cylinder is cylinder speed, or more accurately peripheral

cylinder speed. Morrison (25), Goss ^ (8) and

Pickard (27) have shown that damage increases exponentially

with increasing cylinder speed. Waelti (32) has shown

that plant variety is an important factor in determining

how the grain reacts to the cylinder speed. Arnold (1)

showed for wheat that as concave clearance is diminished,

damage increases. Pickard (27) found that concave

clearance has little effect on damage to the corn.

Morrison (25) reports that the percent of unshelled corn

increases exponentially as concave clearance is widened.

New Methods to Reduce Damage

Pickard (27) conducted tests on different types of

cylinders and concave arrangements. The following

combinations of cylinder and concave bars were studied;

1. Steel angle cylinder bars and rubber concave

bars

2. Rubber-covered angle cylinder bars and rubber

concave bars

3. Steel angle cylinder bars and rasp-type concave

bars
 4. Rasp-type cylinder bars and rubber concave bars

5. Rasp-type cylinder bars and channel concave bars

6. Rasp-type cylinder bars and rasp-type concave

bars,

The corn ears with husks were fed into the combine

cylinder down an inclined chute. A 12/64 round hole sieve

was used as a measure of damage in comparing the various

configurations. The angle bar cylinder (No, 3) with

rasp-bar concave shelled so poorly it was left out of

the results* From this work i t was concluded:

!• Rasp-type cylinders appear to cause less

crackage than the angle-bar cylinders.

2. Rubber on the cylinder or concave bars appeared

to cause higher crackage.

Cylinder to concave clearance appeared to have

little influence on crackage, 5/8 inch compared

to 3/4 inch.

4. Cylinder speed increases crackage.

Morrison (25) conducted similar work comparing
rubber coated rasp—bars to steel bars. He also compared
the conventional or cage sheller to the combine cylinder
with both covered (front section only) and open concave.
He concluded:

1. Crackage increased rapidly as the cylinder

speed was increased above 45 3 RPM (2600 FPM)
 8

and was higher at the higher moisture contents.

2. Concave cover plates, covering the front section

of the concave, reduced ccackage slightly, but

not enough to justify the special installation.

3. The rasp-bar resulted in less kernel damage than

did a special rubber-covered cylinder bar.

4. Crackage was similar for the two mechanical

actions, the conventional sheller versus the

combine cylinder studied.

It was reported by Hopkins and Pickard (10) that

closing the open space between cylinder bars or increasing

the number of bars decreased shelling losses. A 22 inch

diameter cylinder with eight rasp-bars was tested at 17.5

percent kernel moisture and 37 percent cob moisture.

The shelling losses were found to be reduced at all cylinder

speeds when filler plates were used. The effect of the

filler plates upon kernel damage, however, was investigated

and found to be negligible,

A cone shelling device is being manufactured by Union

Iron Works in Decatur, Illinois. A conical cylinder rotates
inside a stationary conical cage. The two cones are in
non—parallel planes so that as the ear passes through the
machine, there is only room for the shelled cob 3t the

discharge end. The sheller is used mostly for seed com,

and no reports of high moisture tests or field tests are
available.
 Saul, Hukill and Waelti preceding and concurrent

with this investigation studied another type of shelling

device called the "squeeze sheller". This machine

consists of two endless belts rotating in opposite

directions. The ear is fed into the machine between

these two belts. The ear is rotated as i t passes through

the machine because of the ratio of the driving speeds

of the two belts. Recordings of low damage have been

obtained, but capacity seems to be a limiting factor.

 LO

OBJECTIVES

To find characteristics of the corn kernel and/or ear

that would lend themselves to reducing kernel damage

during shelling.

2. To design and build a shelling device in an attempt to

redace damage.

3. To test and evaluate the sheller with respect to a

conventional threshing cylinder.

 LI

KKRNEL DETACeiMENT TESTS

Kernel detachment force is defined as the maximum

tensile strength of the rachilla which connects the

kernel to the cob. The hypothesis was that the kernel

detachment force could be reduced by compressing the

ear.

Test Procedure

Ears used in the test were randomly selected and

hand picked from a corn harvesting experimental plot.
The variety was Pioneer 3306. Each ear of corn was
broken into two sections. A Rinc?<-Mcllwaine valve spring
tester. Figure 1, was used to put one section under 250
pounds of force in such a manner as to put each row of
kernels under an approximately equal force an equal
number of times. This was done by rotating the ear
and compressing each row when it was perpendicular to the
line of action of the applied force. A rotation of 180
degrees was considered one cycle. The detachment force
for each section was measured with a Chatillon dial
push/pull gage. Model DPP-10. The sharp hook at the end
of the pull rod was pushed into the soft embryo region of
each kernel, shown in Figure 2, and pulled rapidly.
The maximum force required was registered on the dial of
 12

Figure !♦ Rinck-Mcllwaine valve spring

tester used to compress ears

Figure 2, Chatillon Dial push/pull gage

as used in kernel detachment tests
 13

the gage. This force was recorded, the dial was released

and the next measurement of kernel strength taken. The

kernels close to the broken section were removed by hand

so that any physiological changes due to breaking along

the length of the ear would not enter the results. The

kernels were pulled radially from the cob in one series

of tests and in the direction tangent to the circumference

of the ear in another s e r i e s of tests. To eliminate friction

forces between kernels, a row of kernels along the axis of

the ear was removed before readings were taken, thus

eliminating the wedging effect of kernels. This was

repeated throughout the harvest season to obtain the

effect of moisture content on the force required to rupture

the rachilla.

Results

The results of the radial tests are shown in Figure 3

where each point represents an average of 10 replicates.

There is a marked difference between the uncompressed and

compressed sections. This may be due to a fatiguing

effect on the rachilla. Figure 4 shows that the tangential

force required to break the rachilla was also greater for

the uncompressed section of the ear than the compressed

section. Also as the moisture of the grain decreased, the

radial force decreased as well as the tangential force.

 14

e UNCOMPRESSED

e
O COMPRESSED

. 4
w
m
ij

pa
o
P5
O

I-:]
o
<
H
Q

j.
20 25 30 35 40

MOISTURE CONTENT, %
Figure 3, Effect of moisture content on radial detachment
force
 15

UNCOMPRESSED

O COMPRESSED

U5

S 4
w
0
01
o

w
o

<
E-<

20 25 30 35 40

MOISTURE CONTENT, %
Figure 4 Effect of moisture content on tangential
detachment force
 16

The rachilla apparently changes physiologically during the

drying period losing its tensile strength.

The radial detachment force at varying numbers of

cycles for 28 percent moisture corn is shown in Figure 5.

The repeated application of the compressive force on the

kernel could possibly cause rachilla failure if enough

cycles were applied to exceed the rachilla fiber endurance

limit.
 17

7 -

6 - 28% MOISTURE

5 -

w 4
ffl

w
u
oi
o
b
3 -
ij
<
H
Q

CYCLES

5, Relationship between radial detachment force

and number of compression cycles at 28%
moisture content
 18

DESIGN AND PRELIMINARY TESTS

OF EXPERIMENTAL THRESHER

Ancestor of the Design

The endless belt sheller developed in part by Saul

and Hukill as sighted in "Hugs and Squeezes" ears (11) was

a direct ancestor of the rubber roller sheller. The major

differences are that rollers are used instead of belts, and

the velocities are higher in the rubber roller sheller.

Principal of Operation

A schematic diagram of the rubber roller sheller

is shown in Figure 6. The corn is fed into the machine

under the feeder plate. The left roller causes the ear

to rotate in a counterclockwise direction until the ear

hits the bottom edge of the plate. This causes the

kernels at that location to receive a slight impact.

When the ear hits the right roller, it is caused to undergo
angular acceleration. The length of time it stays between
the rollers (dwell time) depends upon the velocity of the
left roller and the relative angular velocity of the two
rollers. The hypothesis is that this combination of com

pression, low impact and centrifugal force reduces the

strength of the rachilla. The wedging action of the kernels
 FEEDER PLATE

INLET

LEFT ROLLER RIGHT ROLLER

OUTLET

Figure 6. Schematic diagram of rubber roller sheller

 20

and/or centrifugal force causes failure of the weakened

rachilla and kernels are shelled with little damage.

Description of Design

The machine consists of a channel iron frame on

which two 1 1/4 inch shafts are mounted for holding the
rubber rolls as shown in Figure 7. The right shaft was
adjustably mounted so that it could be slid back and forth
to change the clearance between the two rollers as shown

in Figure 8- This was accomplished by two threaded shafts

located on each bearing support. Located along the threaded
shafts were 3/16" X 1" X 3" plates. These were used to
mount SR-4 strain gages as shown in Figure 9. From these
transducers force measurements were taken. This requires
some assumptions that are explained in the force deter

mination section, page 63.

The rollers were 10.50 X 15 M and W rubber tires
(called racing slicks), which were 10 1/2 inches wide
and 30 inches in diameter- The surface of these tires
was smooth and made of a very soft rubber compound.
Since special rims for mounting the tires were not
available, a narrower rim was used. Because of this
the surface was slightly rounded at the center as shown
in Figure 10. This probably had some effect on the shelling
efficiency. This rounding of the tread was overcome to
 21

.•a**

mp
i'
Figure 7 Frame and rubber t i r e s shown
under construction

Figure 8. Portion of frame and drive

assembly showing adjustability
of the right shaft
 22

Figure 9. SR-4 strain gage mounting

Figure 10. Rounded surface of tire as

seen from inlet
 23

some extent by shielding the edges of the rollers with 1/8

inch sheet steel to keep the ears on the shelling surface.

The rollers were driven by means of a tractor PTO

coupled directly to the left shaft. The right roller was

driven by means of a No. 60 roller chain from the left

shaft as shown in Figure 11. Relative speed of the two

rollers was adjusted by the use of drive sprockets with

varying number of teeth.

The feeder plate was designed to orientate the ears

into the machine as well as provide low levels of impact.

It was fabricated from steel deck plate shaped to the

contour of the tire as shown in Figure 12. The lower

edge of the plate was ground to a smooth parabolic shape

so as to damage as few kernels as possible. The plate

was adjustable by means of threaded rods located on the

sides of the plate.

Another feature of the feeder plate was the

implementation of an eccentric drive to "wobble" the

plate transversely across the surface of the tire as

shown in Figure 13. This was done to provide stress to

the kernel attachment point in more than one direction and

to provide a cross shear to the rachilla.

The machine was mounted on 32 inch legs and enshrouded

in a V-shaped sheet metal shell as shown in Figure 14.

The grain and cobs were collected in a metal tray under the
 24

Figure 11. Main drive assembly and electric

motor to power the "wobble" plate

Figure 12. Feeder plate covered with rub

ber belting to cushion ears
 25

Figure 13. "Wobble" plate and eccentric

drive

Figure 14. Complete shelling unit with

shield removed
 26

machine. No means were made for separating the grain from

the chaff or cob since the investigator was only interested

in the threshing ability of the machine.

Theoretical Considerations

The angular velocity of the ear passing over the

tire and under the feeder plate can be found expediently

by using the analogy of a planetary gear set- The ears

of corn act as planets, the rubber roller as the sun

gear and the plate as the ring gear as shown in Figure 15♦

At any given RPM, assuming rolling contact between the ear

and the feeder plate, the ear velocity can be calculated

using the following equation:

N 1 = • N (1)
2r

where

Nei = angular velocity of the ear, stage 1, RPM

r = radius of the ear, inches

= radius of the roller, inches

= angular velocity of the roller, RPM.

For roller speeds of 200 and 400 RPM and an ear

with a radius of one inch, the RPM of the ear will be

1500 and 3000 respectively.

As the ear is rolled into the area between the tires

(stage 2), the ear undergoes angular acceleration. As

 27

Figure 15. Angular velocity diagram with ear at

stage 1

Figure 16. Angular velocity diagram with ear at

stage 2
 28

shown in Figure 16 the resulting angular velocity depends

upon the speed of the drive tire as well as the differential

speed of the tires. The angular velocity of the ear can

be found from the following equation:

N (1+d) R
Ne2 " ^ ^ (2)

where

N^2 ~ angular velocity of the ear, stage 2, RPM

= angular velocity of the drive roller, RPM

d = differential speed ratio of the rollers

R^ == radius of the rollers, inches

r = radius of the ear, inches.

The dwell time can also be calculated from the

following equation:

^t -—Nr R
19.IL
(1-d)'
r ^
where

t = dwell time, seconds

L = distance the ear travels between the tires, inches

= speed of the drive roller, RPM

= radius of the rollers, inches

d = differential speed ratio of the roll e r s ,

For a differential ratio of 5:6, and = 200 RPM

and 400 RPM the angular velocity of the ear between the
tires is 2750 and 5500 respectively. For L = 4 inches
and R^ = 15 inches the dwell time is ,15 and .075 seconds
respectively.
 29

The theoretical maximum throughput can be calculated from

the dwell time. It is assumed that the ears feed continuously

between the rollers. The following equation applies;

3600 • W
Q= ^ (4)
where

Q •= throughput, pounds per hour

t - dwell time, seconds

= weight of grain on ear, pounds.

Under the assumption of 0.5 pound ears, the through

put is 230 bushels per hour for 200 RPM and a differential

speed ratio of 5:6 and will increase as the dwell time

is decreased.

The centrifugal force exerted on the ear can be

found from the following equation:

^ 0.0000284n^^ • r • (5)
where

= centrifugal force, pounds

N = angular velocity of the ear, RPM

r = radius to center of gravity of the kernel, inches

= weight of one kernel of grain, pounds.

The kernel weight average for the corn using in the
experiment was approximately 0.2 9 grams or .000633 pounds
and the radius to the center of gravity of the kernel was
approximately 1 inch. The centrifugal forces for drive
 30

roller speeds of 200 and 400 RPM are 0.14 and 0.55 pounds

respectively. A plot of centrifugal force versus angular

velocity of the ear is shown in Figure 17.

The derivations of the previous equations can be

found in Appendix A.

Preliminary Tests

Preliminary tests of the machine began early in

the fall of 1967. The following parameters were adjusted

1. Tire clearance

2. Speed

3. Differential speed of the rollers.

The effect of the eccentric drive was also evaluated.

The preliminary evaluation was based on the shelling

efficiency of the machine. Shelling efficiency of 95

percent was considered adequate. Shelling efficiency

for this test and all following tests is defined as:

E = I X LOO (6)
where

E = efficiency in percent

L = amount of corn left unthreshed, pounds

T = total throughput of corn, pounds.

The corn used in the preliminary tests was Pioneer

3510 and Pioneer 3306 taken from a field harvesting

 31

1.2 -

1.0 -

0.8 -

0.6 -
M
U
Pi
O
[iH

<
go.4
fa
H
X

m
u

0.2 -

2000 4000 6000 8000

ANGULAR VELOCITY OF EAR, RPM

Figure 17 Theoretical relationship of centrifugal force

(radial detachment force) and angular velocity
of the ear
 32

experiment. The moisture content varied from between 32

percent to 35 percent. The corn was hand harvested and all

husks were removed. Any insect or mold damaged corn was

discarded.

During the initial tests it was discovered that

pressure in the tires had the same effect as adjusting

the tire clearance and did not affect the relationship
of the feeder plate to the adjustable roller. For the

damage tests it was decided to use tire pressure as a

pertinent parameter. It was found that shelling efficiency-

remained constant between 5 and 15 PSI. The differential

speed of the rollers was varied to six different ratios.

At close ratios it was found that feeding the corn between

the tires was a problem. At large ratios, it was found

that the dwell time was too short to get complete shelling.
The final ratio selected was a compromise of the two that
set the differential speed at a ratio of 5:6. The fi '
shelling efficiency was consistently 100 percent over
speed ranges from 150 to 400 RPM. A "satisfactory run"
is defined as the event when the ear of corn fed into the

machine with its axis parallel to the axis of the roller.

This occurred in approximately 90 percent of the runs.

In the other 10 percent of the runs the corn wedged along
the edge of the rollers. This was caused in part by the
narrowness of the tires and also because the rollers were
 33

slightly rounded in the center.

The eccentric drive on the feeder plate was tested

and from preliminary damage counts had l i t t l e effect.

Shelling efficiencies were also unaffected by the use

of the back and forth motion of the plate.

After repeated tests of the above combinations, one

modification was made. The roughness of the metal feeder

plate seemed to be causing excessive damage. A rubber

surface was bolted on the plate which did reduce the damage
caused by the impact and rubbing of the metal surface.

From these preliminary tests which involved both

qualitative and quanitative measure a set of variables

was set forth for the damage tests as shown in Table 1.

Table 1. Variables used in damage testi

Speed Moisture Variety Tire

Content Pressure

200 RPM 22% Pioneer 3306 5 PSI

250 RPM 24% Pioneer 3510 15 PSI

300 RPM 26%

350 RPM 28%

400 RPM 32%

 34

DAMAGE STUDIES

A major problem in studying damage caused by

shelling devices is determining which kernels are injured

and which are not. Whether or not a kernel is damaged

depends to a large extent on the final use of the grain.

If the grain is to be stored, CO^ production was shown

to be a good measure of extent of damage. Saul and

Steele (29) studied the length of time damaged corn

can be safely stored by measuring CO^ given off. If

the seed is to be used for seed, various germination

tests are available. These tests, however, are time

consuming. A relatively new method of damage determination,

first used by Lakon (15), has been extensively used by

R. P. Moore (23) at North Carolina State College. He

used a chemical, 2,3,5 triphenyl tetrazoliura chloride,

which stains the viable portion of the seed a dark red

color. The staining occurs because of an oxidation

reduction reaction of the colorless solution within

the live portion of the seed. If the embryo of the seed

is dead, the staining does not occur. This procedure
requires a good knowledge of the seed parts and a great
deal of experience. More common means of measuring seed
damage are:

1. pass the grain through a sieve to screen out

 35

the broken k e r n e l s

2. Visual inspection

3, Soak the seeds in sodium hypochlorite to swell

minute cracks otherwise invisible-

Damage Determination Trials

In an attempt to find a reliable method of testing

for damage, tetrazolium staining was used. Various

degrees of damage were inflicted upon individual kernels

to see if the difference could be detected by the use

of tetrazolium. The corn was first presoftened by soaking

in wet paper for 12 hours. The corn was then dissected

longitudinally and soaked in the tetrazolium solution

for 6 hours. The outcome of this staining is shown in

Figures 18» 19, 20 and 21.

The heat damaged kernels showed a marked difference

in the stains, but all the damaged kernels were stained

a dark red. With close examination, an experienced

researcher may see some difference, but the author

could find none. Since tetrazolium has been well

correlated with germination tests, germination tests

were not tried.

Sieving the grain for damage determinations was shown

to underestimate the number of kernels that were injured

by the machine. Split seed coats and hairline cracks

 36

Figure 18. Healthy seeds after staining

 37

Figure 19. Mechanically damaged seeds

after staining
 38

vO
Figure 20. Kernels dried at 110 F.
after staining
 39

Figure 21. Kernels dried at 150 F

after staining
 40

through the endosperm as well as some broken grain were

all left as damage-free grain.

Soaking the grain with sodium hypochlorite did not

give any higher damage levels than unsoaked grain. The

soaking also proved to be time consuming.

In the final analysis of existing techniques, visual

inspection proved to be the simplest and most effective

method of measuring the injury occurring to grain during

shelling.

Damage Tests

The purpose of the damage tests was to find which

of the machine parameters selected have a significant

influence on the injury of the seeds, and to compare the

rubber roller sheller to a combine cylinder.

S t a t i s t i c a l mode1

Since nothing was known about the effects of the

variables on damage for the new machine, a factorial

experiment was used. The same model was used for both

the rubber roller sheller and for the combine cylinder.

The treatment combinations are given in Table 2. For

all damage determinations, four subsamples were taken

from each treatment.

Equipment

Combine cylinder test stand This piece of

Table 2. Statistical model showing each experimental u n i t — f o r A - moisture
content, B - variety, C - cylinder or roller speed, ana D - concave
clearance or tire pressure

^2 ^^3 ^4 ^5
^2 ^2 ®2 ^2 ^1 ^2
>^1 ^1111 "^1211 "^2111 "^2211
V Y
°2 '^1112 ^1212 "^2112 ^2212

^2 '^1121 ^1221 "^2121 "^2221 >1^

V
^2 ^1122 "^1222 "^2122 ^2222

°2

^4 '^4141 "^4241 ^5141 "^5241

°2 "^4142 "^4242 "^5142 "^5242

y y
^4151 ^4251 ^5151 ^5251
^5
°2 ^4152 ^4252 ^5152 "^5252
 42

equipment, shown in Figure 22, was built and used by

Waelti (32) for laboratory shelling investigations. He

found a good correlation between damage levels caused

by test stand and field shelling. All pertinent parts

of the shelling mechanism were John Deere 95 combine

parts and relative positions of the cylinder, beater and

concave were identical to those in a John Deere 95 combine,

A removable grain collecting pan, located at the bottom,

as shown in Figure 23 was used to collect all grain and

cob parts. A trough type ear inlet was constructed at the

front and a canvas curtain at the rear was used to absorb

energy from kernels deflected by the beater. A chain

drive between cylinder and beater increased beater RPM 40

percent over cylinder RPM. The cylinder was driven by a

tractor PTO. The variables used in the damage tests are

shown in Table 3.

Table 3. Variables used in damage tests for combine cylinder

Cylinder Moisture Variety Concave

Speed Content Clearance

300 RPM 22% Pioneer 3306 5/8 inch

400 RPM 24% Pioneer 3510 3/4 inch
500 RPM 26%

600 RPM 28%

650 RPM 32%

 43

Figure 22. Combine test stand

Figure 23. Grain pan located under test

stand
 44

Rubber roller cylinder This machine was discussed

in the section Design of Experimental Sheller, page 18.

It was driven by a tractor PTO.

Sample dryer The sample dryer was used to reduce the

moisture content of the grain so that it could be stored for

damage determinations at a later date. Figures 24 and 25

show the construction of the dryer which utilized two small

squirrel cage fans and two thermostatically controlled 250

watt heaters. The dryer has compartments for 36 samples,

however 72 samples could be dried simultaneously by placing

one sample above the other.

Procedure

The test began on October 27, 1967, and continued

throughout the harvesting season. The corn was randomly

selected from the Date of Planting experimental plots.
Waelti (32) found that date of planting had no effect on

kernel damage; this made it possible to select the ears

of the desired kernel moisture content range.

The samples were hand harvested and the husks were

removed. (Insect or mold damaged ears were discarded).

The ears were taken to the laboratory and immediately
shelled. Six ears of corn, with axis parallel to the axis
of the cylinder, were fed into the combine cylinder test
stand at one of the predetermined concave clearance settings
(5/8 and 7/8 inch). For each clearance, runs were made by
 45

Figure 24, Sample dryer showing squirrel

cage fans and sample compart
ments

Figure 25. Inside of dryer showing heaters

and thermostat control
 46

randomly choosing varieties at five different speeds and

feeding the ears laterally into the machine at approximately
two ears per second. The concave was then changed to the
other setting and the procedure repeated. This gave 20
runs for each machine at a given moisture content.

The grain was collected and placed into wire mesh

containers for drying. Another sample, taken from each

variety, was oven dried at 100° Centigrade for 48 hours for

moisture content determination. The grain samples were
dried in the sample dryer to between 15 and 17 percent
moisture content (for approximately 48 hours).
The kernels left on the cobs were shelled to secure

an estimate of shelling efficiency. This value was later

divided by the estimated number of total kernels for the six

ears and subtracted from 1 for an approximate shelling
efficiency.

Kernel damage

The damage was defined as the percent by weight of all

kernels having breaks, cracks or other injuries in their
seed coat plus the fines that would pass through a 12/64
round hole sieve. Each kernel was thoroughly inspected for
cracks or breaks in the seed coat and the damaged ones were
spearated from the whole sound kernels. A magnifying lens
and mirror aided in these observations and are shown in
Figures 26 and 27. The mirror reflected the under side of
 47

Figure 26. Author using magnifying lens

and mirror to count damaged
samples

Figure 27. View of kernels through

magnifying lens
 48

the kernel which made i t unnecessary to turn each kernel.

Results

The ability of the rubber roller sheller to reduce

damage was quite apparent. The overall comparisons at

the five moisture levels are shown in Figure 28.

The combine cylinder injured the grain in the

characteristic manner, except that damage did not increase

exponentially, but as a straight line for both speed and
moisture content, as shown in Figures 28 and 29. The concave

clearance showed no significant effect between the two

settings. None of the two factor interactions was significant

The analysis of variance is shown in Table 4.

For the rubber roller sheller, an unexpected result

was observed; injury to the seed was higher at 26 percent
moisture content than at 28 or 34 percent. See Figure 23.
The only other significant factor was the effect of speed
on injury as shown in Figure 30. As for the combine

cylinder, no two factor interactions were significant.

The analysis of variance is shown in Table 5,

The overall shelling efficiency of the combine

cylinder versus moisture content and speed is shown
graphically in Figures 31 and 32, These curves should

be kept in mind when looking at the damage results.

The shelling efficiency of the rubber roller sheller was
relatively constant at 100 percent. Exceptions to this
 49

30

25

O CYLINDER

20 Q RUBBER ROLLER

>?

w
o

e
10

"9"
e

20 25 30 35

KERNEL MOISTURE, %
Figure 28. Relationship of kernel damage and kernel
moisture content for the cylinder and
rubber roller sheller
 50

35

28% Moisture Content

/
30
/

25

^ 20

•q 15
/

10

200 400 600 800

CYLINDER SPEED, RPM

Figure 29. Relationship of kernel damage and cylinder
speed
 51

Table 4. Analysis of variance for combine cylinder

Source Degrees of Sum o f Mean

freedom squares square

Moisture Content, A 4 11664.93 2916.23*

Concave Clearance, B 1 168-18 168.18

Error (a), AB 4 907.84 226.96

Cylinder Speed, C 4 6998,23 1749.56**

Variety, D 1 41.16 41.16

AC 16 530.51 33.16

AD 4 737.11 184.27

BC 4 107.06 26.76

BD 1 34.85 34.85

CD 4 32.88 8.22

Error (b), 3 & 4

factor interactions 56 1843.23 190.38

Total 99 23065.93

** Significance at 1 percent level

* Significance at 5 percent level

 52

Table 5. Analysis of variance for rubber roller sheller

Source Degrees of Sum of Mean

freedom squares square

Moisture Content, A 4 744.03 186.01**

Concave Clearance, B 1 22.81 22.81
Error (a), AB 4 16.87 4.27

Cylinder Speed, C 4 658.42 164.60**

Variety, D 1 27.56 27.60
AC 16 357.22 22.33

AD 4 105,39 26.35

BC 4 33.96 8.49

BD 1 8.75 8.75

CD 4 61.35 15.34

Error (b), 3 & 4

factor interactions 56 293.63 22.51

Total 99 2334.99

** Significance at 1 percent level

 53

35

30

25
28% Moisture Content

20

vo
ON

w
a
15

10

j. _L

100 200 300 400 500

ROLLER SPEED, RPM

Figure 30. Relationship of kernel damage and roller
speed
 54

100

95

o
2: O
w
H
u o
M

90
W

O
!3 \
M

(fi
85

20 25 30 35

MOISTURE CONTENT, %
Figure 31. Relationship of shelling efficiency and
moisture content for combine
 55

28% Moisture Content

100

95

S3
W
H
O
H
b
90
U

O
!2;
M

ij

C/3 85

200 400 600 800

CYLINDER SPEED, RPM

Figure 32 Relationship of shelling efficiency and
cylinder speed for combine
 56

were when the ear did not feed under the feeder plate with

the ear axis parallel to the roller axis or when the ear

was so long that the tips were not in contact with the

rollers or shelling surfaces.

Cob breakage comparisons between the two machines is

illustrated in Figures 33 and 34, The rubber roller sheller

seldom broke the cobs. This would of course make separation

an easy task as well as keep the shelling efficiency close

to 100 percent.

Problems of the rubber roller sheller

The principle of the rubber roller sheller shows promise

in reducing corn damage. However, there are two design

factors that must be overcome.

All test results in this thesis were run with husked

ears of corn. Feeding problems were encountered with

unhusked ears. In very high moisture corn (above 30 percent)

the husk sometimes wrapped tightly around the ear so that

the kernels, while shelled, remained wrapped in the husks.

Other times the husks caused the ear to wedge between the
plate and the roller.

The second design factor that must be overcome is

involved with feeding the corn into the machine. In the

tests reported in this thesis all ears were hand fed into

the machine. A method of feeding the ears with their axis

parallel to the roller will have to be developed before

satisfactory field shelling can be accomplished.
 57

ft

Figure 33. Cobs following run through

rubber roller sheller

mm
Figure 34. Cobs following run through
combine cylinder
 58

TORQUE AND FORCE STUDIES

In order to compare the two machines and to learn

more about the shelling process, force measurements were

taken from six supports holding the concave of the combine

cylinder sheller and from the two supports holding the

right roller of the rubber roller sheller. Strain gage

transducers mounted on these supports were used in conjunction

with a Visacorder Oscillograph to record the measurements.

Review of Literature

A technique of measuring the forces on a concave has

been given by Krutikov (14) for cereal grains. He assumed

that the cylinder bars were continuous, that is the surface

of the cylinder was made entirely of cylinder bars. The

method involved measuring forces at six locations on the

concave, resolving these components and then calculating

a torque.

Csuk^s (7) used force measurements along with sound

recordings to compare methods of feeding cereal crops into
the thresher and found a strong correlation between the two.

Equipment

The shelling cylinder as shown in Figures 35 and 36,

 59

Figure 35. Combine shelling cylinder

Figure 36. Close-up viev; of concaves

and cylinder
 60

equipped with concave supports, was built by Waelti.

Epoxy SR-4 strain gages were mounted on the 3/16 X 1 inch

aluminum support plates as shown in Figure 38. This bridge

arrangement provided both temperature compensation as well

as cancelling any moment that might be present. No attempt

was made to combine the influence of the strain gages

within the circuit; i t was thought that six individual

channels would be much simpler to calibrate and keep

operative. Six Dana DC amplifiers provided a power supply

for the transducers. The signals were passed through six

Endevco Data Conditioners and the results were recorded

on a Model 906-C Visicorder Oscillograph shown in Figure 37

This recorder was capable of recording up to 3000 cycles

per second.

Procedure

After a 30 minute warm up period, the strain gages

were calibrated by means of a 50 pound spring balance.

One end of each plate was unbolted from the concave and

a 50 pound force was applied to the plate. Each channel

was calibrated once, and then the entire system was

rechecked. The plates were then rebolted to the concave

and each channel was zeroed in on the proper gridline.

The Pioneer 3306 variety of corn used for the tests

was visually graded so that no unusually small or large

 61

Figure 37. Model 906-C Visicorder

Oscillograph
 62

o GAGE 1 bi
a
<
o
o

TOP VIEW A
B

C
D

n->

o GAGE 2
W
o
< o
u

BOTTOM VIEW

Figure 38. Schematic diagram of strain gage

0,50

0,50 ^

0,50 -

0,50

0,50

.10 .09 .08 .07 .06 .05 .04 .03 .02 .01
TIME, SECONDS
Figure 39. Reproduction of an original force recording
for the combine cylinder
 63

ears were shelled. The husks were removed, and each ear

was individually fed into the machine at the center of

the concave. At the same instant the recording paper

drive was engaged? this brought the paper past the ultra

violet light beams at 25 inches per second. The curves

shown in Figure 39 are reproductions of the original

recordings. The two outside channels are the forces at

the rear of the concave. The next two are the center

mounting points and the innermost recordings are the

forces from the front of the concave. From these curves,

one can see that Krutikov's (14) assumption concerning

a continuous cylinder could not be used here. Peak

force values were used to find the resultant force.

A computer program was written to resolve the forces and

also to calculate the torque which is free from all para

s i t i c forces.

The same recording equipment was used to measure

the forces involved between the rolls in the rubber

roller sheller, except that a hydraulic torque-meter was

used to determine the torque. The sheller was equipped
with two strain gage plates, identical to the ones for

the combine cylinder, mounted in the right roller bearing

supports as shown in Figure 40, The roller chain that

drives the right tire was assumed to be in the same plane

as the bearing to facilitate summing moments to find the
 64

Figure 40. Strain gage location in

rubber roller sheller
 65

force between the rollers.

Results

The computer output for the combine cylinder is shown

in Table 6. The resultant force and its angle with the

vertical was calculated along with the X and Y coordinates

of a point on the line of action as shown in Figure 41.

The results with respect to moisture content show that the

forces and torque more than doubled from 18 percent to 28

percent moisture content. There is a wide range in values

for replications of the same moisture content. This may

be due to the chance involved with the angle at which the

ear is struck by the rasp-bar. It is proposed that the

increased force required to shell at high moisture contents

is dae to the higher resilience or toughness of the ear.

Thus impact shelling at low moisture contents causes less

damage simply because the ear need absorb only a small

quantity of energy to remove the kernels due to the

brittleness of the ear. Lamp (16) has shown that more

energy is required at higher moisture contents to shell

corn.

The resulting peak torques are shown in Table 7 for the

rubber roller sheller. However, no rational results were

obtained for the force measurements. The torques remained

fairly constant for the moisture contents studied. The
Table 6. Results of force studies for cylinder sheller showing moisture content,
force vector, angle 0, locating coordinates X and Y and computed torque

Moisture Force Angle 9 X Y Torque

Content (lbs.) (Degrees) Coordinate Coordinate (in.-lbs.)
(in.) (in.)

18 39.3 18.3 0.0 0.0 157.25

18 20.5 11.6 0.0 0.0 122.81

18 63.1 12.3 2.0 0.1 217.55

18 57.5 20,1 0.0 0.0 205.20

18 64.2 24.2 0,0 0.0 146.66

18 59.7 24.6 0.4 0.0 127.98

o^
18 46.3 31.3 -1.2 -0.1 84.58

18 36.8 25,8 0.0 0.0 190.88

18 78.4 30.2 0.0 0.0 88.17

18 85.9 22,2 0.0 0.0 267.25

22 75.3 26.3 6.8 -0.5 477.30

22 33.1 20.8 -1.8 -0.1 176.51

22 81,4 15.0 4.5 -0.2 507.96

22 65,2 8.3 1.9 0.0 286.36

22 70.0 17,6 0.9 0.0 235.84

22 73.2 19.7 0.0 0.0 268.36

22 70,0 29.5 0.0 0.0 211.03

22 56.4 9.2 0.0 0.0 341.28
22 142.5 23.6 6.1 -0.4 919.48
Table 6, continued

Moisture Force Angle 6 X Y Torque

Content (lbs.) (Degrees) Coordinate Coordinate (in.-lbs.)
(in.) (in.)

22 133.0 5.8 4.3 0.1 338.32

22 160.6 27.9 6.7 -0.5 1085.60

22 107.6 43.9 5.5 -0.7 1102.20

22 146.6 23.6 4.8 -0.3 1090.70

22 154.1 16.8 2.3 0.1 324.38

28 259.5 29.5 4.3 -0.3 2347,00

a>
28 208.4 24.6 3.6 -0.2 1844.03 -si

28 120.1 31.8 5.0 -0.4 1076.06

23 235.7 13.8 4.2 -0.1 1520.65

28 222.8 24.2 3.1 -0.2 2060.42

28 205.3 23.2 4.6 -0.3 1600.13

28 259.5 29.5 4.3 -0.3 2347.00

28 243.1 25.1 4.7 -0.3 1824.84

28 129.0 19.7 4.6 -0.2 916.56

28 231.4 28.0 4.4 -0.3 2020.82

 68

CYLINDER

8 in. 10 in.

Res.

CONCAVE

Figure 41. Schematic view of cylinder sheller showing

X, Y coordinates as used in force studies
 69

Table 7. Moisture content and resulting torque for the

rubber roller sheller

Moisture Content Torque (in.-lbs.)

18 181.2
CM
18 111.1
•

CO
18 132.6

18 101.2

18

22 91,6

22 102.1

22 130.2

22 107.1

22 120.2

23 153.9

28 93.6

28 162.1

23 150,2

28 141.5
 70

torques for the rubber roller sheller all include parasitic

torque where the combine cylinder did not. When the

two shellers are compared we see that shelling torque

does not increase as much with increasing moisture

content for the rabber r o l l e r sheller as it does for the

combine cylinder. Also the torques are substantially less

for the newly designed machine.

 71

HIGH SPEED PHOTOGRAPHY STUDY

The two shellers were further compared by means of

high speed photography. Up to this point no actual evidence

was available for determining just how the kernels were

separated from the cob therefore the pictures obtained

were valuable in evaluating both shelling methods. The

equipment and services of the Civil Engineering Department

and Engineering Research Institute at Iowa State University

were used in this study.

Review of Literature

There is little written material on the subject of

high speed photography of threshing grains. However, various

industrial organizations have done work in this area.

Schulze (30) has studied the threshing process of a beater

drum using a Fastax camera with a 200 fold time extension.

The upper, middle and lower sections of the concave were

arrested in photographic enlargements. Proportions of rye,

wheat and barley oats mixtures along with different types
of feeding were mainly studied. He found it was possible
to make evaluations with sharp resolution 30 to 40 cm. into

the threshing cylinder. From his photographs he was able

to determine velocities of stalks and grain as they were
struck by the beater and compare these at different concave
 72

adj ustments.

Equipment

Both the cylinder sheller and the rubber roller sheller

were used in this study. The equipment used in taking

the pictures consisted of a 16 mm. Wollensack Fastax

camera, J-410 Fastax actuator and control and high intensity

photo flood lamps. The camera was capable of taking 1000

to 8000 frames per second. At 5000 frames per second the

entire 100 foot roll of film was accelerated and decelerated

through the camera in 3,1 seconds. Kodak Tri-X and Plus-X

black and white film was used during the filming.

Procedure

Panels of both shellers were removed to expose the

shelling surfaces. No transparent protector was used to

eliminate glare and refraction from the high intensity

photo flood lamps. Lighting was most critical; the

intensity of light had to be adjusted from approximately

12,000 to 18,000 foot candles depending on film speed and

type. F~stop settings were varied from 3.5 to 8.0 in an

attempt to capture the best depth of field and contrast.

In all, 1400 feet of film were used. Speed varied from 2000

to 6000 frames per second.

The cylinder was set at 600 RPM and concave clearance

 73

was 5/8 inch. Ears were hand fed into the sheller

simultaneously as the actuator button was pushed on the

control unit. This involved good timing on the part of

both operators since the film ran through the camera in

a matter of from 2.5 to 8.0 seconds. A more sophisticated

method of feeding the sheller was not attempted since

good results were obtained with hand feeding.

Since the filming took place in early spring only 18

percent moisture content corn was available for shelling.

The ears were free of husks and were sorted to be uniform

in size.

The films were processed at the Film Production Unit

located at Iowa State University. The films were reviewed

with a Graflex-16 projector which allowed individual

study of each frame.

Results

By measuring velocities after a kernel was struck by

the rasp-bar, a measure of the elasticity of the grain can
be made. During impact, the kernel is deformed until it

reaches the velocity of the rasp-bar. The resulting

stresses and strains may lead to kernel failure. The

of the seed to withstand this impact is an important

characteristic affecting mechanical damage.
From studying the film for the combine cylinder it was
 74

noted that the corn was completely shelled after being hit

approximately seven times by the rasp-bar. The kernels

that were not directly hit by the rasp-bar or concave

seemed to be shelled by a combination of the deflection

of the cob and by the force due to acceleration imparted

to the ear by the rasp-bar.

Velocities of individual kernels taken a t random

were measured after striking the rasp-bar. Table 8 shows

the bar velocity, kernel velocity and an index of elasticity.

Elastic index can be calculated by using the following

equation:

^kf ~ "^rf
Elastic index = — —rr— (7)
ri ki

where

= initial velocity of rasp-bar

= initial velocity of kernel

= final velocity of rasp-bar

= final velocity of kernel.

Since the kernels did not hit obliquely to a surface, and

compound angles were involved, the elastic index is only

an indication of the coefficient of restitution.

Accelerations were impossible to obtain since the

kernels were accelerated in less time than the time

represented by one frame.

Figures 42 and 43 represent individual frames taken

at 3000 and 5000 frames per second.

 75

Figure 42. Individual frame taken at

5000 frames per second

Figure 43- Individual frame taken at

3000 frames per second
 76

Table 8, peripheral speed, velocity of kernel and

elastic index as found by high speed photo
graphy study

Peripheral Speed Velocity of Kernel Elastic Index

in./sec. in./sec.

690 1250 .812

690 1125 .630

690 975 .400

690 1125 ,630

690 1142 .656

690 1184 .713

690 844 .208

690 1030 ,456

The film of the rubber roller sheller was quite revealing.

The most obvious characteristic was the way in which the ear

moved through the rolls. The ear rolled with little slippage

between the feeder plate, and the roll, but when the ear

ca-ne to the end of the plate a great deal of slippage

occurred and much of the shelling occurred at the point of

the feeder plate. Individual frames in Figures 44 and 45

show this characteristic. As stated previously, there was

no chance of filming high moisture corn and for that reason

it can only be inferred that this was happening for other

moisture contents.
 77

Figure 44. Individual frame showing ear

passing under feeder plate

Figure 45. Individual frame showing ear

being shelled at the point of
the feeder plate
 78

SUMMARY

Kernel detachment tests revealed that compression of

the ear lessened the radial and tangential force required

to remove the kernel from the cob. The force continued

to dacline as the number of compression cycles increased.

A machine was designed and built to use the effect

of compression to separate the kernels from the ear. It

basically consisted of two rubber rolls mounted in a steel

frame with the axis of the tires parallel. The two tires,

both turning in the same direction with a differential

speed of 5:6, were powered by means of a PTO connection.

The ear was fed between the two rolls by means of a

feeder plate wrapped over the top quadrant of the faster

rotating tire.

Damage studies were run using both the rubber roller

sheller and a combine cylinder. Moisture content, cylinder

speed, concave clearance and variety were varied for

both shellers. Damage was determined by visual inspection

after investigations with various techniques such as

tetrazolium staining. The rubber roller sheller was more

effective in reducing damage than the cylinder at all

settings, being approximately three times as effective

at the 30 percent moisture content level. Concave clearance

and variety were not significant in the damage tests.

 79

Force and torque for the cylinder were studied by

means of six strain gage transducers located on the

concave support brackets. These strains were transformed

into force and then by computer into torque readings.

For the rubber roller sheller, torque values were recorded

by means of a hydraulic torque meter. The recorded

torques revealed that the rubber roller sheller took less

torque than the conventional cylinder at 28 percent

moisture content level—1700 inch-pounds for the con

ventional sheller compared to 150 inch-pounds for the

rubber roller sheller.

High speed photography was also implemented in an

attempt to learn more about the threshing process in

both machines. For 18 percent grain it was found that

mvost of the shelling occurred in approximately one

revolution of the combine cylinder. The grain that did

not actually come in contact with the cylinder or the

concave seemed to be removed by acceleration force and

the deflection of the cob. The films of the rubber

roller sheller showed that the ear did not feed between

the rollers as had been thought. Much of the shelling

occurred at the feeder plate rather than between the

two rollers for the 18 percent grain.

Velocities were measured for the cylinder sheller

and an indication of the coefficient of restitution was
 80

found. These ranged from .208 to .812, a value of 1.0

being perfectly elastic. Compound angles mads an exact

value impossible to obtain.

 81

CONCLUSIONS

1. In the tests conducted compressing an ear of corn

reduced the necessary force to remove its kernels.

A repetition of this compression will further reduce

kernel detachment force.

2. The rubber roller sheller caused less kernel damage

at the moisture contents studied than a conventional

combine cylinder.

3. Cobs were infrequently broken during the shelling

process for the rubber roller sheller.

4. Husks and stalks caused problems during the shelling

process for the rubber roller sheller.

5. In a combine forces between the ear and the concave

vary widely at the moisture contents studied, being

greater at higher moisture contents.

 82

RKCOMMEIS[D?^TIONS FOR FURTHER RESEARCH

The author makes the following suggestions for further

research on reducing threshing damage to corn.

1. Investigate the use of wider rollers with different

surface textures for shelling.

2. Investigate methods of removing husks and foreign

material from the ear before shelling.

3. Investigate the use of ultra high speed telephotic

photography on high moisture corn in an attempt

to find the exact cause of damage on individual

kernels.

4. Investigate the theory that the rachilla may fail

by fatigue under a repeated compression load.

 83

BIBLIOGRAPHY

1. Arnold, R, E. Experiments with rasp bar threshing

drums, journal of Agricultural Engineering
Research 9; 99-131. 1964.

2. Arnold, R. E. and j. R. Lake. Experiments with

rasp bar threshing drums. II. Comparison
of open and closed concaves. Journal of
Agricultural Engineering Research 9: 250-251.
1964.

3. Barkstrom, Ray. Further development and requirements

in field shelling of corn. American Society of
Agricultural Engineers paper 63-647. 1963.

4. Beldin, R. L. and A, B. Skromme. Shelling attachment

for mounted corn picker. Agricultural Engineering
40: 87-91. 1959.

5. Burrough, D. E. and R. p. Harbage. Performance of

a corn picker-sheller. Agricultural Engineering
34: 21-22. 1953.

6. Cochran, William G. and Gertrude M. Cox. Experimental

designs. 2nd ed. New York, N,Y., John Wiley
and Sons, Inc. 1968.

7. csuk&s, L. Examination of the flow of the crop

with combine harvesters with special respect
to the possible increase of performance
(translated title). Anyagiramlfts vysg&lata
aratfecsfeplbgfepeknfel tekintettel a teljesitmfeny-
nftvelfes lehetbsfegeire. Jirmiivek, Mezbgazdas&gi
Gfepek 11, No. 3: 90-97. 1964.

8. Goss, j. R,, R. Bainer, R. G. Curley and D. G.

Smeltzer. Field tests on combines in corn.
Agricultural Engineering 36: 794-796. 1955.
9. Higdon, Archie, William B. Stiles, Arthur W. Davis
and John A. Weese. Engineering Mechanics
Statics and Dynamics. Vector Edition.
Englewood cliffs, N.J., Prentice-Hall, Inc.
1962.
 84

10. Hopkins, D. F- and G. E. Pickard. Corn shelling

with a combine cylinder. Agricultural Engineering
34: 461-464. 1953.

11. "Hugs and squeezes" ears. Agricultural Research

16, No. 5: 7. 1967.

12. Johnson, W. H. Corn harvesting performance at

various dates. American Society of Agricultural
Engineers Transactions 6: 268-272. 1963.

13. Johnson, William H. and Benson J. Lamp. Principles,

equipment and systems for corn harvesting.
Wooster, Ohio, Agricultural Consulting Associates,
Inc. 1966.

14. Krutikov, I. A. The development of criteria for

the assessment of rational selection of parameters
of threshing mechanisms (translated title).
K razrabotke kriteriev dlya otsenki i obosnovaniya
parametrov molotilnykh ustroistv. Vest, sel.-
khoz. Nauki, Mosk. 9, No. 11: 138-144. 1964.

15. Lakon, George. The topographical tetrazolium method

for determining the germinating capacity of
seeds. plant physiolog^y 24: 389-394. 1949.

16. Lamp, B. J. Corn harvesting 1959-1960: departmental

report, Columbus, Ohio, Agricultural Engineering
Department, Ohio State University. 1960.

17. Miles, S. R. Maturity of corn in relation to field

shelling. Conference on Field Shelling and
Drying Corn Proceedings 1956: 27-39. 1956,

18. Mischke, Charles R. Elements of mechanical analysis.

Reading, Massachusetts, Addison-Wesley publishing
Company, Inc. cl963.

19. Moore, R, p. Mechanical injury, physiological

aging and seed quality. Unpublished mimeographed
paper. Raliegh, North Carolina, Department
of Field Crops, North Carolina State College.
1967.

20. Moore, R. p, previous history of seed lots and

differential maintenance of seed viability and
vigor in storage. International Seed Testing
Association Proceedings 28: 691-699. 1963.
 85

21. Moore, R. P. Seed and seedling decay in germination

tests. Seed Technologist News 33: 9-10. 1964.

22. Moore, R. P. Seed vigor or soundness and corn

improvement. Unpublished mimeographed paper
prepared for the Eighteenth Annual Hybrid Corn
Industry Research Conference, Chicago, Illinois,
Dec., 1963. Raliegh, North Carolina, Department
of Field Crops, North Carolina State College.
1964.

23. Moore, R. P. TZ checks your seed for quality.

Crops and Soils 15: 10-12. 1962.

24. Moore, R. P. and Ewald Smith. Tetrazolium—A useful

research tool in studying causes for seed
germination difficulties. Unpublished mimeographed
paper presented at the Thirteenth Annual Meeting
of the Association of Seed Control Officials of
the Southern States, Nashville, Tennessee, Nov.,
1956. Raliegh, North Carolina, Department of
Field Crops, North Carolina State College. 1957.

25. Morrison, C. S. Attachment for combining corn.

Agricultural Engineering 36: 792-794. 1955.

26. Perry, C. C. and V. C. Lissner. The strain gage

primer. 1st ed. New York, N.Y., McGraw-Hill
Book Company, Inc. 1955.

27. Pickard, G. E. Laboratory studies of corn combining.

Agricultural Engineering 36: 792-794. 1955.

28. Saul, R. A. Rate of deterioration of shelled corn.

Iowa Farm Science 22: 21-23. 1967,

29. Saul, ^^obert A. and James L. Steele. Damaged shelled

corn costs more to harvest and dry. American
Society of Agricultural Engineers paper 65-146.
1965.

30. Schulze, Karl-Heinrich. Kinematic study of the

threshing process in a beater drum (translated
title). Kinematographische untersuchung des
dreschvarganges in einer schlaglerstentrommel.
Grundlagen Der Landtechnik 7, No. 13: 113-120.
1956.
 86

31. Steel, Robert G. D. and James H, Torrie. Principles

and procedures of statistics. New York, N.Y.,
McGraw-Hill Book Company, Inc. 1960.

32. Waelti, H. Physical properties and morphological

characteristics of maize and their influence
on threshing injury of kernels. Unpublished
Ph.D. thesis, Ames, Iowa, Library, Iowa State
University. 1967.
 87

AC KN0WLED3MENTS

The author wishes to express appreciation toi

Doctor Wesley F. Buchele, major professor, for his

valuable suggestions and guidance throughout this research

endeavor and his patience during the preparation of this
thesis.

Doctor Steve Marley for his guidance in completing

this thesis and also for his assistance as a committee member,
Doctors C. W. Bockhop and Roger Mitchell for their

assistance as committee members.

The author's fellow graduate students for their

suggestions and criticisms.

Mr, Robert Fish and Mr. Rex Whitcanack for their

assistance in construction of the rubber roller sheller.

Dave Rix, undergraduate, for his help as project
technician.

My wife, Barbara, for her assistance in typing and

her patience during my long hours away from home.
 88

APPENDIX A: DERIVATION OF THEORETICAL EQUATIONS

N
el, =
^ . N
r (I)

I1
1 I

v/here

N angular velocity of the ear, stage 1, RPM

el

r radius of the ear, inches

radius of the roller, inches

N angular velocity of the roller, RPM.

now:
=
Nr • R
r

also: V + N • :
c el

and: V = N , • r
c el

bo: N - • r + N • r = 2N ,
el e
I l el
equating: Nr - R
r
=

^el == 2
^r .
or t

r
 89

O
>
II

R_
Ne2 = (2)

where

^2 = angular velocity of the ear, stage 2, RPM

= angular velocity of the drive roller, RPM

= differential speed ratio of the rollers

= radius of the rollers, inches

- radius of the ear, inches.

now: s= N •
r

and: = • d •

also: = V +
c ^e2 ' r

and: = V - • r
c

equating:
\2
 90

and: N 'd-R = }^ • r - V
r r e2 c

- N^2 . r = N^2 . r - . d -
from this I N_^
'e2
= ^r * ^ * ^r"*" ^r * ^r
2r

or: Ng2 = ^

N^R^(l-d)
where

t s= dwell time, seconds

L = distance the ear travels between the tires, inches

= speed of the drive roller, RPM

R^ = radius of the rollers, inches

d = differential speed ratio of the rollers,

from equation 2x V = N R - N (1+d) Rj- ^

err ^ . I.
2 r
rV

V = W - ^r\^
^2 2
N R (1-d)
= r r ^ '
2

then: ^ ^ ^ 2l ^
N R (1-d
c r r ^ '
converting units: t =
r r

3600 • W
Q= 1
where

Q = throughput, pounds per hour

t = dwell time, seconds
 91

W = weight of grain on ear, pounds.

The number of ears per hour at a continuous feeding

rate would bet

3600

If their average weight is W^i

3600 • W
Q = z ^
 92

APPENDIX B: DATA RECORDED FOR DAMAGE STUDIES

FOR THE RUBBER ROLLER SHELLER

Key to damage studies for the rubber roller

sheller

Moisture Content Key

22% 1

24% 2

26% 3

28% 4

32% 5

Variety Key

Pioneer 3306 1

Pioneer 3510 2

Roller Speed Key

200 RPM 1

250 RPM 2

300 RPM 3

350 RPM 4

400 RPM 5

Tire Pressure Key

5 PSI 1

15 PSI 2

Figure 46. Damage studies for the rubber roller sheller

(Each damage recording is an average of 4
replicates)
.^OISTu«E CONTENT CONCAVE CLEARANCE CYLINDER SPEED VARIETY KEPLICATES DAMAGE
PtKCENT

1 i 4 9 22
1 2 4 & 93
2 I 4 6 07
2 2 4 6 48
3 1 4 7 86
3 2 4 6 87
4 1 4 8 89
4 2 4 7 41
6 1 4 9 60
5 2 4 15 67
1 1 4 ti 25
1 2 4 d 55
V£>
2 1 4 6 05 OJ
2 2 4 6 63
3 1 4 7 94
3 2 4 6 83
4 1 4 8 60
4 2 4 9 40
6 i 4 il 13
5 2 4 14 57
1 1 4 8 23
1 2 4 5 3^
2 1 4 9 96
2 2 4 4 86
3 1 4 10 96
3 2 4 6 93
4 1 4 11 19
4 2 4 8 63
5 X 4 10 05

Figure 46, continued

MOISTURE CONTENT CONCAVE CLEARANCE CYLINDER SPEED VARIETY REPLICATES DAMAGE
PERCENT

2 1 5 2 4 6 80
2 ? 1 1 4 6 03
2 2 I 2 4 5 92
2 2 2 1 4 9 82
2 2 2 2 4 6 50
2 2 3 1 4 10 93
2 2 3 2 4 10 36
2 2 4 1 4 11 2b
2 2 4 2 4 a 5b
2 2 5 1 4 10 28
2 2 2 4 y 89
3 1 1 1 4 b 62
\£>
3 1 1 2 4 7 13
3 1 2 1 4 10 61
3 1 2 2 4 13 42
3 1 3 1 4 10 49
3 1 3 2 4 7 07
3 1 4 1 4 10 17
3 1 4 2 4 11 97
3 1 5 1 4 12 76
3 I 5 2 4 10 30
3 2 I 1 4 a 24
3 2 1 2 4 7 IB
3 2 2 i 4 10 05
3 2 2 2 4 11 95
3 2 3 1 4 11 15
3 2 3 2 4 7 23
3 2 1 4 10 49
3 2 4 2 4 10 19
Figure 46. continued
 {E CONTENT CONCAVE CLEAKANCE CYLINDER SPEED VARIETY KEPLICATES damage
PEKCENT

*>
3 5 1 4 12.76
3 2 & 2 4 14.62
4 1 1 1 4 7 • 64
4 1 2 4 6. 73
4 1 2 1 4 &«84
4 1 2 2 4 6.84
4 1 3 1 4 6.3y
4 1 3 2 4 7.27
4 1 4 i 4 12.06
4 I 4 2 4 9.49
4 1 5 1 4 7*66
4 1 5 2 4 10.57
V£>
4 2 1 1 4 7.39 cn

4 2 1 2 4 7.90
4 2 2 1 4 7.84
4 2 2 2 4 7.68
4 2 3 1 4 9.31
4 2 3 2 4 8.87
4 2 4 1 4 10.56
4 2 4 2 4 9.53
4 2 5 1 4 10.35
4 2 5 2 4 10.35
5 1 1 1 4 3.54
5 1 i 2 4 4.07
5 1 2 1 4 5.69
5 1 2 2 4 5 • 44
5 1 3 1 . 4 6.09
5 1 3 2 4 6*23
5 1 4 1 4 7.84

Figure 46. continued

MOISTURE CONTENT CONCAVE CLEARANCE CYLINOEK S^^££D VAKIETY KEPLICATES DAMAGE
PEKCENT

7.07
7. 78
7 • .16
50
4. 19
4.05
5.40
5.60
6.74
7.32
5. 14
8.85
6.b5 cr>

Figure 46. continued

 97

APPENDIX C: DATA RECORDED FOR DAMAGE STUDIES

FOR THE COMBINE CYLINDER

Key to damage studies for the combine

cylinder

Moisture Content Key

22% 1

24% 2

26% 3

28% 4

32% 5

Variety Key
Pioneer 3306 1

Pioneer 3510 2

Cylinder Speed Key

300 RPM 1

400 RPM 2

500 RPM 3

600 RPM 4

650 RPM 5

Concave Clearance Key

5/8 inch 1

3/4 inch 2

Figure 47, Damage studies for the combine cylinder

(Each damage recording is an average of 4
replicates)
 MOISTURE CONTENT CLEARANCE CYLINDER SPEED VARIETY replicates DAMAGc.
PERCENT

1 1 1 4 ^2.94
1 1 2 4 26*45
1 2 1 4 32.6a
1 2 2 4 29«7tJ
1 3 1 4 37.90
1 3 2 4 37.53
I 4 I 4 36.21
1 4 2 4 35.26
I 5 1 4 37.61
1 5 2 4 39.05
2 1 1 4 Iti.lO
2 1 2 4 25. £»9
VD
2 2 i 4 24. 57 CO

2 2 2 4 26.99
2 3 i 4 23.59
2 3 2 4 24.61
2 4 I 4 25«54
2 4 Z 4 34. 12
2 S 1 4 33.47
2 5 2 4 33.77
1 1 1 4 20.54
1 1 2 4 19.27
1 2 1 4 22.31
1 2 2 4 23.84
1 3 1 4 • 23.54
1 3 2 4 28.86
1 4 X 4 28.63
1 4 2 4 25.66
1 5 \ 4 36.70

Figure 47. continued

 MOISTURE CONTENT CONCAVE CLEARANCE CYLINOEK SPEED VAKIETY REPLICATES DAMAGE
hekcent

2 1 5 2 4 30.01
2 2 1 1 4 20*83
2 2 1 2 4 23.81
2 2 2 1 4 29.49
2 2 2 2 4 23.42
2 2 3 I if 30. 14
2 2 3 2 4 23.42
2 2 i* 1 4 23.68
2 2 U 2 4 22.67
2 2 5 1 4 34.21
2 2 b 4 29.5b
3 1 1 4 19.74
3 1 1 2 4 10.00 yD

3 1 2 1 4 17.54
3 1 2 2 4 19.04
3 3 i 4 23.63
3 1 3 2 4 19.3ti
3 1 1 A 23.11
3 1 4 2 4 22.07
3 1 5 1 4 28.59
3 1 5 2 4 31.20
3 2 1 1 4 17.91
3 2 1 2 4 11.12
3 ?. 2 1 4 23.97
3 2 2 2 4 17.76
3 2 3 1 4 23.69
3 2 3 2. 4 23*01
3 2 4 1 4 26.38
3 2 2 4 26.97

Figure 47. continued

 MOISTURE CONTENT CONCAVE CLEARANCE Cylinder speed VARIETY REPLICATES DAMAGE
PtKCtNT

3 2 5 i 4 27.99
3 2 5 2 4 26.94
4 1 1 i 4 15.45
4 1 2 4 17.21
4 1 1 4 6 # 67
4 1 2 2 4 20.32
4 1 3 1 4 10.33
4 1 3 2 4 23.60
4 1 4 4 16.99
4 1 4 2 4 24.86
4 1 5 1 4 23.90
4 1 5 2 4 2«.37
4 2 1 X 4 14.09 o
o
4 2 1 2 4 14. 24
4 2 2 1 4 15.44
4 2 2 2 4 16.16
4 2 3 1 4 20.57
4 2 3 2 4 23. 89
4 2 4 1 4 23. 18
4 2 4 2 4 24.8i
4 2 5 1 4 23.89
4 2 5 2 4 24.48
5 1 1 i 4 6.81
5 1 1 2 4 9.23
5 1 2 1 4 11.49
5 1 2 2 4 12.65
5 1 3 1 4 15.69
5 1 3 . 2 4 17.84
5 I 4 1 4 20.83

Figure 47. continued

MOISTURE CONTENT CONCAVE CLEARANCE CYLINDER SPEED VARIETY RtPLlCATES DAMAGE
PERCENT

5 1 4 2 4 21.22
5 1 5 i 4 21.50
5 1 b 2 4 21.27
5 2 1 1 4 5.40
5 2 I 2 4 6.56
5 2 2 1 4 9.
5 2 2 2 4 11.26
5 2 3 I 4 14.18
5 2 3 2 4 14.70
5 2 4 1 4 16.63
5 2 4 2 4 16.6b
5 2 5 1 4 17.37
5 2 5 2 4 18.62 o

Figure 47. continued