WI

{

l

H

W

11

WW

$

W

M

M!

L
l

 

--"I_p_.\ I
Imp ll
(0034:.

‘9 £9 {2‘3“ . t. ' R. Q" 8?- ~‘='- “zeo- "a: 2,?- “'3

but Ci: 3": S {Page'fi‘céefi‘u‘ i u “€ch E H: 1L“ géfit: 'f‘ "
Q. ' " 4‘ C. :1.“ a 41-31:“
.35?qu «3% ?Y 3?, 1-K! ”J E '“e‘i‘E $915.1.“ 5

":6 3 Kurt“? flammvn ‘:"‘"”4"! Eff-E‘f'
64‘ a“: 5‘ :2; {9; ‘E \-“4‘:éitfl’ 3L5: ‘15 ‘a'9

'E‘hssis see she wgsee o
uiSCHE‘1‘H Sf‘éi": E UMWELSET‘?
you 1 c. - rt ‘ an a
3mm mama swamp-baa?

   
     

  

LIBRARY

Michigan State
University

THE DETERMINATION OF THE DEGREE OF ANGULARITY OF,
AND THE DEVELOPMENT OF.A VELOCITY CONTROLLING FEATURE FOR,
STORAGE CHUTES IN AN AUTOMATIC ORDER PICKING DEVICE

By

Erwin Milton Campbell

AN ABSTRACT

Submitted to the College of Agriculture of
Michigan State University of Agriculture and
Applied Science in partial fulfillment of
the requirements for the degree of

MASTER OF SCIENCE
Department of Forest Products

1960

Approved}::§1gi, LAC5 «(EEYI
\

 

     

AN ABSTRACT

This project was undertaken in an effort to determine the degree
of angularity of and the development of a velocity controlling feature
for storage chutes in an automatic device that dispenses various
products in corrugated cases.

The factors studied consisted of five degrees of elevation (130
through 17°), four weights of products (2.2, 9.h, 21.9, and 24.6
pounds per linear foot), and three types of plastic covered runners.

The test results indicated that 170 was the optimum choice for
the angle of incline. Runner type C was far superior to types A and
B as a runner for light packages, but was not efficient enough to
be used on heavier packages. Both runners A and B were too efficient
to be used as runners for heavier packages.

In this report runners are rated according to their ability to
resist the movement of packages sliding down the chute. The most
efficient runners being those offering enough resistance to prevent

any unassisted package movement.

THE DETERMINATION OF THE DEGREE OF ANGULARITY OF,
AND THE DEVELOPMENT OF A VELOCITY CONTROLLING FEATURE FOR
STORAGE CHUTES IN AN AUTOMATIC ORDER PICKING DEVICE

By

Erwin Milton Campbell

A THESIS

Submitted to the College of Agriculture of
Michigan State University of Agriculture and
Applied Science in partial fulfillment of
the requirements for the degree of

MASTER OF SCIENCE

Department of Forest Products

1960

ii

ACKNOWLEDGMENTS

This study was done while the author was an employee of ADMOS,
Inc. of Detroit, Michigan and at this time he would.like to extend
his sincere gratitude to all members of that organization for all
the help and considerations received.

Thanks are also due to Dr. J. W. Goff and Dr. H. J. Raphael
for their guidance and consultations and to Dr. W. D. Baten for his

help on the statistical interpretation of this study.

ACKNOWLEDGEMENTS . .

LIST

LIST

I.

II.

III.

OF TABLES . . .

OF FIGURES

INTRODUCTION .

PROCEDURE . .

ANALYSIS OF DATA

CONCLUSIONS

TABLE OF CONTENTS

SUGGESTIONS FOR FURTHER WORK . . .

OF REFERENCES .

iii

Page
ii

iv

16
22
23
2h

Table

II.

III.

VII.

VIII.

Dimension and
Elapsed Times
Elapsed Times
Elapsed Times
Elapsed Times

Elapsed Times

LIST OF TABLES

Weights of Test Specimens

at 130 of Elevation .
at 1&0 of Elevation .
at 15° of Elevation .
at 160 of Elevation .

at 170 of Elevation .

Summary of Test Results Showing Average

by Main Effects . . . . .

Final Analysis of Variance . . . .

iv

Page

9

. . ll
. l2

. . l3
. lh

. . 15
. . l7
. . 18

LIST OF FIGURES

Figure Page
I. Diagrammatic View of Typical Runner Installation . . . . . 6
II. Laboratory Testing Rack . . . . . ... . . . . . . . . . . 7
III. Regression Curves of Total Elapsed Time Vs. Degree of
Incline for Runners A, B, & C . . . . . . . . . . . . 20
IV. Regression Curves of Total Elapsed Time Vs. Pounds Per

Linear Foot of Package for Runners, A, B, & C . . . 21

I. INTRODUCTION

The problems involved in a shipping warehouse are, as one would
imagine, many and quite complex. Not the least of these are the high
costs of labor and the total time required to assemble for shipment to
a customer any given order. As is often the case in industry, the
shipping center is viewed as a manufacturing operation that produces
customer orders as a product. In this respect the shipping manager is
given a budget with which he must manufacture this product. As in all
manufacturing operations, spiralling costs force the manager to con-
tinually strive to be more efficient by the reduction of any number of
such things as space, labor, paper work, inventory or any combinations
of these and other items in order to stay within the allowed budget.

Over the past few years there have been.many great advances in
the techniques of materiéé handling now in use in the shipping centers
,of American industry. Mechanization and automation are the key words
of progress and as a result such things are seen as smaller, more
powerful fork-lift trucks doing a wider variety of jobs, miles and
miles of conveyors of various types transporting goods from the more
remote sections of the warehouse, closed circuit T.V. systems greatly
facilitating location and inspection functions, and electronically
guided towing systems that move whole trains of loaded cars without
the benefit of an operator. These are only a few of the hundreds of
items that could have been.mentioned.

Perhaps the most revolutionary innovation in recent years in the
materials handling field was the design and construction of an auto-

matic device for the mechanized selection of orders. This machine

2
consists essentially of a hugh steel structure supporting several tiers
of inclined storage chutes. From.these gravity-fed chutes, cases of
various products are diSCharged by an electro—mechanical release
mechanism. A slightly modified standard business machine interprets
data cards, into which the customer's order has been punched, and
through a complex timing network provides the electrical impulse to
cause the chute containing the right product to release the required
number of packages. These packages are then conveyed to the shipping
dock and into a highway trailer for the trip to the buyer.

Whole volumes could be written about the many possibilities and
economies presented by this type of system. The scope of this thesis
will be limited to cover some of the problems encountered during the

design of the storage chute.

______P..-p 2: m

The purpose of this study is to develop a workable method to be
used in controlling the velocity of packages sliding down the storage
chutes. It is proposed that this be accomplished by coordinating the
effects of two unique variables. They are as follows:

1. The angle of incline of the storage chute.

2. The surface condition of the chute on which the

package must slide.

The interrelationship of these two variables on the velocity of a
sliding package is readily apparent. That is to say that, for a given
package, the steeper the angle of incline the easier it is to overcome
inertia and sliding friction. And, on the other hand, any surface

change resulting in a reluction of static or kinetic friction encountered

3

by the package could bring about a prOportional reduction in the angle
of incline, if all other factors are held constant.

The storage chutes are, of course, filled from the upper end. In
order for the corrugated cases to freely slide to the lower or dispens-
ing end of the chute a certain amount of friction must be overcome.
This friction occurs in two situations. The first situation being when
a particular chute is empty and the first package must slide the entire
length of the chute and then the next package must travel one package
length less than the preceding package and so on until the chute is
completely filled. The second situation occurs then when individual
packages are released from a full chute and the remainder of the
stored packages must slide down one package length to take up the
space previously occupied so as to be in a releasing position for the
next selection impulse should a package from that chute be required.

The design of the overall steel structure must conform to, among
other requirements, the available space within the warehouse and the
angle of incline at which the storage chutes are to be supported.

While this paper is not necessarily concerned with the overall height
requirements, the angle of the chute has a direct effect upon the

velocity of the corrugate case as it slides down the chute.

 

 

 

II. PROCEDURE

Preliminary investigations showed that on a bare metal chute with
no surface improvement, other than dusting, an angle of at least 200
was required to allow packages to easily overcome both frictional re-
strictions. This condition gave no control over impacts and it was
therefore decided to test at 130 to 170 with some type of surface
improvement so that packages could be slowed down or speeded up as
required.

Due to the wide variation in the dimensions and weights of
packages that would be handled in this type of machine, it was
originally felt that controls for at least two situations would be
needed. Light weight packages would, in most cases, need a surface
condition that would tend to keep friction at a minimum and offer as
little resistance as possible to package movement. Heavier packages,
conversely, would need a surface condition that would tend to slow
them down in order to prevent the impacts from causing product damage.
The delineation point between a light and a heavy package, based on

previous experience, was set at twelve pounds.

Runner Specifications

 

In order to keep friction at a minimum for light packages, it was
felt that some type of runners, either triangular or circular in cross-
section, would be most suitable. These two shapes would, in theory,
allow for only a line of contact between the runner and the packages.
The area of contact would be slightly modified due to runner indentation

on the package or by relaxation of the package structure caused by high

humidity or longer than usual storage time.

The material selected must either be rigid or fastened in a manner
so as to allow little lateral motion, thereby preventing packages from
becoming wedged between the runners or between one runner and the side
of a chute and, of course, the material used must be easily installed,
physically and chemically stable, and economical to obtain.

A wire, stretched the length of the chute, on which could be
strung various types of plastic tubing, appeared to fit all specifi-
cations. It was decided to test and compare three types of such
plastic covered runners as follows:

Type A - High Density Polyethylene with 5% Carbon Black
Type B - Nylon
Type C - High Density Polyethylene

Runners are to be installed in pairs, as shown in Figure One, with

a minimum of two runners per chute. Wider chutes will have as many

pairs of runners as are required to support the wider packages.

Testing Device

 

A laboratory testing device was constructed as shown in Figure
Two. The various cross members upon which the chutes are supported
are adjustable so as to accommodate several degrees of inclination.
A short, powered belt-conveyor was located at the discharge end so that
packages could automatically be released and removed in order to test
the action of the runners during intermittant operation. The runners
in the chutes, as well as the chutes themselves, are wholly inter-

changeable on this testing device.

 

 

 

 

 

DIAGRAMATIC VIEW ‘OF
RUNNER INSTALLATION '

 

 

 

LABORATORY

TESTING RACK

Products Tested

 

The packages used for testing were selected to be a representative
group. Both the length and the weight of the package are critical
factors, therefore, the number of pounds per linear foot of carton was
used as a measuring classification. The lightest package tested
weighed 2.2 pounds per foot. The next heavier package weighed 9.4 pounds
per foot. The third package weighed 21.9 pounds per foot and the
heaviest package weighed 2h.6 pounds per foot (See Table I). It was
anticipated that the more troublesome items would be the heavier
packages and, therefore, a lowrheavy package was tested as well as a
tall heavy package. No attempt was made to classify cartons by either
the Mullen test or by the carton.manufacturers, except that all test
cartons for any given product were made of similar material by the same
manufacturer. The effect of these qualities on the tests is not known

but is assumed to be negligible.

Statistical Design

 

An automatic machine such as described earlier in this paper would
be capable of selecting and dispensing cartons at speeds of fifty
cartons per minute or faster. In order to fully utilize these speeds
to advantage, the runners should not be so efficient1 so as to cause
the packages to take more than twenty seconds to travel the length of
the chute. On the other hand, the packages should not be allowed to
gain sufficient momentum that would cause heavy packages to damage the

release mechanism or light packages to jump out of the chute on impact.

 

1For the purposes of this paper, the most efficient type of
runner is defined as one whose physical and/or chemical make-up is
such that packages will not slide on them unaided.

 

 

V!
we

TABLE I

DIMENSIONS AND WEIGHTS OF

TEST SPECIMENS

 

 

 

 

 

 

DIMENSIONS WEIGHT
12 wt
L w H Wt.
(Inches) (Pounds) (lbs/ft)
PRODUCT
1. 12—3/h 11 6-1/2 26 24.6
2. 11-1/2 9-1/h 7-3/h 9 9.u
3. 13-3/h l2-3/u lO 25 21.9
h. 16-1/u 13-1/h 10-1/2 3- 2.2

 

 

 

 

 

 

 

10
As a measure to test the effectiveness of combinations of runner,
product, and degree of incline for this report, the time required for
a package to travel the length of the chute was used.
For each three way combination of product, runner, and degree of

incline, five time trials were observed and recorded.

Testing Procedure

 

The test packages were placed on the upper end of the chute and
held until a signal was given. At that time the packages were re-
leased and started down the chute. When the package reached the
bottom of the chute the elapsed time was recorded to the nearest
one—hundredth of a minute. If at the end of five minutes, however,
the package had not yet reached the bottom, that particular trial
was stopped and a time of 5.0 minutes was entered. A total of three
hundred time trials were recorded.

Testing was started with the rack set at 130 with type A runners.
After the required number of time tests, the runners were changed to
type B and then to type C. After finishing the tests on the C
runners, the angle of inclination was raised one degree at which point
all three runner types were tested again. This process was then re—
peated until all three hundred tests were conducted (See Tables II

through VI).

 

 

TABLE II

ELAPSED TIMES

 

 

 

 

 

 

13O
PRODUCTS
RUNNERS l 2 3 h
2.80 1.70 5.00 5.00
2.00 5.00 5.00 5.00
A 1.10 5.00 5.00 5.00
1.60 5.00 1.00 5.00
2.00 2.00 2.70 5.00
5.00 0.17 0.18 0.22
5.00 0.18 0.17 0.14
B 0.31 0.1M 0.11 0.13
0.31 0.12 0.12 0.13
1.60 0.11 0.13 0.13
0.11 0.19 0.16 0.25
0.17 0.11 0.10 0.28
C 0.13 0.11 0.11 0.30
0.17 0.16 0.12 0.23
0.17 0.13 0.11 0.19

 

 

 

 

 

 

 

TABLE III

ELAPSED TIMES

 

 

 

 

 

 

14O
PRODUCTS
RUNNERS 1 2 3 4
5.00 0.89 0.14 5.00
1.60 0.92 0.12 0.95
A 1.70 5.00 0.20 0.88
5.00 5.00 0.12 0.75
2.30 1.70 0.21 0.87
0.10 0.10 0.18 0.23
0.12 0.13 0.13 0.15
B 0.10 0.16 0.10 - 0.14
0.12 0.10 0.12 0.11
0.16 0.13 0.16 0.20
0.10 0.10 0.10 0.11
0.11 0.09 0.10 0.10
C 0.14 0.10 0.11 0.10
0.12 0.10 0.07 0.12
0.15 0.11 0.13 0.12

 

 

 

 

 

 

TABLE IV

ELAPSED TD/EES

O

 

 

 

 

 

 

15
PRODUCTS
RUNNERS l 2 3 4
0.18 0.10 0.17 0.48
0.25 0.15 0.16 0.75
A 0.15 5.00 0.11 0.21
5.00 5.00 0.18 0.68
5.00 0.27 0.30 0.13
5.00 0.20 5.00 5.00
0.58 0.10 0.12 5.00
B 5.00 0.10 0.50 5.00
0.11 0.11 0.14 5.00
0.09 0.95 0.11 0.13
0.09 0.09 0.08 0.12
0.08 0.09 0.08 0.09
C 0.10 0.10 0.10 0.10
0.09 0.09 0.08 0.10
0.10 0.10 0.09 0.09

 

 

 

 

 

 

TABLE V

EIAPSED TIMES

O

 

 

 

 

 

 

l6
PRODUCTS
RUNNERS l 2 3 4
0.16 0.10 0.12 0.14
5.00 0.09 0.12 0.15
A 0.28 0.11 0.11 0.13
0.13 0.11 0.10 0.15
0.17 0.15 0.13 0.17
0.22 0.15 0.18 5.00
5.00 0.12 0.15 5.00
B 0.36 0.09 0.07‘ 5.00
0.09 0.08 0.08 0.09
0.10 0.08 0.08 0.08
0.08 0.08 0.07 0.08
0.09 0.09 0.07 0.08
C 0.09 0.08 0.08 0.08
0 10 O 06 0.07 0 08
0 11 0 13 0.10 O 09

 

 

 

 

 

 

TABLE VI

ELAPSED TIMES

 

 

 

 

 

 

l7O
PRODUCTS
RUNNERS 2 3 h
.13 0.09 0.10 0.12
.13 0.09 0.10 0.13
A .13 0.11 0.09 0.13
.12 0.12 -0.09 0.16
.11 0.08 0.09 0.11
.16 0.08 0.07 5.00
.08 0.07 0.10 0.24
B .09 0.08 0.07. 5.00
.09 0.08 0.07 0.08
.09 0.07 0.09 5.00
.07 0.07 0.07 0.09
.08 0.08 0.07 0.07
C .05 0.10 0.09 0.09
.06 0.07 0.06 0.05
.10 0.07 0.05 0.07

 

 

 

 

 

 

16
III. ANALYSIS OF DATA

The data presented in Tables II through VI were subjected to a
three factor analysis of variance including products (1, 2, 3, and A),
degrees (130, lho, 150, 160, and 17°), and runners (A, B, and C). The
techniques and procedures used are those described in references
(2), (3), and (A). The results of the analysis are Shown in
Tables VII and VIII.

The analysis of variance revealed that the three—way interaction,
degree x runner x product, to be significantly different from the
error term. This led to the use of the three-way interaction mean
square as the error term to obtain F test scores. As a result of
this test it was found that only two of the three main effects,
namely degrees and runners, were significant. Using the Studentized
Range Test, the averages for degrees and runners (see Table VII)

were tested for significance between themselves.

Angle g: Incline

Of the five degree levels tested, only the average at 170 is
acceptable for being less than the time limit of twenty seconds as
described on page seven. Statistically, however, 170 was not signi-
ficantly different than 160. Further cause for rejection for 160,
as well as the more Shallow angles, is the number of five minute time
trials in each case. This is an unallowable situation. The author
is at a loss to explain why lho of inclination appeared to function

better than 150.

TABLE VII

SUMMARY OF TEST RESULTS
SHOWING AVERAGE ELAPSED TIMES
BY MAIN EFFECTS

 

 

 

 

A 1.47

Runner B 0.94
C 0.11

130 1.54

Degree of 14° 0.72
Inclination 15° 1.07
160 0.52

17° 0.34

1 0.99

Product -2 0.67
3 0.47

1 1.23

 

 

 

 

17

TABLE VIII

FINAL ANALYSIS OF VARIANCE

18

 

 

 

source DF ggfiaiis HZEEre F F0.95 F0.99
Total 299 790-99

Degrees 4 54.90 13.73 3.12* 2.78 4.22
Products 3 ' 25.95 8.65 1.96 3.01

Runners 2 95.10 47.55 10.08** 3.40 5.61
D.P 12 18.21 1.52 0.35 2.18

D.R 8 148.21 18.53 4.53** 2.36 3.36
P°R 6 42.65 7.11 1.61 2.51

D-P-R 24 105.81 4.41

Error 240 300.16 1.25

 

 

 

 

 

 

 

 

*Significant at the 5% level.

**Significant at the 1% level.

l9

Runners

When the runner averages are compared, it is very obvious that
only type C performed adequately and that the other two were far
too efficient to permit their use in this manner. The studentized
Range Test showed that C was Significantly different from A and B
and that B was significantly different from A.

Two regression curves were calculated for each runner (See
Figures 3 and A). One curve plotted total elapsed time against
degree of incline and the second curve plotted total elapsed time
against the pounds per linear foot of package. It was hOped that
these would result in fan-Shaped arrays and except for the erratic
behavior of type-B runner this was the case. When other materials
are tested, similar curves can be constructed and runner selection
could eventually, under ideal conditions, become a graphical

procedure.

 

I. In III‘I' .’ .‘I’IIII .Iv.".l.."ll||u '.

 

 

 

 

 

 

 

.. L O
70 FIGURE ONE
RUNNER A
,1 so-
If:
h: 50..
40
I" O A
RUNNER B
‘ 30...
O
A
201.
A
A
IO.
0 RUNNER c
_.EL A E j
0 I3 I4 I5 Io I1

DEGREES or ”um um:

TOTAL TIME " SECONDS

50

4O

3O

20

IO

  
 

FIGURE TWO

r—RUNNER A

RUNNER B

 

 

 

 

' A
rRUNNER P.
19 "' J
1 J J J—
IO I5 20 25

POUNDS/ FOOT or PACKAGE

22

IV. CONCLUSIONS

1. Only two of the three main effects were found to be Signi-
ficant. Of the two, only the effect of the runners was found to be
Significant at the 1% level while the effect of degrees was signi-
ficant at the 5% level only.

The effect of products did not seem to be Significant.

2. Of the three runners tested, only type C performed in a
manner that was considered acceptable. It consistently gave times
below the maximum time limit. In fact the runner was so inefficient
as to be considered unsafe to use on heavy packages. Runners A and
B, however, were found to be so efficient as not to be usable for

either weight classification.

 

23
V. SUGGESTIONS FOR FURTHER WORK

1. Conduct a similar series of tests to investigate the possi—
bility of using a C runner and an A (or B) runner in the same chute
instead of two A (or B) runners for situations requiring a more

efficient slow-down than type C.

2. Determine if there is any correlation between the co-
efficient of friction for a given material and its performance as a
runner in an effort to find a rule of thumb to accept or reject

potential runner materials.

2h

LIST OF REFERENCES

Blake, H. C., Heinlen, R. L., McClelland, J. F., and Nagy, A. J.
unpublished notes, files, and miscellaneous papers of staff
members of ADMOS, Inc., Detroit: 1959-60.

Brownlee, K. A. Industrial Experimentation. New York:
Chemical PUblishing Co., Inc., 1952.

 

Dixon, W} J. and Massey, F. J. Introduction to Statistical

 

 

Analysis. New YOrk: McGraw-Hill Book Co., IKE., 1957.

Shedecor, G. W. Statistical Methods. Ames, Iowa: Iowa State
College Press, l9h6.

 

RGBIII USE ONLY

 

MICHIGAN STATE UNIV R

III | ILIIIIIIIIIIIIIIIII“

3 1293

 

b" "Nflo v.“ -— ~df aw, “-59 ‘ - ‘o- « -.-..-.\. to .,- .