NPz^^

Bulletin of the University of Utah

Vol. 53 November 1964 No. 26

STORAGE AND FLOW 0 F SOLIDS

by

Andrew W. Jenike

Bulletin No. 123

of the
UTAH ENGINEERING EXPERIMENT STATION

Sixth Printing
(revised)
March 1970

University of Utah
Salt Lake City, Utah

'fts^'"'^'"''
 DISCLAIMER

This report was prepared as an account of work sponsored by an

agency of the United States Government. Neither the United States
Government nor any agency Thereof, nor any of their employees,
makes any warranty, express or implied, or assumes any legal
liability or responsibility for the accuracy, completeness, or
usefulness of any information, apparatus, product, or process
disclosed, or represents that its use would not infringe privately
owned rights. Reference herein to any specific commercial product,
process, or service by trade name, trademark, manufacturer, or
otherwise does not necessarily constitute or imply its endorsement,
recommendation, or favoring by the United States Government or any
agency thereof. The views and opinions of authors expressed herein
do not necessarily state or reflect those of the United States
Government or any agency thereof.
 DISCLAIMER

Portions of this document may be illegible in

electronic image products. Images are produced
from the best available original document.
 PREFACE

There hardly is an industry which does not store and handle solid

materials in bulk form. When the volume of the solids is large, gravity

is invariably relied upon to cause the solids to flow out of storage,

through channels and reactors. Such materials as ore, coal, ccanent,

flour, polypropylene, clay, soil, to which a general term of bulk solids

is applied, flow by gravity or are expected to flow by gravity in

thousands of installations and by the billions of tons annually. Mining

relies on gravity flow in block-caving, in ore passes, as well as in

storage and loading bins. Agriculture uses gravity flow in elevators, in

feed plants, and in silos. The chemical and process industry depends on

gravity flow of solids in process and out of storage.

This bulletin follows Bulletin No. 108 which was published in 1961.

The present aim is to provide the engineer with enough information to

enable him to design storage plants and flow channels for unobstructed

flow. Only an outline of the theory of flow is given.

While the design procedure remains essentially unchanged. Important

refinements have been Introduced In the past three years. These refine-

ments have been based on both, further developments in the theory of flow,

and an expanded background of industrial experience. In particular, the

1
calculations of the flow-factor for no-doming have been refined and the

formula for the width of an outlet is more conservative. The interpreta-

tion of the tests of flow-functions has been modified, better to account

for the effect of capillary forces. The evaluation of the angle of

friction between a flowing solid and a wall material has also been altered

and permits a less conservative design.

A simplified method of measuring the flow-function of a solid has

been introduced. This method saves a great deal of time In separating

free-flowing from non free-flowing solids, and in intricate testing

problems.

The subject matter is organized in three chapters. Chapter I

introduces the concepts of flowability of bulk solids and of channels, and

the flow - no flow postulate. Chapter 11 describes the testing apparatus

and procedure, and Chapter 111 contains applications to design. An out-

line of the derivation of channel flow-factors and flow - no flow formulas

is given in an Appendix.

While the concepts of this theory were formulated by the author in

1953, most of the work was carried out at the Bulk Solids Flow Laboratory

(1957-1962) of the Utah Engineering Experiment Station, University of Utah,

where the author had the satisfaction of working with Dr. Jerry R. Johanson,

who was then a graduate student. The author is grateful to Dr. P. J. Elsey,

Assistant Director of the Station for his constant and sympathetic Interest

in the project, to Professor R. H. Woolley of the University of Utah for

his assistance in setting up the Bulk Solids Flow Laboratory, and to

Professor R. T. Shield of Brown University for the many long discussions of

the topics of plasticity.

11
 The cost of this project was substantial and the author grate-

fully acknowledges the initial support which he received from the

American Institute of Mining Mineraloglcal and Petroleum Engineers,

whose Mineral Beneficiation, Coal, and Research Committees were in-

strumental in providing AIME endorsement of the project in 1954. This

was followed by a grant of money from Engineering Foundation and in 1956

by the support from research funds of the Utah Engineering Experiment

Station. The AIME and the Engineering Foundation remained sponsors of

the project to its completion in December, 1962.

Sincere thanks are due to Dr. S. S. Kistler, Dean of the College

of Engineering, and to Dr. Carl J. Chrlstensen, Director of the Utah

Engineering Experiment Station, for their support in initiating the

project and for keeping it alive through times of financial stress.

The main support for the applied part of the project, entitled

"Bulk Solids Flow" came from the American Iron and Steel Institute, to

whom the author Is most grateful.

The mathematical concepts described In this report, as well as

other work which has appeared elsewhere, were worked out under a 1959

grant from the National Science Foundation to a project entitled:

"Flow of rigid-plastic solids in converging channels under the action

of body forces."

Andrew W. Jenike
Winchester, Mass.
November, 1964

ill
 CONTENTS

£££•
PREFACE 1

NOTATIONS viii

CHAPTER I, FLOWABILITY OF BULK SOLIDS AND OF CHANNELS 1

Introduction 1
Obstructions to flow 2
Angle of repose 6
Effective area of an outlet 6
Flow patterns 10
Flowability of solids 10
Comparison of solids and liquids 10
Effective yield locus, EYL 11
Yield locus, YL 14
Kinematic angle of Internal friction, ^ 16
Time yield locus, TYL 18
Unconflned yield forces, F and F 18
Static angle of Internal friction, ^ 20
Wall yield locus, WYL 20
Kinematic angle of friction between a solid and a
channel wall, ^' 20
Flow-function, FF 22
Particle size 26
Flowability of channels 28
Flow-factor, ff 28
Solids in a channel 31
Flow - no flow postulate 31
Dimensions of an outlet 31

CHAPTER II, TESTING THE FLOW PROPERTIES O'P BULK SOLIDS 33

Apparatus 33
Purpose of tests 35
Test of specimens 38

V
Continuous tlow 40
Consolidation 40
Shear 42
Example 45
Time effect 48
Density 49
Kinematic angle of friction between a solid and a
wall material, ^' 49
Simplified testing procedure 52
Slip-stick solids 54
General classification of flowability of solids 56

CHAPTER 111, DESIGN 57

Funnel-flow bins 57
Dimensions of the outlet 64
Live capacity 68
Mass-flow bins 68
Conical and plane-flow hoppers 71
Hoppers with a vertical wall 74
Rectangular, full-slot bins 76
Pyramidal hoppers 77
Dimensions of the outlet 78
Stresses on walls 82
Slip-stick solids 84
Dischargers and feeders 109
Dischargers 109
A vibrating hopper 109
An air hopper 110
Feeders 112
A screw feeder 112
A belt feeder 112
An apron feeder 116
A side-wise reciprocating feeder 118
A star feeder 118
A rotary table feeder 118

vi
 Page

An air-slide feeder 120

Summary of dischargers and feeders 122
Feeder loads 123
Promotion of flow 132
Impact pressure 132
Circulation 133
Solids conditioning 133
Flow promoting devices 134
Small hoppers 135
Segregation and blending 136
Flooding 140
Gas counterflow 140
Bin failures 142
Ore storage 146
Coarse ore (-8 inch) 146
Fine ore (-1 inch) 146
Examples of design for flow 148

APPENDIX ' 169

Introduction 169
Doming 170
Consolidating pressure in converging channels, o^ 170
Radial stress fields 172
Convergence of general stress fields to radial stress
fields 176
Critical pressure in a dome, 9. 180
N6-doming flow-factor 184
Piping 188
188
Consolidating pressure in a pipe, o.
190
Critical pressure in a pipe, o^
No-piping flow-factor 190
Pressure on a wall, a' 191
Resultant vertical force Q across a channel 191

BIBLIOGRAPHY 194

vii
 NOTATIONS

area of a shear cell (1/13 sq ft)

width of a rectangular outlet, side of a square

outlet, diameter of a circular outlet, ft

diameter of a pipe, major dimension of an outlet,

diameter of a circular cylinder, ft

effective yield locus

unconflned yield force, lb

flow-function of a solid

moisture content, %

coefficient

length of a rectangular outlet, ft

vertical force acting on a feeder, lb

shearing force applied to a shear cell during

consolidation, lb

shearing force applied to a shear cell during shear, lb

temperature

time yield locus

normal force applied to a shear cell during

consolidation, lb

normal force applied to a shear cell during shear, lb

major consolidating force, lb

major force in a dome or a pipe, lb

vertical force applied to a shear cell during twisting,

wall yield locus

yield locus

unconflned yield pressure of a solid, psf

viii
flow-factor of a channel

coefficient: m - 0 for plane flow, m - 1 for conical flow

feeder load coefficient

plane radial coordinates

spherical coordinates

stress function

time

plane cartesian coordinates

cylindrical coordinates

bulk density of a solid, pcf

effective angle of friction

hopper slope measured from the vertical

kinematic angle of internal friction of a solid

kinematic angle of friction between a solid and a wall

static angle of Internal friction of a solid

pressure, psf

major consolidating pressure, psf

minor consolidating pressure, psf

pressure on a wall in mass flow, psf

major pressure in a dome or a pipe, psf

minor pressure in a dome or a pipe, psf

pressure in a radial direction, psf

pressure in a direction normal to radial, psf

shearing stress, psf

shearing stress on a wall in mass flow, psf

shearing stress in a coordinate direction, psf

angle between the direction of major pressure and a

coordinate ray

ix
 CHAPTER I

FLOWABILITY OF BULK SOLIDS AND OF CHANNELS

Introduction

The first significant studies related to the storage of bulk

solids were reported at the end of the nineteenth century. That

xrork originated from the need to store large quantities of grain and

was concerned mainly with wall pressures affecting the structural

design of silos and bins. The Janssen formula [1], the work of Airy

[2], Jamieson [3], and Ketchum [4] form the highlights of that period.

Structural problems of bin and silo design have been the subject of

texts written more recently by Caquot and Kerlsel [5], M. and A.

Reimbert [6] in France; Nowacki and Dabrowski [7], Sienlcki,

Domaszewski and Klos [8] in Poland, and of numerous articles [for

example, 9 to 14]. Other related subjects which have been studied

extensively, though mostly empirically, are the rate of flow of bulk

solids through orifices [15 to 22], methods of preventing segregation,

and patterns of flow [23 to 30]. Only representative articles are

listed in the bibliography. Zenz and Othmer in their book [31] refer

to most of the above topics even though their work is mainly concerned

with fluidized materials. A bibliography on "Flow of Bulk Materials

from Storage" has been published by the Research Department of the

American Society of Mechanical Engineers.

One problem which is not resolved by the above literature and

which is the subject of this publication is whether or not a given

solid will flow out of a given bin. It is well known that tlow out of

bins and hoppers is often unreliable, that time and money are spent -

with or without success - on flow-promoting devices, that solids segregate

in storage, feed erratically, flood, arch, pipe, and stick to the bin

walls reducing the live capacity below specified values [32].

The theory of flow of bulk solids and the design of storage bins

and channels for flow were the subject of studies by the author, first

on his own (1952-1955) [33-36], then at the Utah Engineering Experiment

Station, University of Utah (1956-1962). The work led to the postulation

of a flow - no flow criterion and ot tlow properties of solids and of

channels. The author has continued this work and the present bulletin

contains an up-to-date report of the results. Since this bulletin is

written for the design engineer, the theory is kept to a minimum. For a

complete derivation of the formulas the reader is directed to references

37, 38, 39, 41 and 42.

It will be useful to consider the typical flow patterns of bulk

solids in gravity flow and to define the terminologyn Since relatively

new concepts, such as "flowability of bulk solids" are used, it will be

necessary not only to redefine some of the existing terms with greater

precision but also to Introduce new terms. However, first of all let us

consider why some bulk solids do not flow by gravity but arch and pipe.

Obstructions to flow. While the number of obstructions to flow

which may develop in a bin is infinite, two types will be analyzed here.

They are: doming, Fig. 1; and piping. Fig. 2. It will be assumed that,

if the design of the bin is such that these two obstructions cannot occur.
satisfactory flow will result.

When a stable arch or dome forms across the outlet of a hopper,

it is apparent that the solid has enough strength to support its own

weight. Yet, the same solid when fluffed up has no strength.

Evidently, the strength of a solid varies, and depends on the degree

of consolidation of the solid. While some exceptions are possible, in

general, the strength of a solid increases with consolidating pressure.

Various solids develop various values of strength for the same

consolidation. This is illustrated in Fig. 3 by the curves (a) and

(b) which represent two different solids. Evidently, the stronger

solid represented by line (b) will be less free-flowing than the weaker

solid represented by line (a). Intuitively, one feels that for a given

hopper there must exist a critical line - for Instance as shown by the

dashed line in the figure - such that, as long as the pressure-

strength curve of the solid lies below the critical line, the strength

of the solid is insufficient to support a dome, and the solid flows;

whereas, when the pressure-strength line lies above the critical line,

the solid domes.

The work which follows develops this concept: the pressure-

strength curve of a solid is referred to as its flow-function, the

dashed line is called the flow-factor of the channel. When these two

lines intersect, the point of intersection usually determines the

minimum size of the outlet required to assure unobstructed gravity

flow. A similar reasoning is applied to piping.

-3-
Fig. 1

Doming

Fig. 2

Piping

-4-
strength
of a
solid
A Hopper
'*~^ flow-factor

Point of
intersection
Solid flow-functions

Consolidating pressure

Fig. 3

Strength versus consolidating pressure

 Angle of repose. When an unconsolidated (loose) bulk solid is

deposited on a horizontal surface so as to form a pile, and the

velocity of the stream onto the top of the pile is negligible, the

particles of the solid roll down the pile and the slope of the pile

forms an angle of repose with the horizontal. The angle of repose

assumes values between 30* and 40" and is not a measure of the flovr-

ability of solids. In fact, it is only useful in the determination

of the contour of a pile, and its popularity among engineers and in-

vestigators is due not to its usefulness but to the ease with which it

is measured.

If a solid contains a wide range of particle sizes, it segregates

the fines collect along the trajectory of the charged solid while the

coarse fraction rolls to the periphery of the pile. When the solid

drops onto a pile from some height, the fines along the trajectory pack

under the Impact of the larger particles, gain strength, and form a

slope angle steeper than the angle of repose. Fig. A.

If a fine powder or a flaky solid drops from a height, it aerates

and spreads at an angle smaller than the angle of repose.

Effective area of an outlet. It is necessary to differentiate

between the physical size of an outlet of a pile, bin, or hopper, and

its effective area because, in the development of a flow pattern of a

solid within the pile or bin, it is the effective area which is signi-

ficant. The effective area of an outlet is that part of the total area

through which the solid actually flows when the feeder is in operation

or the gate is open. It is Important to realize that in many cases the

effective area forms only a part, sometimes a small part, of the total

-6-
 Trajectory

Angle of
repose

Fig. 4

Segregation and packing in a pile

-7-
outlet. Typical examples are provided by apron and belt feeders of

constant width, Fig. 5, and by screw feeders of constant diameter and

pitch. Fig. 6. If the length of the outlets exceeds, say, twice the

width of the outlet at the apron or belt, or twice the diameter of the

screw, it is a practical certainty that the solid will feed at one end

of the feeder only. In an apron or belt feeder it is not possible to

say, off-hand, which end it will be: it depends on the friction

developed between the solid and the surface of the feeder. In a screw

feeder the solid fills the screw at the end opposite the discharge

(the back), the screw runs full over the remainder of its length and

is unable to accept any more material. If the solid is at all cohesive

it packs into a firm stable arch over the remainder of the screw. A

similar situation occurs in a rotary table feeder when the skirts seal

the solid around the table. Fig. 7: the solid then flows onto the table

only through a channel over the plow; outside of that channel the solid

remains stationary.

Stationary solid sliding over the feeder packs hard and creates

large normal and frictional forces on the feeder contributing a major

share to the power consumption and wear. This situation is therefore

detrimental not only from the standpoint of flow but also from the

standpoint of maintenance and consumption of power.

For a feeder to be effective along its full area, the capacity

of the feeder must increase in the direction of flow of the solid with-

in the feeder. Suitable feeder designs are described in Chapter III.

In the discussion that follows it is assumed that the feeders are ade-

quate and the outlets are fully active.

-8-
Effective
area

Fig. 5 Fig. 6

Effective area of outlet in an Effective area of outlet in

apron feeder a screw feeder

Table

Effective area

Fig. 7

Effective area of outlet in a rotary table feeder

-9-
 Flow patterns. Two types oi flow pattern will be considered:

(1) Funnel-flow which occurs when a solid flows toward the outlet of the

container in a channel formed within the solid itself. The solid outside

of the channel is at rest and the shape of the walls of the container

has no Influence on either the shape of the channel or the velocity

profile of the solid within the channel. (2) Mass-flow which occurs

when the flowing channel coincides with the walls of the container. In

perfect mass flow, all the solid in a bin is in motion whenever any of

It is drawn out of the outlet- Mass tlow hoppers are smooth and steep.

Flowablllty of solids

Comparison of solids and liquids. The word-"flow" is more often

associated with fluids than with solids, and when the "flow of solids"

Is mentioned, one is inclined to assume - by association - that the solid

will behave much like a liquid. Such an assumption is incorrect. The

properties of solids and of liquids differ so much that the mechanisms of

flow of these phases are quite unlike. First of all, solids can transfer

shearing stresses under static conditions - they have a static angle of

friction greater than zero - whereas liquids do not. This is why solids

form piles whereas liquids form level surfaces. Secondly, many solids,

when consolidated - that is after pressure has been applied to them,

possess cohesive strength and retain a shape under load. They can form a

stable dome or a stable well; liquids cannot do that. Thirdly, the

shearing stresses which occur in a slowly deforming (i.e. a flowing) bulk

solid can usually be considered Independent of the rate of shear and de-

pendent on the mean pressure acting within the solid. In a liquid, the

-10-
situation is reversed, the shearing stresses are dependent on the rate

of shear and Independent of the mean pressure.

These fundamental differences are listed to explain why all

attempts to derive flow - no flow criteria for solids from fluid flow

concepts have been fruitless. A bulk solid has to be looked upon as

a plastic and not a visco-elastic continuum.

Effective yield locus, EYL. Consider a solid flowing in a

channel, Fig. 8, and, in particular, follow an element of the solid

along Its path through the channel. The particles of the element are

assumed to be deposited at the top of the mass without impact. As

the element flows down from the surface and is covered by new layers

of the solid, pressures acting on the element Increase, at first, then

pass a maximum, and decrease toward the vertex of the channel

[33-36,45]. Simultaneously, the shape of the element changes,

causing the particles of the element to slide on one another. The

particles are first brought closer together, as pressures increase,

then farther apart, as pressures decrease.

During this process, which is called flow, the major pressure on

the element is denoted by o. and the minor pi-essure by o^. Since these

pressures cause a change in the density of the solid - in fact, during

flow, density is a function of the last set of pressures - they will be

called consolidating pressures.

Experiments carried out on hundreds of bulk solids have demon-

strated that the ratio o^/o„ varies but little as a solid of given

moisture content H flows under a constant temperature T. This property

of bulk solids is usefully expressed by the relation

-11-
 a
1 1 + sin ^ /„ ^ m .. \ ft\
__ • _ —_ (H « const., T - const.). (1)
o» 1 - sin o

This relation is due to Jenike and Shield [37] and is referred to as the

"effective yield function." Angle 6 is called the "effective angle of

friction," and for a given solid has been found to vary only within a

few degrees for the range of pressures which occur in gravity channels.

Angle 6 has been measured at between 30" and 70° for various

solids. In general, fine and dry solids have low values of £ while coarse

and wet solids have large values.

In the coordinates: pressure, a, shearing stress, "^, Eq. (1) with

a constant 6 implies that, during flow, the Mohr stress circles* possess

a straight envelope which passes through the origin. This envelope is

called the "effective yield locus," EYL for short, and is Inclined at an

angle 6 to the o-axls. Fig. 9(a).

In the interpretation of tests, it is often more convenient to

plot this and other relations in force coordinates. The corresponding

plots are similar and differ only in units. For instance, it is convenient

to replace the coordinates o, i by coordinates

V - Ao, S - AT, (2a)

where A is the area of the cross-section of the shear cell used in the

tests. Fig. 10. The forces applied to the shear cell can then be plotted

directly.

*See any text on strength of materials.

-12-
 Fig. 8

Pressure acting on a flowing element

•13-
 Relation (1) in principal consolidating forces

^1 " •'^^l' ^2 " ^^2 ^^'^^

becomes
V
JL . ^ "^ ^^» ^ , (H - const., T - const.). (3)
V„ 1 - sin 0

This is shown in Fig. 9(b).

Yield locus, YL. In discussing the flow - no flow criteria of

solids, it is necessary to consider the yield strength which a solid

develops as it flows in a channel. An unconsolidated solid has no yield

strength: in fact, an aerated solid may behave like a fluid. But, as

the solid is placed in a channel, pressures arise under the weight of the

superimposed mass, and - in the presence of a wetting liquid in the pores

also due to surface tension. Some air is forced out, the particles are

brought closer together, and molecular forces develop: the solid consoli

dates and gains strength. The higher the pressure, the greater the

consolidation and the strength of a given solid.

All solids do not gain the same strength under equal pressure.

Gravel, dry sand are cohesionless; they gain practically no strength

within the range of pressure which acts in storage plants. In order to

cause such solids to shear, it is sufficient to overcome their angle of

internal friction <^. If dry sand is placed in the shear cell. Fig. 10,

loaded by a normal force V, and a shearing force S is applied, no

continuous deformation occurs as Iqng as S < V tan ()>, and slip takes

place for S « V tan *. This is Coulomb friction. For values (V,S)

lying below that line, Fig. 11, the solid can be considered to behave

-14-
Shearing
stress, T

•Mohr stress semi-circle

>
Pressure, a

(a)

S - AT

Mohr force s e m i - c i r c l e

V2 • A02 Vj » AOj V •= Aa'

(b)

Fig. 9
E f f e c t i v e y i e l d locus

-15-
rigidly (or elastlcally); for values lying on the line, slip or yield

occurs. This line then is the locus of yield values of the normal and

shearing forces and is therefore called the yield locus, YL for short.

For free-flowing solids like gravel and dry sand, ^ •• 6 , and the yield

locus YL coincides with effective yield locus EYL.

Other solids are cohesive, they gain strength as pressure is

applied to them. In V,S coordinates one such solid is now represented

by a family of yield loci, each locus corresponds to a different con-

solidation. Consider a cohesive solid flowing in the channel. Fig. 8.

If the flow of the element were stopped at the higher location, the

element removed - undisturbed, placed in the shear cell. Fig. 10, sheared

under a small normal load V, and the shearing force S was measured, the

point (V,S")- would define a point on the yield locus, YL, Fig. 12. If

the experiment were repeated but with a somewhat larger normal load V, a

different shearing force S would be measured and point (V,S)2 would also

lie on the yield locus. Since, during flow, the element is continuously

at yield, the Mohr force circle of flow defined by the principal consoli-

dating forces V. = Ao. and V„ - Ao- is also tangential to the yield locus.

If the above experiment were repeated from a different point of

the path, where the principal consolidating pressures are smaller, say,

o| and o', and the forces are V| and V', the solid would be less consoli-

dated, and would possess a lower yield strength, as Indicated by the yield

locus YL' in Fig. 12. In each case, the yield locus terminates at a point

E, or £', on the Mohr circles determined by the consolidating forces of

flow. The effective yield locus EYL forms an envelope of these circles.

Kinematic angle of Internal friction, <|>. This is the angle between

-16-
 Bracket
Cover

Loading pin
Ring
Plane of shear
area A

777777777777777777777777

F i g . 10

A shear c e l l

-17-
the yield locus and the axis V, Fig. 12. In general, the yield loci

form a family of convex-upward, curved lines and, therefore, the angle

of friction of a solid varies with both, the particular yield locus and

the normal force V under which shear is performed. The curvature of a

yield locus is especially pronounced at low values of normal force. In

fact, there are good reasons to suppose that angle ^ increases all the

way to 90° at the point of intersection of the locus with the V-axls

[38,39]. Fortunately, the angle of friction ^ does not appear in the

analysis of flow.

Time yield locus, TYL. If the flow of a solid in a channel is

interrupted for a period of time t, the forces V., V^, which have acted

on an element during flow, are assumed to remain essentially the same

for that period of time, and tend to consolidate the element giving it

greater yield strength. This is illustrated by a location of the time

yield locus, TYL for short, above the YL, Fig. 13.

Unconflned yield forces, F and F . It is reasonable to assume

that, when an obstruction to flow like a dome or a pipe is about to

collapse, failure starts at an exposed surface of the obstruction and

propagates from the surface into the mass. Therefore, the stress condi-

tions which occur at an exposed surface during failure are of particular

interest. They are also particularly simple because the stresses acting

on an exposed surface are zero, the surface is a principal plane, and

the major pressure within the solid is tangential to the surface. When

this pressure causes yield, it is referred to as Lhr unconfined yield

pressure and denoted by the symbol f • The corresponding unconfined

yield force is denoted by F • Af . At failure, a Mohr force circle is,

-18-
 Fig. 11

Yield locus of dry sand

Fig. 12

Family of yield loci of a cohesive solid

•19-
of course, tangential to the yield locus. The value of F is therefore

readily obtained, when the yield locus is known, by drawing a Mohr

circle through the origin, V = S « 0, tangentially to the yield locus.

The yield locus, YL, and the time yield locus, TYL, Fig. 13, define

values of unconfined yield forces F and F , respectively. As a solid

gains strength with time of consolidation at rest and its unconfined

yield force F increases, the solid can support more extensive

obstructions to flow; hence, it becomes less free-flowing.

Static angle of internal friction, • . In the analysis of plug

flow, it is also necessary to know the angle of Internal friction which

has developed within the solid at an exposed surface of a pipe. This

angle is determined at the point of tangency of the TYL with the Mohr

circle which determines the unconfined yield pressure F , Fig. 13.

Wall yield locus, WYL. In mass flow, a solid slides along the

wall of the channel and the stress conditions along the wall enter the

analysis of consolidating pressures. The normal and shearing forces at

the wall during flow are denoted by V',S', respectively. They lie on

the wall yield locus, WYL for short. Fig. 14. The WYL is another

Coulomb friction line and is often in the form of a curved, convex-

upward line.

Kinematic angle of friction between a solid and a channel wall, •'.

Since a wall yield locus is usually represented by a curved line, the

kinematic angle of friction between a solid and a channel wall, ^', varies

and is a function of the pressure at the wall. Yet the mathematical

-20-
 Fig. 13

Tine yield locus

Fig. 14

Wall yield locus

-21-
analysis requires a constant value of <J>'. It is therefore necessary

to select a critical value of 4)' lor a sale design. In practice, the

largest value of <t' is usually critical, and, for a convex-upward WYL,

it occurs at the outlet of the hopper.

In the design process the values of the effective angle of

friction 6 and of the major consolidating torce V. at the wall of the

outlet are usually determined first. This permits the drawing of the

EYL and the tangential Mohr semi-circle through V., as shown in

Fig. 14. The point (V',S') is now determined, and the kinematic angle

of friction 4)' is found by drawing a straight line from the origin to

point (V',S'). Then

|I- - tan *'. (4)

While the Mohr circle provides two points of intersection with

the WYL, the observed tlow patterns require the point (V',S') shown in

Fig. 14.

For asymmetric, plane-flow channels, it is simplest to assume

that V^ is the same at both walls. If the walls are made of the same

material, then d)' is the same at both of them; if the materials are

different, two WYL's are drawn in Fig. 14 and two angles <(>' are deter-

mined .

Flow-function, FF. Consider again the solid flowing down the

channel. Fig. 8. In general, moisture content H and temperature T

affect the strength F of the solid. During continuous flow, T and H

can usually be assumed constant, while the time of consolidation at

rest t is, of course, equal to zero. Under these conditions, the yield

-22-
locus and the unconfined yir1d force F become a function of the consoli-

dation of the solid which, in view of relation (3), and neglecting the

effe(t of the third principal pressure, is fully determined by the major

consolidating force V., thus

F • f(V^), (t - 0, H = const., T «=> const.) (5)

This relation Is referred to as the instantaneous flow-function of a

solid or, in brief, its flow-function.

Three typical flow-function curves are denoted in Fig. 15 by

letters FF(a), FF(b) and FF(c). Evidently, the highest curve (c) marks

a solid with most strength, greatest ability to support obstructions to

flow and, hence, the least free-flowing one. On the other hand, dry,

round sand, which develops no unconfined yield strength and is free-

flowing has a flow-function curve coinciding with the V.-axis. The

curves In Fig. 15 could represent three different solids; they could also

represent one and the same solid for three values of, say, moisture

content.

When the flow of a solid is interrupted for a period of time t,

the solid remains at rest under pressure and may gain additional strength.

Flow-functions measured under these conditions are referred to as time

flow-functions, they are of the type

F » f(V.,t), (H « const., T - const.) (6)

A further complication arises if, during the time of consolidation

at rest, the moisture content or the temperature changes. This can be

represented by the general relation

-23-
 Fj. - f(Vj^,t,H(t),T(t)). (7)

The gain in strength at rest may be caused by any one or a com-

bination of the following factors:

1. Escape of entrained air with corresponding increase of

density.

2. Migration of water.

3. External vibrations due to operating machinery, railroads,

or wind which cause a rearrangement of particles and an
increase of density.

4. Evaporation of free water with concurrent precipitation of

dissolved salts which cement the particles.

5. Break-up or softening of particles or crystals under

pressure, causing an increase in the surface of contact
and cohesion.

6. Changes in the surface of particles, e.g., crystallization,,

fermentation.

Experience Indicates that functions of the type (6) are measured

most frequently under ambient temperature for two to four moisture

contents and one or two times of consolidation. Such tests apply to

metallic ores and concentrates. Functions of the type (7) are usually

measured for trace values of moisture, which is considered constant,

and either for a constant temperature or for a decreasing temperature to

reproduce cooling in storage. The latter condition usually leads to the

critical design values for temperature-sensitive solids like plastics.

The flow-function curves are constructed from the yield functions.

For instance, the values of F, F , and V. in Fig. 13 determine one point

on each, the instantaneous (t - 0) and the time (t > 0) flow functions

-24-
 F
ai^
V,

P - V
1 '

Fig. 15

Flow-functions

•25-
of Fig. 16(a). In the absence of a wetting liquid, a solid has no

strength when unconsolidated; the flow-function curves pass through

the origin, V, = F = 0 and the two points: (V.F), (V ,F ) permit to

draw two straight lines through the origin to approximate the flow-

functions. When a wetting liquid is present, or its presence is sus-

pected, the tests have to be repeated for a different value, say, V|,

and the straight lines through the two pairs of points now determine

the approximate locations of the flow-functions. Fig. (16b). These

lines are usually sufficient to expose the non-free-flowing solid and,

in intricate cases, to show the direction in which the critical combina-

tion of parameters should be sought. Additional points are obtained

later if it is necessary to define the flow-functions with greater

precision.

Particle-size. As a general rule, solids which do not contain

particles smaller than, say, 0.01 inch are free-flowing. There are

exceptions to this rule. For instance, grain ferments under adverse

moisture and atmospheric conditions and can develop sufficient strength

to cease being free-flowing. Soybean meal contains oil which under

conditions of high moisture and temperature binds the particles into a

non-flowing mass. Flaky or stringy materials, like wood shavings, mica,

asbestos, interlock and also form obstructions to flow. Most ores, coal

etc. come under the general rule.

The flowablllty of a solid containing a range of sieve sizes,

which include both fine and coarse particles, is invariably governed by

the flow properties of the fine fraction. This is explained by the fact

that during flow the shearing takes place across the fines. The coarse

-26-
 (a)

(b)

Fig. 16

Construction of flow-functions

-27-
particles are a passive agent in this process. However, the size of

the coarse particles will affect the tendency to Interlock at the

outlet, and the impact of the heavy, coarse particles charged into a

container may cause compaction of the solid along the trajectory of the

falling stream.

Flowablllty of channels

Flow-factor^ ff. While there is an infinite number of possible

obstructions to flow, extensive observations point to the conclusion

that a solid will flow if a dome does not develop across the channel.

In mass-flow channels, this is sufficient. In plug-flow channels, it

is also necessary to assure that the solid be unable to sustain an

empty vertical pipe oi excessive height. Fig. 2.

For an obstruction to occur in a channel, the solid has to be

consolidated to such a degree that it develops sufficient strength to

support the weight of the obstruction. Hence, the higher the consoli-

dating pressures a. in a channel, and the lower the pressures a. which

act in an obstruction, the lower the flowablllty of the channel. This

is expressed by the flow-factor,

ff - Oj^/Oy (8)

In corresponding force coordituites V. « Aa., V. - Ao,, the flow-

factor becomes

ff - Vj^/V^. (8a)

The smaller the value of ff, the better the channel.

-28-
 Fig. 17

Conical channel

Syimnetric Asymmetric - one v e r t i c a l wall

Fig. 18
Plane-flow channels

-29-
 The no-doming flow-factors are computed for two horizontal shapes

of channels: the conical channel. Fig. 17, and the plane flow channel.

Fig. 18. It is evident that plane flow does not exist in reality;

however, the results are useful because plane flow is closely approached

at the sides of channels with rectangular (slot) outlets whose ratio

of length L to width B exceeds 3. Such channels are oi great practical

importance.

In mass-flow, values of the flow-factor are plotted in Figs. 45

to 49 for conical channels; in Figs. 50 to 54 for symmetric, plane-flow

channels; and in Figs. 55 to 57 for asymmetric, plane-flow channels.

ff is a function of the slope angle of the channel 6', the kinematic

angle of friction between the flowing solid and the wall i^', and the

effective angle of friction 6. Contours of constant values of ff are

drawn in 6', (>' coordinates for constant values of 6 = 30', 40", 50°,

60" and 70° for the sjrmmetric channels. This covers the useful range

of parameters. For the asymmetric plane-flow channels, calculations

have been carried out only for 6 '^ 50", and a kinematic angle of friction

at a vertical wall (ji^ = 20", 30", and 40"; 6' and *' refer to the sloping
P

wall. This is sufficient to provide a comparison of the fiowability of

symmetric and asymmetric plane-flow channels.

In plug-flow, the flow-factor depends on 6 and the slope of the

channel, Fig. A-7. For design purposes ff «• 1.7 is recommended. Con-

tours of the no-piping flow-factor are plotted in Fig. 35 as a function

of the static angle of internal friction ^ and the effective angle of

friction 6.

-30-
Solids in a channel

Floy - no flow postulate. Gravity flow of a solid in a channel

will take place provided the yield strength which the solid develops as

a result of the action of the consolidating pressures is Insufficient to

support an obstruction to flow.

For an obstruction to fail, the stresses in an obstruction must

exceed the yield strength of the solid at a critical location. It is

assumed that the critical stress occurs at the exposed surface of an

obstruction. The critical stress is then a major pressure and is denoted

by o., while the corresponding major force is V. - Ao.. The yield

strength is clearly the unconfined yield pressure f . The flow postulate

is expressed by

Dimensions of an outlet. The usefulness of the concepts of the

flow-function and the flow-factor will now become apparent, o^ is

eliminated between expressions (8) and (9), leading to

Oj^/f^ > ff. (10)

The pressures are replaced by the corresponding forces V. •> Ao., F « Af ,

resulting in

Vj^/F > ff. (10a)

Now, in Fig. 15, coordinate V. is plotted along the axis F. ff,

Eq. (8a), plots as a straight line in the V.,V. coordinates and, for

expression (10a) to be satisfied, the flow function FF of the solid must

-31-
lie below the ff-line. In the example, the solid represented by FF(a)

will flow in the channel represented by the flow-factor ff; the solid

represented by FF(c) will not flow in that channel, while the solid

represented by FF(b) will flow in that part of the channel in which

F - V^ (11)

Generally, the flow-function is convex-upward, as shown in the

figure, and expression (11) determines a minimum dimension of the

channel, i.e. the dimension of the outlet of the channel, B. This

dimension is the minor dimension of an outlet, e.g. the side of a square

outlet, the width of a rectangular outlet, the diameter of a round out-

let, and is

B - H(e')Vj^/AY, (12)

where H(e') is plotted in Fig. 43.

For no-piping in plug-flow, the analysis is similar. This time

the no-piping flow-factor is drawn, and another point of intersection

(V..,V.) is determined. For the solid at the surface of the pipe to fail.

Inequality (11) has again to be satisfied. The formula for the minimum

diameter of the pipe is

D - G(*^)V^/AY, (13)

*
where G(^ ) is plotted in Fig. 36. Since the diameter of the pipe

approaches the largest, i.e. major, dimension of the outlet, D determines

this dimension. In a circular outlet, D is the diameter; in a square or

rectangular outlet, D is the diagonal.

In Bulletin 108, G(.<^) = 4(da)/dn) ,

t n"i

-32-
 CHAPTER II

TESTING THE FLOW PROPERTIES OF BULK SOLIDS

Apparatus

The field of soil mechanics was well developed at the time when

this study originated (1952). Since soils are bulk solids, it was

natural first to investigate the applicability of the available soil

testing equipment to the measurement of the flow properties of bulk

solids.

Direct shear apparatus and unconfined compression apparatus were

tried. The results obtained with shear apparatus were not satisfactory,

primarily, because the range of values of cohesion for which the soil

test machines were designed was outside of the useful range for bulk

solids. By the terminology of soil mechanics, bulk solids can be con-

sidered cohesionless, while soils under their usual degree of consolidation

would not flow by gravity in any storage system. Unconfined compression

apparatus was also found unsuitable because of the impossibility of

applying a known, uniform consolidation and because the stresses due to

the weight of the sample were in the range of the measured yield stresses.

These s£unples often disintegrated during handling.

It was decided to develop a direct-shear tester especially for

bulk solids. The direct shear method was selected in preference to other

methods, and in particular to the triaxial method, because of several

reasons: (1) Shear tests are easier and quicker to make than triaxial

tests. In view of the large number of points which often need to be

-33-
obtained, the difference in the time required for testing is substantial.

(2) All tests are made on loose solids which are consolidated under known

pressures in the tester; hence uniform, easy and quick consolidation is

important. The shape of the shear specimen lends itself better to uniform

consolidation than the shape of the triaxial specimen, because it is

easier to consolidate a shallow specimen than a deep one. The ratio of

height to diameter of a shear specimen is below 1:2 as compared to 2:1,

or more, required for a triaxial specimen. To simplify and speed up con-

solidation, the shear cells are made of circular cross-section. (3) The

measured stresses are so low that the influence of the weight of the solid

within a specimen must be taken into account. This Is easily done in a

shallow shear specimen whose plane of shear is horizontal but is practi-

cally impossible to do in a deep triaxial specimen with an oblique plane

of shear. (4) The objectionable non-uniformity of stress distribution

in a conventional shear cell has been overcome by a new design of the

cell.

The shear cell, Fig. 10, is composed of a base located on the

frame of the machine, a ring resting on top of the base, and a cover. The

bottom of the cover and the inside of the base are roughened to increase

adhesion of the tested solid. The base and the ring are filled with the

tested solid. The vertical force V is applied to the cover. The hori-

zontal shearing force S is applied by means of a stem which acts on a

bracket attached to the cover. The stem acts in the plane of contact

between the ring and the base. A part of the shearing force is transferred

from the bracket to the ring through a loading pin. This ensures a suffi-

ciently uniform distribution of the shearing force across the cell. The

-34-
standard shear cell is 3.75 Inches inside diameter.

The tester which has been used since 1958 is shown in Fig. 19. It

is equipped with a shear cell, a gravity vertical loading system, and an

electro-pneumatic shearing force applicator. The applicator has a

shearing rate of 0.106 in. per mln. The shearing force S necessary to

maintain the strain-rate is continuously recorded on a pneumatic recorder.

This arrangement produces a permanent record of the stress-strain relations

for each test.

The shear tester is used in conjunction with a six-cell consolidating

bench In which the cells are placed for the time of consolidation at rest,

Fig. 20, and covered to prevent moisture changes.

The bench shown in Fig. 21 is used for temperature sensitive solids.

Here the consolidating bench Is enclosed in a heating chamber which permits

the control and recording of the temperature of the tested solid.

Purpose of tests

The purpose of the tests is to measure the effective angle of

friction 6, the necessary flow-functions FF, the bulk density Y> and

either the kinematic angle of friction (|>' between the solid and a material

of the wall - for mass-flov bins, or the static angle of internal friction

^ of the solid - for the plug-flow bins.

-35-
1

Fig. 19

Direct-shear tester

-36-
 • ' • ' ' ' '

,OI

^•m^^^m^m'
il

Fig, 20
Consolidating Bench

-37-
Test Specimens

In the execution of the tests it is necessary to ensure that the

specimens of the tested solid and of the materials for the hopper walls

are uniform and representative ot the field conditions. In general,

the yield locus of a solid for a given consolidating pressure and a

given rate of shear is affected by the water content, temperature, and

particle size distribution of the specimen. As a rule, the tests are

conducted on solids containing water below the point of saturation.

In view of the considerable influence which water content has on the

yield loci of solids, it is necessary to reproduce the required content

accurately, and to handle the samples and shear cells rapidly in order

to minimize errors due to evaporation. When testing is done at a

temperature other than ambient, handling of the cells should also be

rapid.

The particle size distribution is not as perplexing as it might

appear at first sight, in the flow of bulk solids, our interest lies in

the conditions that lead to stoppages of flow. Stoppages occur when the

solid reaches an excessive yield strength. Thus, the critical conditions

to be considered are those of highest strength. During the flow of a

mass of mixed particle sizes, the large particles move bodily while the

material shears across the fines. Therefore, the yield strength of the

mass depends on the properties of the fines. The coarse particles are a

passive agent and, like aggregate in concrete, they do not develop yield

strength without fines to bind them. Experience shows that screened out

coarse solids are invariably free flowing; they do not develop any

cohesion. To determine the yield loci of a mixed material, the fines are

screened through an arbitrarily chosen No. 8 mesh (0.0937 in. aperture)

-38-
 Fig. 21

Consolidating bench in a temperature controlled chamber

-39-
sieve, and tested.

Continuous flow

Consolidation. The consolidation of a sample is carried out in

two stages. The first stage is called preconsolidation and its purpose

Is to prepare a uniform specimen. With the cover off the test cell, a

packing mold is placed on top of the ring and both the mold and the ring

are placed In an offset position on the base, as shown in Fig. 22. A

sample of the tested solid is then placed in the cell. One layer after

another is slightly packed with the fingers, up to the top of the mold.

The excess material is scraped off level with the top of the mold. A

twisting top is placed over the solid. A vertical force V is applied

to the top by means of the system of lever and weight shown in Fig. 19.

Force V causes a vertical pressure a in the material. By means of a

special wrench, a number of oscillating twists is now applied to the

cover. This preconsolldates the solid and assures a unifoinn specimen.

The value of V is discussed below in the example.

Consolidation is completed in the second stage by causing the

specimen to flow under given stresses until a steady state is reached,

or closely approached. This is attained in the following way: Load V

is taken off, the twisting top and the mold are removed, the excess

material is scraped off level with the top of the ring, and the test

cover is placed on the material. A load V is now placed, and the stem

of the shearing device is advanced against the bracket. Fig. 10.

The shearing force S divides between the cover and the ring. This

method of applying the shearing force seems to assure a satisfactory

distribution of stresses across the specimen. Prior to the adoption of

-40-
 \

.Twisting top

buld

Ring

Base

Frame

// y/ /////
Fig. 22

Preconsolidation of a shear c e l l

-41-
this method, the shearing force used to be applied dlcectly to the ring

causing stress concentrations at the loaded ed^^es ot the rln^ and of the

base and frequently leading to the shear of the solid w i t h m the ring

rather than in the plane between the ring and the base. Furthermore,

the unloaded side of the ring would often rise off the base, exposing

the sheared specimen and weakening it significantly. It will be noted

from Fig. 10 that the loading pin bears against the ring at its mid-

height: this reduces the tendency of the ring to rise off the base. As

an additional safeguard against the rising of the ring, it is sometimes

pressed down with the fingers, at intervals, during the process of flow.

As shearing proceeds, a condition is reached when a layer of the

solid across the whole specimen is caused to flow plastically: the

recorded shearing force reaches a steady value S. The stresses which

occur in the solid during this stage of consolidation have been described

in references 38 and 39. Consolidation determines point (V,S), Fig. 23.

Shear. When consolidation is completed, the stem of the shearing

force device is retracted. The vertical compacting load V is replaced by

a smaller load V to locate a useful point (V,S) of the yield locus,

Fig. 23. The shearing force is now applied until a failure plane has

developed. This fact is indicated on the recorder by the force S passing

a maximum value. After shearing, the plane of failure of the specimen

is checked. The plane of failure should roughly coincide with the plane

of shear of the cell. If the planes deviate, it means that the measured

point (V,S) does not lie on the yield locus and the test is repeated.

The determination of one yield locus requires the measurement of

two or three points of the locus (V,s")j^, (y,!)^* and (V,?)^, Fig. 23.

-42-
 Fig. 23

Construction of a yield locus

-43-
For each point, the specimen is first consolidated and then sheared.

The first specimen is sheared under an intermediate value of normal

force V; for the second specimen, force V is stepped up to about 2/3

of the force V. The location of the unconfined yield force Mohr semi-

circle is estimated from the two (\r,S^) points.^ This semi-circle Is

drawn-through the origin tangentlally to the yield locus. A third

value of V is selected so as to lie somewhat to the right of the

point of tangency between the yield locus and the semi-circle. Any

point (V,S) lying to the left of the point of tangency should be

rejected, because it implies tensile stresses within the shear cell,

to which the cell is obviously not adapted. It is necessary to

obtain the values of these points for the same steady consollfatfng

shear each time. This is accomplished by running a sufficient number

of tests to permit interpolation to a suitably selected value of S.

The yield locus is now drawn and extrapolated toward the higher

values of V and a Mohr semi-circle is drawn through point (V,S)

tangentlally to the yield locus. The point of tangency E locates the

terminus of the yield locus. The point of intersection of the semi-

circle with the V-axls determines the value of the major consolidating

force V.. The Mohr semi-clrcle for the unconfined yield force F Is

now drawn. Tl\e force F Is determined by the point of intersection of

the circle with the axiS'V, as shown in the figure.

The reader will notice the difference from the earlier procedure pub-

lished In references 38 and 39, which recommended Interruption of

consolidation at 95% of a value of S obtained in a preliminary test.

-44-
 Examyle; A typical recorder chart is shown in Fig. 24. The

curves ar« nuatbered and represent the following steps:

1. Preconsolldation under load V « 8.9 lb.

2. Ccmsolidation of tlie first specimen under a vertical load V « 8.9 lb.

3. %earl]^ of the first specimen under a vertical load V - 3.9 lb.

4. Consolidation of the second specimen under the same vertical load V.

S« Smm as 2. Test 2 is repeated because of the sc-flttcr in the VMIIWH of S,

6. Gonslldatlon of the third specimen under tte same vertical load V.

7. ^tearing of the third specimen under a vertical load V • S.9 lb «

approx. 2/3 x 8.9,

8. Cimsolldation of the fourth specimen under the same vertical load V.

9. %«aring of the fourth specimen under a vertical load V * 2.9 lbs.

1%e iieasur«d and prorated values are shown below:

Measured Prorated

V 8.9 8.9 8.9 8.9 8.9 8.9 8.9

S 8.9 8.5 8.7 8.8 8.7 8.7 8.7

V 3.9 3.9 5.9 2.9 3.9 5.9 2.9

S 5.5 5.3 6.8 4.6 5.4 6.8 4.55

The prorated values are those shown in Fig. 25. In drawing the yield

locus, it should be b o m In mind that, on theoretical grounds of

ctmvexLty of yield surfaces (46), the yield locus cannot be concave

tjpwards. In this case the locus defines V - 22.4 lb and F • 9.3 lb.

Thtfsc* valttes dctcrminu u puint of the flow-function In FIR. 25.

A tangent to the V.-circle is also drawn through the origin in

Fig. 25. The slope of this tangent determines the effective angle of

-45-
 F i g . 24

A t y p i c a l rtcutdwr ch«rt - neafcureDent of fl'jw«b£llty

46-
 15 1

S Vb

rh
10

.9,8.7)
(5.9, 6,
^
p.9, 5.4>J \
( 2 . 9 , 4,55)1 Jl^^/

\ / \ ^ - 41
L
b 5 ^Xlo 15 20 / 25 ' V lb 3|o

^F - 9.3 ^ • V^ - 22,4

TO*
_ 60*
*50«
-^• - ( 2 2 . ^,48")
40*
30*

r lb

10
-H • - ( 2 2 . 4.9.3)

1
5 10 15 20 25 V lb 30
1
Fig. 25 Example

-47-
friction 6 - 48* for Vj^ - 22.4 lb.

In preconsolldation, a vertical force V •• V is used at first.

V is increased above V If the force S during consolidation keeps rising

and does not reach a steady value within ± 1/16 inch of the central

position of the ring on the base. V should not be so large as to cause

force S to go over a maxlnum, since this Indicates ovet-<onsolldation in

the preconsolldation stage, and the development of a weak plane across

the specimen.

Tine effect

When the flow of a solid in a channel is stopped for an interval

of time, the solid remains under the action of static pressure and con-

solidates with tine. Since consolidation at rest significantly increases

the strength of some solids and, thus, reduces their flowabllity, it is

necessary to measure the time effect in order to know whether or not the

flow of a solid can be restored after storage at rest.

In these tests, the solid is first consolidated in the shear

cells under conditions described above for continuous flow, and then

the cells are placed in the consolidating bench under a static load for

the prescribed interval of time. No shearing force is applied in the

consolidatix^ bench and the vertical force is V.. The load V. is selected

on the assumption that the static pressures In the channel during storage

at rest are the same as the pressures which prevailed during flow.

Experience indicates that the results of tests are not sensitive to an

error of 5 to lOZ in the value of V..

After the lapse of the prescribed interval of time, the cells are

placed in the shear tester, one at a time, and sheared as described

-48-
above for continuous flow. The yield locus is plotted through the

points (V,S^ and the value of the unconfined yield force F is obtained

in the sane way as F was obtained for continuous flow. The value of V^

naasured for continuous flow applies here in plotting the tine flow-

fuiMtion.

During the static consolldetlons it is necessary to preserve the

noisture content of the specimen. This is done by covering the shear

cells with the plastic cups shown in Fig. 20. The bottoms of the cups

are lanerged in baths of a sealing liquid.

Density

Since the analysis shows that the influence of the compressibility

of a solid (change of density with consolidating pressure) on the process

of flow is negligible [391, it is not necessary to obtain a function of

density versus pressure. However, the density y of > solid at the out-

let of a channel, or within a pipe, enters the flow formulas and, hence,

iMeds to be obtained.

The easiest way to measure the density is to weigh the whole cell

after it has been sheared to failure, subtract the weight of the cell

itself and divide the net weight by the volume of the cell.

Kinematic angle of friction between a solid and a wall material. •'.

The angle of friction ^' between a solid and a material of the wall

is measured on the same apparatus. The arrangement of the cell is shown

in Fig. 26. In this case, a sample of the wall material is placed on a

filler so that the top surface of the sample is level with the center of

-49-
 Weights

Cover

^Sample of wall
\ / material
Stem Filler

reme

TTTTTTTTTTJ

Fig. 26

Measurement of the kinematic angle of friction between a solid

and a wall material

-50-
 Fig. 27

A typical recorder chart - measurement of fj'

•5l-
the stem. The ring is placed over the sample, filled with the solid,

compacted slightly, axtd enclosed with the cover. All the tests necessary
4
to determine angle i' are now run without replacing the solid.

Several (say, six) one or two-pound weights are placed directly

on top of the cover of the shaar cell to give the largest required load

V. The stem is advanced. As the force S levels off, the ring is

twisted slightly with the fingers in order to assure that it does not

transfer any significant vertical force directly to the wall material.

When S has levelled off, one weight is removed, after a while S again

levels off, another weight Is removed and so on, till all the weights

have been removed. The cover, ring and the enclosed solid are then

weighed. Thair weight plus the superimposed weights determine the

vertical loads V.

A typical recorder chart Is shown in Fig. 27. The points (V,S)

are plotted In Fig. 28. A smooth line is drawn through these points and

this is the wall yield locus, WYL. Typically, the WYL Is convex-

upward .

Simplified testing procedure

On the assumption that: (1) the yield locus is straight, and

(2) point E, Fig. 23, coincides with the measured point (V,S) an

approximate yield locus. Fig. 29, and one approximate point of the flow-

function (V^,F) can be obtained from one consolidation/shear test. The

relations are

-52-
 Fig. 28

Plotting the WYL

-53-
 W^' V + KS, (14)

F - 2K(S - V ^ - ^ , (14a)
V -V

2KV -t- (K'^-l)S

tan ^ - ^ - ^ (15a)
V -V
where

^^Vi.<Hl>^.
K - S ^ 1 4 + V 1 + (^-^^^.
V -V
(16)

A repetition of this test for a different value of V produces a second

point (V.,F) and a straight line through the two points is an approximate

flow-function. Fig. 16(b)

The time if low-function can be obtained similarly, with the addi-

tional assumption that: (3) TYL is parallel to YL. The unconfined yield

force F is computed from the formula

F^ - 2K(S - T ^ - ^ . (17)
t t t V -V
By this method the free-flowing solids are readily identified and

no additional time is wasted on them. This method is also useful when a

range of two or more parameters (moisture content, temperature, time of

consolidation at rest) needs to be examined. It permits a rapid narrowing

down on the range to those values for which a solid is not free-flowing.

If the solid is to be handled for these values of the parameters, then

complete tests are carried out and minimum outlet dimensions are established.

Slip-stick solids

Some fine powders exhibit slip-stick properties which may appear

during consolidation and/or during the frictlonal tests. The shearing

force does not level off but oscillates.

-54-
 Fig. 29

Simplified testing procedure

Vj^ & F

F - 5

Fig. 30

General classification of fiowability of solids

-55-
 In order to consolidate such a solid, a normally-closed, spring-

loaded button-switch is placed in the electric supply cord of the tester.

Consolidation proceeds until the peak value of the shearing force S has

been reached, say, three times; then, on the next approach, the stem is

"Inched" toward the peak value by depressing and releasing the switch.

When the peak has been reached, the stem Is retracted manually so as

not to disturb the cell.

A safety factor is Imposed on the flow-factor in the design for

these solids. This Is discussed in the next chapter.

General classification of fiowability of solids

Several numbers and curves are required for a precise definition

of the fiowability of a solid. At times it is convenient to forgo some

precision for a gain in brevity of definition. For instance, it is

evident from Fig. 30 that for F > 5, the instantaneous flow-function lies

below the line ff - 2, and the time flow-function lies below ff - 1.

The fiowability of this solid can thus be written briefly as: FF > 2,

FF > 1 for F > 5.

The fiowability of solids can be classified according to the

limiting FF-values thus:

FF < 2 very cohesive and non-flowing

2 < FF < 4 cohesive

4 < FF < 10 easy-flowing

10 < FF free-flowing

-56-
 CHAPTER III

DESIGN

Funnel flow bins

A solid drawn through an outlet of a flat bottom bin forms a

channel within the stored mass and flows toward the outlet within that

channel while the mass around it remains stationary. This flow

pattern is referred to as fnnnel-f low. The channel tends to assume the

shape of a vertical, circular cone. At the bottom, the diameter of

the cone may approach the largest dimension of the outlet. Thus,

with a rectangular outlet, that diameter may approach the diagonal

dimension of the outlet. As the solid is discharged and its level

within the channel drops, solid at the outside sloughs off the top of

the stationary mass and slips into the channel, Fig. 31(a). As more

and more material is withdrawn, flow finally ceases and a crater is

exposed within the remaining, stable mass. When the solid is com-

pletely cohesionless, the slope angle of the crater is equal to the

angle of repose (b). For all other solids, the crater (c) is composed

of three regions: From the outlet - whatever its shape - the channel

expands upward to a circular cross-section; when the outlet is

rectangular, the walls are steeper at the ends than at the sides; in

fact, at the ends the walls may overhang the outlet. In the middle

region the circular cone extends over some height. At the top, the

crater expands along a cone of slip.

The included angle 26' of the middle cone is always quite small

-b7-
and frequently zero; the middle part is then a vertical cylinder, or a

pipe. The height of the vertical cylinder varies. In extreme cases It

can extend tens of feet and through the whole height of a bin (d). This

condition is referred to as piping, or rat-holing, and severely reduces

the live capacity of a storage plant. As long as the height H of the

vertical cylinder or middle cone is less than its diameter D, the

operation of the bin should be considered satisfactory. In estimating

the live capacity of a fume 1-flow bin it is unreasonable to expect a more

complete discharge unless either the stored solid is free-flowing, or

final discharge is enforced by means of vibration or some other flow-

promoting device.

A pyramidal hopper. Fig. 32, has in-flowing valleys which provide

firm support to the solid and tend to hold it stable. As a result. In

the vicinity of a valley, the solid is forced to form a channel across

Itself and that leads to funnel-flow unless the solid is sufficiently free-

flowing and the hopper sufficiently steep and smooth. Mass flow In a

pyramidal hopper is discussed later.

When an emptyfunnetflow bin is filled, the solid which first falls

into the bin rolls to the periphery and remains there until the bin is

again completely emptied out. This is a serious drawback in the storage

of solids which deteriorate with time; it also leads to obstructions to

flow of solids which pack and gain strength when left undisturbed under

pressure.

Like all bins, funnetflow bins are subject to arching or doming of

solids with complete cessation of flow. Fig. 32(a). In this respect,

fuinel-flow bins are erratic because a solid does not flow uniformly in a

-58-
 Fig. 31

F l a t bottom bin
vertical or very steep channel. A solid flowing down such a channel

tends to form voids topped with domes, (b). As a dome breaks, the solid

falls into the void; under the impact, the solid tends to pack and dome

again. Surging flow is the rule In fumel-flow bins. In the storage of

powders, surging results in flooding which occurs when the void under a

dome momentarily extends all the way to the feeder, (c). The falling

mass of solid aerates and acquires fluid-properties as it impacts the

feeder. A falling mass may also act as a piston, and develop high air

pressure in the region of the hopper, causing a break through a hopper

wall, or a detachment of the hopper from the bin structure. These

accidents have been serious and have led to the loss of life.

A bin or hopper is often required to act as a gas seal. Funnel-

flow with its irregularities is quite unsuitable for this purpose.

The flow pattern which obtains In funnel-flow bins is also detri-

mental when storing materials which segregate during charging into a

bin because there is very little remixing as the solid flows down the

vertical channel. When the trajectory of the charged material is

located over the outlet, fines congregate within the region of the flow

channel, while coarse material rolls to the sides, Fig. 33. The bin

draws a preponderance of fines when the level of the solid in the bin is

rising (a), and of coarse - when the level is falling (b). When the rate

of draw is the same as the rate of chaige, the level remains steady and

a last-in first-out flow pattern obtains; in effect, the bin acts as a

spout (c).

When a solid domes In a funnel-flow channel, it is difficult to

break up the dome because there is no direct way of reaching the

-60-
 (b)

Mcmentary void Void extending

to the feeder

Fig. 32

Funnel flow
obstruction. Vibrators or inflatable panels applied at the walls must

first disturb the stable, compacted solid around the channel. This calls

for a substantial input of energy which carries the danger of additional

consolidation in other parts of the bin and, subsequently, more flow

problems. Since it is seldom possible to observe the conditions Inside

a hopper, it is easy to continue operating a vibrator or an inflatable

panel after the initial dome has collapsed and the void in the bin has

been filled. This leads to serious obstructions to flow requiring

manual prodding to dislodge the severely packed solid.

From the aforesaid it follows that funneLflow is acceptable only

under certain conditions:

(1) Segregation has to be unimportant either because it is insignificant

as is the case with solids of uniform particle size and density, like

sand or gravel, or because it is slight and can be tolerated, as is the

case in small bins, like weigh hoppers.

(2) Deterioration does not occur during the scheduled time of storage.

(3) The outlet is sufficiently large so that the solid flows without

flow-promoting devices.

Rjnnel-flow bins are useful for the storage of hard, abrasive,

lumpy solids because in funnel flow there is little wear of the hopper

walls. However, it should be remembered that in mass-flow hoppers the

wall pressures are relatively low and the wear is seldom unreasonable.

If the solid is very cohesive after it has been stored at rest and re-

quires very large feeders in funnel flow, it may be more economical to use

mass-flow bins, save on feeders and build heavier hoppers to provide for

wear.

-62-
 i
Charge Charge

i
OV,
i^ol'^bgi)

o
'••6<>

t8i

Level rising Level falling Level steady

Draw mostly Draw mostly Draw same
fines coarse as charge
(a) (b) (c)

Fig 33
Flow pattern in funnel-flow bins

_ ^ A
^

[^ ^
T

L «j.
0 ^
J<
(a.) (c)

Fig 34

Dimensions of an outlet

-63-
 Funnel-flow bins are also useful for the storage of mine-run ores

which contain large rock. Here the argument against the mass-flow bins

is that large rock falling from a considerable height tends to pack the

ore in a steep hopper and to cause doming. Coarse ore storage is dis-

cussed in another place in the chapter.

However, the main reasons why funnel-flow bins are so commonly

used today are: the general lack of knowledge of the existence and the

advantages of mass-flow bins, and the notion that funnel-flow bins are

cheaper per unit volume. That "economy" is more apparent than real: It

often becomes quite expensive in terms of excessive startup time, disrupted

production schedules, additional shifts, lowered product quality, unrealized

storage, redesigns, modifications, flow-promoting devices, and time spent

by operating and supervisory personnel in keeping the solids flowing.

Dimensions of the outlet.* In order to assure satisfactory flow in

a funnel-flow bin, it is necessary that the outlet be large enough so that

piping and doming do not occur. Flow should not rely on flow-promoting

devices. Typical shapes of outlets are shown in Fig. 34. The major

dimension of the outlets is D, and the pipe which tend to form will have

that diameter. To make this pipe unstable, D must be selected at least

equal to the value computed in the following way:

(1) Using the height, h, and hydraulic radius, R (= ie°gth^of perimeter ^

of the bin, estimate the effective consolidating head:

h . 2 ^d-e-O-^l^-*') (17A)
e tan (|)'
(lA) Compute major consolidating load V^

Vi=-^A (1710

(2) Using this value of V , find V and (t>^ from the time flow-function.
1 1 t
*For more details, see J. R. Johanson, "Effect of Initial Pressures on
Flowability of Bins," ASME J. Engr. Ind., V. 91, Ser. B, No. 2, May 1969,
p. 395-400.
-64-
 (3) Compute D from Eq. (L3). For a shearing cell with an

area A » 1/13 sq. ft, this equation becomes

D - 13V^G(*^)/Y (18)

Function G(* ) is plotted in Fig. 36.

In a square or circular outlet computed for no-piping, doming

cannot occur. In a rectangular outlet of diagonal D, the minor

dimension of the outlet, which is the width of the outlet B, Fig. 34(a),

must also be computed so that doming does not occur. A flow-factor

ff •" 1.7 la used here. This assures a slope ,.-.ngle 6' not less than 30°,

Fig. A-7, at the sides of the rectangle. Such an angle is necessary to

expand the crater from a width B to a diameter D, Fig. 31(c), otherwise

a stable pipe of a smaller diameter will form. B is calculated as

follows:

(1) Draw the flow-factor ff = 1.7 over the time flow-

function of the solid and determine the value of

V^ at the point of intersection. Fig. 15.

(2) Compute B from Eq. (12). For A * 1/13 sq. ft, and

H(30') = 1.15, this equation becomes

B - 15V,/Y. (19)

-65-
 12

G(^^)

10

30'- 40' 50* 60» i 70*

F i g , 36

Function G(^ )

-67-
 Live capacity. A funnel-flow bin with a sufficiently steep pyra-

midal or conical hopper can be cleared completely of the stored solid by

means of vibration, aeration, or manual rodding and air-lancing.

If the walls of the hopper are sufficiently smooth as well as

steep, a funnel-flow bin will clear by gravity. To attain gravity clear-

ance, it is essential that the material of the hopper does not corrode

and retains smoothness during storage. Stainless steel, ceramic tile,

resin coatings seem to provide satisfactory hopper materials. Unless

the stored solid is quite free-flowing, the hopper slope 6' should pro-

vide a margin of safety, fox instance,

8' < 65- - •'. (20)

The total storage capacity of these bins is easy to determine

from the geometry of the bin.

It is more difficult to estimate the capacity of a pile or of a

fuhnelflow bin without flow promoting devices, and without a smooth

hopper. A safe estimate results from the shape of the stable crater

shown in Fig. 31(c), provided the outlet has been designed for no-

piping and no-doming, and H is taken equal to D.

Mass-flow bins

When a bin has a sufficiently steep and smooth hopper then,

whenever the feeder is set in motion, all the solid in the bin flows;

there are no Inactive or dead regions. Under these conditions, mass-

flow is said to occur. Mass-flow bins can have a variety of shapes, as

shown in Fig. 37, but they are all characterized by steep hoppers and,

-68-
(a)

(b)
MASS-FLOW BINS

Fig. 37

69
usually, also by the absence of in-flowing valleys and sharp transitions.

Mass flow has the following properties:

(1) Channeling, hang-ups, surging and flooding are absent.

(2) Flow is uniform, and steady state flow can be closely approached.

As a result, analysis based on steady flow can be applied to

design with a high degree of confidence.

(3) The bulk density of the drawn solid is constant, and practically

independent of the head of the stored solid. This is very

advantageous in all cases of controlled flow rate, and essential

when the rate is controlled volumetrically. It should be noted

here that highly aerated powders should have a sufficient

residence time to deaerate and form a contact bed.

(4) Pressures throughout the mass and at the walls are relatively

low, with resultant low consolidation and attrition of the solid

and wear of the walls.

(5) Pressures are relatively uniform across any horizontal cross-

section of the hopper, causing uniform consolidation and uniform

permeability, which is Important in gas seals and gas-solids

reactors.

(6) There are no dead regions within the bin, hence, there is a

minimum of consolidation at rest, degradation and spoilage.

-70-
(7) A first-in first-out H o w pattern can be obtained, il desired.

This is useful in the storage of solids which deteriorate with

time.

(8) Non-segregating storage is obtained in a first-in first-out bin

because, while segregation takes place during the charging of a

solid into such a bin, re-mixing occurs within the hopper.

(9) By circulating a mixture around a suitable mass-flow bin, blending

may be obtained.

Conical and plane-flow hoppers. Hopper slope angles 0' which

lead to mass flow have been computed for conical channels, Fig. 17, and

for plane-flow channels. Fig. 18. For both shapes of channels the slope

angle 6' depends on the effective angle of friction of the solid 6, and

the kinematic angle of friction between the solid and the wall $'. The

regions within which mass flow can prevail are those covered by the flow-

factor contours in Figures 45 to 57. Each figure is drawn for a constant

value of 6. It will be observed that, for conical channels. Figures 45

to 49, the regions of mass flow are quite restricted. For instance.

Fig. 38 shows the cross-sections of three conical channels from Johanson'

work [44]. For the steepest channel, 6' = 15°, mass flow obtained; for

6' "• 21", flow at the walls was observed at first but then plug flow

developed; for 9' - 30", plug flow developed from the beginning. In this

case, 6 was measured at 50* and <^' at 24*. In Fig. 47, the corresponding

points (15%24"), (21",24"), and (30",24") are marked with crosses. The

first point lies within the mass-flow region, the second at the boundary

between mass and plug flow, and the third in the plug-flow region.

-71-
 In designing a conical hopper, Fi}',. 37(b), tor mass flow, it is

necessary to stay within the boundary line by some 3 to 5 degrees in the

value of 9' in order to allow for deviations which may occur between the

laboratory value of the angle <^' and the real value which obtains in the

hopper.

In designing plane-flow hoppers, Fig. 37(a), the

regions of mass-flow are much wider, Figs. 50 to 57. However, It is

recommended that the slope angle 9' be selected to the left of the

dashed lines in order to prevent the development of excessive non-

flowing regions on the sides and at the top of a bin. Such regions

take place and extend upwatd to the surface of the solid when the head

of the solid in the vertical portion of a bin is low and the transition

from the vertical portion to the hopper is sharp. If the transition is

smooth with a radius of curvature R >^ D/3, Fig. 39, then 9' may be

increased by, say, 5".

These limits on the angle 9' are not rigid like those on 9'.

It is not necessary to stay 3" to 5" to the left of the dashed lines

for 9*. In fact, the actual value of ^' in plane flow will often be

smaller in the bin than was measured in the laboratory because the

plane walls will tend to polish as the solid flows on them during the

periods when the head of solid in the bin is high. In a conical channel,

flow will never start if ^' is too large and point (9',^') is outside of

the region of flow and the walls will never have a chance to become

polished.

The end slopes 9' of the transition hopper. Fig. 37(a), are

selected from the charts for conical hoppers. Figures 45 to 49.

-72-
 (J. R. Johanson)

^c ^ ^^* c 9» = 30"
c
f£g. 38

Flow in conical chamelSs 8 =* 50°, / ' = 24'

-73-
 Within the mass-flow regions, the flow-factor contours indicate

the flowability of the hoppers; the lower the value of ff, the better

the hopper. It will be observed that for all shapes and values of

angle 5 there is a trough of low ff-values. Thus, for a given value

of friction at the walls ^ \ there is an optimum slope angle 9'.

Making the slope steeper, as well as less steep, decreases the flow-

ability of the hopper. The decrease of flowability toward the steeper

slope is computed conservatively; the analysis is based on the assumption

that the domes which might form across the channel would fall across the

solid and not along the walls. This assumption is justified by the fact

that the angle of friction at the wall after storage at rest, i.e. the

static angle of friction, is usually greater than the kinematic angle y ,

and that even a slight increase in ^' is sufficient to eliminate the

weakness at the wall and to require failure across the solid [33].

Hoppers should be built of materials which do not corrode or scar

and permanently increase the angle of friction <))' during operation. A

temporary increase of (^' may occur during storage at rest. This is not

harmful as long as i' reverts to its original value when flow is restored

Hoppers with a vertical wall made of a smooth material seem to

have a slight advantage over sjmmetric hoppers in starting the flow of a

solid after storage at rest. This is probably due to the fact that

flow tends to start in a vertical plug along the smooth wall. Such plug

flow develops more readily than either plug flow within the solid in a

symmetric hopper, or mass flow. The advantage in the flow-factor due to

the vertical wall is probably in the range of .1 to .3. The drawback of

-74-
 Fig 39

Bin with a smooth transition to a wedge

hopper

Included angle

Fig. 40

Conical hopper with one vertical side

•75-
the asymmetric design is that, once plug flow has developed, it may

persist and may fall to transform into mass flow.

Flow-factors for mass, plane-tlow hoppers with one vertical

wall have been computed and are shown in Figures 55, 56, 57 for

6 - 50" and for angles of friction along the vertical wall re-

spectively, •^ - 20", 30" and 40". 9' and (>' in the figures refer

to the sloping wall.

To date, only the axi-symmetric, conical and the straight,

plane-flow channels have been analyzed. For other shapes, the

theory becomes less certain and the computational difficulties

increase out of proportion to the likely practical benefits. This

does not imply that no other mass-flow hoppers are possible.

A conical hopper with its axis tilted from the vertical, e.g.

with one side vertical. Fig. 40, will lead to mass flow but its

included angle should be only 1.25 9', where 9' is the slope angle of
c c

a symmetric cone for mass flow. If the included angle is too large

plug-flow will develop along the vertical wall.

Rectangular, full-slot hopper. Such a hopper can lead to mass flow

provided its valleys are filled or rounded to prevent material from adhering.

It is advisable to design the outlet length greater than six times the

width in order to guard against the possibility of funnel flow higher within

the hopper.

-76-
 (DELETED)

Pyramidal hoppers. A pyramidal hopper, Fig. 32(a), may lead

to mass flow In spite of the In-flowing valleys; however, observations

Indicate that the flow-factor of this hopper Is more than A for mass

flow, and the valley angles should be within the mass-flow region for

conical hoppers. Since the flow-factor of a plug-flow channel Is less

than 1.3, Fig. A-7, plug flow will tend to develop in a pyramidal

hopper unless the solid Is sufficiently free-flowing and the walls of

the hopper sufficiently steep and smooth.

-77-
 (DELETED)

Dimensions of the outlet. The proper selection of the shape

and slope of a hopper leads to mass flow - if flow develops. To

that effect it is necessary to assure that doming does not occur by

making the outlet sufficiently large.

In mass-flow hoppers only one dimension of the outlet. Fig. 34,

is determined. That dimension, B, is either the width of a rectangular

outlet, or the side of a square outlet, or the diameter of a circular

outlet. In the design of a rectangular outlet, it is assumed that

plane flow occurs at the sides of the outlet. For this assumption to

-78-
be valid, the ratio of length L to width B of the outlet should be not

less than 3.

The procedure is outlined below, step by step. It is also

illustrated in Figures 42(a) to (e). Figures (a) to (d) are drawn in

v., F & V. coordinates, while Fig. (o) is drawn in V,S coordinates.

(1) Estimate 5, ((t'and y at the outlet.

(2) Select 9'. For circular and squate outlets, use the conical-

flow charts. Figures 45 to 49. For rectangular outlets, use the

plane-flow charts. Figures 50 to 57.

(3) Read ff from the appropriate chart.

(4) Plot 6, FF and if. There are four possibilities:

(a) There is no intersection, and FF lies below tf.

The analysis does not determine a minimum dimension B.

Select B from considerations ot tate ot flow, particle

size to avoid interlocking, standard designs and, using

Eq. (12) with A * 1/13 sq. ft, compute

Vj^ « BY/13H(9'). (22)

The function H(9') is given in Fig. 43 tor circular,

square and rectangular outlets. Locate point V. on the

ff-line to determine the value of V. at the outlet.

(b) There is an intersection at point (V.,V. «FJ. This analysis

determines a minimum dimension B from formula (22), thus

B > 13V^H(9')/Y- (23)

-79-
(c) There is no intersection and FF lies above ff. The solid

will not flow. If lower values of ff are available, select

6' and, possibly, ^' by using a smoother wall material to

reduce ff, and start again from step (1). If lower values

of ff are not available, and if FF is an instantat^eous

flow-function, either solids conditioning or non-gravity

storage is required.

(d) The time flow-function FF lies entirely above ff but the

instantaneous flow-function FF lies below ff. External

vibrators are specified to start flow after storage at

rest, and the outlet is designed with a safety factor to

allow for the unfavorable effects of vibration. This is

accomplished by selecting V. so that, at the outlet,

Vj^ = 1.5F. (24)

The construction is shown in the figure. The width B is

then computed from Eq. (23).

Now check •' at the outlet. This is shown in Fig. 42(e) in the

following order:

- draw WYL,

- read 6 from the appropriate Fig. (a) to (d) for the selected

value of V., and draw EYL,

- mark point (V.,0),

- draw the ir.dicated Mohr semi-circle through (V^^.O) tangentlally

to the EYL,

- the indicated point of intersection of the semi-circle with

-80-
 (a)

13V^H(0')/r v^ (b)

vi n
8)
F & V
1: k ^ ^ ^ F F
. ff (c)

—7r>

P & V,

B - l3V^H(0')/r Vj
(d)

(e)

(Vi.O) V

Fig. 42

Calculation of outlet dimensions in mass flow

-81-
 the WYL is (V',S'),

- draw a straight line through (V',S') and the origin; the angle

between this line and the V-axis is ^'.

The above procedure assumes:

- A straight or convex-upward FF. In the exceptional case of a

concave-upward FF, the point of intersection (Vj^,V--F), would determine

1
the largest width, side, or diameter of the hopper.

- A straight or convex-upward WYL. In the exceptional case of a

concave-upward WYL, it may be necessary to check the dimension B at a

few elevations within the hopper, as well as at the outlet. At each

level, the procedure would be the same as described above for the

outlet.

Since the values of ff are Insensitive to small changes In the

angle 5, it is not necessary to define this angle with great precision.

Stresses on walls. . The pressure o' which a floVlng solid of bulk

weight Y exerts on the wall of a channel of width B, Fig. 44, is

computed directly from the |>lots presented in Figures 58 to 67. For

example, in a conical hopper of 9' - 15", ^' - 22*, for 6 - 40",

Y " 100 pcf, and B - 2 feet, O ' / Y B is read from Fig. 59 at .50.

Hence o' - .50 x 100 x 2 - 100 psf. Similarly, for a plane-flow

channel for the same values of 9', ^', 5, Y and B, from Fig. 64,

O'/YB - 1.04, and a' » 1.04 x 100 x 2 - 208 psf.

It will be observed that these pressures are quite low and that

they increase upwards from the outlet as the width B of the hopper

increases.

-82-
 J.O
, , -,
• "1
L—._. .. H — . - -

H(e») \ -
J
2.5 L._—
— ,—«"
Mtco^ ti—-^•'"""'^ ^
" ^
^^^^ ce_,.-i - — " ^
1—Z^
2.0
^ ^ ^
- •- - -
. _
L- - - -

1.5
1
J
i » . (L> J B i ^

1.0
"^ 1 _— •J —
L_&S£t IIIK^

10* 20* 30" 40' 50" 0' 60'

Fig. 43

Function H(0')

-83-
 The shearing stress T ' which acts at the wall under the same

conditions can be computed from Eq. (4), with the substitution

V - Ao', S' - A T ' , thus

r' - o' tan *'. (25)

The above stresses occur in contact beds in mass flow. If an

aerated solid is charged into a bin, the pressures are hydrostatic

until a contact bed has developed. In plug-flow bins, within the

flowing channel, stresses are non-steady. Around the channel, within

the non-flowing mass, they are statically indeterminate, hence, highly

dependent on the method of charging and draw, time in storage, and the

relative deformation of the stored solid and the containing structure..

Thus, in plug-flow bins, the wall stresses cannot be computed with any

precision. This situation is reflected in the large number of publica-

tions on pressures in silos and bins and in the general lack of agreement

among the authors.

Slip-stick solids

Solids which exhibit slip-stick properties during the shear tests

are evidently incapable of steady flow, and the consolidating pressures

during flow are likely to be higher than those computed on the assump-

tion of steady state flow. To account for the over-pressures, the

flow-factor used in design should be increased in accordance with the

amplitude of the oscillations of the pen, as shown in the following

1
Oscil- 2S/3 and S Design 1.2ff
lations S/3 and S flow- 1.4ff
between 0 and S factor 1.6ff

-84-
 ^ B

^ • '

V
/

\ /

Fig. 44

Stresses on hopper walls

-85-
30*
|||]]]^]ljjjjijP|]]^#jljfj]jl
20*
HIBH
ftpji 1 ]]^^ii i -1111pjipj j i] pii^^
jni^^usJJjT^i^HJJJ J1 n\ |1 nJI M-tj] niIT i u"^ii IT
10* J^^^MHOT \\/\\\\\ liiiflllH+ffi
•^^T-jjaTi^ii iM 11 nJJ 111 111 T JJJ I W J I J J J I I lijJJil"

0"
0"
i;l|i-|ilii||p
10" 20" 30' 40" 50" 0' 60"
c

Pig. 45

ff contours for conical channels, 6 =» 30*

-86-
40'

^^

30« :!;iy; ^

:s;
•1
"<;
-V
20' ^
IS:
:-53! ==^r
-S^^r
:E2I
:^^ 5^3
:5: 2v ^ :
10* ;!Es; : ?i; :i
:!5=
^5:
:5::^is ;iii!;i; :E:

^^
"«•
^:
!!i: :5; •h^: :::

0« H
:^^=
H H:
5S: ;^
:v. :s:
I0« 20* 30" 40' ^c 50'

Fig. 46

ff contours for conical channels, 6 •= 40*

-87-
 11 i H 1111 H 11 1111 i 11111'H'i i i i i i i i i i i i i i i i i i i i i i i t i t
I 1 11 1 I 1111 1 I 1111 I I 11 1 111 11
*o n > j
r >[
1
r 17 1i M
1 ^^v I l l i l l i l l i l i t l l l l
J1 'j>i
'' J N 1 11111 1 1 11 1 1111 11 1 111 1 1111
LA I r r k l itt'iti'illilMitiilliittiit
11 1 1 1 1 1 1 t 1 t 1 t 1 1 1 1 1 1 1 1 1 1 1 1 1
I t l l l l t l l l l l l i l J l l l l J l l l
in* I TTI „ M M 1^ 1 1 1 t 1 1 1 1 t 1l lI l 1i t l M l t t i i t l i i l t l i M l l
30 1 1 1 1 ^ 1 11 1
1 1 >J 1 1 I1 1 1 1 1 1 1 1 1 I
t t i i i i i i i i i i i i i i i i i i i i i i i i
1 1 1 1 I 1
rf u r'J.
iTS
1 I r r - L t 1 1 M 1 1 1 t 1 1 M : TH. 1 1 1 11
1 1 11 rN-LI 1 1 1 1 t 11 n 1 Nrvt 11 1
KJ 1 n44T-H; 1 rSl I ! 1 1 1 1 1 1 I 1 1 1 1 1 1 1 1 1
\\TA m l IT -/ PR l i l l l l l l l l i l l l l l l l l l l l l l l l
MT-L T 1 1 m ^ 11111t 1I 11 1I i 11111111111I I
,n* 1 T> L rrs 1
20 U P NL 1 I I 1 I1 1 i
|>>J 1 1 1 1 1 1 1 1 .1 t 1
1 1
1 1
n>[ 1 TSJT.
L iHi 1 wiJ |vj 1 1 1 1 1 1 1 1 1 1 1
Iril nl 1 n^ 1 1 1 p4 1 1 t 1 1 1 1 1 1
t i l l Tx. 1 1 1 1 1 t 1 1 1 1 1 1 1 1 1 t 1 1 1 1
'>Li>L n>Lh im 144
M^l
1 rKJ iS-J 1 r r i L n >L
M>LrN \JW\ \Tr\ 1 ^ >L
>1 m J iS-PJ T l i 1 m 1 1 1 11 1NiS 1 1 \ t Tx 11 i t 1 1 1 1 1! 1 1 I I
^" rlJ i>l n-ll>[ u r i ) 1 >J
f>LrN UlPfflH rlJ 4 M M; fiM
K N ITl LrvJ l> M>L 1 rrrTl iSL
i NTM.n >Ll t<T
4 bi t rt l I H Trs.NMM I Hr k iH
IN
1VTH 1^4 r"M 1VL "Si H t 1 t t M 1 1 1 1 1 1 t 1 M 1 N. 1 1 1 1 1 I 1
meiii>LM4 I N J riJ M' AK ^ ^1 I 1I 1! m 1
IN.
111111 PFI 1 r4 TK ["ki r>
r>l i>l r m nA r>J
n«l 1 1 1 M 1 M Kl 1 1 rJ 1 n-.! 1 1111> H I1r-ki 1 1 M I 1 1 1 1 1 M t M H 1 1 1
10' 20' 30* 40' 9' 50'
c

Fig. 47

ff contours for conical channels, 5 = 50"

-88-
 Fig. 48

ff contours for conical channels, 6 = 60'

-89-
 U 1 1 1 I 1 1 1 1 1 1 1 144 4-4'HHH \'\'{-\H'\
[ Hii"! \ \ j4-| 11111111 i
IN.
N
/.n* iSL 1 1 4-1 11
KN t
1 1 1 11 1 II 1 II 1 1 1 1 M
'^•iiiiiirNiiiiim
^ N iTiTrTiT II 1 II 1 1 1 1 II 1 11 1 1 1 1
H II II 1 1 11 1 1 II 1 1 1 l l 1 1
1
rmi ri 1 1 1 II 1 1 1 11 1 M 1 1 1 1 1
3o« n T44 U ikN
PJ N
rTrK+ r> •
I I 1 I I II 11 II 1 1 1 M 1 1
U '94 r>J I1 1
1 11111 t1t 1t I1It 1111111 1I I11 1
r1 H
1 i1THJ 1 1T1 1 1 11 >J 1 1 1 1 1 11 tvl
1 11 11 11 11 11 11 1 1 1 1 1
Kl 1 1 ['"t'i 1 I I I I 1 M 1 M I N J
rKJ
L r>L i K i>4-
J NN 1 1 1111111111 11 1 1
riJ i K J n>L N IIIIMMII
m 1N 1 H I L N
20* R l n >I lN mMJM Tr4(fH MTS'ffJ M
i N 1 M 1 M M 1 1 1 11 1 M 1 11 I I I t 1 M
mM>1 n>l iN 1 >LMrs N. l i t 1 M M T M
1 N L irhL r Kr
risl 1441
T4"t t t l t 1111 i t t 11 1 1 1 1 1
i>j r K >J r--^J PrKL
^r>j >N
liN
M>J
1 N L >ri-
rvkrW
1 H i MN 1 1 M t rt M IT 1 1 It 1 1 M t 1 1
t
1 i Hv 4 I^-NJ
i m H 1VL f W i>j 1 rrviy [ rvl | 1 | 1t 1t 1 1
r m 1 >J N r T l M M
N n>J>L
io«M> i>74 r i iH
i> n-fcr^ iHlTT'
HIM rrKK* N r M .T 11111'M r>j
M 'K M I I M II M II M
M f> SU T>£NL M > Mi h 1 M-
iVJ m WSL iKi N Mb i i S,T>1 "•
Pi
N
M 1 Vr-TTcf^J i R l H M H >
K 4 M.rH MS 1 > N rr J \H Si
N-'l t^pi > > Ki H "l-ffN 1 PKl
NTci r L1 >J r>rN H
MR H I M I N ] > li-
no 11111, J1 Is! 1 N 1 Kl 111 IN. r kl 1 1-kl
1 144 1 H I i 1 II 1 1 1 II Kl 1 II 1 1 1
10' 20' 30' 40° 9' 50'
c

Fig. 49

ff contours for conical channels, 6 = 70"

-90-
 Fig. 50

ff contours for symmetric plane-flow channels, 6 = 30'

•91-
 11111 11114U t M M 111 M 1,11 11 111 MM}} 1 1 1 1 1 1 ' 1 11 1 1 [ M 11 { \u
• no 1 1 1 i ! 1 ' ' 11 ! 1 J 1 ! M !1t 1 1 ' ' ' ' ' ' ' ' ' u '1 1 M1rrt
M 11 i- j 1 M M i jl I
^i Li ' M > l
1 r 1 111 T-4 1 1 IN. 1 1 I 1 144 I 1 11 1 1 1 M T 4 rL1
1 ITH H 1 t M ITT
^' U 1 r14-Ln4. M i-1 1 H4
MrkL M 1 1;J_M 14 1 1 1 1p*!M M1- l l iM k^f^J II 1 •1 Tx
NL1 M
M 1i 1l I l 1
r " ^ i 1 i^L M l
14 11 rs kl M l M 4
N irN M 1 1 n 1 M M rKrtt 1 1 1 r k i 1 I NI 1 1 I 11
1 It t M 1 1 t 1 [>=>.. T 1 MM
iJ THi
30" L^WJ M N M H I M • M ' M 1iT'H-i M H i 1' M ' 1 IN Ml M
t^Ttfyin n
1 M M 1 I I 1 11 1 11 11 M M 1 1 1 1 Svi'U M M
' N 1 - 'r'Kki 1 M > i hi
11 t M 4 t t t t 1 1 1 t 1 1 1 1 1 1 1 1 1 1 M'iJ >^ s M M 1H "MriNci
T>, PN
t iSsJ N1 1
[
f-JJ 1 1 t H'i-
|N ' J
1 M L l l Hd? r H 1 r4.
MNfiVt TM ri r-
It rHn Jrt-MJ I NiV- L MTr'' M> 1 KM M
i t '^ IH N
1 I T W U Hw J MN HJ 1 N M 'rM^M M >l
20" rLn j M1 r H1 fM
l ^ M 1 111 '4^11 klN M MS 1 I I M l i^J
'N M
M>1
1 Sk M M^J M 1 11 NIS'MM ' > M
1 "^l
riXllf m-L M M M 1 M1 Hi R MVl
IITIHL
rH-L IrHL rtiS^J 1 1
riL-hl iM I Ml 111 1 1 r>L HJ h
M S SIH
[ 1fKJ ir~L r m irs^ M4 n4^4
1 Tsn M HL MS 1
1 sLJ 1 rkLi iPsr-lJ l l M i 1 I M HJ
ri-L H
MM
10" 1 rhU rtrs^'LLrW MTH 1 11 i 1 tv N 1
HI iHl Ll M'^vm Mr J
1 1 t r 1 PH^frl X I I I 1 i""-^ 1 MJ psij
jSJ rkn
1 i"4-l 1 T'^ii'l 1 I r ^ l"l 1 M r4. M l ! 1 1 t l It t 1 1 I I 1 1 1 1 M M rSt t 1 It 1 M 1
M i4 U l^^•«h^ rk ll ] 1 rJ M M M pk 1
n-rHTTTbt W W 1 t T -' 1 i Hi J14^1I 1 1 1 1 1 1 L [ .H. 1 M i
M
>L
M1 M1 1 M
M l U \\\\ 1 1 1 I M ~
n-HttHHHIfHllttilJiliJ IIIMJI M14I I 1 1IIIt1r M 11 t M It 1 I 11 1 1 t 1 11 1 1 1 I
M rM 111 50' II II II0'II1160"
M II 1
10' 20' JO" 40'
P

Fig. 51

ff contours for symmetric plane-flow channels, 5 = 40

-92-
 Fig. 52

ff contours for symmetric plane-flow channels, 6 = 50"

-93-
 Fig. 53

ff contours for symmetric ilane-flow channels, 6 = 60'

-94-
70" iil'lil UJJ i l LL
hbtiTffffbTOfLTl'H^l
l"N-l 1 T~W 1 r+-L PH 1 l"H-_tarTT^ II 1 1 II
T s H iHn1 rirHnsH i^rE-rN
^i • nA
I J I >1 L
H J NM ISJ
INH1M1
J I > PHiyiS-LN
H n IN
M>L 1 N 1>L M>1 H T T M 1H N 1 1 1 II 1J 1 M 11 1 1 jl 1 1 1 1 1 I 1 1 t
rriJ i>rHi > ' ' nN.iwVN 1 >K1
^na\H1iISI 1 ITHIN
II IN. 1 IX 1 Ni MTJ n MS1 l \ 11i 1m I^W
1 J 1 "Vlr-rS
°" HJ iHJi>L
rM iMrM rvrii> M1>J
rSl >L«i5tU r i Hi
r k r nsP NN M 1
H MSJ 1M r>l W R MS IV I N 1
I H J IM H IN.
M> HJ ]> 1 HJ^MIS i>wS H i M>J iM N NMX
MS 1 M M
1 1 N 11 1 1 1 11 1 1 11 1 1 1 1 1 1 1 1 1 1 M 1 II
L n>i 1 >1 IN y 1 M I S J K 1 h
1 M I N M T I : M S 1 T^rS M 4 H H n s IS n
Hi MS MS HTr\ MS I N N 1 H Hi IN
1 1
LIN
in'r+L ISJ iKJ H > M> ^
M > iTq^TM I>iiS fSJ S 1
H^^pi'i N Hi
KlN
•.J
MSMN
>iN 1 1 II 1 1 1 1 1 1 M 1 1 1
M N 11 1 1 M 1 1 1 i M 1 1 M M 1 M
^ H1 1 >1
>rU MS MridS MS
N k4J 1iwISJ14M W H J
N iSM N f
IN
M>L> 1 m l v iN
441IN
MM RMS. N N.K41
MSm
NiSN 4MN
M R
r- H S. H N
MN
M
M 1
1 1INiSclJ-1MKVi M^ I 1 1 LNIN. MSIX Hi IS
S > SIHMS
J't i
H l l
Lrl
P i T J Ki MJ Pv iJ kI M
1 TsJ
x r
M
^ i N TVI 1 1 4*4
1 1 1 M 11 1 N'lfL 1 114 11 11 "N 1 N 14-J i 1 "N^1r>4
H HJ ISMS
kM Smr i HJ
NM MR
N iR
M
II 1 1 II 1 1 M T*CI 1 M M II M 11 "hM T M 1 - ?>
/.n*I 1 1 1 1 l i 1 ' ' M-l 11 1 N 1 1 1 1 Ml 1! 1 1 Nl 1 iS V > H H HMS
H M^MH N H- H >M
MSr v t lMS
LIIMtlllllM l i x M l l N l M I rH^ Hv R H IX R1 >J R
rW pf^ H i t 1N H S, V HH.RHiS H HJ HMSSI 1 SJ
N MS
MS M1 MHRS
pw 1 T4 H I M 4 H4HJ > S S M NH hJ-Rffl44
TijtHMS m r o W MNl MX\\
HI
i1 >J 11NN r>4 V r>-i S M14 14J
MfiMsi- HI N ri^ R rSi Hi H IS H iS > MSN N H MR R IN
iVM 1 >L iri MS
R M^i rSl H H Hi MHHR M mlN
•»n» N N M x rtJ - MS N S N S Nr-HHHH Hi
MMS, HJ S N
MS M>1 \\\\ MS
M M>i
1pi1 N 1 1 H
rN
1 1 N 1 1 1MS
MS*.
M11
11 1 M-1 1 1 1 11
MS
SvIVI M1 1 SfflfflflffflfflB^MH
Kl MS N rk! W MSJ >HJ TH n
H S kVl H14 H R 1
HJ I rk N H H Hixr IJ
nS
L MSH H4 MS M>J MSrIS
M ki4 i I\NN H k N MX Hv Nv H MS M 4HNHN 1 M l
N MS R M>M 1 Hi M^
M > H-L 41
MS MS^ T4
"0" M 11V > IPIMS
MN M MS 4S
>N iSIHSriisitTiJ
M1 ^M M SMSHIt4xwt4im4tii'i
HH H RRHl HH M1 l
>
H N 1 >1 MT^M 1 >i -
>1 N MS \?^H-\ M N J MS- NHH MM J R 1SJ H R H H
N S H MS H11
R MS 1 >J JnRH-1 M>1 1S i l f r t t t \ t R m f R H ± p R N ± | 4
KMSv HJ vH^R"
IS'TH rrr MS
kHJH lHHJ 1 I k m - K HJ1 U i MSHJ M 'Fn RTRIHI H T n i r H r ^
1*'0"PWXii>v^l
S ^ t m ^HM H^^ IR nk'Hs M H ^
M f"^ MS 1 HI M ''H 1 1 M M^ M 1 N 1TN MNJ N
il M ! imM LHH i
™ RR SWI 1 H
HJII 11 MISMR Ml M
no Hi MSWf-tSHf 1 rNHTH(1 n s MsN R R
0" 10' 20' 30' 40'
M M ;C' K M e' '60'H
[ i H l f R H f i M l R M H H f W tpH t M
Fig. 54

ff contours for sytimetric plane-flow channels, & = 70'

-95-
 Fig. 55

ff contours for asymmetric plane-flow channels with

one vertical wall, 6 = 50", f* = 20"

-96-
 1 111 11 1 11 1 1 1 1 11 1 1 111 M 1 1 1 1 1 H [ 1 1 1 1 1 1 M 1 1 1 1 1 1 1
MUUHITYH^^ PR 1 "H~
I 1 1 1 1 1 T*T*4tei| ! M 1 ! r4J. 1 1 Tr-1 1 1 1 1 T r>-J 1 I 1 1 l MW m*A 1 Tx 1 1 1 1 I 1 1 1 I t 1 1 1 1 M 1 ! 1
hw rVii M i n - J MH4iiHSJ nSSL
TrHLiH-Ln r i U iHS
TSH1 HL
T H iHi'TdJ
1 MrfJ M>1
H N MMVP H
1 rrU
irvlJ 1HJJ 1MSSJ1 H I PrmS PHJS.
rMTii M LM n l MISITRULI 1 rrU HmJ R-
i \\nm
iNt
1 1PH L 1 THIDTO
iHJ IrHJ rW
tl^i IM 1
1R 1 iS.
trti
M >-<*U
IMSNL
I MS-J
N M>V[ rN
I N ' M>i
MSi1M
MHMx
M
MS^HJ
HnSJ iR
M>NJ
kl M>i IN
I N MSJ
MSH iMSMS
R PRrfi
MS IHL 1T r41i i MSJ
HL HLH PMSSL MS MS
LL
rT+4 1 1 1 1 1 1 1 1 1 1 i 1 1 M M M X I 1 1 "1 H^i 1 1 1 M M N 1 1 1 1 M "I^J 1 1 X I 1 i1 >L
M M H I iHJ RR N J M>J H 11 X M N 1 11
TITT-^M M H MrL 1 r^l MKL MS N MN
]Tn444 R
nrK-H-U LM MS H HPHI nH i HHJiSJH R RHJi
1 ITMI I I I r H4^
M l R >L '1 MNMR MS I N M 1 M
MS M l ' '
m 111111111111 H"l 111111111H H'LH 11 11 FTHd
RITWJ
11 IH^I 111 MHiJ
TTRW 1 Txl
W^\ 11
H '11'
1111 ITtftliHSi
111111 i 11111 nil 1 ~i H Ml TlM111 iFhHdtfHitHHItiT
1 txl Hi N
U 1 MLL M iM^JrL 1
Ni4LI 11111111 ITHsJ 1 ILiI I M 1 H H r LLLHL^t i1 MTI I IT IThl'BxfTTri
rSSi
MrR M T-ri'L
Mi^J M1SfS
MTM
1H 1 rl Rl
M 1 1 1 1 1 it 11 1 MSJ M M 1 t 1 1 1 T x l 1t t t 1t I 1 11M 1M M 1
M"RMi4 L
r«34 liSi
M1SJ Ml
MrVL MiSJ
THL
1 RL MiSJ
MM
1 11 1 1 MS-L IX-L
Irs 1 i^L I M 11l M"
11 1 1 1 1 1 1 1 1 1 1 1 l l 1 1 1 1 11 1 1 I I 1 11 1 M S J I 11 1 1 M I 1 N^l 11 I I 1 1 I I 11 1 M M 1 1 1
PHL PHi M
M I I M M 1 M 1 1 II II II M 1 1 II 1 II 1 I I 1 1 MxL IN
I I PN.! 1 1 II 1 11 J I I II 1 1 M I I 1 II
MSvl
10" 20' 30' 40' 50" 0' 60'
P

Fig. 56

ff contours for asymmetric plane-flow channels with

one vertical wall, 6 = 50", /' = 30"

97
 Fig. 57

ff contours for asymmetric plane-flow channels with

one vertical wall, 6 = 50", ^^ = 40"

-98-
 tillRtiitmi 11111II111II1111111 M H 1 M } 1111 R R i R f i M 1 i • H '
1 1 1 j 1 j 1 1 1111111II111 1 II II j M M
K-tiiiTIJJ^nTiJrilfll I 11 Ml M 1
I1 1I 1M
I11 M 1
1 I
Jl rrvl
MpHM M M 1 1 1 IIM 11 1 1 1 1 1 11 1 M I I
MM rVif'a 1 1 rfS^
Ml 1 PI P—^H Hrl 1 Irxl M l
20 jk[U1I^rH.J.
M l ^N^/i 1 [ MTTT T H i
1 1 1 M M 1 i! II 1 i M 1 I
M M MT*l^^ti i J J f
i ^ * J 1 ll^T*4^Li A I I 1 1 PW. 1 1 1 1 I I 1 1 1 1 1 1 1 1 I I 1 1 1 1
l TxL
1 1 ^TT^(DM'J 1 I TT^ILI 1tl 11M 1 1t 1 1
! 1 M 11 TTML H i 1 ITRJ
PKI 111 l i f l i l j i ' L
M
ii^WJ M i l l
i PHJ
rsu
1 rTrSJ. r l J l l r tff' M l Tr4j. M l MSH
.prrrwbRL n r r L L THTTS.1f ti+H- -kl+HiJ IT
10" rRRR-T-pTfechpRFpn rTrfkH- 1 TITl N
J] Pfr-L iTrS-
M i l l nPh«L 1 1 1 1 1 1 i^KU TrHL PHJ
LJ'SL
r l I PKUI-'UI 1 1T
1 rW HL
rhkL rrld J rrh'TL TTrri-"WJ M hR
rxL PNJ
rXi M
rMPP-HZhi
hMS^LJrKHn
IFRf^pkI;r^^ 4UTxll
THl i rhR
irrhL
M t r 4 i rrH^sl MM L M
MTK. M i l l TXi^ 1 TXL X<1.
ffi^rWrrPL
TH THLTH MriSd
rHsLI 1 IIr II
H4iiTXll M i l PHJ
l IT<J 1 11x1i Pkl 1 1
0" 10' 20" 30' 40' 50' 0' 60'
c

Fig. 58

O'/TB contours for conical channels, 6 = 30'

-99-
 1 1 1 1 1 1 1 I I 1 1 1 1 I 1 1 I { 1 I 1 1 1 1 1 1 1
1 1 1 I I I I 1 1 1 1 l l 111 1 I I 11111 1 1 1
AO*

i(' r x L M i l l
p THiL irHL M
R 111II1111 1 1 1 M 1 1I
MJ l y H i
Iff/ R s 1 b t X U
30* m-lLPri M xL
TTTTi HrtPHL 1 11
1 1 1 1M
1 1 1111111
1 1 1 I I 1 1 111M
1 I I1111 11 I IM1 1 M
11 1 1 1 MM
l l M l 1 Tl liJ
rHJ
1PHJ
xL 11 1111
1 1
1 M
1 1 1 1
1
1
1m i
PirNlriSJ HR
PIT TiH
PNJ
14rxiiR LN
MrrxlJ+f-L Rl Irul
HJ
n>LL Ml rx1 H
mS R PRH PRi
20" \\r\A 1K L
r4, 1 MThl^ R 'I INS 1 i<Ma' X I 1 1 1 1 1 1 1 1 1 1 1 1 M 11 1 1 M 1 1 1 1 l l
PRMSv 1 M
w. Mxl
iHLMiSiR Mr>i
iPrRi rxT J
i H l M n3^ 11 Ml rtL
1 1 1 1 1 XMI (i-'LX*!^^^! 1 1 1 1 rXL H 1J 11HL
1PR1
MSJ M
•
MJ M T T M 1T^ PHL M-L HS.
nS^U MS^tkJjxL
MxL o ' \nii P H L PTXL
Prx 4^ Tsl t i l
10* k PHr>J 1MxLlS PN-J
m 4 MS^L PHL
lis r H P >L
iH J PHi ii(
\vt\ H i \\\U.\\\\\\\\\\\\
rHJ
U HS i i
ci IKJ M>J H 4
T THJ HH I n MSV
HJ •
INMxi^rTRJ
1 MpfJi I NM>i M H i>L J MMM-L MS-LPHi HPRI M
M
N JS^m 1>1^ r N iHL H i
1 I'X'trfjlXWi-Ni-l M i l I'Xi 1 I 11 1 N 1 1 HJ 1 HJ H4 i H
HJ
Pm--mi ixL PR M> iXL H-L 1x1
o» L L MT1WMi ^ lMIP 1 1 1 r k l 1Srs11
x LHIJr L I 1MXL L 1 H
x l 11 1H
f x t t M 1 iHJ
M 1 r x l 1 PR
1 1 1 1x1 1
0^ 10" 20" 30" 40'' 0' 50"

Fig. 59

O'/Y"B contours for conical channels 5 = 40"

•100-
 Fig. 60

o'/rB contours fot conical channels, 5 >* 50°

-101-
 Fig 6'

a'/TTB contours for conical channels, j = 60"

-102-
 Mill l"t~Hift'iifhtli'tlfilt't 11 i-t i-i J 1 1 J 1 i j 1. n 1 i 1 1 i 1 I u ; i
50* 1 11 11 1t11t1 I1111t 11 1
1111111111111111! 111 111
ilIlLilllltlXllltlllllll! 1
1 1 1 1 [ I 1 1 I 1 1 1 I I 1 I 1 I ' 1 1 I I ! 1 1il lI ll li lr li ll tl ll lt lJl ll T
i ll il li tIllllltlllt-tlllli' l
If LliilLMIlllllltlltil i l l l l ] I l l ! l L l | l t i l l l l i i l i l i
bl 1 I 1 I
ho* rrtiTTri T i l l 1 1 11111M1111I111[ 11111111 1
11111 i\i 1111111111111111111
1 1' I N
1^'rH mHWmH t
IlilMIIN.IiiIlillllillil
1 1 111t11111
1 1111111t 1It 111 1111111111 1
111I 1 1
l l t l l l t l l l l t t l l t l l l l L l I l l l l
t 1
1 Nv 1 •
L11 1 I I t M t 1 rx 1 1 1 1 1 1 1 1 1 1 1 1 1
ITN PK t i t ' T i T i i i i i i i i i i i r i i i i i i i i
1ft* 1 1 i ^ 1 H '1 1 1 1 >
114 rN»rri rkl It
I I 1 I I r pH/T-J^ l i l t IrJ 1 1 1 1 I 1 1 1
1111111 ppiiyj 1111 [[v] 1111111
111rJ 1 1 1 1 t 1 T4-4 I 1 Wp-J I t !
>v| 1 1 1 1 1 i 11 r T 1 r i 11 ^ j ! 1 1 1 1
M> N
ri 1 T p* I t 1 11 1 1 t t 11' 1vl 1 1
U 1 s . ^ i ^ 111I1[1 j11111 j11t 1] j Ij 1 I
mJ rilil-i-M-l -I UKl| i | [ j | t j l | l l l [ l | t J | | | | | | l l
U Mrs KJ r4pj. 111 1
1 111 11 1I1 11 111 1
1 1II11 1I1I1 II 11 I1
Li r U i i IN. 1 rrvl «*J- 1^v l l t l t t t l l l t l L t l l l l
1 rl TiJ rmil 1 1 i v l 1 i 1 1 t11 1111 111
TiJ 1 r l 11 rrJ 1 1 1 iHi 1 1 11 1 t 1 1 1 1 1 1 1 1
liiiiiiHIimjjj t'H Til I T ^ T I T I T n T n TTIT
io*miTtT MfrWiT? tMr]Jj|(|jf||j]j|[}J4J|-t|^
rttttHkJ i M ™m 1w11 HI +'
B i SmN ^ tmitfl
H^^^iJm T iTl1 1 iH1 1 1 ffU J|14l[[ ] llj
L T> W

10" 20 •
INJ

ilillJliltlmiMiittli
30* 40 9'
c
50'

Fig. 62

o'/rB contour* for conical channels, 6 = 70*

-103-
 •

30*

20*

10"

o ' " i II n II11IIIII11 r u n IIII r s i i i IIIII11IIIIII II iT-hiTTTn r n ri ^

0* 10* 20' 30* 40* 50* 9* 60°
P

Pig. 63

o'/fB contours for synmetrlc plane-flow channels, 5 " 30*

10^

•
 - 1II I I I I — 1 hh 1—H-j-^
Aft* _ . . > i _ _ - _ _ _ _ - _ _ - _ - _ - _ _ _ _ - - - _ _ _ J
*" : : : : : : : : : : : : : : : : : s : : j i : : : :4a \ \ T ^^ r
:: : :_ _ . z i i . i _ _ J p K I^
:::_::::::::: : : : : S : : J : : :_j *i i \ \ '^
^/.' :: :: :: :: :: ::: :: :: : : :: : ^ ::z: _ :_ Jj
\ \ \ I ^
:: ^ \ V X
'^ : : : : : : i^y j y ^ ^ 1^ I
::::::::::: "wife'S? SI
^ ^ i K , - - . _ j , - 3j j
_ _
\ _\ V ^ S, L—
:__::_j::_f^ ^ .._%___s ^ ^
3ft* : : : : : : : : : : : : : : : : : _ : : _ _ S _ : : : : Z J:::::±|::::^::::::^:::::s±:^shir
"*' _: ;I !: i =: =: -:i "> .:. i« _—
_ < ;• J ^ ^ 9 5 : L.^*^.-_J
—71 _Ll ^ __2r^----». - - - - l - L
_.:.__L._.^^
y. ^ \
i^^.J_H. s ^..^^L.i£
h N"; V* sM
« ,-_.v
/ \'»
:\*.
s ^
>i:
0* '
4s;4-
11II M r r t 4 4 J I N I mTTi-iH=4dd-i-i-ti-ist=H^it-hH-ti-ti-ht-^rr 111111 TM 1111 r:
' ^ . . ,1, .,,, ., . ^,, .L .1
'" 5S., ~^^^..-. _ J
a,_ *,_ _!i„„ J
___:._ 4] !.-|+J
"" s . *• ^ J 2^ *" »-^ ^' > . -
. "^'t •
X1 vin_
! 1^ '
1 1 1 11 I rSU M 11 l > 4 . l M l 1 1 11 rWiP,--kJ 1 1 1^1 M 11 M 1 1 1 M . I l l
:>^ ^ i ^ .. T : _ n ^ "" "i ^ ^ 4 1 1
l i ^! i . . _ _ : : _ _! '_- ^>_* ,. _: _ _ > „ ,:^-^°^rt
. . , i . . . -1
_i "^ *•fc^ ^ ^ "
s
— -'Sk ^ ^
/ •• , 1
r
i: _
iii!E..::::_'=a_ ==a"/J 1 i.
ift* ::::::-S^:::_::_:!b,i^±_!:5.^a
*" :i Z
! 5i ~i i: i: i: i:^: :E : >s
: : : :: : : :I ' i ^ ^-"^L^J^
r i i^j
:: :_ : _! s: :" ! i :" : _: _ _ _ "^K n i - _ ^ i - ". * "* . j] "" •> I 111 1 1
. i K . ; : . ^ s, T. 1 III
::.:::.!^"_:_-:^i2j.:_j_:_!:d •- - ^ ;^ " I I I
" i z i i - i - ^ ^ i c . ' . : : : Z=«£<.: 1 ' \ ""< 1
: : : : : : _ : : - ! 5 S _ k r : i.^^., _-. 1 " > ^ i» 1 * 11-'
::_::__::_-:S4;.
::_::__: : _ ^ s ; ^ ^: . !>,,.
- " s , .. rJ
] '^ *•'* S "i" ^. ^"T**
ill ^ 1I1I1I
ft* 0*
: : : : : : : : : _10«
_ : : _ - : : L _20'
± : __z^_J30' 40' :o' 0' 60*
p

Fig. 64

O'/TB contou».r t c . syrrnnetric plane ^low ->annt's B = 40'

-165-^
 II111111 l i 111111 i MM M l .
-•^ Lt^^ttc^t"ttrfcrrrrr T t r t ^• L' ' 1 • ; ' L 1 • P 1 t 1 Mi
U- 'W:|\F:;"N,jfkL :[^
Lti:±t_LtH.pta,.LL_: LLL4 1 vt
, [±uxt-.tuututw--t.-rti
LL±±LLLLLLL±5LL^LL^ - •11 Itr
/• r h rrrri-rrrTrrL?
'^ LL_ILLLLLL 41 , M . L N LU. , |
IL T ±|L ^^i.-iJ^^-LL'-klLL'
l l L \ L LT r' r '"P' :• •; "k 1 1 ] 1 1 1 [ M l \\1 ['
LLI;_ZLL[:PLJ
11 k[ 1 \[ ^ N(L 1 1 1 M l M M
I t 1 L L11" L bil> Ui\] 1 1 t i i 1 t 1 11
/A.E^^-EE-E^'t^btE^Effir-ELbL±L: d f [j Y:.• L::rKL-ii-^KFh.^-fffi\\\\\
rrr"^rrrr"r'^nT'r"'rcrkrh M l ' ' M ^
LitrLLttL'EEticrrS" T ^. n 1
L'tCTuttkOT^rt-
tttil^'^ t r n t tl 'l^r;
p t z t t t t u 'Wf^1-f • Lr4
^bln
-Ml
.i1
t U
H ' tU
•}
[ill tF?4iT^[ii¥mttn[i!•
LMsLLiLLHrLi^mH
L|:L:bLj:|:"bLpbpL:t;_:.R^ f lL-:kM:fU:|.:Sj|]j^|:
h^b-^-p-F^^pff^^ajv^gr- L r ~ \ r p m
L . I f ,-8', FbKJ :.Khh::FSH 1
p_pa.,p.pp:ip.p-^Lp>pL^LLL_pp:,J^ t^pj byi.aL±.sL.._.tst-_tj.s ii|L
LLL L Li-L • 1 I LI Lf_L\LL —[-L.-PLh-LLL-iM 1 j
, t. .. L i ^kLi l i i^h-LL 1 LL, s
CCt't'LuiLps: L/ NLt [ M l
r r r r i ~ r i - ~ s ^Liirrr^k^sr^ | r [^Lktl^t^P-PLpf
LI INL
^.^<"^Li r r : r , r r . rrrFk.L'i,i$i PPPP^-PP^
L L L t j f k L l 1 t l i i [ r n L l 1 1 1 1 1 1["""1 '^LrLEr.if rE'i LLLS;LLL:bErS:i 1 ; :
E^t:EEL:TE-r Lii. "rrh^tl'
LLLUuLUL uLu^n'
Pi & -hht 'fSL" [• "hi WtFhFtBM
I Li l i S l j i i i l l I'M I 1 i r r i x 1 M t 1 1 M I " [ " W i 1 T i T T~" M ''
LLL^Lt&l-_LL...^^(s^L i I i , ub-Li/iaL i M 1L 'sL -1 M . M k 4444 l
LLLLLL^^iF • -1 hk^Ll .LLLr^ i I 'MX - - • M4-tt-+-* < j
tSLLLLtt^bkL ] r^kt'U T- ra
L ^ t ^ L L i k - u ^ K . i f L h"K,*,.- lit LM p ffr^ LM.lpLpptL 1
DTpr PKJ PR r^ki'j-L r^FnT rti
" • "rrFpf "priLi"; LrfjL i.Lpp._pw!., MM
c^LtiLLtip-'T, r k.r-,1- .•-^bLLlt
t*&LttLt:^bir^U.LK-L-&^[ r SiLl r I ***'LL I SL • ULS I 4 '
LLfjzsttF^rrh^t-Khurhkir LLrfe
LI. Li I uHiij- hkj iLji-L- -L ..fc^k . . . 11 p<LXp[:pL-Lpp"^^L[-pp-p-s^LpLL:]slj,
L^ LL_t:Liv«,,' ,_rm 1 ! . !kL. Li h^kL t i L1 ^F»Li 1 i 4—r M M '
tL:^±t:tp>Lj'o,-tr^sL t . hsit
LLLxLLLLrH^? r l , J ^Lfi ,. rNJ II II II [^L M i l l r'FiL-L i—M i M '
tLu-Ukt---Lr-^44- • - h hk^ 4 rSr k i [ \ I M p'Lr M M1 I...L.i,..i..,!r!tiiLl„t
'^sL "^cs M l M M
M », i
i»«r:b:~ba.::L:L-kL L,,., 4.r^,L ill 9'
0* 10' 20* 30* 40' 50* 60'
P

Fig. 65

o'/rB contours for synmetric plane-flow channels, 6 = 50*

•106-
 Fig. 66

a'/rB contours for symmetric plane-flow channels, 6 = 60*

-107-
 Fig. 67

O'/rB contours for syometric plane-flow channels, 5 = 70*

108
Dischargers and feeders

The outlet computed in the previous sections must be effective,

i.e. the solid must flow through the whole area of the outlet. A dis-

charger or a feeder should draw the solid across the whole area of the

outlet to make it fully effective. It is not necessary to provide a

constant velocity of the solid across the whole area of the outlet.

As long as all the solid above the outlet is in motion, the outlet is

effective.

The size of the outlet is determined by the flow criteria.

Frequently, the computed dimensions of the outlet are many times greater

than might be required either by the rate of flow or by the largest

particle size. When this occurs, the use of a discharger should be

considered to loosen up the solid in the region of the outlet and to

cause the solid to flow through a smaller outlet into a smaller feeder,

valve, or gate. Such a combination may be cheaper than a large feeder.

Dischargers. A discharger is a device which permits the stored

solid to flow out of a hopper but does not meter the rate of flow.

A vibrating hopper. Fig. 68, is a discharger with a circular

pan, a central outlet, and a baffle fixed to the pan above the outlet.

The pan is suspended from the hopper and is flexibly connected to the

hopper. Torsional oscillation is applied to the pan about its vertical

axis. This motion loosens up the solid above the pan and causes it to

flow around the baffle and into the outlet. Like all vibrating devices,

the vibrating hopper is applicable to solids whose instantaneous flow-

function satisfied the safety factor expressed by Eq. (24).

-109-
 An air-hopper. Fig. 69, is a discharger which uses air panels

to reduce the angle of friction 4>' at the hopper walls close to zero,

and to aerate the solid in a small region above the outlet. It will

be noticed from Figures 45 to 54 that for <t>' = 0 the troughs of low

values of the flow-factor occur for slope angles 9' • 40° to 50" for

conical hoppers, and 6' = 50" to 55" for symmetric plane hoppers.
P

The air-hopper works best at these slope angles. The whole area of

the air-hopper should be lined with air panels.

It should be stressed that the air-hopper does not rely on an

expanded bed for operation, the aerated region is limited to the vicinity

of the outlet and air flows out of the hopper through the outlet. In an

expanded bed, air flows upward through the bin. It also follows from the

figures that a reduction ot <^' to zero along steeper hopper walls, say,
6' - 12", or 6' » 20" would lead to tlow-tactor values of 2 to 4, hence,
c ' p

to a reduced flowability ot the hopper. Therefore, air panels should not

be placed on steep hopper walls. An air-hopper is particularly well suited

for discharging into an air conveyor.

Both the vibrating hopper and the air-hopper have the advantage of

requiring little headroom.

Other dischargers like stirrers, vibrating panels are effective in

smAll hoppers. In large bins it is advisable not to have any devices

inside the stored solid as they are likely to lead to major maintenance

problems. When a device is located in the central channel over the outxet,

its efficacy is limited because it cannot break up a piping solid.

Internal dischargers are undesirable in mass-flow hoppers because

they usually prevent mass-flow from developing.

-110-
 Pan

Oscillation

Flexible
connection

Fig. 68

Vibrating hopper

-111.
 Feeders. A feeder is a discharger which permits the regulation

of the rate of flow of the solid out of a bin.

A screw feeder of screw diameter B and length L substantially

greater than B should have a rate of flow of the solid increasing in

the direction of flow. This is attained either with a tapered screw,

Fig. 70, which widens the stream in the direction of flow, or by an

increasing-pitch screw. Fig. 71, which speeds up a stream of constant

cross-section in the direction of flow. A stepped up screw. Fig. 72,

with steps of 1/4 pitch and 1.5B length, will generally work as well as

a screw with a continuously increasing pitch. Two or more screws can be

placed side by side to widen the outlet.

For mass-flow hoppers, the clearance between the screw and the

trough should be small. In particular, for tapered screws, the trough

should also be tapered. Otherwise plug flow may develop.

A belt feeder can also be uised with a tapered outlet. However,

here the height of the flowing bed within the hopper is uncertain and,

therefore, the ratio of the widths of the outlet B^/B^, Fig. 73, should

be as large as possible and preferably greater than 2. It should be

born In mind that in a tapered outlet the smaller width of the taper B.

should be equal to the computed value of B so that doming does not occur

For greater uniformity of flow, the front of a hopper with a tapered

outlet should be vertical, as shown in the figure.

If the hopper outlet is raised above the belt so that a gap of

height a •• 1/4 to 1-1/2 Inches is formed. Fig. 74, the outlet becomes

fully effective without a taper. The gap is of constant height and the

solid flows through it into the side pockets as well as through the

-112-
 • Air
panel
50" to 55'
depending on 5
m" to 50°
depending on 6
(a) Conical (b) Slot

Fig. 69

Air hopper

-113-
Tapered t r o u g h

Plan
Z5KS

Fxg. 70

Taporcd sctew le-ader

miw^^^y
F.g. 71

Increaaing pilch screw feeder

Fig 72

Stepped pit-ch ocrew ffe-:jdBr

L14..
 TtZ
Fig. 73

Tapered-outlet belt feeder

PocKet

Fig. 7^

Raised-hopper belt feeder

-115-
front opening of the hopper. This is a very efficient design for solids

which do not contain material over 1 inch size because, obviously, the

size of the particles must be smaller than the height of the gap for

this design to work. If the gap is too large, flow becomes difficult

to control. Sealing skirts can be provided on the outside of the hopper.

The rate of feed is controlled by a variable speed drive. The front

gate should be adjusted to provide unltorm flow into the slot and the

support of the feeder should permit initial adjustment of the height of

the gap. Since the pressute on the sealing skirts is small, this

feeder, when properly supported, leads to feeder loads which are close

to the theoretical values (see below); the hoisepower and wear are also

at a minimum. A raised hopper can be placed directly over a belt con-

veyor, dispensing with the feeder and sealing skirts altogether. An

example is given in the section on Fine Ores.

A slide-belt, that Is a belt supported by a continuous, smooth

plate - either troughed or flat - is preferable from the standpoint of

flow to Idler support. It Is also cheaper and simpler.

An apron feeder is normally used for large size solids, e.g.

mine-run and primary crusher ores, and its dimensions are usually

governed by the largest rock size. Whenever the length of a rectangular

outlet over the feeder is greater than the width, this feeder does not

provide an effective outlet and mass-flow is unlikely to develop. The

outlet can be made effective by either tapering out the feeding area of

the outlet or by feeding across the side of the rectangle. The latter

arrangement, Fig. 75, has not been used so far - to the knowledge of

the writer - probably because it is more expensive and its advantages

-116-
-"^=3^
Fig. 75
Side .discharge apron feeder

-117-
are not known to industry.

A side-wise reciprocating pan feeder is shown in Fig. 76. This

feeder [39] seems to be particularly well suited to provide an active

outlet for large size solids like mine-run and primary crusher ore.

It differs from the orthodox reciprocating feeder in its direction of

oscillation. The orthodox feeder oscillates length-wise with respect

to the rectangular outlet and its stream does not increase in the

direction of flow. The side-wise reciprocating feeder is capable of

drawing along the full length of the outlet. If necessary, such

feeders can be placed side by side to activate an outlet of any desired

length. The feeders discharge onto a parallel belt conveyor as shown

in the figure.

A star feeder. Fig. 77, provides positively uniform draw along

the length of a slot-outlet. This feeder is particularly useful for

solids which require but a narrow outlet, say, B •• A to 8 Inches, to

assure flow. The feeder discharges into a chute or gathering conveyor

as shown in the figure. The space between the feeder and the conveyor

should never be full of solid.

A rotary table feeder. Fig. 78, has a tendency to feed above

the plow with the rest of the solid sliding on the table. In order to

make the outlet fully effective, it is necessary to provide a gap be-

tween the hopper outlet and the table, as shown in the figure. This

permits the solid to flow out of the hopper along the whole circum-

ference; the plow then need not enter the hopper. Outside skirt can

-118-
 Outline of pile

Gathering conveyors
Troughs

Jb A- J^ I
••Reciprocating Reciprocating
feeder feeders placed
side by side
•Gathering under trough
conveyor

Section A - A
Fig. 76

Side-wise reciprocating feeder

-119-
ba provided for sealing purposes. This feeder has much similarity to

the raised hopper belt feeder.

An air-slide feeder is shown in Fig. 79. The air In introduced

into the bottom chamber, flows through a porous diaphragm, aerates the

solid above the diaphragm and conveys the solid out of the hopper into

the air conveyor. The bottom chamber is divided into separata com-

partments of length equal to 2 to 3 times the width, and air In-flow into

each compartment is controlled Independently in order to assure that the

feeder is effective along its full length.

An air-slide feeder is particularly well suited to operate with

a slot-outlet air-hopper. In this combination, the solid flowing into

the feeder is already aerated and it is not necessary to divide the

bottom chamber.

-120-
 Star
feeder

Ji.

1^-7^s:
.
>. * \ f
A^
UL - r Y V
^Gathering
screw conveyor

Fig. 77

Star feeder

Sealing
skirt

<^_Plow

Sealing
/ skirt
.Table

Air in-flow

Fig. 78 Fig. 70

Rotary table feeder Air-slide feeder

--121-
 Summary of dischargers and feeders. The applicability of the above

feeders is summarized in the following table:

Type of Shape of Maximum

Discharger of Feeder outlet particle size

D 1 1 < 1" l"to8" > 8" 1

0
Vibrating hopper X X

1 Air-hopper X X X X

Single screw X X X X x^
Screw feeder
Multiple screw X X X X x^

Orthodox X X X x
Belt feeder Tapered outlet X X X X

Raised hopper X X X X

Length-wise X x2 X X
Apron feeder
Side-wise X X X

Length-wise X x2 X X X
Reciprocating
feeder Side-wise X X X X

Star feeder X X X X

Rotary table feeder X X X

Air-slide feeder X X X

Notes: Largest particle size must be less than the smallest pitch.

Not effective for L > B unless outlet is tapered.

-122-
 Feeder loads. The ideal vertical force Q which a flowing solid

exerts on a feeder can be computed from the formula

Q - qYB^ (26)

for circular outlets; and

Q - qYLB^, (27)

for rectangular outlets. Both these formulas are derived from

Eq. (A-53).

Contours of the parameter q are plotted In Figures 80 to 85 for

conical and plane mass-flow hoppers, for 6 « 30°, AO' and 50*. For

solids of 6 larger than 50°, the 6 - 50° charts can be used. For oblong

outlets, the plane flow formula gives conservative results.

As an example: assume a hopper with conical walls inclined at

an angle 6' - 12°, angles <^' - 28°, 6 - 4 4 ° , y - 80 pcf, and an outlet

of diameter B - 1.5 feet. It is safe to use the chart for the lower

value of 6 - A0°, Fig. 81; the point (12°,28°) specifies q - 0.20, and

Eq. (26) yields Q - 0.20 x 80 x 1.5^ - 5A lb.

And another example: a hopper with a rectangular outlet

1 ft X 3 ft., side hopper slopes 6' - 25°, angles (>' - 30°, 6-60°,

and Y • 110 pcf. From Fig. 85, q - 0.35 and from Eq. (27)

Q - 0.35 X 110 X 3 x 1^ - 116 lb.

One conclusion is apparent: these Ideal forces are small.

They have been computed on the assumption that the feeder is perfectly

efficient, and that no sliding of packed solid occurs. These forces

do not allow for the weight of the solid which would collect in the

-123-
pockets outside the hopper (see Fig. 7A), for the confining pressures

from skirts, and for sturting loads.

The starting loads in particular may be several times (a factor

of ten has been measured [50] as much as the running loads. The

reasons are as follows: the low pressure which a solid exerts on a

feeder during flow is due to the arching stress pattern which develops

within the flowing solid and which causes most of its weight to be

transferred directly to the hopper walls. This occurs whenever the

feeder is in motion and the solid is flowing. When a solid - especially

a coarse solid - is charged into an empty bin, while the feeder is at

rest, the Impact of the falling stream produces pressures which may be

many times those which obtain during flow. This situation is often

aggravated by the fact that, as the solid is charged into a bin, its

total weight increases, and the bin structure deflects. If the feeder

supports are rigid and the feeder does not deflect in unison with the

bin, the feeder may take up a considerable part of the weight of the

bin and of the stored solid. Of course, as soon as the feeder has

started and flow in the bin has developed, the vertical force drops,

but the initial force on the feeder may be very large.

The peak loads are particularly detrimental to belt feeders

which rely on pre-tenslon for traction. In order to assure traction

during the peak loads, the pre-tenslon must be high - much higher than

is required during normal flow. This high tension contributes to the

loss of horsepower and to the wear of the feeder.

These starting peaks can be eliminated by running the feeder,

however slowly, whenever the bin is being filled or, more positively,

-12A-
by providing flexible support to the feeder, for instance, a spring

support. Feeders or dischargers, which are suspended from the hopper,

lend themselves particularly well to such relief because springs can

be incorporated in the supporting rods. The reciprocating feeder and

the vibrating hopper provide examples. In floor-supported feeders,

it may be necessary to build springs into the support of the feeder.

The springs should be so designed that they do not deflect until a load

2Q has been reached. A total spring constant of 1/4 inch deflection

per load Q will be found satisfactory.

Laboratory tests [50] have shown that a very small deflection:

In the range of a few hundredths of an inch Is sufficient to relieve

the peak loads.

-125-
 Fig. 80

q contours for conical channels, 6 = 30"

126
 rtiHHHMHMHMHI Ir-:r:11
11 H 1 1 '"ti""*"' MM'"Mii'^tM'"
w
1 MM i l 1 M
n4j
JTH_ M IH ' ' 1
>LM
r>i > L
r>l M M Ml
U rsj 1 rt4 1 1 11 1I 11 i t
iKJ I N M kL M 1 j JI t1 11 1 11 1 . 1 1 I 1j 1 11 i I , 1j 1j
30* rKL M>1 i H l M l
>1 i H J 1 N 1 N M M M! i M
iN^ 1 kM 1
tf-L'U U-I44 I i ItI' H
U-ij U; l-R-i M111UHi
IVL M ^ M
V rhiTTOw^m TiHfflHH 1 j Hi
^' SL 1 N I H J I IN I N M
^ NJ
ITW
M>J MS4 ISJ MSJ
I K I N MIN I N 1 \A U-l-lli 1111 III J i l , 1 i l i
f 1 1 I " M M P4. I ' rv 1 1 1 rs. 1 111
1 iN.M1I M 1 1 M M 1 1 II 1 I t
20* M n>lMN M>1
M N M>JrR IKKJ 1i> k l rrt 1117 r 11 r f f l t T 11 r l fM1 Ifr
n>M i>l rN-JM>liM>l 1 1""^ 1 M 1 M M M 1 1 l l I I I I M i I1
1 1 HJ IrH 1^1 1 N MJ Spsj 1 MMM M1 1M1 M11M11
KJ M>L \\TK I N M>J ' TN 1 rv
1 >L n > l 1 R MH "^N 'rhS^'N
L iHJ I N J M>J M > r MSVK / • » ! |>s>'| PNL 1 M 1 1 I H
KJ M IN 1 >L 'r PR £^ i 1"N^i 1 \ HJ i'K MI I1 l iM1 M 1 1 Ml M i M 1 1
M>M J n>l iSJ 1 SJ T^i-1 1 1 1 1 ' ^ 1 1 1 >vi 1 rs. t i i 1 . 1 1Mi ', ' 1M111 ,'1 i 1 M
IN
M>1 MNM>L TVM 1^1 o'n4h ,M-T1 r-
10* L rHM>i M^Li'M>/'r MN'L-l i>j
I N M1 S^r^ 1 >J nSv 1 >j IVL
KJ M IN L i M i JiSJ M> M>1 M>J I N I iNu M M
b>L M iS^-' iVi?' TN M KWM MN N rN MN INMHJM^JTis^ M
KffKJ r> f^h ^ M M>1
LSI7*>1 c
>i>i\rN^>'
SJ
i>M
IN IN T'4 ' M>J M>J rSliRi'
LN T> 9r>Vk 1 K M>JriN 1 ^R1 riJ r rU 1 rU M>1 M>1 M
^b^is^^Fm^
«• 11 Tvr H 1S>L 1 Nvi M i l l i>LI 1 1'1f^ 1M
1 * uil% UTH J liiJ T u4itlsl'^U
I r s . M M Px 1 1 1 i V l 1 1 1 iN 1 IN1 I 1
0* 10" ^0 30* 40" 0' '.0"
c

Fig. 81

q contours lor conical chanr Is, 6 = 40"

-i2/-
 1 j H j 1 i-fH-M'M 11IHHI ll'M" 44 1 I'M M li M MM li 1 M l"H 1 \ \
40" N 1 H 1-444 1i M 1 I Il I 1 1 lI I 1 j I j 1_14
'" 44SJ44
4iJJJ4444JJ444JJ JJJJJ-liii J J M
J444444 J4114444441
4 4>J MJ J J J JJJ J J J J J J J L i i 1
, |J:JJJ21iJ_ t 1 MM
,,M MS N M
^ 444444-1344-114444J44
JJj^JJj^JJXMJJ JJMU-ii iul M i l M l M i i M M I t l M I M l M
R N i
H -J II II I I M It M I I M M M M M M
43444443444^144144444
30' 444344 4^.1444 11 1 11 t tf tt T1
•^ Jj44^4J44.5J Msl M-H-i-ll- I I I
SJ444S44444444NJ M J l 4 - t t t
4 5 , 4 4 4 ^ J 4 4 J 4 ^ 1 J ^ J .1 J T J I M II 1 1 111IIM
4 4 j i j 4 4 4 4 4 l l i J J1JII44 t l i
4444 iH • ' M i l 1'"N*!^! 1 1 IT 1 II II 1 1 II II M M M II II 1 1
11 1 M 1 1> M 1 iN^ bl 1 1 IN 1 IN 11 I I
4Jd444ll4444I44J i S j n J J t'
7n* Ml
• IN. > ilNf.I^^,.J > M
MIIHTJ IN N
INL
^" JJiji^JJJ-J-U'wJ-lH.,
il^H MSt 114 RITS
IN S'iHid'1~H HHM 1 H
34444444444.1 n J N l l r i J ]\j J i J -l 1 I l L I . I 1 J. 1 -1 -1 1
1> 1 M 1>J N M^IS.I M ^ t 1 1 1 1 I 1 1 1 1 I 1 1 1 1 1 1 1 1 1 1
H T 4- 114 MS
^J _^J4444^>^il.
M-J 114 J !>'! 1 MvJ 1 1tiJ 1
> 1 RiSJ
MN M iV 1 K i R M 1 1M
MH
N M' N
' r a j S l B j q j l l . i T S :JXR444C siIN
i n - N r n I I ^ I T M - T T I H M T I T H I I 1 :J 1 H3Hd']^d.H
1 11 V I I INj 11 1 1iiJidddii-LIl
II M 1 1 II 1 1 1 i
*•" Uj^JJJ^^ikljJJ M j J X l l J x U IHJJIJ N I I X I Ill-iUJiJ-U-l
4 M> M n 14 iW r>J r> J1 M x l m 1
14 J H4.H> I N M? r>j sj4 ! J11 rim
11sl 14JN4
11 1144 I
1 I N
IN I I 11 1 1 1 1II
r f l ^ i q N i ^ l l ^ q 1 444TI4I4 11 f N MN- 1 > MN
4Ja Jdj^X^N4
l4.9t^'-l^rl,TN 1 1 I^J UA4 11N JNJ
i l 1"1^4J ^ ^ IN
1 ITli
IN 1 MsJ PV
r IN'SH 1 1 IN 1 1 PKI M N 1 1 1 isld 1 Mx
.41N44-
1r
4-R44 11 N 14
1X MN P N
0«1,M n>?NMsl IN 1>MSI M Hn^l I>1 isl
M44 M I>J^n iMS
H n30" R H N IN
i i T T i iiiHdiiin'^iTli^n^rT
0" lO" 20" 40" 0' 50"

Fig. 82

q contours for conical channels, 5 = 50"

•128-
 M 1 M II1II }"H"'t4 M M M M M HHHHlHHHHlHH'M I : M : • *
M J 1 MJ 1 M M 1 1 i J 1 L- t--—

•*° 1Phi
TN
hf-RI rH4
r r R
TT
Ml rlJ M NJ MITTN
1 1 T+^l^l 1 1 V I 1 I I 11 1 M M i l N i 1 H MN J MM Mirkl M 11 M}M "'^
/. llTRri M>L M l ! 1 1 rsl 1 1 1 1 T r J 1 , 1 I 1 11 1 MS». 1 1 ! ! i
P JJflTl
L I uTHiMtkM
H^ ITT. Mr
rr r r r r r M i r i lAri M l >IjNT. r i 11 r r KITS.
i T r ^ H1rpik iM1 ^ i"rrH^i [-' 11'
1 T€j i N .•'". _i
PL M>' rVM i^A
rrrLL' tu LH H rH-L RUi-tiS r L M i ±U-i-
L L i U-l
Ti 1 : N^i
> L 1U UsJ4HU
T - P ^ * U f • i -NT.
xj
r r i r " ! [ nnr4*-M i 1 1 1 1 ! 1 1 1 i ! I nS^i/
P H '' M N ? ' ' ' ' RJK' M
ptprFFhrffRM rnrn'M^rT hl-Fi^RI f MHM+4 f; t4~H["W"hi~i"?'^"4 ^ • •»
LrrrM r r ^ H r n ^ R H T i M Ml tRttTO'FftitftH -R^^^^^ L
p k H 1 1 1 1 \\ r4-44 1 I I I rr-L-L 1 1 1
JJ M f-L ' M r\' ' '
TTl irW- M ' ^ R M i r M l M M I i l i "^j1 —

r T r T - 111 I I M ph-LJ Mi~

| | 1 1 14 T^U 1 I I I I 1 r r r - L l M M f i pf-EJ An rT r~t~4-J 1 i">^ M
I M I I I 11 ir-l M M I r 4 - 1 UM 1 M T H ^ l 11 I I I I M 11 M 1 l > ^ L 1 I I ! M ! 1 • ,
i r l T T I M i i M r r4--L i i 11 I I 1 i r•f4?L4- B h rr+4 L f>A M t-
rrr«yo
1 I 1 1 1 r r s j i M FT
M r-M4
14 11 Ti^kl^i M 1 1 -l T'
MV 't.
' "^i f
Hi rrrKfl
THJ 11 nrnT+ij
T TT+J
pM"^H^iJ]'''"^ i
M M T H J M M 11 I M n-f^.1 1
n«hh H M H 10'
HHH H HH
20'
H I 1 30* 40' 50' 9' 60'
L.H. ....

F i g . 83

q contours for symmetric plane-flow channels, 6 = 30'

•129-
 MM iH~HHH~i 1 i li"l 11 i M"11 \ f l"f R1 111 HRH H1H {i 1M U Ri l i t M tl ^
d O * I n J - H J ,ul M. J _ l 1 J
1
]n<\ RNJ HKiH M M H 1!
1^44 M i 1 M 1 114 i 11 l^^l M ' 11IN11 T R'Uv M it H R
J 1 ]s4 ] 1 INJ r1">
M i l l 1 IN 11 1 11 I I 11 1 1 1 1 rJ. 1 M 1 t 1> i ' Pjsi M l Tisj 1N. 1 M 1
/' 4 IN IN MN IN'J b ri"'l Ni t i l l
1 isl 1 iv r IN I N M l
j 1 i 11
1TR
1 isJ
n rJ.
Ms!
MsJ
TrsJ
[N^ ?•]
' M r
|N^JN. M i l
INN.
'J 111 M1 1 iJ 1 IIN t I N . R IN M M MNSU 1 1
vi» JiJJJ J J J I J J I J44 M iN,„ M i l N. 1 [ M M I N f j I t M 1 i MS L M j 1 [SSl i
1 ]sj J 1 134 H 4 M IJ Ij'JjJ—L i 1. 1
pjJ .M 4 • 1 M M M ' iJ M IS JM M J i J S 3 . 1 J ' U J iSJ-t v4 1- U '
M'^sj 1J1 1 TT" 1 N J l i M i i i M s l l n i t l
1 t 1 1 r4 1 I t 1 1 I I I I I I 1 M 1 INsJ 1 1 1 1 1 1 MiNj I N R 1 RRJ' 1
111' 1 IN. 1 " M l IN' T
M ?^^<L *^ M ll>JIN^'J 1 11 M11nJTiT M INI MMM H1"^ i '
4 J l T l ] J N 4 ,
f 1 f M 1 1 [N^
l S n r
1 Ms >
1NJ
R n>H
r Pn-H RIJN
?n' J..,j..J,J.J.J.J.J.1J.. 1 H^L 1 I.MMI. 11 N
N
1
^" 3344^444^4444444 n ^ j l 4 4 l l 4 4 ' ]^J R U Ml.t ilS^J.M Mn-L
1 is-J M•
M44J4'
nS M M MM ^TN M M^N'
1 1 Ti'^ M 1 ^f~>^
MM MN_ i t iKl>l 1 M M I ' c v i U l U U l i l U I LL-ISN 11 M i
" I I I 1 J4 1 IS M M 1 lis: 11 It 1 i N M l l t l l i l M I rrT 1 iSvL 11
rt 1 M 1 IN.- M iN' i TN
1 IN M M M l 1 T^I M Il ' ' 1
M J i 1 M r 11 1 iN. ivi
1 S>J M U M T I N I^^1 M
4 MS«J M U M 1 IN J nt I t 11 I I1t 1 MiSJ
ITS I I I I 1 1 1 1M1 I M 1Mt 1
10" M^JJ-.-WJ. 1 1 INJ.. iJJ. > 1 4T 4 1 1 4 NJ 11 I I T T I T M I I
li isJ N4 1 n-L
D4:j43J"444g4HUl44-l44l M ^ J 4 vJI
I I 11 1T^ lUI-l-LII
I 1 M I I 1 IM
I M NJ n I4k.
R1- I 1I 1 1-H
1 1 IAI H
M 1
>L n4 rn-i I N J I J 11 M M l<i 11 11 11 1 1 11 1 IN4 I I I I 1 M M
PNJ N J 1 iHJ n IN- 4 n^kU MNJ
n^J fN^l 1 1 1 N J 1 1 1 N-J Ni M 11 I I IS-L I 1 T 1nNi4. J
M I N J l>L I n J 4 1 H4 T 6
MNJIR 41NI1JJ > T
TRM isJ 1 \nA
M IN ?H-t • U N-3^ nH4 1 MINJ I N
MiN ' H-n^ 1NJ
n" n7li1111liH:^n4snTliH-R44 iT<M
n-L 1 n-IJ
R1S-L M
10' 20' 30' 40' 50" e' 60"
P

Fig. 84

q contours for symmetric plane-flow channels, 6 = 40"

-130-
 4--- — "j "RiR RRU R l i'^Ri—1 H-t-j 11 \ '•••'-
1 , M
:"::"::::"=^: ^ v^"'^"":":"""^:
_:-"5,., _u--:5s, - ^ ^ v =^:"":"'sr -" ~~
- i t i b 4- __ i t ~ ±i :t" ~ :
""-^ s.
>-^>. v. ^^ V.
"^^ N ^ . ^s ^S S
^- ^- ""J ^ ^ ^
s., J I>^ -V >. S
S V. ^^ S 'v \
^. V ^. V. '^N, \
^ . ^N^ ^^ ^'4 ''s, S \ '
s ^__ ^s sj r- >., , ^s N 1 .„ 1,
A ^v,' P s^ . ^'-v,
X . '-^
'^'N
>.^ V 'NV--, - "s. A
: _ ^ ^ . . JT ^r^ ' ^s ^ ^ ^ ? ; i S;r''s: : _
^*- - T - ^* ^ I-R.4: ^S ^ - ^ ^ S - ^ * -Sr-
">. 'v. ""s ^s , v?o ^, V.
•s -s >s S?.., N N \
iti ^^
s. u ^tq-t--
^> -+-\ ^v^.
-^ s<s
s.°l R t ^ N
i \ N S^\ - -
5s 4 I - %T ^ ^N1 is^b" ^v, !!.^ ^S S ^ T
:^v±: - i."it '_4_ ^ j i ^ s ° _ '_^v : s ^ :Sr_s,_-
i45^
: .Ml...:..d_'s'^^ MT^if^^r
R ^^ H
"^s^!.^ '"'^v i;^
-S, ^ ^' ^ v is:^*4M
"^4^
M'- "^ ' ^- "^ ^ -J ^
_::b^4:
^^n
^5."^. ^5.
^
1 1 •'• ^ ^^ _i r"^, ^-
t T ^ sS \ S I -
t "i^:
^^> s~4i
^. T'" ^Ji>r
'^ r . , "IxT "^^.
^ s— ^N
i^s silt -^-^^ s
N ^ 4 '^kp,^ V Ss \
^s^ q ^ ^ -.
^ Tr-^ . - . 1-Ss;- RP ^ ""..
-> ^-u
S ^ -T1+
I^c- _ .. ^1. + Mi. -^ ^^ \ 4-5s-4_ U
^^ - ^ _ s^ --• ^>^- -hs^ Ns; - + - * v , -
_^vi^. _ ^ ^ ^ ^ ^ "• ' '^.^^ ^s. '5":::'=::
ii '^^ ? i-iN " ' R - • it"' V — s ""Ziii
,^ s l±t _
_ _ ..^vj^
s . i / 4n i^^^u l ,• -R": ' • i t"^^^
lu^^ ' it i ' ?- ^^ , _I I:I^S :I _ _
^s '->:^ ^ '^. ''v >>
"^^ -^-i: I V ^s
^. N S 1
. , , 1 ^.... ,
S;
+- -f ^
-t-4 N - - - h - ^ - ^ h -^ - + T ^^, ^
'^^ "sLs f- ^L R i _+. "^^ \
^ N^? + r - f.^-"- - - H 1 - - --^ - t ^ ^ -+- - ^ d -
^v Vv^o' ^-> ' s ^
^^5^L, S N -.• ' ^ \ - M — . —+ ^- :i>^ - - -
T- "^^.^ '^^.M t + - ^r -f- -+- ^ -h
"^s' N ' '\- ^v
-H s s, H . ,-4 |/t. -.|_4.4_.^ ^ _+_ -h-t-f-
-4 --. s j - . t -4i>^-4- 'U , i-4 - . . - ^vJ ,4.4,4-1
>i "^v ^i/, • N
10' fk =^s
20' • 30'
---^^ 1+ 40'
-+- 50"
-^s-rtt
e'
T
60"
p

Fig. 85

q contours for syametrlc plane-flow channels, 6 = 50"

-131-
Promotion of flow

Impact pressures. The pressures which consolidate a solid as

It flows In a hopper are roughtly proportional to the width of the

hopper and are practically independent of the head of the solid in the

bin. That means that the consolidating pressures In the region of the

outlet of a hopper are quite low during flow.

On the other hand, when a solid is charged Into a bin, the

pressures at the point of impact Increase with the height of fall, and

in a tall bin may be much higher than the consolidating pressures of

flow. The impact pressures may be sufficient to cause doming. This

condition is common in the storage of solids which contain both coarse

and fine particles, because the large particles produce high Impact

pressure while the fines develop the strength to sustain the dome.

When a solid is carried into a bin by a high-velocity air

stream, there is also the likelihood of high Impact pressure with

possible doming.

Impact pressures can usually be kept below a dangerous level

by: (a) Maintaining a minimum level of solid in the hopper, so that

the sensitive part of the hopper is not exposed to impact, (b) In-

creasing the impact area. A baffle or impact breaker placed in the

stream spreads the particles over a large area of the bin.

(c) Directing the stream at the wall of the hopper and not at the

hopper outlet, (d) Expanding the air stream of an air conveyor to

reduce Its velocity prior to discharge into the bin.

132-
 Circulation. There are many solids which flow as long as the

particles are in relative motion with respect to each other but cake

when left in storage at rest. Some freshly produced solids behave like

this for a period of hours until the surface conditions of the particles

have reached equilibrium.

In those cases, circulation of the solid around the bin can be

considered. Circulation causes the particles to move with respect to

one another and break up the bonds before they can gain full strength.

Evidently, circulation is effective only in mass-flow bins. There is

usually no need to keep the solid in motion all the time; 10 seconds of

flow every, say, 30 minutes may be sufficient.

Solids conditioning. Good mass-flow hoppers have a flow-factor

of 1.1 to 1.6 and can accommodate a vast majority of the solids handled

by Industry. However, there are solids whose instantaneous flow-

functions are high and which require Impractically large outlets for

gravity flow. The choice then lies only between a non-gravity form of

recovery from storage, and solids conditioning.

Solids conditioning means changing some property of a solid in

order to improve its flowability.

It is, of course, much easier to specify the proper conditions

of a solid during the design stage and to design the process accordingly,

than it is to alter the conditions in a completed plant. Flowability

tests carried out at the design stage will point out such factors as the

maximum acceptable moisture content and, hence, the required filter or

dryer capacity; the permissible range of temperatures and, hence, the

needed cooling capacity; the preferred crystalline structure of the

-133-
solid; the need to screen out the coarse particles from the fines.

The two most Important parameters are moisture content and

temperature. As the moisture content of a solid is increased, its

flowablllty decreases reaching a minimum at 85 to 95 percent of

saturation. If a saturated solid is stored in a bin for any length

of tine, water drains out, and the solid tends to reach the moisture

content of minimum flowablllty. It is always desirable to store

solids at a moisture below 85 percent of saturation. If a solid Is

allowed to lose moisture through evaporation, while in storage at

rest, the particles may become cemented by the salts precipitated

out of solution, causing a marked decrease in flowablllty.

Temperature has a significant effect on the flowablllty of

many solids. As temperature Increases, the particles tend to develop

greater adhesion and the solid becomes less free-flowing. Some solids

will flow at a certain temperature, provided it is kept constant, but

gain strength and lose flowablllty if the temperature Is allowed to

drop during storage at rest.

Flow promoting devices. For mass-flow bins, the recommended

device Is an externally mounted vibrator to start flow after consoli-

dation at rest. However, the safety margin prescribed by Eq. (2A)

should be satisfied. It is sound practice to have mounting pads for

vibrators on all hoppers likely to carry solids which cake during

consolidation at rest. It is then easy to mount vibrators, should

that become necessary. In a mass-flow bin the energy from a vibrator

is applied right at the abutments of the doming solid and, therefore.

-134-
the vibrator la very effective.

Infunnel-flow bins designed for no piping, devices are used

for the final clearance of the solid from the hopper. Fig. 32(a).

Either vibrators or air panels can be used. A simple air panel can

be formed from filter cloth attached to the hopper walls with air

Inlets spaced under the cloth.

There Is, of course, a large number of other flow-promoting

devices such as: Inflatable panels, Internal vibrating panels and

frames, arch breakers, flexible rubber cones, etc. These devices are

not required In bins designed for flow, though they find numerous

applications in existing Installations which were not so designed.

When an existing bin does not perform adequately, the operating

engineer often finds it easier to add a flow-promoting device -

sometimes several of them - than to replace the existing hopper with

a new one. The new hopper, as likely as not, would have to be of the

mass-flow type and would provide less theoretical capacity. This is

looked upon as wasteful of plant space. However, an analysis of the

gains In product uniformity, quality, reliability, savings in mainten-

ance, and total availability of the stored solid will usually show

that a hopper redesign provides the more economical and certain

solution.

Small hoppers. In small hoppers, used to feed small quantities

of reagents and additives, the feeders have to be small and it is often

not practical to use gravity alone to cause flow into and through a

feeder. Here, dischargers and flow-promoting devices are Indeed mcst

-133-
useful. Vibration of the whole hopper asssmbly, tilting hoppers,

vibrating panels, aeration are used to fluldize the solid and cause

It to flow.

Segregation and blending

During the storage and handling operations, a bulk solid tends

to segregate (separate) according to the particle size, shape and

density. Since segregation is usually detrlmeiital In the proeasaing

and nercluindislng o£ a solid, the conditions leading to segregation,

neans bf minlnizing segregation, and methods of remixing during flow

will nov be discussed.

Host of tha segregation takes plao^ in the following two

situations: (1) In a freely falling stream which has a horizontal

component of velocity.. In free fall, each particle is acted upon by

two forces: its weight, and air resistance to motion. The former Is

equal to the product of the volume and density of the particle, and

acta vertically down; the latter is & function of the size and shape

of the particle and of its velocity, ana acts In the direction

opposite to the velocity. In vertical free fall, both these forces

are vertical and, while the terminal velocities of the various parti-

cles may vary, their tra.iectoriesi do not, hence there is no significant

segregation. Winen a solid is dj:scharged over a conveyor pulley or

spouted through an inclined chute, an Initial horizontal component of

velocity is Introduced, the two forcee: are not aligned any longer and,

as their ratios differ for the various particles, so do the trajectories

of the particlesi which then segregate.

-136-
 (2) When a stream hits a sloping surface. Whenever a stream

hits a sloping surface, the particles roll down the slope: the large,

heavy, and more nearly spherical particles tend to roll farther than

the fine, light, and flaky particles which tend to remain at the

point of impact of the stream. As a pile of the solid builds up, the

fines congregate in a column which has the shape of the trajectory;

the size of the particles gradually Increases with the distance from

the column toward the periphery of the pile or the walls of the con-

tainer. If the outlet of a pile or of a plug-flow bin is located

directly beneath the trajectory, the draw will be heavily segregated:

an excess of fines will be drawn every time the feeder is started

after charging the solid while the feeder was at rest. When the

container has several outlets, the outlet located under the trajectory

will consistently draw an excess of fines. This phenomenon can be

exploited when some size separation is desired, otherwise the outlets

should be located away from the trajectory.

The aforesaid indicates that segregation is a dynamic effect

and occurs inevitably when the solid is in free fall. It also indi-

cates methods of minimizing segregation, which are: reducing the

velocity of the particles by shortening the height of free fall,

reducing the horizontal component of velocity, and limiting the area

of the slopes over which the particles roll. The last method calls

either for narrow bins or for means to spread the stream over the top

surface of the container. Spreading can be accomplished either by

placing a shelf in the stream or by moving the point of discharge. A

combination of the two can be very effective in minimizing segregation.

-137-
 When segregation in storage is to be avoided, a mass-flow bin

with a high vertical portion should be used. Fig. 86, and the range of

operation should be as shown in the figure. This causes the solid to

flow like a rigid plug in the vertical portion of the bin. A flrst-in,

first-out flow pattern is enforced and, while the solid segregates in

charging, it automatically remixes within the hopper.

In slot-outlet bins charged by means of a tripper-conveyor,

segregation is much smaller if the slots are perpendicular to the

conveyor. Fig. 87(a), than if they are parallel, (b), because In the

former case the slot draws a cross-section of the segregated solid

which is remixed in the feeder.

In a multiple-outlet bin, segregation is reduced by drawing

simultaneously through all the outlets and blending the draw.

A series of bins can be used to even out the grade of a run-

of-mine ore. Fig. 88, by charging the bins in sequence and drawing

them simultaneously.

A bin can be used for the blending of a solid by recirculating

the solid in a closed circuit about the bin. Fig. 89, provided the

following conditions are satisfied: (a) The solid is non-segregating,

e.g. it is composed of uniform size, shape and density particles, or of

very small size particles; (b) a mass flow bin is used, so that there

are no dead regions of the solid within the bin; (c) there is a sub-

stantial velocity gradient in every horizontal cross-section of the

channel, so that flrst-in, first-out flow does not occur.

The points (b) and (c) imply a steep, smooth bin with a short

vertical portion, whose height does not exceed one-half the diameter

-138-
 range of
operation

Fig,. 66

Non-segregatxng bin

Charging conveyor Charging conveyor

Outlets Outlets

(a) Perpendicular layout (b) Parallel layout

Fig. 87

Layout of slot-outlet bins

•139-
of the cylinder or no vertical portion at all. A higher vertical portion

would Induce flrst-in, flrst-out flow, like the non-segregating bin. An

Insufficiently steep hopper would cause dead-regions to develop at the

transition.

Flooding

Fine solids which aerate readily should be stored in mass-flow

bins In order to prevent flooding. In a funnel-flow bin, flow is non-

•teady, momentary voids form within the flow channel. As the domes over

the void fail, the solid falls, and aerates. Unless the feeder is

completely air-tight, flooding occurs.

Gas counterflow

A contact bed is assumed. In channels which serve ae chemical

reactors, the solid flows down by gravity, while a gas flows upward

under the action of a gas pressure gradient. This gradient opposes the

forces of gravity and buoys the solid. In a converging channel, the gas

pressure gradient Is greatest in the narrowest part of the channel, i.e.

at the outlet. The maximum pressure gradient which can be applied with-

out disturbing the gravity flow of the solid is about 70% of the bulk

weight of the solid, y.

It Is important that the gas pressure at the outlet be maintained

at a uniform level without sudden changes, either up or down. When the

process is steady, the consolidating pressures In the solid are reduced

by the gas pressure gradient in the same ratio as the pressures of

failure, and gravity flow is not affected. If gas pressure fluctuates,

the solid may consolidate under the pressures of gravity unrelieved by

-140-
Charge in sequence
)que] _

r k. i.
Draw simultaneously

Fig. 88

Blending of mine-run ores in bins

Fig. 89

Blending by recirculation

-li^l..
the gas pressure gradient, and dome when the pressures of failure are

decreased as the gas gradient becomes high. If the process is inter-

mittent, then, during the stoppages of flow of the gas, the pressure

gradient vanishes and the solid consolidates under the full pressures

of gravity. If the gas gradient is then applied prior to the re-

establishment of gravity flow, the gas gradient supports the solid,

reduces the pressures of failure, and flow may not start. It is better

to start gravity flow first, and then gradually apply the gas pressure.

Since the gravity flow of a solid in a vertical channel tends to

be unsteady and erratic, narrow vertical channels should not be used as

reactors. Vertical spouts connecting sections of a reactor, as shown

in Fig. 90, should have by-pass pipes for the gas, otherwise doming in

the spouts may occur.

Only mass-flow channels should be used in conjunction with gas

counterflow. Punnel-fiow channels lead to piping and uncontrolled flow.

Bin failures

Failures of bin structures can be divided into two categories.

In the first category belong those failures which are due to recognized

structural causes, such as: foundation failure, uneven settlement of

footings of a multi-column bin, wind loads, earthquakes. In the second

category are the failures caused by irregularities in the flow pattern

of the stored solid. The latter are of particular interest in this

work.

The most dangerous failures occur when a stored solid domes at

or above the transition between the vertical portion of a bin and the

-142-
Gas
i Solid

, A A ^ ^ ^ ^ 7v \ A
£^
:!i^ V V >^ V V V >^ V ^

Solid
•I Gas

n A

V V
\ A

V
A

V
A

V
A

V
A A

V
A

V
' ^

V \

Gas
tl ^
I T Solid
Fig. 90

Gas by-passes of spouts

-1^+3-
hopper, while the hopper empties out. A large void then forms under

the doming solid. If the whole mass happens to break away simultaneously

it acts like a piston. Impacts the hopper, compresses the air in the

hopper, and either splits the bin open or tears the hopper off its

suspension. Solids whose time flow-functions approach a straight line,

are most likely to dome at or above the transition. Bins for such

solids (many agricultural products belong here) should be built with

rounded, smooth transitions and with rough vertical walls to reduce the

consolidation at the transition.

While a break-away of a large doming mass develops a high

pressure In the hopper, it also develops a vacuum at the top of an en-

closed bin and, unless the top is vented, the roof of the bin may be

sucked in. A vent may be closed by a diaphragm to keep the dust in;

the diaphragm being designed to fail if the pressure differential

exceeds a safe value. Such vents are also recommended to safeguard

against minor dust explosions [51].

Less dangerous failures occur typically in asymmetrically loaded

bins of circular horizontal cross-section. In a circular cross-section,

the hoop stresses are tensile and the walls require no bending rein-

forcement, provided the stresses acting on the walls from the solid are

uniform all-around. This condition is satisfied if the solid is charged

and drawn at the center of the bin. Therefore, the hopper, if any, must

be circular and symmetric with respect to the vertical axis: only one

outlet is permitted. If the hopper is offset to one side. Fig. 40, the

pressures from the solid on the vertical side are decreased and the hoop

stresses in the hopper may pull that side in. Bending reinforcement is

-144-
 Fig. 91

Uneven draw in a cylindrical bin with

two outlets

-li^5-
required to retain the shape of this hopper.

In a cylindrical, flat-bottom bin with two (or more) outlets,

Fig. 91, the solid empties out of two craters centered at the outlets

and the pressures on the walls at points A fall to zero while,

simultaneously, at B the solid presses against the walls. As a result,

the walls at A are pulled in. To counteract this effect, the cylinder

should be reinforced by rings between the A and B levels.

Ore storage

Coarse ore (-8 inch). Flat bottom storage. Fig. 31, is generally

preferred for coarse ore because of the possibility of wedging of the

solid in a steep hopper under the Impact of the falling rock, and also

because of the wear of hopper walls. However, if a layer of ore is

always retained over the feeder to cushion the impact and reduce con-

solidation, a short mass-flow hopper can be used to advantage in

expanding the channel beyond the dimension D required for no-piping.

The design of such a hopper is described in Example 8. The layout shown

in Fig. 76 is well suited for the storage of cohesive coarse ore, but

this layout is expensive. It may be cheaper to screen out the -1 inch

fraction. The coarse fraction then becomes free-flowing and will flow

through outlets just sufficiently large to prevent interlocking, say,

30 X 30 inches, while the fine fraction can be handled as a fine ore.

Fine ore (-1 inch). Fine ore is usually stored as a mill feed,

and size segregation has to be minimized. To achieve almost perfect

non-segregation, the bin shown in Fig. 86 should be used. However, the

layout of Fig. 87(a) is usually satisfactory provided the outlet is

-146-
fully effective. This can be accomplished quite economically by using

raised-hopper feeders in troughed belts without skirts. Fig. 92.

Tests of fine ores frequently show that a width B of 6 to 12

Inches is suCflclcnL to assure Clow. Slot lengtlis ol 12 to 20 fort

can then be used on belts 18 to 42 inches. The long slots can be

broKen at Intervals by structural members tying the sides of the out-

let.

Charging
^ conveyor

Fig. 92

Fire ore bin

-147-
Examples of design for flow

1. Provide storage capacity for a dry sand. The bin will be

emptied out completely at regular intervals. The sand is free-flowing

(FF and FF > 10) for F > 0. Segregation is not a problem because the

range of particle size is narrow. The material of the hopper should

be black steel. The angle of friction <|)' • 3 0 * to 32' on black steel.

The funnel-flow bin. Fig. 93, is selected with a hopper of slope

8' - 45*. A flatter hopper, say, 9' - 6 0 ' can be used with vibrators

for final clearance. The outlet dimensions are not determined from

this analysis; they should be sufficient to assure the required rate

of discharge.

2. Design a square storage bin for a granulated solid. The

bin is charged at Intervals during five days of the week and drawn

continuously 24 hours a day, 7 days a week. The solid contains about

3% dust, which should not be allowed to segregate out. The granulated

solid is free-flowing for F > 0, but the dust is not. The angle of

friction ^' - 20" on epoxy-coated steel. The effective angle of

friction 6 is between 40* and 50".

The mass-flow bin, Fig. 94, is selected. Mass-flow is required

to minimize segregation. The valley slope angle is determined from

Figures 46 and 47 at 23", which gives a wall slope angle

tan 6' - ^f"_l^', or 9' - 17"

V2

As in Example 1, the outlet dimensions are determined by the required

rate of discharge.

-148-
 Fig. Sfi
Bumelofloir bin for ^ey

Fig. 9*»* Fig. 95

Kftss-flow \&.n for m tTtm^^ llBSft-flotf bin for a free-

fIo«dng grmimlAt«d solid flowing granulaced solid
for continuous dr»w for continuous c»arge and
a S-day a week draw
-1*>9*
 3. Design a square bin for the granulated solid of Example 2

to supply a bagging machine. The bin will be charged continuously

and drawn only during the day shift, five days a week.

As the solid is charged during the weekend without being drawn

out, some dust will percolate through the granules to fill the voids

in the region of the hopper, and the solid may cease to be free-

flowing. If the hopper is made of steel, the type shown in Fig. 94

can be used with vibrators mounted on the hopper to initiate flow on

Monday morning. If the hopper is made of concrete, it should be

built without valleys in order to have an adequately low flow-factor.

A square-to-round transition is recommended as shown in Fig. 95. The

flow-factor of this transition hopper in its lower region is likely

to be equivalent to that of a conical hopper. Hence, the slope angle

of the hopper walls needs to be 9' » 23".

If headroom is at a premium, a discharger can be used, as shown

in Fig. 68. The rate of flow out of the discharger should not exceed

the demand of the bagging machine by too large a margin as this may

cause disintegration of the granules.

4. Design a mass-flow bin for the solid whose 6, FF, FF^ and

WYL on the selected wall material are given in Fig. 96. The bulk

density y " 50 pcf.

There are several possibilities:

(a) Unassisted gravity flow. This requires that the hopper

outlet be designed for the point of intersection of FF with the

hopper flow-factor. Assume ff - 1.3 as shown in Fig. 96. At the

point of intersection, V, - 4.7, and for this value 5 - 5 0 ° . The

-150-
F & V

4^7,0)

Fig 96

Flow properties. Example 4<.a)

l5i-
corresponding EYL Is drawn over the WYL and a Mohr semi-circle Is drawn

through point (4.7,0) tangentially to the EYL. •' - 24° is now found

as the angle between axis V and the line connecting the origin with the

appropriate point of intersection between the semi-circle and the W Y L ,

a s shown in the figure.

For these values of ^ and ^', the largest acceptable values of

B' are selected from Figures 47 and 3 2 , for conical and plane-flow

channels, respectively. The corresponding values of ff are also read

from these figures. H(d') is obtained from Fig. 43, and B is computed

from Eq. ( 2 3 ) . The values are tabulated:

Channel 5 •' 9' ff H(e') B

^1
Conical 17» 1.28 2.27 2.16
50« 24 • 3.65
I Plane-flow 28* 1.20

For conical flow, the obtained ff-value of 1.28 is slightly on

the safe side from the assumed 1.3 and no further analysis is necessary.

For plane flow, ff " 1 . 2 differs by a larger margin and a new trial is

made: ff > 1.2 is drawn in Fig. 96, a new point of Intersection is

found, and the corresponding values of ^ and ^' are again obtained. The

corresponding value of ff is interpolated between ff •= 1.19 for ^ • 5 0 ° ,

Pig. 5 2 , and ff - 1.09 for 6 - 6 0 ° , Fig. 53. These values as well as

H(e*) are tabulated. B Is computed and also tabulated:

Channel 6 *' 6' ff H(e') B

\
Plane-flow 52° 26° 3.3 25° 1.17 1.13 .97

-152.-
For the plane-flow outlet, the length L should be at least 3 feet.

In a transition hopper, the end slopes are selected at 6* > 17*.

Possible bins are shown In Fig. 37. The diameter, or width,

and height of the bins are determined from the required capacity.

(b) A vibrator Is to be used to start flow after consolidation

at rest. 6, FF, FF , WYL, and ff - 1.2 are redrawn in Fig. 97. In

accordance with Eq. (24), a value V. • 2.25 is selected for which

V. " 1.9 " 1.5F, where F - 1.25. Proceeding as before, 6 and ^' are

obtained, and 6' and ff are Interpolated for £ - 55° between

Figures 47 and 48 for the conical hopper, and between Figures 52 and

53 for the plane-flow hopper. The following table shows the results.

Channel 6 •' \
e' ff H(e') B .

Conical 12° 1.21 2.20 1.09

55* 28° 1.9
Plane-flow 23° 1.14 1.12 .55

In view of the safety factor in Eq. (24), there is no point in refining

this analysis further. The possible bins are the same as computed at

(a), except for smaller outlets.

It should be noted that, as the outlet dimensions are reduced,

the analysis calls for a steeper hopper. This is due to the convex-

upward shape of the WYL and hence an increasing angle of friction ^'

as the forces V. and V. decrease. When $' is constant, the slope

angle 6' Is independent of pressure and, therefore, of the width of

the hopper. In the example, the optimum hopper would have a slope

angle 6' Increasing upward from the outlet as shown in Fig. 98(a).

While such a hopper is, generally. Impractical, a hopper made up of

-153-
 70° 1
60* .
8 50* pr. ,._ —
"* ' !•
1 —
ho* 1

?0* /
^
-^

^ <f
^
F & VI ^ f ,,,

*y

v^ » 1,9 y
r^
\^

/
7
^
/
r - 1.25 TCFF
X
, /
1 4

i \ Cirli.
^ v., 5
r

1
—
2 —

f L / 8 - 55'

.
/ "
>
u \
\
t

t ^

^ 1

L 2 3 i^ V 5
F.g. 97
Flow propertia:.. Example ^(b)

iS^
 (b)

F i g . 98

"Hyperbolic" hopper

155-
sections, Fig. 98(b) may save some headroom, especially where the

curvature in the WYL is substantial. In fact, hoppers of that type

have been recommended by some investigators [28,52,53,54], who arrived

at the same conclusion by different and several lines of thought, which

this writer does not follow.

5. Some solids will discharge out of the hopper but dome at the

transition from the hopper to the vertical part. This case is Important

because it may lead to a structural failure of the bin when the doming

mass collapses.

Consider the solid whose 6, FF, FF and WYL are shown in Fig. 99,

Y • 20 pcf; and design a circular mass-flow bin with a vibrating hopper

(discharger). Fig. 100. Assume ff - 1.3. In accordance with Eq. (24),

V. is found equal to 3.3. For this value, V. - 2.5 - 1.5F, where

F - 1.65. This determines « - 50" and <J»' - 23". From Fig. 47, there

is 9' > 18" and ff • 1.29 — close enough to the assumed value.

From Fig. 43, H(9') - 2.29, B is then computed from Eq. (23). The

values are tabulated:

Channel 6 *• 9' ff H(9') B

^1 c
Conical 50" 23" 2.5 18" 1.29 2.29 3.7

A 4-foot diameter vibrating hopper is adequate to start flow.

Consider now the conditions in the vertical part of the bin

where 9' - 0, and H(9') - 2. From Eq. (22), with B - D - 12, there

is Vj^ - 12 X 20/13 X 2 - 9.2. Assume ff - 2.0. This determines

Vj^ - 18.4, 6 - 40" and •' - 15". From Fig. 46 for 9^ - o, •' - 15",

ff checks out at about 2.0. It is evident that the FF^ lies above

-156-
 70"
60"
50"
40"

30"

14

12
F & V^
10

10

Fig. 99

Flow properties. Example 5

157-
the ff In the vertical part of the bin, and that doming will occur.

To correct this situation it is necessary to roughen the vertical

wall so as to Increase its angle of friction ^* toward, say, 30", and

bring the point (6',^') into the trough of low ff-values, Fig. 46. It

is also advisable to fill In the corner between the vertical part and

the hopper.

These conditions prevail whenever the critical FF < 2 for

F • A D Y / 2 , where D Is the diameter of the cylindrical bin, or for

F *• ABY, where B Is the width of the vertical portion of a rectangular

bin.

6. An Iron concentrate Is to be stored. The question is; what

should be the moisture content of the concentrate?

Assume a mass-flow, slot-outlet hopper. A function will be

developed between the moisture content and the width of the slot needed

for gravity flow. The price of the feeders can then be balanced

against that of the filters needed to extract the water for a minimum

overall cost.

Iron concentrate at high moisture develops full strength

Immediately upon consolidation: FF *• FF. Hence the time effect need

not be considered. Vibration Is not reconnended for these materials.

Three sets of 6, FF, and WYL curves ere drawn In Fig. 101,

Y " 150 pcf. The sets are marked by lettere (a), (b), and (c),

respectively. The ff-values are estimated at 1.2, 1.075, and 1.05 for

the sets (a), (b), and (c), respectively, and the V., 6 and 0' values

are obtained. The largest acceptable values of 6' and the corresoondlng

ff-values are reed and Interpolated from Figures 52 to 54.

-158-
 Vertical walls
should be rough

Fig. 100

Bin, Example 5

-159-
 Parameter H(e') is read from Fig. 43 and B is computed from Eq. (23).

6' angles are read and interpolated from Figures 47 to 49. All the

values are tabulated:

Set V 6 •• ^1 P
ff H(e') B c
(a) 5.3X 50* 29. 14.4 22" 1.2 1.12 1.4 11'
(b) 7.6Z 65- 32* 32.3 20« 1.07 l.ll 3.1 7"
(c) e.7X 70' 35» 52.2 18' 1.03 1.10 5.0 5*

B is now plotted versus the moisture content H.O in Fig. 102. Any one of

the slot-outlet, mass-flow bins shown in Fig. 37 can be used for storage.

Depending on the selected width B, a raised hopper slide belt or an idler-

supported belt feeder can be used.

7. A solid whose flow properties are shown in Fig. 103 and whose

bulk density y " 80 is to be stored in a square bin. Segregation is not

a problem and the bin will be emptied periodically. From required volume

and available space considerations, it has been determined that, at the

bottom of the bin, the effective consolidating head h^ = 20'. Two alternatives

are considered:

(a) A funnel-flow bin with a vibrating hopper (discharger) and a

small screw feeder, Fig. 104(a). The outlet diameter D of the static

hopper is designed for no-piping. From Eq. (17B) , V = 80 x 20/13 = 123 lb.

From Fig. 103, we read V - 14 lb and <^ =45°. A value of G(<t)^) =4.3

[R rend from Fig. 3h and D is computed from Eq. (18). This results in

U - 9.8'.

-160-
 Fig, 101

Flow pi-opertie^„ Example 6

-J4|lr'
 6

5
6

2 |—i—iiiw—il-*!™—!—ii-« .nil. ••! fc..i.i.i. • — * • — 1 1 - ^ — . A . I . I I . . - - I . ••i-|.i lyW • II * ... «i—^—.,1—II-.I i|

1
I I I I r I I I I I
0
0 1 2 3 <» 5 « 7 8 9
«2°
Fig. 102
EMi9>le 6: width B of slot outlet as a function of HgO

-162-
 70"
60"
U--i —i
o«a... ... T a: A 8x 1

40* W4 —-—L ^JZL

u
30* r\ >u
14 — 1 .

12 1

F & Vi
10 1

H
L25
h==1l
V, - 7.2 FF
L-L..-. \ ^
iT

r
1

u1^ 8 10 12 l^
\
0 or

(a)

50

F &V 25
•

50 100 150 200

V.

(b)
Fig. 103 Flow properties. Example 7

163
 A 10-foot vibrating hopper is required. If the walls of the

square-to-round transition hopper are sufficiently smooth and made of

a non-corroding material. It is likely that with the asslstsncs of

the vibration from the discharger, the bin will be self-cleaning. A

wall slope of 35" Is recommended. Fig. 104(a).

(b) A funnel-flow bin with a screw feeder, Fig. 104(b). The

intersection of ff - 1.7 with FF , Fig. 103, determines V^ - 7.2.

The width of the outlet la computed from Eq. (19) at B - 15 x 7.2/80 -

1.35. A 2 4 " stepped-pitch (1/2, 3/4, 1 and 1-1/4) screw is selscted

for a total length L - lO.Oi'

The diagonal D Is in

excess of 9.8 found above to be sufficient to prevent piping. Two wall

vibrators are located on the side walls for final clearance only.

8. Design a storage pile. Fig. 76, for coarse ore weighing

130 pcf. The flow properties are given In Fig. 105, the layout around

the outlet is shown in Fig. 106. The hopper is of the mass-flow type;

the height of the hopper Is designed so that the diagonal D at the top

of the hopper be sufficient to prevent piping from developing within

the flat-bottom part of the bin. During operation the hopper will

always be kept full of ore to minimize the effect of Impact. The

hopper is designed for the tlme-flow-function and, hence will give

unassisted gravity flow, and vibrators are not required.

At the outlet ff " 1.1 is assumed, the point of intersection

-164-

f
 ^ Discharger
Screw feeder

(a) (b)

Bin with a discharger Bin with a screw feeder and wall

and a small screw feeder vibrators for final clearance only

Fig 104

Bin, Example 7

-165-
 with FF^ determines V^ - 27.3, 6 - 60" and •' - 32". From Fig. 53,

9' « 20" is assumed. This checks out the assumed value of ff.

H(9') is read off Fig. 43 and B is computed from Eq. (23). The

values are tabulated below.

1 6 *' 9' ff H(9') B

"i P
60" 32" 27.3 20" 1.1 1.1 3.0

Above the hopper, funnel-flow analysis applies. An effective

conHolldatlng head It ^ « 25' I H assumed. From Eq. (17B), V >« 130 x 25/1 \

250 lb. From Fig. 105, we obtain V = 45 lb and (J) = 4 0 ° . A value of

1 t

G(<|)^) = 3.6 is read from Fig. 36 and D is computed from Eq. (18). This

results in D = 16.2' .

A hopper with straight ends is used, as shown in Fig. 106.

This widens the hopper width to 15 feet at the top and produces a

dlagontl H - y 9^ -f 15^ - 17 .5 feet.

The side-wise reciprocating feeder is racommsndsd for this

installation.
75

50
•
F & V

25

I) •)() 100 150 200 250

V

Fig. 105(a) Flow properties. Example 8

- 166
 70"
1 6 « 65-> ...
60"
8" - 5 ) ^ — , L ^ ),
50'
40
— r
L

30" V-- '. ) ^ "~V^

.( A
70 ... 1

60 - -- /
F & V
1
50
KV
/

40 h^y^ 5 ~y ..^^ r^ "FF^

— ^ • ^ ^
30 V - 27 3 /
r^—
r ^
20 ' : ^
•/

10 /

W L _.^—

50

40

30

20

10

0 10 20

Fig 105 (b)

Flow p r o p e r t i e s , Example 8

16 7.-
Side-wise
reciprocating Conveyor
feeder

Fig. 106

Layout around the outlet, Example 8

-168-
 APPENDIX

Introduction

The solids which are considered in this work are rigid-plastic.

In the plastic regions, the solids are assumed isotropic, frlctional,

cohesive and compressible. During Incipient failure an element of a

solid expands, while during flow, the element either expands or

contracts as does the pressure along the streamline. Steady flow is

assumed.

The determination of the flow-factors for no-doming and no-

piping requires the solution of steady state flow stress fields in

converging and vertical channels. The fields are solved in plane flow

(plane strain) and in conical flow (axial symmetry). These fields are

statically determinate and are solved without reference to velocity

fields. Since the prevailing stresses are compressive, pressures are

considered positive and tensions negative. Plane-polar and spherical

coordinates are used in converging plane and conical flow, respectively.

In order to handle both types of fields with one set of equations,

coefficient m is introduced and assigned the values:

m - 0, in plane flow, I
} (A-1)
m - 1, in conical flow. I

-169-
Doming
Consolidating pressure in converging channels, o. - A solid
consolidates under the pressures which occur during flow. The analysis
of these pressures is based on the following equations:

(a) Two equations of equilibrium

3o 3T

+ Y cos 9 - 0. (A-2)

^ + 7^+7t»<''e- V cot 9 . K 2 + m)T^3] -

- Y sin 9 - 0. (A-3)

The coordinates and the stresses are defined in Fig. A-1.

(b) The equation of state, Eq. (1), written in the above component stresses

(o^ + Og) aln 6 - ]/(^^ - O Q ) ^ -I- 4 T^^ - 0. (A-4)

(c) In axial symmetry, the Hear and von Karman hypothesis [47] which, in
converging flow sets the value of the circumferential pressure o^ equal to
the major pressure o of the meridian plane r,9.

o^-vO^. (A-5)

In plane flow, pressure o does not appear.

(d) A one-to-one relation between mean pressure,

0 - 1/2(0^ + o j . (A-6)
r 0

and density, Y
Y - yia). (A-7)
-170-
 Top boundary

Wall Wall

> y

Fig. A-1

A general channel

-171-
These equations together with appropriate stress boundary conditions

define a general stress field.

Radial stress fields. First of all, particular stress fields,

called "radial stress fields" are obtained. The radial fields satisfy

the functions

o - Y r 8(9), (A-8)

• - •O), (A-9)

in which s(9) is a stress function and i|i is the angle between the

direction of the major pressure a. and the coordinate ray, Fig. A-1.

Density

Y - Yo (A-10)

Is constant, i.e. the solid is Incompressible.

In accordance with Fig. A-2, the component pressures are expressed

as follows:

o - 0(1 -f- sin 6 cos 2ij)), (A-11)

0- - 0(1 - sin 6 cos 2\j)), (A-12)

9

T^Q - 0 Sin 6 sin 2^, (A-13)

a^ - 0(1 + sin 6). (A-14)

The substitution of the above expressions and their derivatives for the

component stresses In Equations (A-2) and (A-3) yields after transforma-

tions .

-172-
 Fig. A-2

Mohr stress circle

-173-
 f|=F(e,«,s) -

- 1 - [m s sin 6(1 + sin 6) (cot 6 sin 2i() cos 2^ -1) + cos 8 -

2
sin 6 cos(e + 24)) •^ s cos 6]/2 s sin 6(co8 2i|» - sin 6), (A-15)

ff-G(e.^,s) -

s s i n 2rp + s i n (9 + 2i<)) + m s s i n 6[cot 6(1 -f cos 2^) - s i n 2t|)]. (A-16)

cos 2\\> - s i n 6

It will be noticed that r has cancelled out and does not appear in the

equations. These equations are equivalent to the integral equations

*(6) - H^') + A[t,i|»(t),s(t)]dt, (A-17)

6o
s(9) - sO") + A[t,*(t),s(t)]dt, (A-18)
9o

which have been solved numerically from initial conditions (9* - 0,iK9") =

Tr/2,s(9°) - s") for a range of useful wall conditions (9;iii'), for symmetric

channels. Fig. A-3(a); and from initial conditions (9° - 0,ii'(9")-i|)^,s(9°) - s'

for wall conditions (9>|i^ ) , for asymmetric plane flow channels.

Fig. A-3(b). The results are plotted in 9', ^^ coordinates. This requires

a change of variables from i)i' to ^' and from ^ to ^ . Since

T^Q/O^ " - tan •', and T^Q/OQ - tan (>'', (A-19)

it follows from Equations (A-12) and (A-13) that

(A-20)

-174-
(a) (b)

Fig. A-3

Channels

175-
Equations (A-17) and (A-18) are then in the form

U» - *(m.6,6',<^'), (A-21)

s = s(m,6,6',(^'), (A-22)

in symmetric channels, and

ip - tKm,6,9',•',(!.''), (A-23)

s - s(m.6,9',•',•''), (A-24)

in plane asymmetric channels with one vertical wall. The component

pressures are now readily obtained from Equations (A-11) to (A-14). For

details of the analysis the reader is directed to references 39 and 43.

Convergence of general stress fields to radial stress fields.

The significance of the radial stress fields lies in the fact that real

stress fields in plane strain -wnd in axial symmetry converge to the radial

stress fields as the vertex of the channel is approached, i.e. for r ->- 0.

This convergence is of fundamental importance in the derivation of flow-

no flow criteria because most obstructions to flow originate at the out-

lets of channels, i.e. in the vicinity of the vertex. Since in real

channels, convergence to a radial stress field is very rapid [44], the

stresses in the region of an outlet are usually sufficiently well

represented by the radial stress field for design purposes. Therefore,

the stresses in the region of the outlet can be predicted on the basis

of the slope and angle (or angles) of friction of the walls at the out-

let without reference to the top boundary conditions, such as the head

-176-
of the solid over the outlet, the shape of the top, traction-free

boundary, and the shape of the walls away from the outlet.

This Independence of the head and boundaries away from the

vertex has been observed by Investigators and engineers for a long

time [1, 20, 48]. An analysis of the equations shows how this

convergence takes place.

It is compatible with experience and with the equation of state

(A-4) to assume that density is represented by the function

Y - Y. + Y ( O ) . (A-25)

where Y(0) " 0> and the function is continuous. In general fields,

the functions (A-8) and (A-9) do not hold but can be written in the

form

0 - r[Y, + Y(o)]s(r,9). (A-26)

* - •(r,9). (A-27)

Expression (A-26) is substituted for a In ths expressions (A-11) to

(A-14) which are, in turn, substituted for the component stresses and

their derivatives in the equations of equilibrium (A-2) and (A-3)

leading, after transformations, to

||-4- 8 f(r,9) + g(r,9) - 0, (A-28)

r |5--h 8 h(r.9) -t- j(r,9) - 0, (A-29)

where

-177-
 f ( r . 9 ) - 2 ( | | -^ l)5iH_« s i n 2i|» + 2r | ^ 5i2_« ( s i n 6 + cos 2*) +
cos fi COS 5

-I- - l ^ - K m ^ i ^ (1 + s i n 6) [ s i n 2i|i - cot 9(1 + cos 2 * ) ] , (A-30)

cos o

g(r,9) - - ^ s l n ( 9 + 2*) - ^ ^ , (A-31)

cos fi COS fi

h ( r , 9 ) . 1 + 2 ( | | - H ) 5 i 2 _ i (cos 2* - s i n «) - 2r | ^ ^ ^ s i n 2* +
cos fi COS fi

•I- - 1 ^ -f m 5^—^ (1 -I- s i n fi) (cot 9 s i n 2<j» + cos 2* - 1 ) , (A-32)

' COS fi

j ( r . 9 ) - - S i i L i c o s ( 9 + 2*) •^ ^ ^ . (A-33)
cos fi COS fi

Equations (A-28) and (A-29) have r e a l c h a r a c t e r i s t i c s whose

d i r e c t i o n s are

l l - r cot [t± ( J - f ) ! . (A-34)

The equations are hyperbolic for r yl 0, but parabolic for r •• 0

(i|< - ir/4 -fi/2and i|i - 3ir/4 + fi/2 lie outside of real channels, because

•' and i < 5 ) . A typical field of characteristics for the general channel

of Fig. A-1 is shown in Fig. A-4 in orthogonal axes 9,r. It is apparent

that the effect of the variable r decreases as r -> 0. At r •• 0, the

parabolic solution is independent of the derivatives with respect to r and,

in fact, is the same as the radial stress field. Hence, barring dis-

continuities, the general stress field converges to the radial stress field

at the vertex.

The concepts of plasticity permit stress discontinuities. In

-178-
 Top boundary

Wall

Stress
characteristics

Fig. A-4

Field of characteristics

-179-
particular, a discontinuity can follow a streamline. This type of

discontinuity does not affect the present discussion because such a

discontinuity can be looked upon as a wall and the regions on each

side of the discontinuity can be considered as separate channels. Any

other discontinuity would reach a wall, reflect from It, and continue

reflecting from wall to wall since the field Is hyperbolic. A solid

flowing across these lines -of stress discontinuity would undergo

successive, discontinuous expansions and contractions. It is unlikely

that these conditions can satisfy the second law of thermodynamics.

In physical channels, discontinuities do not seem to appear in

flowing solids. If a discontinuity is Introduced at the wall of a

channel, the solid either does not flow at all, or remains rigid

(elastic) in a part of the channel, and forms new walls across itself.

Surface discontinuities along a top boundary are invariably contained

within rigid regions.

Results of numerical solutions of general fields have been

presented by Johanson [44] and an example from his paper is shown in

Fig. A-5. Mean pressures, Eq. (A-6), are plotted: o** is the value

along the axis of symmetry and o" along the wall. The dashed lines

represent radial stress values. The wavy lines represent the values

of the general field which arises from a high vertical head of the

solid above the conical part. The convergence is very rapid.

Critical pressure in a dome, g.. While the stress field in a

doming mass of a solid may not be uniquely determined, it appears

possible to establish a lower bound a. for use in formulas (8) and

(9) to assure the yielding of the dome. The reasoning is as follows:

-180-
 Axis of
symmetry

(J. R. Johanson)

Fig. A-5

Example of convergence

181-
 (a) Consider a mass of a solid forming a stable dome across

the channel, Fig. A-6. The traction-free, bottom surface of the

dome has assumed some shape expressed by the function

r - a f(e). (A-35)

Assume that the shape of the dome is such that the critical pressure

is at a minimum. Evidently, such a shape will be smooth. The

critical pressure is assumed to occur at the abutments of the dome.

Take f(0) - 1 and the parameter a constant along the surface of the

dome.

(b) Consider the mass of the solid as being made up of super-

imposed layers, each bounded by surfaces similar to that of the

traction-free, bottom surface, i.e. given by Eq. (A-35) with parameter

£ different but constant on each surface.

(c) Assume that each layer is self-supporting, i.e. no stresses

act between adjacent layers. The pressures 0. at the abutment of each

layer are then directly proportional to r', the distance from the

vertex.

(d) The above conditions lead to a lower bound on the pressure

in the bottom layer of the dome because, In fact, pressures do act

between adjacent layers and, in particular, on the bottom layer from

the mass above.

The pressures between adjacent layers arise because, if pressure

0, - f in the bottom layer, then o, > f in the higher layers which

1 c 1 c

therefore cannot be self-supporting, o, > f in the higher layers

because, while o. is proportional to r', f Increases with r' at a

-182-
smaller rate due to two reasons:

(1) In the region of radial stress, consolidating pressure o^ Increases

linearly with r'. But flow-function curves are, generally, convex-upward

and/or have a positive intercept on the f axis. Hence, f Increases at

a rate smaller than linear with o..

(2) At some height of the channel, the real consolidating pressure a^

falls below the radial consolidating pressure. Above that height, the

strength f falls still further below the value required for self-

support, in fact, f may fall to zero at the top, traction-free surface

of the solid.

The problem of the lower bound thus resolves to finding the shape

of a layer bounded by surfaces a f(d) and (a + da) f(6), such that, for

a solid of given bulk density y spanning a distance B, the non-dimensional

ratio

H(e') - BY/OJ^ (A-36)

Is at a maximum. Density y is assumed constant in this analysis which

is described in detail in reference [42]. Function H(6 ) has been

computed and is plotted in Fig. 43 for conical and plane-flow channels.

The value of H(d') for square channels is estimated at 0.9 of H(e') for

conical channels.

Replacing o^^^ by V. » Ao., the critical width B of a dome is obtained

from Eq. (A-36), thus

B - H(e')V^/AY. ^A-3/

This is Eq. (12) in Chapter I.

-183-
 In vertical channels, i.e. for 6' - 0, H(e') - 2 in conical

channel's, and H(e') > 1 in plane-flow channels. These particular

results were published in reference 33.

No-doming flow-factor. It is evident from Fig. A-6 that

B - 2r' sin e'. (A-38)

From Equations (A-36) and (A-38), the critical pressure at the abutment

of a dome is

o^ - 2 Y r' sin 8VH(e'), (A-39)

while the major consolidating pressure Is found from Equations (A-8),

(A-14) and (A-22) for 6 - e' to be

o^ - Y r'(l + sin «)s(m,«,e',4.') (A-40)

for symmetric channels.

The flow-factor is now obtained in accordance with the definition (8)

ff . H(e') a + «ing)sK;>e',f) . (A-41)

2 sin 0'

For asymmetric channels in plane flow with one vertical wall, the

function s'(6,6',•'j^^), Eq. (A-24), has been computed [39] but the

function H(9'), Eq. (A-36) is not available. Therefore, the same value

of the latter function is used as for symmetric channels. This yields

the following safe expression for asymmetric channels

ff . H(8') a + 8 i n g)s(m.6.e'.»'.»^)
H^H ) Tsine'^ (A-42)

-184-
 Wall
Wall

Vertex

Fig. A-6

A dome across a channel

185-
 The no-doming flow-factors are plotted for mass-flow hoppers

in Figures 45 to 57, In 6',^' coordinates. The contours show values

of constant flow-factor. In symmetric channels the analysis has been

carried out for 6 - 3 0 ' , 40", 50", 60° and 70'. In asymmetric plane

flow channels with one vertical wall the analysis has been made for

6 - 50" and •^ - 20", 30° and 40'.

In funnel-flow bins, the solid forms its own channel within

itself. The angle of friction at the walls should therefore equal

the angle of Internal friction ^ of the solid. However, the walls

become polished as the flowing solid slides on them and the angle of

friction at the walls is less than the angle of internal friction.

As was mentioned earlier, the angle of internal friction also varies

within a rather wide range and, hence, it is difficult to determine

its actual value at the walls. In view of these uncertainties it

seems reasonable to assume that the angle of friction at the walls ^'

in funnel flow is somewhat less than ^ and estimate it at

tan •' - sin 6 (A-A3)

This Implies that <|i' - 3ir/4 at the walls, Eq. (A-20). Such an

assumption makes the slope angle 8' a function of fi and of the £.,.ow-

factor ff.

Contours of the no-doming flow-factor for furmetflow are

plotted in Fig. A-7. The significance of this chart is different from

those for mass flow. In funnel-flow, the slope of the channel is not

defined: the channel forms within the solid itself. The chart showj.

that, for any given value of 6, the flow factor increases (tlowabiLwv

-186-
 Fig. A-7

ff contours for doming in funnel flow

-187-
of the channel decreases) with an Inccease of the slope angle 6'. The

chart also gives the bounds for the angle 6' in conical and in plane-

flow. The former lies quite low, hence, a conical channel in funnel-

flow is very steep, often veitical. A plane-flow slope can form only

at the side of a long rectangular outlet; upward from the outlet the

channel approaches a cicculac, horizontal cross-section and the

channel becomes conical, as Illustrated in Fig. 30(a). In the design

of rectangular outlets, it Is recommended that an ff * 1.7 be used.

This permits an angle 6' approximately equal to 30* at the sides of

the rectangle for most values of 6, and, hence, a widening of the

channel from a slot to a ciiculaE channel circumscribing the outlet.

iMS&.
Consolidating pressure in a pipe, g^. A stress field which is

Independent of the vertical coordinate is assumed here and the angle

of friction at the walls is assumed to satisfy Eq. (A-43). The

equilibrium of an arch across a pipe of diameter !D is shown in Fig. A-8:

the weight of the arch causes a shearing stress at the walls

T D/4. (A-44)

It then follows from the relations of the Mohr stress circle. Fig. A-2,

that the major consolidating pressure at the walls of the pipe is

Yp 1 + sin 6 /. /,r\
0 « -ij — (A-45)
1 4 sin 6

-188-
 Critical pressure in a pipe, o.. The conditions leading to the

failure of a solid around a vertical, circular hole were analyzed in

references [39] and [41]. The unconfined yield pressure f and the

angle of friction ^ were assumed constant throughout the failing

region. The stress field was assumed axi-symmetric with a traction-

free inner boundary of constant diameter D. The critical major

pressure at the surface was computed at

g^ - Y D/G(«j.), (A-46)

where function G(^ ) is as plotted in Fig. 36.

Replacing g^^ by the corresponding force V. • Ag^^, Eq. (A-46)

yields the following expression for the critical diameter of the pipe

D - G(*J.)VJ^/AY. (A-47)

This is Eq. (13) in Chapter I.

-190-
Preaaure on a wall, g'

The preaaure o' from a flowing solid on a wall haa been computed

[39] for aynmetrlc channels in mass flow from Eq. (A-12) with the

following substitutions: r - B/2 sin 8' from Eq. (A-38), s(m,6,8',^')

from Eq. (A-22), and ^* from Eq. (A-20). This leads to the aKpresaion

g'/YB . .(m,6,8'.*') ^ - ' ^ ^ ^ ^ ^ ^ g . ^ ^ ' ^ M ' ) (A-A9)

Contours of constant values of O'/YB are plotted in Figures 58 to 67

in 8',^* coordinates for constant values of 6 - 30°, 40°, 50°, 60° and

70°, for conical and plane-flow channela.

Reaultant vertical force Q across a channel

Force Q is obtained by integrating the vertical pressure

o^ - Yrs(m,6,8,*)[l + sin « cos 2(* + 8)] (A-50)

over a horizontal cross-section of a channel. Fig. A-9. The relevant

relatione are

r - B cot 8 7 2 cos 8

- y - (B/2) cot 8' tan 8 ^ (A-51)

d(-y) - (B cot 8 7 2 cos^8)d8

and the total force Q for symmetric channels is

Q - 2ir" L^'^^f^^^a^y'^dy. (A-52)

-191-
Upon integration

Q-q(m,6,e',*')YL'' B''™. (A-53)

Contoura of constant values of q have been computed [39] and are plotted

in Figures 80 to 85, for conical and plane, mass-flow channels, and for

6 - 3 0 ° , 40° and 50°.

-192-
Wall

^ y

Fig. A-9

Vertical force Q

193-
 BIBLIOGRAPHY

H. A. Janssen, "Versuche uber Getreidedruck in Silozellen",

Vereln Deutscher Ingenieure, Zeitschrift, V. 39, Aug. 31, 1895,
pp. 1045-1049.

W. Airy, "The Pressure of Grain", Minutea of Proceedinga of the

Institution of Civil Engineers, V. 81, Paper No. 3049, 1897.

J. A. Jamieson, "Grain Pressures in Deep Bins", Canadian Society of

Civil Engineers, Transactions, V. 17, 1903, pp. 554-654.

M. S. Ketchum, "The Design of Walls, Bins and Grain Elevators",

McGraw-Hill Book Co., New York, 3rd edition, 1919.

A. Caquot and J. Kerisel, "Traite de Mecanique des Sols",

Gauthler-Villars, Paris, 3rd edition, 1956, Chapter 20.

M. and A. Relmbert, "Silos - Traite Theorlque et Pratique",

Editions Eyrolles, Paris, 1956.

W. Nowacki and R. Dabrowskl, "Silosy, Metody Obliczefi 1 Konstrukcja",

Budownictwo 1 Architektura^ Warsaw, 1957.

S. Sienicki, H. Domaszewskl and C. Klos, "Zbiornikl Materjalow

Sypkich", Budownictwo 1 Architektura, Warsaw, 1958.

P. Rogers, "Design of Large Coal Bunkers", Tranaactiona, A.S.C.E.,

V. 117, 1952, pp. 579-595.

R. A. Saul, "Measurement of Grain Pressures on Bin Walls and Floors",

AGRICULTURAL ENGINEERING, V. 34, Apr. 1953, pp. 231-234.

M. North, "Research into the Design of Horizontally Corrugated Grain

Silos", Journal, The Institution of Engineers, Australia, Jan.-Feb. 1934.

0. F. Theimer, "Zur Berechnung von Mehlsllozellen", DIE BAUTECHNIK,

V. 34, Dec. 1957, pp. 458-465.

-194-
R. L. Brandes, "Design of Deep Bins and Silos", Concrete Engineering
Handbook, Section 18, McGraw-Hill Book Co., New York, 1961.

A. M. Turitzin, "Dynamic Pressure of Granular Material in Deep Bins",

Proceedings ASCE, V. 89, No. ST2, April 1963.

J. M. Rausch, "Gravity Flow of Solid Beds in Vertical Towers", Ph.D.

Thesis, Princeton University, Princeton, N. J., 1948.

F. C. Franklin and L. N. Johanson, "Flow of Granular Material

Through Circular Orifices", CHEMICAL ENGINEERING SCIENCE, V. 4,
June 1955, pp. 119-129.

H. E. Rose and T. Tanaka, "Rate of Discharge of Granular Materials

from Bins and Hoppers", ENGINEER, V. 208, Oct. 1959, pp. 456.

F. E. Keneman, "Mechanism of Free Discharge of Granular Solids",

INZHENERNO-FIZICHESKII ZHURNAL, V. 3, March 1960, pp. 69-73,
April 1960, pp. 18-22.

W. A. Beverloo, H. A. Leniger and J. van dk Velde, The Flow of

Granular Solids Through Orifices", CHEMICAL ENGINEERING SCIENCE,
V. 15, 1961, pp. 260-269.

R. L. Brown, "Minimum Energy Theorem for Flow of Dry Granules

Through Apertures", NATURE, V. 191, July 29, 1961, pp. 458-461.

H. Wohlbier and W. Reisner, "Fragen der Bunkerung von mlttel und

felnkornlgem Schvlttgut", CHEMIE-INGENIEUR-TECHNIK, V. 34, 1962,
pp. 603-609.

R. M. LaForge and B. K. Boruff, "Profiling Flow of Particles

Through Hopper Openings", INDUSTRIAL AND ENGINEERING CHEMISTRY,
Vol. 56, n 2, 1964, pp. 42-46.

D. R. Mitchell, "Segregation in the Handling of Coal", AIME

Trans. V. 130, 1937, pp. 107-142.

-195-
H. M. Peacock, The Deaign or Adaptation of Storage Bunkers to
Prevent Size Segregation of Solids", The Institute of Fuel,
Feb. 1938, pp. 230-239.

E. F. Wolf and H. L. von Hohenleiten, "Experimental Study of

the Flow of Coal in Chutes at Riverside Generating Station",
MECHANICAL ENGINEERING, April 1948.

J. K. Rudd, "How Does Material Flow From a Bin?", MILLING

PRODUCTION, Jan. 1954, pp. 5, 7.

J. R. O'Callaghan, "Internal Flow of Beds of Granular Material",

Journal of Agrlc. Res., Wrest Park, Beds, England, V. 4, 1959,
pp. 200-217.

Y. Lee, "Flow of Coal in Hoppers", COMBUSTION, Jan. 1960.

H. Taubmann, "Beitrag zur Speicherung von Schilttgiitern",

AUFBEREITUNGS-TECHNIK, Nr. 1, 1960, pp. 17-26.

W. J. Reynolds and R. S. Carter, "Design of Bunkers for Difficultly

Flowing Fuels", COMBUSTION, Sept. 1962, pp. 32-37.

F. A. Zenz and D. F. Othmer, "Fluidization and Fluid-Particle

Systems", Reinhold Publishing Corp., N.Y., 1960, Chapter 4.

F. Silver, "Storage and Handling of Pulverized Materials",

MECHANICAL ENGINEERING, Sept. 1951, pp. 730-734.

A. W. Jenlke, "Flow of Solids in Bulk Handling Systems", and

"Flow of Bulk Solids in Bins", Bui. 64 of the Utah Engineering
Experiment Station, University of Utah, 1954.

, "Better Design for Bulk Handling", CHEMICAL

ENGINEERING, V. 61, No. 12, Dec. 1954, pp. 175-180.

, "How Bulk Behaves", FLOW, V. 10, No. 1, Oct. 1954,

pp. 74-77, 119-125; No. 2, Nov. 1954, pp. 98-101, 125-131.

-196-
A. W. Jenlke, "How to Keep Solids Flowing in Bins and Hoppers",
ENGINEERING AND MINING JOURNAL, V. 156, No. 3a, Mid-March 1955,
pp. 83-85, 93.

and R. T. Shield, "On the Plastic Flow of Coulomb

Solids Beyond Original Failure", J. of Appl. Mech., V. 27,
Dec. 1959, pp. 599-602.

, P . J . Elsey and R. H. Woolley, "Flow Properties of

Bulk Solids", Proceedings, A.S.T.M., V. 60, 1960, pp. 1168-1181.

, "Gravity Flow of Bulk Solids", Bui. 108, Utah

Engineering Experiment Station, University of Utah, Salt Lake
City, Utah, Oct. 1961.

, "Gravity Flow of Solids", Trans, of the Institution

of Chemical Engineers, V. 40, n 5, 1962, pp..264-271.

and B. C. Yen, "Slope Stability in Axial Syinnetry",

Proceedings of the Fifth Symposium on Rock Mechanics, University
of Minnesota, May 1962, Pergamon Press, 1963, pp. 689-711.

and T. Laser, "A Flow-No Flow Criterion in the

Gravity Flow of Powders in Converging Channels", Fourth International
Congress on Rheology, Brown University, Aug. 1963.

, "Steady Gravity Flow of Frictional-Cohesive Solids

in Converging Channels", J. of Appl. Mech., V. 31, Series E, n 1,
1964, pp. 5-11.

J. R. Johanson, "Stress and Velocity Fields in the Gravity Flow

of Bulk Solids", J. of Appl. Mech., Series E, V. 86, Sept. 1964.

A. W. Jenlke, "Gravity Flow of Frictional-Cohesive Solids -

Convergence to Radial Stress Fields" to be published as a Brief
Note in J. of Appl. Mech. in 1965.

-197-
D. C. Drucker, "A More Fundamental Approach to Plastic Stress-
Strain Relations", Proc. First U. S. Nat. Congress of Appl.
Mechanics, ASME, 1951, pp. 487-491.

A. Hear and Th. von Karman, "Zur Theorie der Spannungszustaende

in plastischen und sandartigen Medien", Nachr. Ges. Wiss.
Gottingen, Math.-Phys. Kl., 1909, p. 204.

R. A. Wilson, "Belt Feeders", MECHANICAL ENGINEERING, V. 79, 1957,

pp. 1042-1043.

W. H. Denton, "The Packing and Flow of Spheres", Atomic Energy

Research Establishment, E/R 1035, Ministry of Supply, Harwell,
Berks, England, 1953.

A. W. Jenlke, P. J. Elsey and R. H. Wooley, "Flow of Bulk Solids,

Progress Report", Bui. 96, Utah Engineering Experiment Station,
University of Utah, Salt Lake City, Utah, Dec. 1958.

R. W. Olson, "Dust Explosions: Prevent or Supress Explosion Losses

in Bulk Materials", MECHANICAL ENGINEERING, V. 83, 1961, pp. 59-61.

0. Richmond, "Gravity ijbpper Design", MECHANICAL ENGINEERING,

Vol. 85, Jan. 1963, pp. 46-49.

G. C. Gardner, "The Best Hopper Profile for Cohesive Material",

Chemical Engineering Science, Vol. 18, 1963, pp. 35-39.

D. M. Walker, "A Theory of Gravity Flow of Cohesive Powders",

Central Electricity Generating Board, South Western Region,
England, Research and Development Report No. 22, May 1964.

J. R. Johanson and H. Colijn, "New Design Criteria Jtor Hoppers

and Bins", IRON AND STEEL ENGINEER, V. XLI, Oct. 1964, pp. 85-104.

-198-