<%BANNER%>

Experimental Investigation of Particle Segregation in Hopper Discharge

Permanent Link: http://ufdc.ufl.edu/UFE0021894/00001

Material Information

Title: Experimental Investigation of Particle Segregation in Hopper Discharge
Physical Description: 1 online resource (47 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Chemical Engineering -- Dissertations, Academic -- UF
Genre: Chemical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Many industrial processes including food, mining, and pharmaceuticals involve handling mixtures of granular materials. The tendency of granular materials to segregate due to differences in particle properties such as, size, shape, and density negatively affects the process efficiency and product quality. Our study focused on segregation of particles in hopper flow with the following conditions: well-mixed initial fill, funnel flow, and cylindrical hopper geometry. Glass beads and steel shots in three particle sizes each are used to investigate the effect of size ratio and density difference on particle segregation. Also, to study the effect of particle shape on segregation of particles, crushed glass in two sizes is used. Results show that size-segregation patterns are initially fines-rich and then fines-depleted at the end of discharge. Segregation increases for larger particle size ratios and decreases for fines fractions that are above 10%. The extent of segregation for mixtures of crushed glass-crushed glass is not significantly different from spherical glass mixtures. Given the same size ratio, mixtures consisting of smaller particles show less segregation. The segregation trend due to density effects is initially rich in particles with the higher density and then rich in the particles with the lower density at the end of the discharge.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Curtis, Jennifer S.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021894:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021894/00001

Material Information

Title: Experimental Investigation of Particle Segregation in Hopper Discharge
Physical Description: 1 online resource (47 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: Chemical Engineering -- Dissertations, Academic -- UF
Genre: Chemical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Many industrial processes including food, mining, and pharmaceuticals involve handling mixtures of granular materials. The tendency of granular materials to segregate due to differences in particle properties such as, size, shape, and density negatively affects the process efficiency and product quality. Our study focused on segregation of particles in hopper flow with the following conditions: well-mixed initial fill, funnel flow, and cylindrical hopper geometry. Glass beads and steel shots in three particle sizes each are used to investigate the effect of size ratio and density difference on particle segregation. Also, to study the effect of particle shape on segregation of particles, crushed glass in two sizes is used. Results show that size-segregation patterns are initially fines-rich and then fines-depleted at the end of discharge. Segregation increases for larger particle size ratios and decreases for fines fractions that are above 10%. The extent of segregation for mixtures of crushed glass-crushed glass is not significantly different from spherical glass mixtures. Given the same size ratio, mixtures consisting of smaller particles show less segregation. The segregation trend due to density effects is initially rich in particles with the higher density and then rich in the particles with the lower density at the end of the discharge.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Curtis, Jennifer S.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021894:00001


This item has the following downloads:


Full Text





EXPERIMENTAL INVESTIGATION OF PARTICLE SEGREGATION IN HOPPER
DISCHARGE




















By

BYUNG-HWAN CHU


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2008


































2008 Byung-Hwan Chu
































To my family and friends









ACKNOWLEDGMENTS

I would like to express my deepest gratitude to the many people who assisted and

supported me through graduate school. I am very grateful to my advisor, Dr. Jennifer Curtis, for

her guidance and encouragement. I am also thankful to Dr. Carl Wassgren, Dr. Bruno Hancock,

and Dr. Bill Ketterhagen for sharing their knowledge in particle science through the weekly

teleconference calls. I would also like to thank my supervisory committee member, Dr. Spyros

Svoronos, and undergraduate assistant, Peter Kovedra.

I would like to acknowledge my research group members, fellow students at the chemical

engineering department, and friends in Gainesville for making my stay in Gainesville such a

wonderful time. Their support, care, and friendship made me a better engineer and a better

person. Lastly, I thank my parents and brother for their support and unconditional love from

home.









TABLE OF CONTENTS



page

A CK N O W LED G M EN T S ................................................................. ........... ............. .....

L IS T O F T A B L E S .................................................................................7

LIST OF FIGURES .................................. .. ..... ..... ................. .8

L IST O F A B B R E V IA T IO N S .................................................................. .............................. 10

A B S T R A C T ......... ....................... ............................................................ 1 1

CHAPTER

1 INTRODUCTION ............... .............................. ............................. 12

2 L IT E R A TU R E R E V IE W ......................................................................... ........................ 14

2.1 Classification of Segregation........................................................................ 14
2.2 Funnel Flow and M ass Flow ......................................................................... 15
2 .3 F in es F reaction ...............................................................................16
2.4 Particle Properties that Influence Segregation................................................................16
2.4.1 Particle D iam eter R atio ................................................ ................ ............17
2 .4 .2 P article D density ....................... ...................... .. .. ... ......... ..... .. ... 17
2.4.3 Particle Shape .................. ......................................... ........... .. ..17
2.4.4 Particle Friction .................. ...................................... ............... 18
2 .5 O th er F actors ............................................................................. 18

3 EX PER IM EN TA L M ETH O D S ..................................................................... ..................20

3.1 Granular M edia ..................................... .................. ....... ....... ........... 20
3.1.1 Glass Bead .................................................................. ........... 21
3 .1.2 Steel Shot........................................................................................................... 2 1
3.1.3 Crushed M edia............................... ... ............... .. ............22
3.2 Sifting Segregation T ester ........................................................................ ..................22
3 .3 A ntistatic B ar.........................................................2 3
3.4 D details of Procedure ............... ............................ .............. .. ...... .... ..... 24
3 .4 .1 F illin g ............... ....................... ......................................2 4
3.4.2 D ischarging and Sam pling .............................................................................. 25
3.5 M easurem ents and D ata A nalysis.......................................................... ............... 26

4 RESULTS AND D ISCU SSION .................................................. ............................... 27

4.1 Effect of V various Particle Treatm ents .................................. .....................................28









4.2 Continuous vs. Discontinuous Discharge.......... ... .......................... 29
4.3 E effect of Particle D ensity......... ......... ........ .......... .......................... ............... 31
4 .4 E effect of F ines F reaction ......... ................. ............................................ ........................32
4.5 Effect of Size Ratio.................. ...................................34
4.6 Effect of A absolute Particle Size............................................... ............................. 36
4.7 E effect of Particle Shape ......... ................. .. ....... ............................ ............... 36
4.8 Comparison with Previous Experimental Results ................................. ................ 38

5 CONCLUSION AND RECOMMENDATIONS ....................................... ............... 42

L IST O F R E F E R E N C E S ......... ................. ...............................................................................44

B IO G R A PH IC A L SK E T C H .............................................................................. .....................47









































6









LIST OF TABLES


Table page

3-1 Summary of mean diameter and standard deviation of granular materials ..................21

4-1 Sum m ary of experim mental w ork......... ..................................................... ............... 27

4.2 Treatment methods applied to glass beads with (D = 2 (2 and 1mm), xf = 5%.................28









LIST OF FIGURES


Figure page

3-1 Glass beads with mean diameter = (a) 0.542 0.045mm, (b) 1.06 0.06mm,
and (c) 2.26 0.13mm at 20 power magnification ............... ................................21

3-2 Steel shots with mean diameter = (a) 0.538 0.044mm, (b) 1.13 0.045mm,
and (c) 2.35 0.01mm at 20 power magnification.......... ..................... ...............22

3-3 Crushed glass with mean diameter = (a) 0.549 0.29 mm, (b) 1.13 0.045 mm
at 20 pow er m agnification. ...................................................................... ....................22

3-4 Schematic and dimensions of Sifting Segregation Tester [12]..........................................23

3-5 Antistatic bar with its power supply (left) and antistatic bar mounted on the hopper
(rig h t) .............................................................................................2 4

4-1 Experimental segregation results for variously treated glass-glass mixtures with DD =
2 (2 and lm m ) and xf = 5% ...................................... ...............................................29

4-2 Glass(coarse)-steel(fine) mixture experimental results with OD = 4, xf = 5%, and
continuous or discontinuous discharge methods..................................... ............... 30

4-3 Segregation patterns for 50:50 glass-steel mixture oD = 1 and continuous and
discontinuous discharge ........................................................... .. ........ .... 30

4-4 Experimental results for mixtures consisted of several combination of materials with
o D = 4 and xf = 5% ................ ............................................................................... 1

4-5 Experimental results for mixtures consisted of several combination of materials with
o D = 2 and xf = 5% ................ ............................................................................ 32

4-6 Experimental segregation results from Ketterhagen et al. [12] for glass-glass
mixtures with DD = 4 and given fines fractions ...................................... ............... 33

4-7 Experimental segregation results for steel-steel mixtures with DD = 4 and xf = 2.5%,
5% an d 10% ...............................................................................3 3

4-8 Experimental segregation results for steel-steel mixtures with OD = 4 and xf = 10%,
2 0% an d 50% .........................................................................................34

4-9 Glass-glass mixture experimental results with given size ratios and xf = 5% ...................35

4-10 Steel-steel mixture experimental results with given size ratios and xf = 5% .....................35

4-11 Glass-steel mixture experimental results with given size ratios and xf= 5% ....................36









4-12 Glass-glass mixture experimental results with OD = 2 for different sized particles and
xf = 5% .................... ............... ......... .................................. .............. 37

4-13 Experimental segregation results for crushed glass mixture and glass bead mixture
with OD = 2 and xf = 5% ......................................................... ..................... 37

4-14 Comparison of results for glass-steel mixtures with OD= 2 and xf= 5%..........................38

4-15 Comparison of results for glass-glass mixtures with (D= 2 and xf = 5%.........................39

4-16 Comparison of results for given mixtures with OD= 4 and xf = 5% .................................39

4-17 Effect of varying particle-wall friction coefficient on predicted segregation profiles
for discharge from a wedge-shaped hopper with OD = 4 and xf = 5%................................41

4-18 Effect of varying particle-particle friction coefficient on predicted segregation
profiles for discharge from a wedge-shaped hopper with OD = 4 and xf = 5%..................41









LIST OF ABBREVIATIONS


d particle diameter [mm]

a standard deviation [mm]

p density [g/cm3]

x, volume fraction of fines in a given sample [-]

xf volume fraction of fines initially in hopper [-]

xfL fines limiting fraction [-]

x,/xf normalized fines volume fraction [-]

OD particle diameter ratio [-]









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

EXPERIMENTAL INVESTIGATION OF PARTICLE SEGREGATION IN HOPPER
DISCHARGE

By

Byung-Hwan Chu

May 2008

Chair: Jennifer S. Curtis
Major: Chemical Engineering

Many industrial processes including food, mining, and pharmaceuticals involve handling

mixtures of granular materials. The tendency of granular materials to segregate due to

differences in particle properties such as, size, shape, and density negatively affects the process

efficiency and product quality.

Our study focused on segregation of particles in hopper flow with the following

conditions: well-mixed initial fill, funnel flow, and cylindrical hopper geometry. Glass beads and

steel shots in three particle sizes each are used to investigate the effect of size ratio and density

difference on particle segregation. Also, to study the effect of particle shape on segregation of

particles, crushed glass in two sizes is used.

Results show that size-segregation patterns are initially fines-rich and then fines-depleted

at the end of discharge. Segregation increases for larger particle size ratios and decreases for

fines fractions that are above 10%. The extent of segregation for mixtures of crushed glass-

crushed glass is not significantly different from spherical glass mixtures. Given the same size

ratio, mixtures consisting of smaller particles show less segregation. The segregation trend due to

density effects is initially rich in particles with the higher density and then rich in the particles

with the lower density at the end of the discharge.









CHAPTER 1
INTRODUCTION

A fascinating characteristic of granular materials is their tendency to segregate due to

difference in particle properties such as size, density, and shape. When a container of different

liquids is shaken, they will mix, but in a blend of particles, they will segregate. A famous

example is the "Brazilian Nut Effect" [1, 2] where a bed of large particles rise above smaller

particles upon vibration. Segregation is of importance in a variety of industries that involve the

handling, processing, or manufacturing of particulate materials. Some examples are mining [3],

food [4, 5], and pharmaceuticals such as in the production of pills and tablets [6]. Segregation

can adversely affect the efficiency of processes, as well as the quality of final products which is

dependent on maintaining a homogeneous blend.

Segregation has been studied over many years by researchers from various disciplines due

to its importance and ubiquitous occurrence. Despite the efforts to explain this phenomenon

through various methodologies such as experimentation, simulation, and modeling, there are still

many unanswered questions.

The segregation of particles in hopper discharge is the focus of this thesis. Specifically, a

series of controlled segregation experiments are conducted, changing one variable at a time.

These experiments provide insight into the fundamental mechanisms influencing segregation of

particles in gravity driven flow. These experiments also serve as a basis for validation of

analytical models and numerical simulations of particle segregation.

A hopper geometry is chosen for the segregation experiments since it is a standard

geometry used in particle processes with many practical industrial applications. In addition,

segregation experiments are relatively simple and straight forward in a hopper geometry.

Furthermore, when segregation occurs in a large-scale particle processes which include many









devices such as mixers, fluidized beds, and hoppers, it is important to characterize the amount of

segregation that occurs in the hopper alone versus the other units. For example, if a perfectly

mixed blend is fed to a hopper, how much segregation is to be expected from particle flow in that

device?

Following this introduction, chapter 2 will review literature that is relevant to the current

work. Then in chapter 3 experimental materials and procedures will be explained. The results

from the experiments are then presented and discussed in chapter 4. Chapter 5 summarizes the

thesis work and makes recommendations for further studies.









CHAPTER 2
LITERATURE REVIEW

This chapter provides an overview of previous experimental research that is relevant to the

present study. Experiments related to segregation in hopper discharge are the main focus of this

review. Section 2.1 illustrates several classification methods of particle segregation. Section 2.2

reviews literature concerned with funnel flow and mass flow followed by section 2.3 which

looks into previous work related to fines fraction. Then, in section 2.4 literature about particle

properties that influence segregation is discussed. Finally, section 2.5 reviews previous work that

has investigated other factors that affect segregation.

2.1 Classification of Segregation

Tang and Puri [7] categorized particle segregation depending on several variables.

Segregation may be classified by physical properties of particles, such as size, density, or shape.

Another way of categorizing segregation is by energy input; vibration [8], gravity [9, 10], or

shear [11] segregation. Segregation can also be classified depending on particle movement

direction; vertical or horizontal. Finally, the device in which particle segregation occurs, such as

hoppers [9, 12-16], drums, and chutes, can be another classification method.

The most common method of classification of particle segregation is by the fundamental

mechanism which gives rise to segregation. As many as thirteen mechanisms by which particles

segregate have been identified and reviewed by researchers [17-21]. The thirteen mechanisms

summarized by de Silva et al. [19] are:

* Trajectory: caused by a greater speed reduction for smaller particles due to air drag

* Air current: fine particles are deposited at silo walls by air currents created by falling
particles

* Rolling: large or rounded particles roll down the surface of a particle heap in formation.









* Sieving: smaller particles flow downward through a sliding or rolling layer of larger
particles.

* Impact: a segregation mechanism where more bouncy particles are found further away
from the center of a heap in formation

* Embedding: larger or denser particles penetrate the surface layer of a heap and become
locked in position there

* Angle of repose: components with lower angle of repose flow more easily toward the
edges of a heap

* Push-away: lighter particles are pushed towards the edge of a heap by equally sized
heavier particles falling on the apex of the heap.

* Displacement: larger particles rise above smaller particles as a result of vibrations

* Percolation: smaller particles fall through void spaces between larger particles, sometimes
as a result of localized shear

* Fluidization: fine or lighter particles are kept fluidized at the surface of the particle
mixture

* Agglomeration: very fine particles form larger aggregates with greater mobility

* Concentration driven displacement: occurs in rotating devices where fine particles
concentrate in zones due to higher mobility

In addition to the mechanisms listed above, electrostatic interactions of particles may also

influence segregation. Of the thirteen mechanisms, percolation and sieving are hypothesized to

be the primary means by which segregation occurs. According to Samadini et al. [13],

segregation primarily occurs near the 'V'-shaped granular free surface where shearing of

particles occur. The other eleven mechanisms should be negligible since the particles used in the

present work are relatively large, free-flowing, and spherical with uniform characteristics. Also,

there is no projectile motion of particles during discharge from the hopper or inside the hopper.

2.2 Funnel Flow and Mass Flow

According to Tang and Puri [7], the terms funnel flow and mass flow were first developed

by Jenike in 1954. Funnel flow has a first-in, last-out flow pattern due to stagnant materials at the









hopper walls. Mass flow has a first-in, first-out flow pattern since materials at the hopper wall

are in motion. Sleppy and Puri [4] performed experiments with binary equal weight fraction

mixtures of granular sugar of size ratios, OD = 2 and OD = 5.7 for both funnel flow and mass

flow conditions. Their results showed that a mixture segregates more in funnel flow than in mass

flow. Markley and Puri [22] conducted similar experiments with identical particle size ratios and

observed that less segregation occurs for scaled up hoppers which have the same orifice

diameter. The generally accepted segregation pattern for funnel flow is that during the first half

of the discharge, fines are predominant and the second half of the discharge is fines-depleted [3,

17].

2.3 Fines Fraction

The concentration of fine particles is another factor that affects segregation. Typically,

more segregation occurs when the fines fraction is lower [7]. This is because there are more void

spaces that the number of fine particles that can move through. Arteaga and Tuzun [15]

developed a term called fines limiting fraction, XfL.

4
L 4+0D

This model is applicable for relatively large, free-flowing, nearly spherical particles. Above the

fines limiting fraction, very little or no segregation occurs. Research results from Sleppy and Puri

[4], and Kettergagen et al. [12] are in qualitatively agreement with this model.

2.4 Particle Properties that Influence Segregation

Properties of particles contributing to segregation include particle size, shape, density,

elasticity, cohesitivity, surface roughness, friction, and size ratio. However, some particle

properties show a more significant impact on segregation than others do. Particle diameter ratio,

density, and particle shape effects will be discussed in the following sections.









2.4.1 Particle Diameter Ratio

Most research on hopper flow segregation has focused on the effect of particle size ratio

since this is considered to be the most dominant factor. Standish [3] examined size segregation in

a Paul-Wurth hopper, a hopper with inclined, rotating pipe at the orifice which is typically used

to feed blast furnaces. Samadini et al. [13] studied segregation in the discharge of bidisperse

glass beads from a transparent, quasi-two-dimensional silo through an orifice. Results from both

of these studies showed particle segregation occurring for a particle size ratio as low as 1.2.

However, for lower size ratios, the extent of segregation was reduced. The results of Ketterhagen

et al. [12] with glass bead mixtures of D = 2 and OD = 4.3 show the same pattern. In contrast,

Sleppy and Puri [4] concluded that there was negligible disparity in the extent of segregation

between the OD = 2 and OD = 5.7 mixture of granular sugar. This may be due to the cubic shape

of the granular sugar used in their experiments.

2.4.2 Particle Density

There are no published works investigating density effect alone on segregation in hopper

discharge. Shi et al. [8] observed segregation trends with experiments in a vibrating glass

cylinder containing equally sized particles of different densities. Their results showed that the

less dense particles gradually formed a layer on top of the denser particles. The thickness of the

layer increases as the density ratio increased. Shinohara et al. [22] investigated segregation

patterns during the filling of a hopper. Denser components settle towards the center of the hopper

and are surrounded by lighter particles. Although the most dominant factor affecting segregation

is particle size ratio, the effect of density difference is still not negligible [23].

2.4.3 Particle Shape

There are no published works investigating shape effect alone on segregation in hopper

discharge. According to Tang and Puri [7], mixtures with different particle shapes are easier to









segregate than mixtures consisted of particles of the same shape. The greatest segregation occurs

when the coarse particles are angular and the fine particles are spherical. This could be due to the

increased void space formed from the irregularly shaped coarse particles or because irregularly

shaped particles may fit into interstitial void spaces more easily. However, they indicated that the

extent of segregation caused by particle shape is less than the effect of particle size ratio.

2.4.4 Particle Friction

The influence of surface roughness on particle segregation behavior has received little

attention in granular flow studies. In general, it has been assumed that surface roughness is an

order of magnitude less important in influencing segregation than particle size or density [25].

Pohlman et al. [25] studied segregation patterns of a binary mixture of smooth and rough

surfaced chrome steel beads in 2D and 3D rotating tumblers. Their experimental and simulation

results showed that surface roughness affected the angle of repose. However, surface roughness

did not cause radial segregation in 2D tumblers or banding in 3D tumblers.

2.5 Other Factors

Ketterhagen et al. [12] experimentally investigated the effect of initial fill conditions on

segregation in hopper discharge. The fill methods that were studied were well-mixed fill and

dual hopper fill. The dual hopper fill was achieved by discharging a well-mixed blend from a

mass flow hopper into a funnel flow hopper directly below. The results from the dual hopper fill

showed an extreme excess of fines towards the end of discharge such segregation behavior is

not observed in most well-mixed fill results.

Shah et al. [6] developed a testing method to predict the particle segregation potential

using a hopper manufactured by Jenike & Johanson, Inc. Their method was consisted of four

passes through the hopper and a pass through a 6 foot polycarbonate tube, approximately 1.5

inches in diameter.









Alexander et al. [26] filled a hopper with a mixture of glass beads or pharmaceutical

excipients. The hopper was then discharged into an identical hopper below, which is then

emptied into the first. Repeating this process several times resulted in an asymptotic segregation

profile significantly different from the segregation profile of the first discharge.









CHAPTER 3
EXPERIMENTAL METHODS

This chapter discusses the details of the experimental equipment and procedure used in the

research. In Section 3.1, the selection criteria and properties of the granular media will be

explained. The sifting segregation tester is then described in Section 3.2 followed by the details

of the antistatic bar in Section 3.3. Section 3.4 covers the procedure for conducting the

experiments. Finally, the methods used to collect and analyze the data are presented in Section

3.5.

3.1 Granular Media

Three types of granular media were used in the experiments. Glass beads and through

hardened steel shots were selected to investigate the segregation trends due to particle size ratio

and density difference. Crushed glass was used to investigate the effect of particle shape on

segregation patterns. The specific particle sizes were chosen so that some of the experimental

results can be directly compared with results from Ketterhagen et al. [12]. The sizes of the

particles comply with guidelines in the operating manual for the hopper as provided by the

manufacturer, Jenike & Johanson Inc. These guidelines indicate that the maximum particle size

should be limited to 1/8 inch and the total volume of materials to be tested should be at least

600mL.

The purchased particle mix was sieved to narrow the particle size distribution (PSD). Then

the particles were placed in an oven for 24 hours and then stored in a desiccator, prior to

experimentation, in order to keep them dry. This was done to prevent moisture on the surface of

the particles from influencing the experimental results.

Particle size information was measured by image analysis via microscope or by Coulter LS

13320 particle size analyzer.









Table 3-1. Summary of mean diameter and standard deviation of granular materials
Merial te Mean diameter Standard deviation RSD
Material type
d [mm] a [mm] [%1
2.26 0.13 5.75
Glass beads 1.06 0.06 5.66

0.542 0.045 8.30
2.35 0.01 0.43
Steel shots 1.13 0.045 3.98

0.538 0.044 8.18

Crushed glass 0.549 0.29 51.4
1.0


3.1.1 Glass Bead

Three sizes of glass beads listed in Table 3-1 (Union Process Inc., Akron, OH) with

density of p = 2.5 g/cm3 have been purchased. Photographs of the glass beads are shown in

Figure 3-1. The glass beads have a relatively uniform diameter and spherical shape.










Figure 3-1. Glass beads with mean diameter = (a) 0.542 + 0.045mm, (b) 1.06 + 0.06mm,
and (c) 2.26 + 0.13mm at 20 power magnification.

3.1.2 Steel Shot

Figure 3-2 shows photographs of through-hardened steel shots (Union Process Inc., Akron,

OH, p = 7.8 g/cm3) with similar sizes to those of glass beads. The steel shot PSD data are also

listed in Table 3-1. The steel shots also have a relatively uniform diameter and spherical shape.

Some rust is present on the surface of the steel shots.



















Figure 3-2. Steel shots with mean diameter = (a) 0.538 0.044mm, (b) 1.13 0.045mm,
and (c) 2.35 0.01mm at 20 power magnification.

3.1.3 Crushed Media

Crushed glass (Strategic Materials Inc., Houston, TX, p = 2.5 g/cm3) with sizes matching

those of = 1.06mm and = 0.524mm have been purchased. As shown in Figure 3-3, the

shapes of the particles are very angular and deviate from spherical shape.











Figure 3-3. Crushed glass with mean diameter = (a) 0.549 0.29 mm, (b) 1.13 0.045 mm
at 20 power magnification.

3.2 Sifting Segregation Tester

A hopper (Jenike & Johanson Inc., Tygnsboro MA) with dimensions specified by the

ASTM standard test for sifting segregation [27] was purchased for the experiments. The hopper

is made of clear acrylic and has a half angle of 550 with respect to vertical. Detailed dimensions

are shown in Figure 3-4. The slide gate in the outlet of the hopper can be opened and closed to

collect samples discontinuously.











12.5 cm I.D.


3.5 cmc

1.3 cm
(tf r--1- slide gate
2.5 cm:
I.D.



Figure 3-4. Schematic and dimensions of Sifting Segregation Tester [12]

3.3 Antistatic Bar

Glass beads, crushed glass and steel shots of smaller sizes become statically charged

during discharge. This causes some particles to adhere to the hopper walls after the bulk of the

material has discharged. Also, clustering of the particles occurs. To reduce this problem, an

antistatic bar (TAKK Industries Inc., Cincinnati, OH) was used. A picture of the antistatic bar is

shown in Figure 3-5. The antistatic bar is 3.5 inches in length and is mounted on the hopper

opening to emit a field of positive and negative ions that neutralize the static electricity inside the

hopper. The antistatic bar was easier to handle compared to the powder antistatic agent (Larostat

HTS 905S, BASF Corp.) which was also used in some experiments. There was less work

involved in filling the hopper and less potential for health hazards with the antistatic bar. In









addition, Larostat showed some adhesive behavior when it was applied to the particles. The

adhesive characteristic increased with the amount of Larostat applied.

















Figure 3-5. Antistatic bar with its power supply (left) and antistatic bar mounted on the hopper
(right)

3.4 Details of Procedure

This section describes the experimental procedure. The filling process is explained in

Section 3.4.1 followed by the hopper discharge and sample collecting procedure in section 3.4.2.

For each experiment the hopper was cleaned and leveled. Experiments were repeated 3 to 6 times

to minimize the effects of any regions in the initial fill that were not well mixed.

3.4.1 Filling

The well-mixed initial state was powused for all the experiments reported in this paper. This

filling method is harder to achieve compared to layered or dual hopper filling methods,

especially when there is a large diameter ratio between the fincollecting procedure and coarse particles. The

following procedure has been used to minimize segregation during filling and to obtain

repeatable results.

First, the appropriate masses of coarse and fine particles are prepared. Particles are added

in the hopper in 17-20 portions of approximately equivalent compositions. One portion contains









50mL of coarse particles and an appropriate amount of fine particles. The appropriate amount of

fine particles for each portion is the total mass of fines divided by the number of portions. A

portion of coarse and fine particles are then poured in a small plastic bag and are mixed by hand

to achieve a homogeneous state. The bag is then carefully emptied into the hopper. This process

is repeated for each of the remaining portions. It is important to minimize the amount of free fall

of particles and heap formation to reduce the amount of particle segregation prior to the initiation

of discharge. Disturbing the particles that are already filled in the hopper will also cause fine

particles to percolate towards the bottom of the hopper.

3.4.2 Discharging and Sampling

Continuous and discontinuous discharge and sampling methods were used in the

experiments. The continuous method is affected by opening the slide gate and collecting samples

in a train of boxes without interrupting the discharge. This method is more relevant to industrial

practices, but there is more particle loss compared to the discontinuous method. The

discontinuous method is performed by opening and closing the slide gate each time a 55mL

sample is collected. Approximately 18 samples are collected for each experiment. The

discontinuous method is specified by the ASTM standard practice and is easier to carry out

without loss of materials. Approximately 2-8% of the fine particles were lost for the continuous

method and less than 1.5% of the fines were lost for the discontinuous method. Standish and

Kilic [14] compared segregation results of continuous and discontinuous methods in a Paul-

Wurth hopper using sinter mixtures of eight different size ratios within the range of 0.25mm to

6.0mm. The two methods were found to produce identical results. Ketterhagen et al. [12] also

concluded that there was little difference in the segregation results between continuous and

discontinuous methods. Most of the experiments done in this thesis work are using the

discontinuous method due to the ease of its use.









3.5 Measurements and Data Analysis

For each sample that was collected, components were separated using a sieve. The weight

of each component was measured using a balance. The weight data were then converted to

volume fraction data, xi using the particle density. Finally, the initial fines fraction, xf was used

to calculate the normalized fines volume fraction, xi/Xf. The composition of the particle mix was

expressed in terms of volume because some experiments involve mixtures of materials with

different densities. Plots of normalized fines volume fraction as a function of cumulative volume

discharged were constructed. A normalized fines volume fraction xi/xf > 1 implies that a sample

is fines-rich and xi/xf < 1 denotes that a sample is fines-depleted. A normalized fines volume

fraction xi/xf = 1 means that there is no segregation and the sample is of the same composition as

the initial charge. The scatter bars represent the 95% confidence interval.









CHAPTER 4
RESULTS AND DISCUSSION

Table 4-1. Summary of experimental work
Particle 2 volume Discharge
Size ratio Particle 1 Particle 2
fraction (%) method

Glass beads (2 mm) Glass beads (0.5 mm) 5 Discontinuous

Discontinuous,
Glass beads (2 mm) Steel shots (0.5 mm) 5
4:1 Continuous

Steel shots (2 mm) Glass beads (0.5 mm) 5 Discontinuous

Steel shots (2 mm) Steel shots (0.5 mm) 2.5, 5, 10, 20, 50 Discontinuous

Discontinuous,
Glass beads (2 mm) Glass beads (1 mm)
Continuous

Glass beads (2 mm) Steel shots (1 mm) Discontinuous
2:1 5
Steel shots (2 mm) Steel shots (1 mm) Discontinuous

Glass beads (1 mm) Glass beads (0.5 mm) Discontinuous

Crushed glass (1 mm) Crushed glass (0.5 mm) Discontinuous

Discontinuous,
1:1 Glass beads (2 mm) Steel shots (2 mm) 50
Continuous



This chapter presents experimental results of the thesis work. Section 4.1 describes the

effect of various particle treatments done on a glass-glass mixture with
Next, section 4.2 validates the use of the discontinuous discharge method. The effect of particle

density on segregation is then covered in section 4.3 followed by the effect of fines fraction in

section 4.4. Then, section 4.5 discusses how particle size ratio affects the segregation pattern in

hopper discharge. Illustrated in section 4.6 is the effect of absolute particle size on segregation.









Section 4.7 then covers the experimental results from aspherical particle mixtures. Finally, a

direct comparison with experimental data from previous literature is made in section 4.8.

4.1 Effect of Various Particle Treatments

Table 4.2. Treatment methods applied to glass beads with (D = 2 (2 and 1mm), xf = 5%
Treatment Antistatic Washed before Discontinuous Continuous
Larostat
# bar experiments discharge method discharge method



2

3

4






The effect of various particle treatments have been investigated for glass-glass mixture

with DD = 2, xf = 5%. Treatment methods used on the particle mixture are listed in Table 4-2.

Figure 4-1 illustrates the experimental results for each treatment method. The segregation trend

is very consistent for all treatment methods: fines-rich at beginning and fines-depleted at the end

of discharge. This trend occurs because of the following reasons. As the hopper is discharged, a

'V'-shaped surface with an incline is formed. Flow down this incline causes fine particles to

percolate downward into the stagnant of slow moving material below while larger particles roll

down the incline to the hopper centerline. There is an excess of fine particles until the fines-

depleted material accumulated at the hopper center line is discharged. Near the end of discharge

the concentration of fines increases slightly as the material close to the hopper walls is

discharged.










The consistency of the segregation pattern indicates that the various treatment methods for

the particles do not significantly affect the results. Treatment 1 is the standard method for all the

experiments in this thesis work unless specified otherwise.

2.5

-Treatment 1
2.0 Treatment 2
r 2.0
.o --Treatment 3
S1.5 A T Tre.5ln'. nl 4
ST

M 1.0

I
S 0.5 ,


0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Discharged Volume Fraction

Figure 4-1. Experimental segregation results for variously treated glass-glass mixtures with D =
2 (2 and Imm) and xf = 5%

4.2 Continuous vs. Discontinuous Discharge

As mentioned in chapter 3, previous work [12, 14] suggests that the discharging and

sampling method has a small effect on the segregation pattern. This can be confirmed through

Figure 4-2 and 4-3. Figure 4-2 shows experimental results for glass(coarse)-steel(fine) system

with OD = 4, xf = 5%, and the two discharge methods. This mixture was chosen because of all the

mixtures investigated, it was predicted to show the greatest tendencies to segregate. The

segregation profiles show the same trend. There is some difference in magnitude, but it is not too

significant. Figure 4-3 shows segregation profiles for a glass-steel mixture with 50:50 volume

ratio and D = 1 and the two discharge methods. The y-axis in Figure 4-3 denotes the normalized

steel volume fraction since there are no fines for this experiment.















2.0 -continuous
C


LL 1.5



,1.0
i.
"oI

0.5
o
z

0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0


Discharged Volume Fraction


Figure 4-2. Glass(coarse)-steel(fine) mixture experimental results with D = 4, Xf
continuous or discontinuous discharge methods


2.5



2.0



L 1.5







0.5


0.0
0.0


5%, and


0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Discharged Volume Fraction


Figure 4-3. Segregation patterns for 50:50 glass-steel mixture OD:
discontinuous discharge


1 and continuous and


discontinuous
--continuous




^^^ -^^^*^^3 ^^










4.3 Effect of Particle Density

The result shown in Figure 4-3 indicates that in a mixture of equally sized particles, the

density difference will cause segregation. There is a slight abundance in steel shots until 90% of

the discharge, while the last 10% of the discharge is rich with glass beads. Heavier particles

moving downward more readily in the hopper than lighter particles is consistent with this

segregation profile.

2.5
Glass- Glass
S--Steel- Steel
2.0 -Glass- Steel(fine)
t- T .... I G"r


S1.5






0.0
0.5
z


0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Discharged Volume Fraction


Figure 4-4. Experimental results for mixtures consisted of several combination of materials with
OD = 4 and xf = 5%

However, when both the particle size and density are different, density became a secondary

factor for segregation. Figure 4-4 shows experimental results for various mixtures with OD = 4

and xf = 5%. The magnitude of segregation in the glass-steel (fine) and the steel-glass (fine) is

much greater than when density alone is the only variable influencing segregation (as in Figure

4-3). In fact, the density of the individual particles in the binary mix shows very little influence

on the resulting segregation patterns when the particle size ratio is 4:1. These results indicate that

the segregation behavior is dominated by geometric effects at this particle size ratio and that the










effect of particle density or differences in the surface friction between glass and steel are not

appreciable.

However, when the size ratio is reduced to 2:1, as shown in Figure 4-5, the qualitative

shape of the segregation profiles of the various mixtures is consistent, but some quantitative

differences in the segregation behavior are evident. These differences are most likely due to

those differences in particle density and/or surface friction between glass and steel these effects

begin to compete with geometric effects as the particle size ratio is reduced.

2.5
Glass- Glass
-- Steel- Steel
2.0 Glass- Steel(fine)


LL 1.5

i I T
1.0 -
,


NT
0.5


0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Discharged Volume Fraction


Figure 4-5. Experimental results for mixtures consisted of several combination of materials with
(DD= 2 and xf = 5%

4.4 Effect of Fines Fraction

It was concluded in previous experimental works [4, 7, 15] that increasing fines content

decreased the magnitude of particle segregation. They hypothesized that as the fines content

increases, the number of available void spaces for fines to percolate to decreases hence,

reducing the ability of the mixture to size segregate. These results were reproduced by

Ketterhagen et al. [12] in Figure 4-6. The segregation patterns for the fines fraction of 5% and










10% are not significantly different. However, as the fines fraction is increased to 20% and 50%,


there is noticeable decrease in segregation.


2.5

-*-5%
2.0 -e-10%
ro (I20%
S_ _50%
U" 1.5



S. 0


0.5
0
z
0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0


Discharged Volume Fraction


Figure 4-6. Experimental segregation results from Ketterhagen et al.
mixtures with DD = 4 and given fines fractions

2.5

T -*-2.5%
2.0 -5%


LL 1.5
1.


S0.5 ''



0

0 .0 -- -- --1.' -- .' .' .' --


[12] for glass-glass


0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Discharged Volume Fraction


Figure 4-7. Experimental segregation results for steel-steel mixtures with (D = 4 and xf = 2.5%,
5%, and 10%














2.0 zu7o
ST --A-50%
0

1.5


1.0



0.5 L 1

z

0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

Discharged Volume Fraction


Figure 4-8. Experimental segregation results for steel-steel mixtures with D = 4 and xf = 10%,
20%, and 50%

In this work, experiments were conducted with steel-steel mixtures with DD = 4 and

various fine fractions, including a 2.5% fines fraction (see Figures 4-7 and 4-8). As in the

previous work, the removal of available void spaces with fines fractions less than 10% does not

noticeably inhibit segregation. The segregation patterns for the 2.5%, 5% and 10% mixtures are

similar. However, as the fines concentration is increased above 10%, significant mitigation of

segregation is observed. The 50% fines mixture shows minimal segregation. These results further

substantiate the geometric argument for the effect of fines concentration on segregation.

4.5 Effect of Size Ratio

Figures 4-9, 4-10, and 4-11 show the segregation patterns for glass-glass, steel-steel, and

glass-steel mixtures respectively with xf = 5%. These results for spherical particles indicate that,

irrespective of material composition, increasing particle size ratio increases particle segregation.

The larger the particle size ratio, the larger is the size of each void space between the particles.












And, fines can much more easily percolate into a larger-sized void space between particles -


hence, a greater tendency for a mixture to size segregate.


2.5


-4:1 Glass- Glass
2.0
2.0 --2:1 (2 & 1 mm) Glass-Glass

.;
LL 1.5



.i 1.0
Li

.N
0.5

z
Z

0.0


0.0 0.1 0.2 0.3 0.4 0.5 0.6
Discharged Volume Fraction


Figure 4-9.


2.5 r


0.7 0.8 0.9 1.0


Glass-glass mixture experimental results with given size ratios and xf


1-, : --


LL 1.5





0
1.0
U-


0.5
z


0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6

Discharged Volume Fraction


0.7 0.8 0.9 1.0


Figure 4-10. Steel-steel mixture experimental results with given size ratios and xf = 5%














2.0 -2:1 (2 & 1 mm) Glass-Steel


L.







z
1.5


1.0
LI


0.5


0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Discharged Volume Fraction


Figure 4-11. Glass-steel mixture experimental results with given size ratios and xf = 5%

4.6 Effect of Absolute Particle Size

Figure 4-12 shows experimental segregation results for glass-glass mixtures with the same

size ratio but different absolute particle sizes. The 2mm/lmm mixture displays more segregation

than the 1/0.5mm mixture. Since the lmm/0.5mm mixture has a higher number of particles and a

lower absolute particle size, surface forces are more important than in the 2mm/lmm mixture.

More particles increases the number of particle-particle frictional contacts and smaller particles

increases cohesive effects. Increasing surface forces tend to inhibit particle migration and

mitigate segregation.

4.7 Effect of Particle Shape

Figure 4-13 shows segregation results with a mixture of two sizes of crushed glass

compared to spherical glass beads. A marked increase in the angle of the particle free surface

during discharge was observed with the crushed glass. Nevertheless, only a minor increase in

particle segregation is observed with the crushed glass. It could be that the angular nature of the












particles preferentially alters the shape of the void spaces such that, in some cases, the fines can


not easily move downward into those available spaces.


2.5

--2:1 (2 & 1 mm) Glass-Glass

2.0 --2:1 (1 & 0.5 mm) Glass-Glass

t;



0









0.0
n. 1.5









z

0 .0 I I I I I I I I I I


0.0 0.1 0.2 0.3 0.4 0.5 0.6

Discharged Volume Fraction


0.7 0.8 0.9 1.0


Figure 4-12. Glass-glass mixture experimental results with DD = 2 for different sized particles
and xf = 5%


.0

LL 1.5



3 1.0

I
N
0.5

z

0.0 I I I
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Discharged Volume Fraction


0.7 0.8 0.9 1.0


Figure 4-13. Experimental segregation results for crushed glass mixture and glass bead mixture
with OD = 2 and xf = 5%










4.8 Comparison with Previous Experimental Results

Figures 4-14, 4-15, and 4-16 show direct comparisons of the present experimental work

with the results obtained by Ketterhagen et al. [12]. For particle mixtures involving glass and

steel, the shape and magnitude of the segregation profiles are similar. In both sets of

experiments, there is a fines rich region in the first half of the discharge, followed by a fines-

depleted region.

However, when the results involving the glass/glass mixtures are compared, significant

differences in the shape of the segregation profile are observed. These differences are observed

at both the 2:1 and 4:1 particle size ratios. In the previous experimentation, after a fines-rich and

then fines-depleted region during the discharge, a fines rich region is again observed at the very

end of the discharge. This final fines-rich region is not observed in the present experiments. The

concentration of fines does increase at the very end of the discharge in the present experiments

but overall the composition of the mixture is still slightly fines-depleted.


2.5
-2:1 (2 & 1 mm) Glass- Steel(fine)
2.0 -*-2:1 (2 & 1 mm) Glass-Steel results from Ketterhagen et al (2007)


S1.5


1.0


= 0.5


0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Discharged Volume Fraction


Figure 4-14. Comparison of results for glass-steel mixtures with (DD= 2 and xf = 5%

















S2.0
.2





1.o
M
LL
1.5


U,
. 1.0



0.5
z


0.0 1
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Discharged Volume Fraction


0.7 0.8 0.9 1.0


Figure 4-15. Comparison of results for glass-glass mixtures with (D = 2 and Xf = 5%

2.5


2.0 -4:1 Glass- Glass results from Ketterh

t;

LL 1.5



1.0



0.5



0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Discharged Volume Fraction


0.7 0.8 0.9 1.0


Figure 4-16. Comparison of results for given mixtures with (D = 4 and Xf = 5%


Given that spherical particles with the same density and size were used in both sets of


experiments, the most likely cause of this difference between the two sets of results is


interparticle forces or particle-wall forces. However, it is unlikely that electrostatic interparticle


-2:1 (2 & 1 mm) Glass- Glass
--2:1 (2 &1 mm) Glass- Glass results from Ketterhagen et al (2007)









effects are responsible for the differences in the results. While the antistatic bar was used in the

current experimentation to reduce/eliminate electrostatic effects whereas the Larostat powder

was used previously, the results shown in Figure 4-1 indicate that the method of electrostatic

treatment does not play a significant role in the segregation behavior. It is more likely that

differences in particle-particle and particle-wall frictional forces are responsible for the

variations in the segregation behavior. The glass beads used in the two sets of experiments were

purchased from different suppliers and there could be variations in their surface characteristics.

The simulation results of Ketterhagen et al. [12] substantiate this hypothesis. Figures 4-17

and 4-18 show the predicted segregation profiles for discharge from a wedge-shaped hopper in

which the particle-wall friction coefficient (Figure 4-17) and the particle-particle friction

coefficient (Figure 4-18) were varied. Decreasing particle wall friction creates a more mass

flow-like behavior in the hopper, minimizing the tendency for fines to be retained in the hopper

until the very end of the discharge. With low particle-wall friction a fines-rich region at the end

of the discharge does not exist. Also, predictions from the simulation indicate that increases in

particle-particle friction produce the same effects as decreasing particle-wall friction. Increasing

particle-particle friction inhibits the percolation of fines through the mixture to the bottom of the

hopper. Hence, a fines-rich region is not present with increasing particle-particle friction.












S- A- =0.01 (N-29,600, n 3)
Lw = 0.15 (N=-29,600, n=3)
-w = 0.50 (N=29,600, n=3)


X
- 2.5


2


S1.5





0.5


0


4:]


0 0.2 0.4 0.6 0.8
Fractional Mass Discharged, M/MTot


Figure 4-17. Effect of varying particle-wall friction coefficient on predicted segregation profiles
for discharge from a wedge-shaped hopper with OD = 4 and xf = 5%


2

x
M

1.5









0.5




0
*^

"d


)i=0.00 (N=31,100,n=3)
_---- g=0.10 (N=29,600,n=3)
g- L=0.50 (N=31,100,n=3)













(b)

0 0.2 0.4 0.6 0.8
Fractional Mass Discharged, M/MTot


Figure 4-18. Effect of varying particle-particle friction coefficient on predicted segregation
profiles for discharge from a wedge-shaped hopper with DD = 4 and xf = 5%


P re_;' _- _









CHAPTER 5
CONCLUSION AND RECOMMENDATIONS

This work examined the segregation pattern of granular materials during discharge from a

hopper. The experiments were carried out in an ASTM standard Sifting Segregation Tester. The

effects of particle treatment, mode of discharge, concentration of fines, particle size and size

ratio, and particle density and shape on segregation were investigated. Various mixtures of glass

beads, steel shots, and crushed glass of several sizes were used. Based on the experimental

results, the following conclusions are made:

* The method of antistatic treatment or particle washing does not significantly influence the
segregation results.

* Segregation increases with increasing particle diameter ratio.

* In general, size-segregation patterns are initially fines-rich and then fines-depleted at the
end of discharge.

* Segregation patterns due to density effects are initially rich in particles with the higher
density and then rich in the particles with the lower density at the end of the discharge.

* The segregation profiles asymptote for fines fractions less than 10%. When xf > 10%,
mixtures with increased fines displays less segregation.

* Given the same particle size ratio, mixtures with smaller absolute particle sizes show less
segregation than mixtures with larger absolute particle sizes.

* The extent of segregation for mixtures of crushed glass-crushed glass is not significantly
different from spherical glass mixtures.

* The segregation behavior for spherical glass mixtures at the end of the discharge is
markedly different than the results reported by Ketterhagen et al. [12].

As discussed in chapter 4, the disparity between the segregation trends in this thesis work

and the work of Ketterhagen et al. [12] is most likely due to the surface effects such as friction.

In order to verify this, experiments need to be conducted which spherical glass beads from the

same supplier as in the work of Ketterhagen et al. [12] to check the reproducibility of their

results. The purpose of these experiments is to verify that there are no systematic differences in









procedure such as method of particle mixing, hopper filling, particle preparation (e.g. antistatic

treatment), etc. If the results from the previous study are reproduced, it is assumed that

differences in surface effects are the cause of the discrepancy between the two sets of data. In

this case, a surface treatment will then be applied to the two sets of glass beads from the two

suppliers. The proposed surface treatment will make the particles hydrophobic and should create

consistent surface characteristics and consistent segregation behavior between the two sets of

particles. If the results from the previous study are not reproduced, any differences in hopper

filling method, method of particle mixing, particle preparation and other particle properties (such

as sphericity) needs to be investigated in detail.

The effect of particle shape should be further explored. In the present work crushed glass

mixtures (2:1 size ratio) did not show much difference in segregation behavior from their

spherical mixture counterpart. However, previous studies of mixtures of non-spherical and

spherical particles did show a difference in segregation behavior from a purely spherical particle

mixture. Experiments with crushed glass mixtures of size ratio 4:1 should be conducted to see if

particle shape effects become more pronounced with increasing size ratio. Additional

experimentation with mixtures of non-spherical and spherical particles, varying size ratio, fines

concentration, and fines material, would give enhanced insight into the effect of particle shape.









LIST OF REFERENCES


[1] A. Rosato, F. Prinze, K. J. Standburg, and R. Svendsen. Why Brazil nuts are on
top: Size segregation of particulate matter by shaking, Physical Review Letter 58,
1038-1040 (1987)

[2] J. Bridgewater. Fundamental powder mixing mechanisms, Powder Technology
15, 215-236 (1976)

[3] N. Standish. Studies of size segregation in filling and emptying a hopper, Powder
Technology 45, 43-56 (1985)

[4] J. A. Sleppy and V.M. Puri. Size-segregation of granulated sugar during flow.
Transactions of the ASAE 39(4), 1433-1439 (1996)

[5] G. Barbosa-Canovas, J. Malave-Lopez, and M. Peleg. Segregation in food
powders, Biotechnology Progress 1(2), 140-146 (1985)

[6] K. R. Shah, S. I. Farag Badawy, M. M. Szemraj, D. B. Gray, and M. Hussain.
Assessment of segregation potential of powder blends, Pharmaceutical
Development and Technology 12, 457-462 (2007)

[7] P. Tang, V. M. Puri. Methods for Minimizing Segregation: A Review, Particle
Science and Technology 22, 321-337 (2004)

[8] Q. Shi, G. Sun, M. Hou, and K. Lu. Density-driven segregation in vertically
vibrated binary granular mixtures, Physical Review E. 75, 061302 (2007)

[9] K. Shinohara, Y. Idemitsu, K. Gotoh, and T. Tanaka. Mechanism of gravity flow
of particles from a hopper, I & EC Process Design and Development 7(3), 378-
383 (1968)

[10] J. Bangtang, M. Lim, C. Monterola, and C. Saloma. Gravity-assisted segregation
of granular materials of equal mass and size, Physical Review E. 66, 041306
(2002)

[11] K. Johanson, C. Eckert, D. Ghose, M. Djomlija, M. Hubert. Quantitative
measurement of particle segregation mechanisms, Powder Technology 159, 1-12
(2005)

[12] W. R. Ketterhagen, J. S. Curtis, C. R. Wassgren, A. Kong, P. J. Narayan, B. C.
Hancock. Granular segregation in discharging cylindrical hoppers: a discrete
element and experimental study, Chemical Engineering Science 62(22), 6423-
6439 (2007)

[13] A. Samadani, A. Pradhan, and A. Kudrolli. Size Segregation of Granular Matter
in Silo Discharges, Physical Review E. 60(6), 7203-7209 (1999)









[14] N. Standish and A Kilic. Comparison of Stop-Start and Continuous Sampling
Methods of Studying Segregation of Materials Discharging from a Hopper,
Chemical Engineering Science 40(11), 2152-2153 (1985)

[15] P. Arteaga and U. Tuzun. Flow of Binary Mixtures of Equal-Density Granules in
Hoppers Size Segregation, Flowing Density and Discharge Rates, Chemical
Engineering Science 45, 205-223 (1990)

[16] U. Tuzun and P. Arteaga. A Microstructural Model of Flowing Ternary Mixtures
of Equal-Density Granules in Hoppers, ChemicalEngineering Science 47, 1619-
1634(1992)

[17] J.C. Williams. The Segregation of Particulate Materials: A Review, Powder
Technology 15, 245-251 (1976)

[18] J. Bridgewater. Segregation mechanisms in condensed granular flow, IUTAM
Symposium on Segregation in Granular Flows pg. 11-29, Cape May, NJ, 2000.
Kluwer Academic Publishers

[19] S. de Silva, A. Dyroy, and G. G. Enstad. Segregation mechanisms and their
quantification using segregation testers, IUTAM Symposium on Segregation in
Granular Flows pg. 11-29, Cape May, NJ, 2000. Kluwer Academic Publishers

[20] J. K. Prescott and J. W. Carson. Analyzing and overcoming industrial blending
and segregation problems, IUTAM Symposium on Segregation in Granular Flows
pg. 11-29, Cape May, NJ, 2000. Kluwer Academic Publishers

[21] R. M. Nedderman, U. Tuzun, S. B. Savage, and G. T. Houlsby. The Flow of
Granular Materials Discharge Rates from Hoppers, Chemical Engineering
Science 37(11), 1597-1609 (1982)

[22] K. Shinohara and S. Miyata. Mechanism of density segregation of particles in
filling vessels, Ind. Eng. Chem. Process Des. Dev. 23(3), 423-428 (1984)

[23] K. Shinohara and B. Golman. Density segregation of a binary solids mixture
during batch operations in a two-dimensional hopper, Advanced Powder
Technology 14(3), 333-347 (2003)

[24] C. Markley & V. M. Puri. Scale-up effect on size-segregation of sugar during
flow, Trans. ASAE 41(5), 1469-1476 (1998)

[25] N. A. Pohlman, B. L. Severson, J. M. Ottino, R. M. Lueptow. Surface roughness
effects in granular matter: Influence on angle of repose and the absence of
segregation, Physical Review E. 73(3) 031304 (2006)

[26] A. Alexander, M. Roddy, D. Brone, J. Michaels, and F. J. Muzzio. A method to
quantitatively describe powder segregation during discharge from vessels,
Pharmaceutical Technology Yearbook pg. 6-21 (2000)









[27] ASTM standard practice for measuring sifting segregation tendencies of bulk
solids, D 6940-03, August 2003









BIOGRAPHICAL SKETCH

Byung-Hwan Chu was born in Seoul, Korea, in 1983. He enrolled at New Mexico State

University beginning in 2002. After receiving his Bachelor of Science degree in chemical

engineering in 2006, the author started his graduate study in chemical engineering at the

University of Florida. Upon receiving his Master of Science degree, the author plans to continue

his education to earn a Ph.D.





PAGE 1

1 EXPERIMENTAL INVESTIGATION OF PARTICLE SEGREGATION IN HOPPER DISCHARGE By BYUNG-HWAN CHU A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Byung-Hwan Chu

PAGE 3

3 To my family and friends

PAGE 4

4 ACKNOWLEDGMENTS I would lik e to express my deepest gratit ude to the many people who assisted and supported me through graduate school. I am very gr ateful to my advisor, Dr. Jennifer Curtis, for her guidance and encouragement. I am also th ankful to Dr. Carl Wassgren, Dr. Bruno Hancock, and Dr. Bill Ketterhagen for sharing their knowle dge in particle science through the weekly teleconference calls. I would also like to tha nk my supervisory committee member, Dr. Spyros Svoronos, and undergraduate assistant, Peter Kovedra. I would like to acknowledge my research group members, fellow students at the chemical engineering department, and frie nds in Gainesville for making my stay in Gainesville such a wonderful time. Their support, care, and friends hip made me a better engineer and a better person. Lastly, I thank my parents and brother for their support and unconditional love from home.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ..........7LIST OF FIGURES.........................................................................................................................8LIST OF ABBREVIATIONS........................................................................................................ 10ABSTRACT...................................................................................................................................11CHAPTER 1 INTRODUCTION..................................................................................................................122 LITERATURE REVIEW.......................................................................................................142.1 Classification of Segregation............................................................................................142.2 Funnel Flow and Mass Flow............................................................................................. 152.3 Fines Fraction............................................................................................................. ......162.4 Particle Properties that Influence Segregation.................................................................. 162.4.1 Particle Diameter Ratio.......................................................................................... 172.4.2 Particle Density...................................................................................................... 172.4.3 Particle Shape.........................................................................................................172.4.4 Particle Friction......................................................................................................182.5 Other Factors....................................................................................................................183 EXPERIMENTAL METHODS.............................................................................................203.1 Granular Media.................................................................................................................203.1.1 Glass Bead.............................................................................................................. 213.1.2 Steel Shot................................................................................................................213.1.3 Crushed Media........................................................................................................ 223.2 Sifting Segregation Tester................................................................................................ 223.3 Antistatic Bar....................................................................................................................233.4 Details of Procedure.........................................................................................................243.4.1 Filling.....................................................................................................................243.4.2 Discharging and Sampling.....................................................................................253.5 Measurements and Data Analysis..................................................................................... 264 RESULTS AND DISCUSSION............................................................................................. 274.1 Effect of Various Particle Treatments.............................................................................. 28

PAGE 6

6 4.2 Continuous vs. Discontinuous Discharge.........................................................................294.3 Effect of Particle Density................................................................................................. .314.4 Effect of Fines Fraction....................................................................................................324.5 Effect of Size Ratio....................................................................................................... ....344.6 Effect of Absolute Particle Size........................................................................................364.7 Effect of Particle Shape....................................................................................................364.8 Comparison with Previous Experimental Results............................................................ 385 CONCLUSION AND RECOMMENDATIONS................................................................... 42LIST OF REFERENCES...............................................................................................................44BIOGRAPHICAL SKETCH.........................................................................................................47

PAGE 7

7 LIST OF TABLES Table page 3-1 Summary of mean diameter and standard deviation of granular m aterials....................... 21 4-1 Summary of e xperim ental work......................................................................................... 27 4.2 Treatment methods applied to glass beads with D = 2 (2 and 1mm), xf = 5%................. 28

PAGE 8

8 LIST OF FIGURES Figure page 3-1 Glass beads with mean diameter < d> = (a) 0.542 0.045mm, (b) 1.06 0.06mm and (c) 2.26 0.13mm at 20 power magnification............................................................ 21 3-2 Steel shots with mean diameter < d> = (a) 0.538 0.044mm, (b) 1.13 0.045mm and (c) 2.35 0.01mm at 20 power magnification............................................................ 22 3-3 Crushed glass with mean diameter < d> = (a) 0.549 0.29 mm, (b) 1.13 0.045 mm at 20 power m agnification................................................................................................. 22 3-4 Schematic and dimensions of Sifting Segregation Tester [12] ..........................................23 3-5 Antistatic bar with its power supply (lef t) and antistatic bar mounted on the hopper (right) .................................................................................................................................24 4-1 Experimental segregation results for va riously treated glass-glass m ixtures with D = 2 (2 and 1mm) and xf = 5%................................................................................................29 4-2 Glass(coarse)-steel(fine) mixture experim ental results with D = 4, xf = 5%, and continuous or discontinuous discharge methods................................................................ 30 4-3 Segregation patterns fo r 50:50 glass-steel m ixture D = 1 and continuous and discontinuous discharge.....................................................................................................30 4-4 Experimental results for mixtures consiste d of several com bination of materials with D = 4 and xf = 5%............................................................................................................31 4-5 Experimental results for mixtures consiste d of several com bination of materials with D = 2 and xf = 5%............................................................................................................32 4-6 Experimental segregatio n results fro m Ketterhagen et al. [12] for glass-glass mixtures with D = 4 and given fines fractions................................................................. 33 4-7 Experimental segregation result s for steel-s teel mixtures with D = 4 and xf = 2.5%, 5%, and 10%......................................................................................................................33 4-8 Experimental segregation result s for steel-s teel mixtures with D = 4 and xf = 10%, 20%, and 50%....................................................................................................................34 4-9 Glass-glass mixture experimental results with given size ratios and xf = 5%................... 35 4-10 Steel-steel mixture experiment al results with given size ratios and xf = 5%..................... 35 4-11 Glass-steel mixture experiment al results with given size ratios and xf = 5%.................... 36

PAGE 9

9 4-12 Glass-glass mixture experimental results with D = 2 for different sized particles and xf = 5%...............................................................................................................................37 4-13 Experimental segregation results for cr ushed glass m ixture and glass bead mixture with D = 2 and xf = 5%....................................................................................................37 4-14 Comparison of results for glass-steel m ixtures with D = 2 and xf = 5%.......................... 38 4-15 Comparison of results for glass-glass m ixtures with D = 2 and xf = 5%......................... 39 4-16 Comparison of results for given mixtures with D = 4 and xf = 5%................................. 39 4-17 Effect of varying partic le -wall friction coefficient on predicted segregation profiles for discharge from a wedge-shaped hopper with D = 4 and xf = 5%............................... 41 4-18 Effect of varying particle-particle fr iction coefficient on predicted segregation profiles for discharge from a wedge-shaped hopper with D = 4 and xf = 5%.................. 41

PAGE 10

10 LIST OF ABBREVIATIONS d particle diameter [mm] standard deviation [mm] density [g/cm3] xi volume fraction of fines in a given sample [ ] xf volume fraction of fines initially in hopper [ ] xf,L fines limiting fraction [ ] xi/xf normalized fines volume fraction [ ] D particle diameter ratio [ ]

PAGE 11

11 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science EXPERIMENTAL INVESTIGATION OF PARTICLE SEGREGATION IN HOPPER DISCHARGE By Byung-Hwan Chu May 2008 Chair: Jennifer S. Curtis Major: Chemical Engineering Many industrial processes including food, mining, and pharmaceuticals involve handling mixtures of granular materials. The tendency of granular materials to segregate due to differences in particle properties such as, size, shape, and density negatively affects the process efficiency and product quality. Our study focused on segregation of partic les in hopper flow with the following conditions: well-mixed initial fill, funnel flow, and cylindrical hopper geometry. Glass beads and steel shots in three particle sizes each are used to investigate the effect of size ratio and density difference on particle segregation. Also, to study the effect of pa rticle shape on segregation of particles, crushed glass in two sizes is used. Results show that size-segregation patterns are initially fines-rich and then fines-depleted at the end of discharge. Segreg ation increases for larger particle size ratios and decreases for fines fractions that are above 10%. The extent of segregation for mixtures of crushed glasscrushed glass is not significantly different from spherical glass mixtures. Given the same size ratio, mixtures consisting of smaller particles show less segregati on. The segregation trend due to density effects is initially rich in particles with the higher density and then rich in the particles with the lower density at the end of the discharge.

PAGE 12

12 CHAPTER 1 INTRODUCTION A f ascinating characteristic of granular materials is their tendency to segregate due to difference in particle properties such as size, density, and shape. When a container of different liquids is shaken, they will mix, but in a blend of particles, they will segregate. A famous example is the Brazilian Nut Effect [1, 2] where a bed of larg e particles rise above smaller particles upon vibration. Segregation is of importance in a variety of industries that involve the handling, processing, or manufactur ing of particulate materials. Some examples are mining [3], food [4, 5], and pharmaceuticals such as in the pr oduction of pills and tablets [6]. Segregation can adversely affect the efficiency of processes, as well as the quality of final products which is dependent on maintaining a homogeneous blend. Segregation has been studied over many years by researchers from various disciplines due to its importance and ubiquitous occurrence. Despite the effo rts to explain this phenomenon through various methodologies such as experime ntation, simulation, and m odeling, there are still many unanswered questions. The segregation of particles in hopper discharg e is the focus of this thesis. Specifically, a series of controlled segregation experiments are conducted, changing one variable at a time. These experiments provide insight into the funda mental mechanisms influencing segregation of particles in gravity driven flow. These experime nts also serve as a basis for validation of analytical models and numerical si mulations of particle segregation. A hopper geometry is chosen for the segreg ation experiments since it is a standard geometry used in particle processes with many practical industrial ap plications. In addition, segregation experiments are relatively simple and straight forward in a hopper geometry. Furthermore, when segregation occurs in a la rge-scale particle pro cesses which include many

PAGE 13

13 devices such as mixers, fluidized beds, and hoppers it is important to characterize the amount of segregation that occurs in the hopper alone versus the other units. For example, if a perfectly mixed blend is fed to a hopper, how much segregation is to be expected from particle flow in that device? Following this introduction, chapter 2 will review literature that is relevant to the current work. Then in chapter 3 experimental material s and procedures will be explained. The results from the experiments are then presented and discussed in chapter 4. Chapter 5 summarizes the thesis work and makes recommendations for further studies.

PAGE 14

14 CHAPTER 2 LITERATURE REVIEW This chapter provides an overview of previous ex perim ental research that is relevant to the present study. Experiments related to segregation in hopper discharg e are the main focus of this review. Section 2.1 illustrates se veral classification methods of particle segregation. Section 2.2 reviews literature concerned with funnel fl ow and mass flow followed by section 2.3 which looks into previous work related to fines fraction. Then, in sec tion 2.4 literature about particle properties that influence segrega tion is discussed. Finally, secti on 2.5 reviews previous work that has investigated other factor s that affect segregation. 2.1 Classification of Segregation Tang and Puri [7] catego rized particle se gregation depending on several variables. Segregation may be classified by phy sical properties of particles, su ch as size, density, or shape. Another way of categorizing segregation is by en ergy input; vibration [8], gravity [9, 10], or shear [11] segregation. Segregation can also be classified depending on particle movement direction; vertical or horizontal. Finally, the device in which partic le segregation occurs, such as hoppers [9, 12-16], drums, and chutes, can be another classification method. The most common method of cl assification of particle segr egation is by the fundamental mechanism which gives rise to segregation. As many as thirteen mechanisms by which particles segregate have been identified and reviewed by researchers [17-21]. The thirteen mechanisms summarized by de Silva et al. [19] are: Trajectory : caused by a greater speed reduction for smaller particles due to air drag Air current : fine particles are deposited at silo walls by air currents created by falling particles Rolling : large or rounded particles roll down the su rface of a particle heap in formation.

PAGE 15

15 Sieving : smaller particles flow dow nward through a sliding or rolling layer of larger particles. Impact : a segregation mechanism where more bouncy particles are found further away from the center of a heap in formation Embedding : larger or denser particles penetrate the surface layer of a heap and become locked in position there Angle of repose : components with lower angle of repose flow more easily toward the edges of a heap Push-away : lighter particles are pushed towards the edge of a heap by equally sized heavier particles falling on the apex of the heap. Displacement : larger particles rise above smaller particles as a result of vibrations Percolation : smaller particles fall through void spaces between larger par ticles, sometimes as a result of localized shear Fluidization : fine or lighter particles are kept fl uidized at the surface of the particle mixture Agglomeration : very fine particles form larger aggregates with greater mobility Concentration driven displacement: occurs in rotating devices where fine particles concentrate in zones due to higher mobility In addition to the mechanisms listed above, elec trostatic interactions of particles may also influence segregation. Of the th irteen mechanisms, pe rcolation and sievin g are hypothesized to be the primary means by which segregation occurs. According to Samadini et al. [13], segregation primarily occurs near the V-sha ped granular free surf ace where shearing of particles occur. The other eleven mechanisms shoul d be negligible since the particles used in the present work are relatively large, free-flowing, a nd spherical with uniform characteristics. Also, there is no projectile motion of particles during discharge from the hopper or inside the hopper. 2.2 Funnel Flow and Mass Flow According to Tang and Puri [7], the term s f unnel flow and mass flow were first developed by Jenike in 1954. Funnel flow has a fi rst-in, last-out flow pattern due to stagnant materials at the

PAGE 16

16 hopper walls. Mass flow has a first-in, first-out flow pattern since materials at the hopper wall are in motion. Sleppy and Puri [4] performed ex periments with binary equal weight fraction mixtures of granular sugar of size ratios, D = 2 and D = 5.7 for both funnel flow and mass flow conditions. Their results showed that a mixtur e segregates more in funnel flow than in mass flow. Markley and Puri [22] conducted similar expe riments with identical particle size ratios and observed that less segregation occurs for scaled up hoppers which have the same orifice diameter. The generally accepted se gregation pattern for funnel flow is that during the first half of the discharge, fines are predominant and the se cond half of the discharge is fines-depleted [3, 17]. 2.3 Fines Fraction The concentration of fine particles is anothe r factor that affects segregation. Typically, more segregation occurs when the fines fraction is lower [7]. This is because there are more void spaces that the number of fine particles that can move through. Arteaga and Tuzun [15] developed a term called fines limiting fraction, xf,L. D Lf,4 4 x This model is applicable for rela tively large, free-flowing, nearly spherical partic les. Above the fines limiting fraction, very little or no segregati on occurs. Research results from Sleppy and Puri [4], and Kettergagen et al. [12] are in qualitatively agreement with this model. 2.4 Particle Properties that Influence Segregation Properties of particles contribu ting to segregation include pa rticle size, shape, density, elasticity, co hesitivity, surface roughness, fricti on, and size ratio. However, some particle properties show a more significant impact on segregation than others do. Particle diameter ratio, density, and particle shape effects will be discussed in the following sections.

PAGE 17

17 2.4.1 Particle Diameter Ratio Most research on hopper flow segregation has fo cused on the effect of particle size ratio since this is consider ed to be the most dominant factor. Standish [3] examined size segregation in a Paul-Wurth hopper, a hopper with in clined, rotating pipe at the or ifice which is typically used to feed blast furnaces. Samadini et al. [13] studied segregation in the discharge of bidisperse glass beads from a transparent, quasi-two-dimensional silo through an orif ice. Results from both of these studies showed partic le segregation occurring for a pa rticle size ratio as low as 1.2. However, for lower size ratios, th e extent of segregation was redu ced. The results of Ketterhagen et al. [12] with glass bead mixtures of D = 2 and D = 4.3 show the same pattern. In contrast, Sleppy and Puri [4] concluded that there was negligible disparity in the extent of segregation between the D = 2 and D = 5.7 mixture of granular sugar. Th is may be due to the cubic shape of the granular sugar used in their experiments. 2.4.2 Particle Density There are no published w orks investigating de nsity effect alone on segregation in hopper discharge. Shi et al. [8] observed segregation trends with experiments in a vibrating glass cylinder containing equally sized pa rticles of different densities. Their results showed that the less dense particles gradually form ed a layer on top of the denser particles. The thickness of the layer increases as the density ratio increased. Shinohara et al. [22] investigated segregation patterns during the fillin g of a hopper. Denser components settl e towards the center of the hopper and are surrounded by lighter partic les. Although the most dominant factor affecting segregation is particle size ratio, the eff ect of density difference is still not negligible [23]. 2.4.3 Particle Shape There are no published w orks investigating shape effect alone on segregation in hopper discharge. According to Tang and Pu ri [7], mixtures with different particle shapes are easier to

PAGE 18

18 segregate than mixtures consisted of particles of the same shape. The greatest segregation occurs when the coarse particles are angular and the fine particles are spherical. This could be due to the increased void space formed from the irregularly shaped coarse particles or because irregularly shaped particles may fit into interstitial void spaces more easily. However, they indicated that the extent of segregation caused by particle shape is less than the e ffect of particle size ratio. 2.4.4 Particle Friction The influence of surface roughness o n particle segregation behavior has received little attention in granular fl ow studies. In general, it has been assumed that surface roughness is an order of magnitude less important in influencing segregation than particle size or density [25]. Pohlman et al. [25] studied segregation patterns of a binary mixture of smooth and rough surfaced chrome steel beads in 2D and 3D rotati ng tumblers. Their experimental and simulation results showed that surface roughness affected the angle of repose. Ho wever, surface roughness did not cause radial segregation in 2D tumblers or banding in 3D tumblers. 2.5 Other Factors Ketterh agen et al. [12] experimentally investigated th e effect of initial fill conditions on segregation in hopper discharge. The fill methods that were studied were well-mixed fill and dual hopper fill. The dual hopper fill was achieved by discharging a well-mixed blend from a mass flow hopper into a funnel flow hopper directly below. Th e results from the dual hopper fill showed an extreme excess of fines towards the en d of discharge such segregation behavior is not observed in most well-mixed fill results. Shah et al. [6] developed a testing method to predict the particle se gregation potential using a hopper manufactured by Jenike & Johans on, Inc. Their method was consisted of four passes through the hopper and a pass through a 6 foot polycarbonate tube, approximately 1.5 inches in diameter.

PAGE 19

19 Alexander et al. [26] filled a hopper with a mixtur e of glass beads or pharmaceutical excipients. The hopper was then discharged into an identical hopper below, which is then emptied into the first. Repeating this process se veral times resulted in an asymptotic segregation profile significantly different from the se gregation profile of the first discharge.

PAGE 20

20 CHAPTER 3 EXPERIMENTAL METHODS This chapter discusses the details of the expe rim ental equipment and procedure used in the research. In Section 3.1, the selection criteria and properties of the granular media will be explained. The sifting segregation tester is then described in Section 3.2 followed by the details of the antistatic bar in Section 3.3. Secti on 3.4 covers the procedure for conducting the experiments. Finally, the methods used to collec t and analyze the data ar e presented in Section 3.5. 3.1 Granular Media Three types of granular m edia were used in the experiments. Glass beads and through hardened steel shots were selected to investigate the segregation tr ends due to particle size ratio and density difference. Crushed gl ass was used to investigate th e effect of particle shape on segregation patterns. The specific particle sizes were chosen so that some of the experimental results can be directly compared with results from Ketterhagen et al. [12]. The sizes of the particles comply with guidelines in the ope rating manual for the hopper as provided by the manufacturer, Jenike & Johanson Inc. These guidelines indicate that the maximum particle size should be limited to 1/8 inch and the total volume of materials to be tested should be at least 600mL. The purchased particle mix was sieved to narrow the particle size distribution (PSD). Then the particles were placed in an oven for 24 hours and then stored in a desiccator, prior to experimentation, in order to keep them dry. This was done to prevent mo isture on the surface of the particles from influenci ng the experimental results. Particle size information was measured by imag e analysis via microscope or by Coulter LS 13320 particle size analyzer.

PAGE 21

21 Table 3-1. Summary of mean diameter and st andard deviation of granular materials Material type Mean diameter d [ mm ] Standard deviation [ mm ] RSD [ % ] 2.26 0.13 5.75 1.06 0.06 5.66 Glass beads 0.542 0.045 8.30 2.35 0.01 0.43 1.13 0.045 3.98 Steel shots 0.538 0.044 8.18 0.549 0.29 51.4 Crushed glass 1.0 3.1.1 Glass Bead Three sizes of glass beads listed in Tabl e 3-1 (Union Process Inc., Akron, OH) with density of = 2.5 g/cm3 have been purchased. Photographs of the glass beads are shown in Figure 3-1. The glass beads have a relatively uniform diameter and spherical shape. Figure 3-1. Glass beads with mean diameter < d> = (a) 0.542 0.045mm, (b) 1.06 0.06mm, and (c) 2.26 0.13mm at 20 power magnification. 3.1.2 Steel Shot Figure 3-2 shows photographs of through-hardened steel shots (Union Process Inc., A kron, OH, = 7.8 g/cm3) with similar sizes to those of glass b eads. The steel shot PSD data are also listed in Table 3-1. The steel shot s also have a relatively uniform diameter and spherical shape. Some rust is present on the surface of the steel shots. (a) (b) (c)

PAGE 22

22 Figure 3-2. Steel shots with mean diameter < d> = (a) 0.538 0.044mm, (b) 1.13 0.045mm, and (c) 2.35 0.01mm at 20 power magnification. 3.1.3 Crushed Media Crushed glass (Strategic Ma terials Inc., Houston, TX, = 2.5 g/cm 3) with sizes matching those of = 1.06mm and < d> = 0.524mm have been purchased. As shown in Figure 3-3, the shapes of the particles are very angular and deviate from spherical shape. Figure 3-3. Crushed glass with mean diameter < d> = (a) 0.549 0.29 mm, (b) 1.13 0.045 mm at 20 power magnification. 3.2 Sifting Segregation Tester A hopper (Jenike & Johanson Inc., Tygnsboro MA) with dim ensions specified by the ASTM standard test for sifti ng segregation [27] was purchased for the experiments. The hopper is made of clear acrylic and has a half angle of 55 with respect to vertical. Detailed dimensions are shown in Figure 3-4. The slide gate in the outlet of the hopper can be opened and closed to collect samples discontinuously. (a) (c) (b) (a) (b)

PAGE 23

23 Figure 3-4. Schematic and dimensions of Sifting Segregation Tester [12] 3.3 Antistatic Bar Glass beads, crushed glass and steel shots of sm aller sizes become statically charged during discharge. This causes some particles to adhere to the hopper walls after the bulk of the material has discharged. Also, clustering of the particles occurs. To reduce this problem, an antistatic bar (TAKK Industries Inc., Cincinnati, OH) was used. A pict ure of the antistatic bar is shown in Figure 3-5. The antistatic bar is 3.5 inches in length and is mounted on the hopper opening to emit a field of positive and negative ions that neutralize the static electricity inside the hopper. The antistatic bar was easie r to handle compared to the pow der antistatic agent (Larostat HTS 905S, BASF Corp.) which was also used in some experiments. There was less work involved in filling the hopper and less potential fo r health hazards with th e antistatic bar. In

PAGE 24

24 addition, Larostat showed some adhesive behavi or when it was applied to the particles. The adhesive characteristic increased wi th the amount of Larostat applied. Figure 3-5. Antistatic bar with it s power supply (left) and antis tatic bar mounted on the hopper (right) 3.4 Details of Procedure This section describes the experim ental proc edure. The filling process is explained in Section 3.4.1 followed by the hopper discharge and sample collecting procedure in section 3.4.2. For each experiment the hopper was cleaned and leveled. Experiments were repeated 3 to 6 times to minimize the effects of any regions in the initial fill that were not well mixed. 3.4.1 Filling The well-m ixed initial state was used for all th e experiments reported in this paper. This filling method is harder to achieve compared to layered or dual hopper filling methods, especially when there is a large diameter rati o between the fine and coarse particles. The following procedure has been used to minimi ze segregation during filling and to obtain repeatable results. First, the appropriate masses of coarse and fi ne particles are prepar ed. Particles are added in the hopper in 17~20 portions of approximately equivalent compositions. One portion contains

PAGE 25

25 50mL of coarse particles and an appropriate amount of fine partic les. The appropriate amount of fine particles for each portion is the total mass of fines divided by the number of portions. A portion of coarse and fine particles are then pou red in a small plastic bag and are mixed by hand to achieve a homogeneous state. The bag is then carefully emptied into the hopper. This process is repeated for each of the remaining portions. It is important to minimize the amount of free fall of particles and heap formation to reduce the amount of particle segregation prior to the initiation of discharge. Disturbing the par ticles that are already filled in the hopper will also cause fine particles to percolate towa rds the bottom of the hopper. 3.4.2 Discharging and Sampling Continuous and discontinuous discharge a nd sampling m ethods were used in the experiments. The continuous method is affected by opening the slide gate and collecting samples in a train of boxes without interrupting the discharg e. This method is more relevant to industrial practices, but there is more particle lo ss compared to the discontinuous method. The discontinuous method is performed by opening a nd closing the slide gate each time a 55mL sample is collected. Approximately 18 samples are collected for each experiment. The discontinuous method is specified by the ASTM standard practice and is easier to carry out without loss of materials. Approximately 2~8% of the fine pa rticles were lost for the continuous method and less than 1.5% of the fines were lo st for the discontinuous method. Standish and Kilic [14] compared segregati on results of continuous and discontinuous methods in a PaulWurth hopper using sinter mixtures of eight different size ratios w ithin the range of 0.25mm to 6.0mm. The two methods were found to produce identical re sults. Ketterhagen et al. [12] also concluded that there was little difference in th e segregation results between continuous and discontinuous methods. Most of the experiment s done in this thesis work are using the discontinuous method due to the ease of its use.

PAGE 26

26 3.5 Measurements and Data Analysis For each sample that w as collected, components were separated using a sieve. The weight of each component was measured using a balance. The weight data were then converted to volume fraction data, xi using the particle density. Fina lly, the initial fines fraction, xf was used to calculate the normalized fines volume fraction, xi/xf. The composition of the particle mix was expressed in terms of volume because some expe riments involve mixtures of materials with different densities. Plots of normalized fines vol ume fraction as a function of cumulative volume discharged were constructed. A normalized fines volume fraction xi/xf > 1 implies that a sample is fines-rich and xi/xf < 1 denotes that a sample is fine s-depleted. A normalized fines volume fraction xi/xf = 1 means that there is no segregation and the sample is of the same composition as the initial charge. The scatter bars re present the 95% confidence interval.

PAGE 27

27 CHAPTER 4 RESULTS AND DISCUSSION Table 4-1. S ummary of experimental work Size ratio Particle 1 Particle 2 Particle 2 volume fraction (%) Discharge method Glass beads (2 mm) Glass beads (0.5 mm) 5 Discontinuous Glass beads (2 mm) Steel shots (0.5 mm) 5 Discontinuous, Continuous Steel shots (2 mm) Glass beads (0.5 mm) 5 Discontinuous 4:1 Steel shots (2 mm) Steel shots (0.5 mm) 2.5, 5, 10, 20, 50 Discontinuous Glass beads (2 mm) Glass beads (1 mm) Discontinuous, Continuous Glass beads (2 mm) Steel shots (1 mm) Discontinuous Steel shots (2 mm) Steel shots (1 mm) Discontinuous Glass beads (1 mm) Glass beads (0.5 mm) Discontinuous 2:1 Crushed glass (1 mm) Crushed glass (0.5 mm) 5 Discontinuous 1:1 Glass beads (2 mm) Steel shots (2 mm) 50 Discontinuous, Continuous This chapter presents experimental results of the thesis work. S ection 4.1 describes the effect of various particle treatments done on a glass-glass mixture with D = 2 and xf = 5%. Next, section 4.2 validates the use of the discontinuous discharge method. The effect of particle density on segregation is then covered in sectio n 4.3 followed by the effect of fines fraction in section 4.4. Then, section 4.5 discus ses how particle size ratio aff ects the segregation pattern in hopper discharge. Illustrated in section 4.6 is the effect of absolute particle size on segregation.

PAGE 28

28 Section 4.7 then covers the expe rimental results from aspherical particle mixtures. Finally, a direct comparison with experime ntal data from previous lite rature is made in section 4.8. 4.1 Effect of Various Particle Treatments Table 4.2. T reatment methods applied to glass beads with D = 2 (2 and 1mm), xf = 5% Treatment # Antistatic bar Larostat Washed before experiments Discontinuous discharge method Continuous discharge method 1 2 3 4 5 The effect of various particle treatments ha ve been investigated for glass-glass mixture with D = 2, xf = 5%. Treatment methods used on the pa rticle mixture are listed in Table 4-2. Figure 4-1 illustrates the experimental results for each treatment method. The segregation trend is very consistent for all treatme nt methods: fines-rich at beginni ng and fines-depleted at the end of discharge. This trend occurs because of the following reasons As the hopper is discharged, a V-shaped surface with an incline is formed. Flow down this incline causes fine particles to percolate downward into the sta gnant of slow moving material be low while larger particles roll down the incline to the hopper centerline. There is an excess of fine particles until the finesdepleted material accumulated at the hopper center line is discharg ed. Near the end of discharge the concentration of fines increases slightly as the material close to the hopper walls is discharged.

PAGE 29

29 The consistency of the segregat ion pattern indicates that the various treatment methods for the particles do not signif icantly affect the results. Treatment 1 is the standard method for all the experiments in this thesis wo rk unless specified otherwise. 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume fractionDischarged Volume Fraction Treatment 1 Treatment 2 Treatment 3 Treatment 4 Treatment 5Figure 4-1. Experimental segregat ion results for variously treate d glass-glass mixtures with D = 2 (2 and 1mm) and xf = 5% 4.2 Continuous vs. Discontinuous Discharge As m entioned in chapter 3, previous work [12, 14] suggests that the discharging and sampling method has a small effect on the segreg ation pattern. This can be confirmed through Figure 4-2 and 4-3. Figure 4-2 shows experimental results for glass(coarse)-steel(fine) system with D = 4, xf = 5%, and the two discharge methods. This mixture was chosen because of all the mixtures investigated, it was pr edicted to show the greatest tendencies to segregate. The segregation profiles show the same trend. There is some difference in magnitude, but it is not too significant. Figure 4-3 shows segr egation profiles for a glass-st eel mixture with 50:50 volume ratio and D = 1 and the two discharge methods. The y-ax is in Figure 4-3 de notes the normalized steel volume fraction since there ar e no fines for this experiment.

PAGE 30

30 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction discontinuous continuous Figure 4-2. Glass(coarse)-steel(fine) mixture experimental results with D = 4, xf = 5%, and continuous or discontinuous discharge methods 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Steel Volume FractionDischarged Volume Fraction discontinuous continuous Figure 4-3. Segregation patterns for 50:50 glass-steel mixture D = 1 and continuous and discontinuous discharge

PAGE 31

31 4.3 Effect of Particle Density The result shown in Figure 4-3 indicates that in a m ixture of equally sized particles, the density difference will cause segreg ation. There is a slight abunda nce in steel shots until 90% of the discharge, while the last 10 % of the discharge is rich with glass beads. H eavier particles moving downward more readily in the hopper than lighter particles is consistent with this segregation profile. 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction Glass-Glass Steel-Steel Glass-Steel(fine) Steel-Glass(fine) Figure 4-4. Experimental results for mixtures cons isted of several combination of materials with D = 4 and xf = 5% However, when both the particle size and dens ity are different, density became a secondary factor for segregation. Figure 4-4 shows experi mental results for various mixtures with D = 4 and xf = 5%. The magnitude of segregation in the gl ass-steel (fine) and the steel-glass (fine) is much greater than when density alone is the only variable influencing se gregation (as in Figure 4-3). In fact, the density of the individual particles in the binary mix shows very little influence on the resulting segregation patterns when the particle size ratio is 4:1. These results indicate that the segregation behavior is dominated by geometric effects at this particle size ratio and that the

PAGE 32

32 effect of particle density or differences in the surface friction between glass and steel are not appreciable. However, when the size ratio is reduced to 2:1, as shown in Figur e 4-5, the qualitative shape of the segregation profiles of the various mixtures is consistent, but some quantitative differences in the segregation behavior are evid ent. These differences are most likely due to those differences in particle density and/or surf ace friction between glass and steel these effects begin to compete with geometric effects as the particle size ratio is reduced. 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction Glass-Glass Steel-Steel Glass-Steel(fine) Figure 4-5. Experimental results for mixtures cons isted of several combination of materials with D = 2 and xf = 5% 4.4 Effect of Fines Fraction It was concluded in previous experim ental works [4, 7, 15] that increasing fines content decreased the magnitude of par ticle segregation. They hypothesi zed that as the fines content increases, the number of available void spaces for fines to percolate to decreases hence, reducing the ability of the mixt ure to size segregate. These results were reproduced by Ketterhagen et al. [12] in Figure 4-6. The segregation pa tterns for the fines fraction of 5% and

PAGE 33

33 10% are not significantly different. However, as the fines fracti on is increased to 20% and 50%, there is noticeable decrease in segregation. 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 5% 10% 20% 50% Figure 4-6. Experimental segregat ion results from Ketterhagen et al. [12] for glass-glass mixtures with D = 4 and given fines fractions 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 2.5% 5% 10% Figure 4-7. Experimental segregation results for steel-steel mixtures with D = 4 and xf = 2.5%, 5%, and 10%

PAGE 34

34 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 10% 20% 50% Figure 4-8. Experimental segregation results for steel-steel mixtures with D = 4 and xf = 10%, 20%, and 50% In this work, experiments were conduc ted with steel-steel mixtures with D = 4 and various fine fractions, including a 2.5% fines fraction (see Figur es 4-7 and 4-8). As in the previous work, the removal of available void spa ces with fines fractions less than 10% does not noticeably inhibit segregation. The segregation patterns for the 2.5%, 5% and 10% mixtures are similar. However, as the fines concentration is increased above 10%, significant mitigation of segregation is observed. The 50% fines mixture shows minimal segr egation. These results further substantiate the geometric argument for the eff ect of fines concentration on segregation. 4.5 Effect of Size Ratio Figures 4-9, 4-10, and 4-11 show the segregati on patterns for glass-glass, steel-steel, and glass-steel m ixtures respectively with xf = 5%. These results for sphe rical particles indicate that, irrespective of material composition, increasing part icle size ratio increases particle segregation. The larger the particle size ratio, the larger is the size of each void space be tween the particles.

PAGE 35

35 And, fines can much more easily percolate into a larger-sized void space between particles hence, a greater tendency for a mixture to size segregate. 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 4:1 Glass-Glass 2:1 (2 & 1 mm) Glass-Glass Figure 4-9. Glass-glass mixture experime ntal results with given size ratios and xf = 5% 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 4:1 Steel-Steel 2:1 (2 & 1 mm) Steel-Steel Figure 4-10. Steel-steel mixture experi mental results with given size ratios and xf = 5%

PAGE 36

36 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 4:1 Glass-Steel(fine) 2:1 (2 & 1 mm) Glass-Steel Figure 4-11. Glass-steel mi xture experimental results with given size ratios and xf = 5% 4.6 Effect of Absolute Particle Size Figure 4-12 shows experim ental segregation resu lts for glass-glass mixtures with the same size ratio but different absolute particle sizes. The 2mm/1mm mixt ure displays more segregation than the 1/0.5mm mixture. Sin ce the 1mm/0.5mm mixture has a highe r number of particles and a lower absolute particle size, su rface forces are more important than in the 2mm/1mm mixture. More particles increases the number of particle-p article frictional contacts and smaller particles increases cohesive effects. Increasing surface forces tend to inhibit particle migration and mitigate segregation. 4.7 Effect of Particle Shape Figure 4-13 shows segregation results with a m ixture of two sizes of crushed glass compared to spherical glass beads. A marked in crease in the angle of th e particle free surface during discharge was observed with the crushed glass. Nevertheless, only a minor increase in particle segregation is observed with the crushed glass. It could be that the angular nature of the

PAGE 37

37 particles preferentially al ters the shape of the void spaces such that, in some cases, the fines can not easily move downward into those available spaces. 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 2:1 (2 & 1 mm) Glass-Glass 2:1 (1 & 0.5 mm) Glass-Glass Figure 4-12. Glass-glass mixtur e experimental results with D = 2 for different sized particles and xf = 5% 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 2:1 (1 & 0.5 mm) Crushed Glass-Crushed Glass 2:1 (1 & 0.5 mm) Glass-Glass Figure 4-13. Experimental segregation results fo r crushed glass mixture and glass bead mixture with D = 2 and xf = 5%

PAGE 38

38 4.8 Comparison with Previous Experimental Results Figures 4-14, 4-15, and 4-16 show direct com p arisons of the present experimental work with the results obt ained by Ketterhagen et al. [12]. For particle mixt ures involving glass and steel, the shape and magnitude of the segrega tion profiles are similar. In both sets of experiments, there is a fines rich region in the first half of the discharge, followed by a finesdepleted region. However, when the results involving the gla ss/glass mixtures are compared, significant differences in the shape of the segregation profile are observed. These differences are observed at both the 2:1 and 4:1 particle si ze ratios. In the previous experi mentation, after a fines-rich and then fines-depleted region during the discharge, a fines rich regi on is again observed at the very end of the discharge. This final fines-rich regi on is not observed in the present experiments. The concentration of fines does increas e at the very end of the discha rge in the present experiments but overall the composition of the mixtur e is still slightly fines-depleted. 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 2:1 (2 & 1 mm) Glass-Steel(fine) 2:1 (2 & 1 mm) Glass-Steel results from Ketterhagen et al ( 2007) Figure 4-14. Comparison of results for glass-steel mixtures with D = 2 and xf = 5%

PAGE 39

39 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 2:1 (2 & 1 mm) Glass-Glass 2:1 (2 &1 mm) Glass-Glass results from Ketterhagen et al (2007) Figure 4-15. Comparison of results for glass-glass mixtures with D = 2 and xf = 5% 0.0 0.5 1.0 1.5 2.0 2.5 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Normalized Fines Volume FractionDischarged Volume Fraction 4:1 Glass-Glass 4:1 Glass-Glass results from Ketterhagen et al ( 2007) Figure 4-16. Comparison of results for given mixtures with D = 4 and xf = 5% Given that spherical particles with the same density and size were used in both sets of experiments, the most likely cause of this difference between the two sets of results is interparticle forces or particle-wall forces. However, it is unlikely that electrostatic interparticle

PAGE 40

40 effects are responsible for the differences in the results. While the antistatic bar was used in the current experimentation to reduce/eliminate elect rostatic effects whereas the Larostat powder was used previously, the results shown in Figur e 4-1 indicate that the method of electrostatic treatment does not play a significant role in the segregation behavior. It is more likely that differences in particle-particle and particle-w all frictional forces are responsible for the variations in the segregation beha vior. The glass beads used in th e two sets of experiments were purchased from different suppliers and there could be vari ations in their surface characteristics. The simulation results of Ketterhagen et al. [12] substantiate this hypothesis. Figures 4-17 and 4-18 show the predicted segregation profil es for discharge from a wedge-shaped hopper in which the particle-wall friction coefficient (F igure 4-17) and the part icle-particle friction coefficient (Figure 4-18) were varied. Decreas ing particle wall friction creates a more mass flow-like behavior in the hopper, minimizing the tendency for fines to be retained in the hopper until the very end of the discharge. With low part icle-wall friction a fines-rich region at the end of the discharge does not exist. Also, predictions from the simulation indicate that increases in particle-particle friction produce th e same effects as decreasing particle-wall friction. Increasing particle-particle friction inhibits the percolation of fines through the mixture to the bottom of the hopper. Hence, a fines-rich region is not pres ent with increasing particle-particle friction.

PAGE 41

41 Figure 4-17. Effect of varying particle-wall friction coefficient on predicted segregation profiles for discharge from a wedge-shaped hopper with D = 4 and xf = 5% Figure 4-18. Effect of varying particle-particle friction coefficient on predicted segregation profiles for discharge from a wedge-shaped hopper with D = 4 and xf = 5%

PAGE 42

42 CHAPTER 5 CONCLUSION AND RECOMMENDATIONS This work exam ined the segregation pattern of granular materials during discharge from a hopper. The experiments were carried out in an ASTM standard Sifting Segregation Tester. The effects of particle treatment, m ode of discharge, concentration of fines, particle size and size ratio, and particle density and shape on segregation were investigated. Various mixtures of glass beads, steel shots, and crushed glass of seve ral sizes were used. Based on the experimental results, the following conclusions are made: The method of antistatic treatment or particle washing does not significantly influence the segregation results. Segregation increases with increasing particle diameter ratio. In general, size-segregation patte rns are initially fines-rich and then fines-depleted at the end of discharge. Segregation patterns due to dens ity effects are initially rich in particles with the higher density and then rich in the particles with th e lower density at the end of the discharge. The segregation profiles asymptote for fi nes fractions less than 10%. When xf 10%, mixtures with increased fines displays less segregation. Given the same particle size ratio, mixtures wi th smaller absolute particle sizes show less segregation than mixtures with larger absolute particle sizes. The extent of segregation for mixtures of crushed glass-crus hed glass is not significantly different from spherical glass mixtures. The segregation behavior for spherical glass mixtures at the end of the discharge is markedly different than the results reported by Ketterhagen et al. [12]. As discussed in chapter 4, the di sparity between the segregation trends in this thesis work and the work of Ketterhagen et al. [12] is most likely due to th e surface effects such as friction. In order to verify this, experiments need to be conducted which spherical glass beads from the same supplier as in the work of Ketterhagen et al. [12] to check the reproducibility of their results. The purpose of these experiments is to verify that there are no systematic differences in

PAGE 43

43 procedure such as method of particle mixing, hoppe r filling, particle prep aration (e.g. antistatic treatment), etc. If the results from the previous study are reproduced, it is assumed that differences in surface effects are the cause of the discrepancy between the two sets of data. In this case, a surface treatment will then be applie d to the two sets of glass beads from the two suppliers. The proposed surface treatment will make the particles hydrophobic and should create consistent surface characteristics and consistent segregation behavior between the two sets of particles. If the resu lts from the previous study are not reproduced, any differences in hopper filling method, method of particle mi xing, particle preparation and ot her particle properties (such as sphericity) needs to be investigated in detail. The effect of particle shape should be further explored. In the present work crushed glass mixtures (2:1 size ratio) did not show much difference in segregati on behavior from their spherical mixture counterpart. However, prev ious studies of mixtures of non-spherical and spherical particles did show a difference in segreg ation behavior from a pur ely spherical particle mixture. Experiments with crushed glass mixtures of size ratio 4:1 should be conducted to see if particle shape effects become more pronounced with increasing size ratio. Additional experimentation with mixtures of non-spherical and spherical particles, varying size ratio, fines concentration, and fines material, would give enhanced insight into the effect of particle shape.

PAGE 44

44 LIST OF REFERENCES [1] A. Rosato, F. Prinze, K. J. Standburg, and R. Svendsen. Why Brazil nuts are on top: Size segregation of particulate matter by shaking, Physical Review Letter 58, 1038-1040 (1987) [2] J. Bridgewater. Fundamental powder mixing mechanisms, Powder Technology 15, 215-236 (1976) [3] N. Standish. Studies of size segregation in filling and emptying a hopper, Powder Technology 45, 43-56 (1985) [4] J. A. Sleppy and V.M. Puri. Size-segregation of granulated sugar during flow. Transactions of the ASAE 39(4), 1433-1439 (1996) [5] G. Barbosa-Canovas, J. Malave-Lopez, and M. Peleg. Segregation in food powders, Biotechnology Progress 1(2), 140-146 (1985) [6] K. R. Shah, S. I. Farag Badawy, M. M. Szemraj, D. B. Gray, and M. Hussain. Assessment of segregation potential of powder blends, Pharmaceutical Development and Technology 12, 457-462 (2007) [7] P. Tang, V. M. Puri. Methods fo r Minimizing Segregation: A Review, Particle Science and Technology 22, 321-337 (2004) [8] Q. Shi, G. Sun, M. Hou, and K. Lu. Density-driven segregation in vertically vibrated binary gr anular mixtures, Physical Review E 75, 061302 (2007) [9] K. Shinohara, Y. Idemitsu, K. Gotoh, and T. Tanaka. Mechanism of gravity flow of particles from a hopper, I & EC Process Design and Development 7(3), 378383 (1968) [10] J. Bangtang, M. Lim, C. Monterola, and C. Saloma. Gravity-assisted segregation of granular materials of equal mass and size, Physical Review E 66, 041306 (2002) [11] K. Johanson, C. Eckert, D. Ghose, M. Djomlija, M. Hubert. Quantitative measurement of particle segregation mechanisms, Powder Technology 159 1-12 (2005) [12] W. R. Ketterhagen, J. S. Curtis, C. R. Wassgren, A. Kong, P. J. Narayan, B. C. Hancock. Granular segregation in disc harging cylindrical hoppers: a discrete element and experimental study, Chemical Engineering Science 62(22), 64236439 (2007) [13] A. Samadani, A. Pradhan, and A. Kudrol li. Size Segregation of Granular Matter in Silo Discharges, Physical Review E 60(6), 7203-7209 (1999)

PAGE 45

45 [14] N. Standish and A Kilic. Compar ison of Stop-Start and Continuous Sampling Methods of Studying Segregation of Ma terials Discharging from a Hopper, Chemical Engineering Science 40(11) 2152-2153 (1985) [15] P. Arteaga and U. Tuzun. Flow of Bina ry Mixtures of Equal-Density Granules in Hoppers Size Segregation, Flowin g Density and Discharge Rates, Chemical Engineering Science 45, 205-223 (1990) [16] U. Tuzun and P. Arteaga. A Microstr uctural Model of Flowing Ternary Mixtures of Equal-Density Granules in Hoppers, Chemical Engineering Science 47, 16191634 (1992) [17] J.C. Williams. The Segregation of Particulate Materials: A Review, Powder Technology 15, 245-251 (1976) [18] J. Bridgewater. Segregation m echanisms in condensed granular flow, IUTAM Symposium on Segregation in Granular Flows pg. 11-29, Cape May, NJ, 2000. Kluwer Academic Publishers [19] S. de Silva, A. Dyroy, and G. G. Enstad. Segregation mechanisms and their quantification using se gregation testers, IUTAM Symposium on Segregation in Granular Flows pg. 11-29, Cape May, NJ, 2000. Kluwer Academic Publishers [20] J. K. Prescott and J. W. Carson. Analyzing and overcoming industrial blending and segregation problems, IUTAM Symposium on Segregation in Granular Flows pg. 11-29, Cape May, NJ, 2000. Kluwer Academic Publishers [21] R. M. Nedderman, U. Tuzun, S. B. Sa vage, and G. T. Houlsby. The Flow of Granular Materials Discha rge Rates from Hoppers, Chemical Engineering Science 37(11) 1597-1609 (1982) [22] K. Shinohara and S. Miyata. Mechanis m of density segregation of particles in filling vessels, Ind. Eng. Chem. Process Des. Dev. 23(3) 423-428 (1984) [23] K. Shinohara and B. Golman. Densit y segregation of a binary solids mixture during batch operations in a two-dimensional hopper, Advanced Powder Technology 14(3), 333-347 (2003) [24] C. Markley & V. M. Puri. Scale-up effect on size-segregation of sugar during flow, Trans. ASAE 41(5) 1469-1476 (1998) [25] N. A. Pohlman, B. L. Severson, J. M. Ottino, R. M. Lueptow. Surface roughness effects in granular matter: Influence on angle of repose and the absence of segregation, Physical Review E 73(3) 031304 (2006) [26] A. Alexander, M. Roddy, D. Brone, J. Michaels, and F. J. Muzzio. A method to quantitatively describe powder segreg ation during discharge from vessels, Pharmaceutical Technology Yearbook pg. 6-21 (2000)

PAGE 46

46 [27] ASTM standard practice for measur ing sifting segregation tendencies of bulk solids, D 6940-03, August 2003

PAGE 47

47 BIOGRAPHICAL SKETCH Byung-Hwa n Chu was born in Seoul, Korea, in 1983. He enrolled at New Mexico State University beginning in 2002. After receiving hi s Bachelor of Science degree in chemical engineering in 2006, the author started his graduate study in chemical engineering at the University of Florida. Upon receiving his Master of Science degree, the author plans to continue his education to earn a Ph.D.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101204_AAAAAB INGEST_TIME 2010-12-04T05:31:04Z PACKAGE UFE0021894_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 25271604 DFID F20101204_AAABYP ORIGIN DEPOSITOR PATH chu_b_Page_39.tif GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
1cee883a0f09a1c219a0bfe6e76b0f21
SHA-1
acd93073923175117b01a417c4457b2696f6945d
1802 F20101204_AAACAP chu_b_Page_31.txt
781cb98283ed4efaf7a561237cb39763
2d1c58998d2e76da2f549e796acad1c1f1cf738c
23997 F20101204_AAABTS chu_b_Page_14.QC.jpg
bc302508589bb95213b1b71e2049dbf4
928b7bc7f4fb351c3050f7f88e1811ca8aa6af88
F20101204_AAABYQ chu_b_Page_40.tif
0f9e26b7ea5beeaf4cb49854948ee4d7
9014e6fca54299b08606aeced4a40bf7b121d29e
1787 F20101204_AAACAQ chu_b_Page_32.txt
4b6bf5c874c81a07c1860f238d9b6c16
0bc1280b57427403eb5a84a7b77539172f03a760
2183 F20101204_AAABTT chu_b_Page_25.txt
c5dd69a0a5a1b73d70d44f0f91618ae2
33e959d676abf781c1a30d58bb028aabb502faa3
F20101204_AAABYR chu_b_Page_41.tif
bb548d9684f757b3f455beeb8e57b80d
a06984048fd1cb848059047f989f2d1a5c84a206
1216 F20101204_AAACAR chu_b_Page_33.txt
157ec45c657042ef2f40d8c08de83778
bf454805eae961601353e18f49d3e46c36a2945d
13641 F20101204_AAABTU chu_b_Page_41.QC.jpg
cd5bd11cf4eae6f51dbfc809fb3da7af
1e40cd31428a3d22e52e361d08876f7140476cb3
F20101204_AAABYS chu_b_Page_42.tif
f1d7105449e0b48f5e506e1f504993f5
64fee7f513d51ad3152b196d5117541c4f739b58
1681 F20101204_AAACAS chu_b_Page_34.txt
66ebf1a11336866d7a0c96565e1fc4c3
645aa317027f70d2035c65756387ffdb2f7c3146
1028802 F20101204_AAABTV chu_b_Page_20.jp2
a37202acf12c1a213b892a25191b57a1
34e09dc327e459f44923f3656a163b62ab650d40
F20101204_AAABYT chu_b_Page_43.tif
e76cb419470fa895bb7d9037d6f778b9
f1fe509f6ca1bde12e4dd9c062b42aff9cc06c64
1072 F20101204_AAACAT chu_b_Page_35.txt
dd33caaa8c1cdd5ab355231e5662b5bb
fa7df15ce37d74618528bdf9a4a15ff2f2c4c286
10031 F20101204_AAABTW chu_b_Page_03.jpg
39d4695ea7cd896351048a8d415ee22a
21faff52ada058fce373e8204395645e2edb1107
F20101204_AAABYU chu_b_Page_47.tif
0defd5f7b4f4ed6bffc33d4f0955126d
67ca03c9911c32591c8f2d8df543a8bf2a8c9920
1198 F20101204_AAACAU chu_b_Page_37.txt
28db3895436888a8f3a8bc0679bdaeac
b9d57497fb786a70fbf694dc8937749bd34f20c6
18481 F20101204_AAABTX chu_b_Page_33.QC.jpg
5b5cbb9a3e74c0dc8042461d2064c71f
f05d28bd88c198c7d0009b521261dcd25416d17c
7726 F20101204_AAABYV chu_b_Page_01.pro
021c68325ae1c1513cc6e48bea2818da
748c9bdd4769648a1faa327d2a60565355f46607
1657 F20101204_AAACAV chu_b_Page_38.txt
dee1f89db857fda47b9db118492f3e23
0f2868ad75e3ba40f738a792386bb526a27e9345
6947 F20101204_AAABTY chu_b_Page_16thm.jpg
4dfa3e1ef15f7b854fa21373e45238a1
ba04a1cb78f79e40752567653bbfc7808cd44b44
906 F20101204_AAABYW chu_b_Page_03.pro
bc9f9991b8a13483ff917213a4abc18b
eb9dd957ac22d0f90de126c6cd6b0e5c3f523557
1709 F20101204_AAACAW chu_b_Page_40.txt
47a1602bbe7d12d659b667a5817b214f
81d7ebeb623fbae981d2dfd20578fb2ecee0b0c9
6841 F20101204_AAABRA chu_b_Page_44thm.jpg
a31f601488dc859ba841a080928bf717
bd036fe985fc02d718784ef65ac9abfe30783f2d
F20101204_AAABTZ chu_b_Page_01.tif
746e5ced51fb9f4f8a36be3f6f81cec3
34f2a0dfe12ddafa2faaed9de48c45793dd20d77
23666 F20101204_AAABYX chu_b_Page_04.pro
d7f9f65fa92f58e67eefd5f10a4b5850
e8eca1ce0504c6bd79ac4ed0dee35a69f8234c81
988 F20101204_AAACAX chu_b_Page_41.txt
d749e15bb5103610bb11faf6e728bbde
6aeac396d7c401edebe532b36d6eee15cfb091cf
23144 F20101204_AAABRB chu_b_Page_29.QC.jpg
ec66a9e37aacd42236220c8fb5535f4b
3b3be35cdbe16ce26e6acb0c549e8f183b1e1e21
88649 F20101204_AAABYY chu_b_Page_05.pro
4a6ce9730929b3f0999b18b403768329
e58780e7db11e2f59bb0b8f2145117f4f680888e
2339 F20101204_AAACAY chu_b_Page_42.txt
719675368da427c0bf01aad2e1e5537b
abcba15910cd1fed2c759fc2fbe039d4df69fce7
6564 F20101204_AAABRC chu_b_Page_10.QC.jpg
3953e67ab96bd379b36d100b81a55b0b
31c38e39424a9fe6880b71156ebeaad1d3d70cd9
81801 F20101204_AAABWA chu_b_Page_18.jpg
e10a5361384627265f25baf2444e1a5d
4e1128419f237070f987f3521361567506724bd7
30626 F20101204_AAABYZ chu_b_Page_06.pro
f944028eb50a17881ead8cf3e0763373
ee6269bbed951412f89b94654ed6b410301a8509
1639 F20101204_AAACAZ chu_b_Page_43.txt
b1cd099156ed85ca3dd5722630e85d20
b2ffb55478ebb84d8c024b3fc3a99a6ef74fbaf3
683 F20101204_AAABRD chu_b_Page_13.txt
9f22984658f4c89c53126320db1d6c39
36423a293ca9c52f56f1d16728c87f93625ad283
74523 F20101204_AAABWB chu_b_Page_20.jpg
e5533fa46ebc9816aa8acf3fdd43805f
5b3d7b7ffdc81b8cc335855c81fafb921616ba85
F20101204_AAABRE chu_b_Page_37.tif
d5507d461689f1e15b64b7dbbefcee17
7b31869a798906358f31892e6cbcbb8681e3142e
71782 F20101204_AAABWC chu_b_Page_24.jpg
6d2a5c2b4f6338eca248bca34955ff69
63928f6ebc3ab1bc2d4fad1c49ad1e750d2f4a50
90 F20101204_AAABRF chu_b_Page_02.txt
81710c1ab4698b5ebd172ccac80dc23b
86b00c92799c51620ae4010ea1b043e9fab34ed5
26481 F20101204_AAACDA chu_b_Page_42.QC.jpg
1511f9ff7562984ee6db0dcecbfef3a8
54be2be661bf1eaf119ff31e2964c1b22c655e73
67866 F20101204_AAABWD chu_b_Page_27.jpg
98fb617ea7215a7edf0b4fd2cc80a595
eb1eebbdf3b1d923514726507adf86facc1344f1
20421 F20101204_AAABRG chu_b_Page_10.jpg
e95c59b798ff4522c08fc372998af46c
4ab7e46f87d0a6723d7e73dff3fd057005ec0a41
68259 F20101204_AAABWE chu_b_Page_28.jpg
8d422ab77ea79416c2f8e0b3a4f746df
09b3509a56c7f7e81fa7a4f217fc3107420126ef
19786 F20101204_AAABRH chu_b_Page_05.QC.jpg
f954920a1188d3f81fe2f60bcce48399
3082279db5f1f19f2fe592c2a94fe6e0e8cfba45
71448 F20101204_AAABWF chu_b_Page_29.jpg
def9e88c36a98130ec355c3f88029120
89725c60bdf8fa46790734665cdd230b1508aec5
2007 F20101204_AAABRI chu_b_Page_11.txt
6d1b97564a0d831a7fdf1493c0ba8019
a59693c44471708f085f52560770d6726d9f4933
72112 F20101204_AAABWG chu_b_Page_31.jpg
628d43db4d6fe5e14f41707143b59992
3778bcb99f6443d9e3e98940015d456221dc83f6
7455 F20101204_AAABRJ chu_b_Page_45thm.jpg
05a2f9ded0bbbc851a1138671e5ab72a
34e5713ffc4dbdab60eebe15756b4666654fbd35
25669 F20101204_AAABRK chu_b_Page_03.jp2
e6b610e267e2580dc3082ffb662da886
168f27856e879d3e92e3383391367dc212b9d7ec
67263 F20101204_AAABWH chu_b_Page_32.jpg
f85a9f5284d1ea2a0067fb092c422949
78c1283a3fcdb93e91f627360fd9998356bad949
69232 F20101204_AAABRL chu_b_Page_40.jpg
ef160614a456bb31f6f71195bb5f9051
da8614aee9640536fccd1123cba0c9904cb8a7c8
66723 F20101204_AAABWI chu_b_Page_36.jpg
043629fc86afb432999dd6e142baa12f
d076799ae37b27c86e339eb8ae5181aacb91dfd8
2124 F20101204_AAABRM chu_b_Page_17.txt
265f5d364475fba9b1df649ad152f5e9
1a185b39f57def5575d99fbae03520f66688bb66
52983 F20101204_AAABWJ chu_b_Page_37.jpg
2f43e0d1189700cc5c43c444df9f43cf
31cd60abfd0b6ab0ffa63cdf67db73f2932b2442
23378 F20101204_AAABRN chu_b_Page_20.QC.jpg
57ecbe3adae72b3604d22599181f7a47
1e5a7fe1955884e22dc7d0f4b38c800e55eaa9cc
65158 F20101204_AAABWK chu_b_Page_38.jpg
34761671b2682b1abe109febbba50ad6
cd52ded2e12c95b32aba030c4716e03bdbbe995e
22218 F20101204_AAABRO chu_b_Page_19.jpg
c69ffa34d46770a0c0a640d79a1233cb
befa1e08c767af9d87faf74808f79e52e80287bb
59009 F20101204_AAABWL chu_b_Page_39.jpg
f225fc58856c706d81edf0b0eac962e3
39b5c65c740a1c7d177b6a2d22a44a451ece285d
F20101204_AAABRP chu_b_Page_18.tif
bf2b9007e802d5f02ecd995b8e742dee
33aa7b9ef736e10a475ad7fa7d237f0a82c8f697
83316 F20101204_AAABWM chu_b_Page_44.jpg
aa37036efd0090e5ecabb6d6b09088b7
ed7b6b763064c6dfb451c3bf640dc990c50fdd49
970601 F20101204_AAABRQ chu_b_Page_34.jp2
f260b738e5c3ee567afb69f6338ca4cc
5d8c4b57823d304e2cf7c1cc8041a726c78abafa
101580 F20101204_AAABWN chu_b_Page_45.jpg
29e990d4a34ad5a1712200455b814811
f860b3fb85047b595fbc0bb2ec440e274bf5262f
F20101204_AAABRR chu_b_Page_15.tif
5ed270395bb353d03368d3372761f69d
090a11a35cf22078332f1883a926ff92f7050a32
13207 F20101204_AAABWO chu_b_Page_46.jpg
2b52cf985efb4b94aaae3a243de37dc5
7d4deccb6c1c97f7710f731f8667d46be08c34b5
66837 F20101204_AAABRS chu_b_Page_43.jpg
c06d85cdbecd666e9b7cd6550794568a
b667dff340789686228aa1024a4d935151379c06
233938 F20101204_AAABWP chu_b_Page_01.jp2
152c35537587c4175abe553d44419f2c
e1dccabea0b13be197240939e4bf7fb319c41030
81295 F20101204_AAABRT chu_b_Page_46.jp2
2ec9de339576497e4d1a6667ef7318c9
e73e475ee878165d35c36f534fc68fdb4c22ae8d
26603 F20101204_AAABWQ chu_b_Page_02.jp2
34c7446b635e6ddbd4de151a7b39fd36
a25b27c278a95f2cde0a1b3a8887380327de49fb
10939 F20101204_AAABRU chu_b_Page_47.pro
a212c0c10355d2bbf884c6ed9c0ae473
afb9052cd08aea3653d39ddd4ff9ed1514420041
546141 F20101204_AAABWR chu_b_Page_04.jp2
9ed4789fb1f1dba18ade7bedd3440a63
77e1cf7c88a3f2b6ffd451b8a351a418f94332b1
2059 F20101204_AAABRV chu_b_Page_14.txt
09a0e372ef7e4fd10d5be17340c8c994
acef4f7a804abbc9b5bd34cf8e4e9aedf91c9b92
1051961 F20101204_AAABWS chu_b_Page_05.jp2
4daf093f4c51cc7310daaf07712281e6
f0b56b785ad85eea1b0ccc45cb28ea90204ae349
F20101204_AAABRW chu_b_Page_30.tif
a2373ece788a53c94fa9121a356048d8
094ac86dedd5c1eadfef533ba6ce0a3dc081840a
342362 F20101204_AAABWT chu_b_Page_07.jp2
b4c66bf3bc3eaa184de0d35826027f7e
513bf77bff2c2b4275f42834e0985b4c7b0636bf
44930 F20101204_AAABRX chu_b_Page_41.jpg
b11d54be717b030c4a128671cb38906b
992f12a6b1e0650d58360446674726014c932ae6
1051977 F20101204_AAABWU chu_b_Page_08.jp2
4dfa8d756d035205df06115da95be5b9
dcda623a1e650824d4fd0f53c22a8d4d99112367
2374 F20101204_AAABRY chu_b_Page_01thm.jpg
4d7d38e93a830349749196831a50a6ed
297fcd5dd60a3a2ec7853755dfb649f5315db0cd
1051964 F20101204_AAABWV chu_b_Page_09.jp2
40988198dcf5a5e95e624b7f5f572713
99138917a78495626061037799b70f1e143532aa
186777 F20101204_AAABWW chu_b_Page_10.jp2
560afc4381d55fc2e12d592b66ded644
aa6d2bcf255b26f9620e2286e0ed20491b102433
3141 F20101204_AAABRZ chu_b_Page_03.QC.jpg
9fb6e5c79be387138159b23ee1a368d2
5afdeadbd1ba95557a7dcdf6a62a17ccc478010d
1051936 F20101204_AAABWX chu_b_Page_12.jp2
0264896c3f00f1be8e98c8a681dcf082
c3efecfe59ae1f88b0310e6a3769ecddc3f76c19
1051960 F20101204_AAABWY chu_b_Page_15.jp2
9118b30e0598b19db1e583c1c1e4f1c2
8f1463870318f86dd9552b7f168056ffc113cbf9
38481 F20101204_AAABUA chu_b_Page_32.pro
345c8e487bdf154d30b163e0915b2e4a
836d0b10ea5887f5ef5304092c71d067f8fa25c6
1051929 F20101204_AAABWZ chu_b_Page_16.jp2
31cac714338901eb3d2aa9e039eaf2a7
5f595030e32b9e6544b1b050de66f0a1f89db77a
1157 F20101204_AAABUB chu_b_Page_09.txt
da83f6f6749325788bbc8641af69b6c3
e64e465a246d85777f42f291a1e8e0e627e82fd9
8507 F20101204_AAABZA chu_b_Page_10.pro
8aa4f0b4e95e53f9cc6121d89f68f38a
240c8788c4c7a825dfcde1c562329c12f720363b
2617 F20101204_AAACBA chu_b_Page_45.txt
f1a7f93415d587c0c974be579f81571b
0c92291aac21f7a914761f1cd231054f3d6a49a7
F20101204_AAABUC chu_b_Page_19.tif
95f8f947685212096141364c9b1b9e0a
2dd58703f5c60f5c5de15e17d09e8b89769181d1
581072 F20101204_AAACBB chu_b.pdf
9728fc6d2577af47b1c1bfd00a90577a
eec905cf1b41a7af108fa3959da52cc5f44f6c96
16903 F20101204_AAABUD chu_b_Page_35.QC.jpg
4ca94d71f41a1dbff6768298a65504ce
44e5c51eb18baaf93c08a9784a1931005266251a
49085 F20101204_AAABZB chu_b_Page_14.pro
fe88d836921977de6ba8ec8ad1adf20e
82d1c4b799dde7207d031571f10aaa6f5324698b
7138 F20101204_AAACBC chu_b_Page_19.QC.jpg
424e689b16cbe00a872defbf2229a626
961d1c624afaaeb8803ccee8921b21feca5c4b60
2256 F20101204_AAABUE chu_b_Page_44.txt
56d9305565c093f096455d6d1ea5ee71
865c3b53f33dd6d622f789e29ca3e2946b92ce88
53132 F20101204_AAABZC chu_b_Page_15.pro
b0e821482234cafd65190f24ad27238e
f8c7f101ae5074852873ebce4bc0b7f4971d4772
16654 F20101204_AAACBD chu_b_Page_23.QC.jpg
8ffb5a9cf7832c1418c082fbb5213f7b
51625dca08382f4c24a5e33e851dfc5bbb3adbd1
1337 F20101204_AAABUF chu_b_Page_02thm.jpg
2f19db7d93bb7a52f4fd80d7552b7dc6
6aa20a981d62b118f833e6cc760b8ed48f0d5a65
5946 F20101204_AAACBE chu_b_Page_36thm.jpg
7dc928a8c1c3a1649f2ff9cbef1e1d49
83496b870b3ce1f6caf5ba7ca46affb1a2a8c71f
44650 F20101204_AAABUG chu_b_Page_30.jpg
01bad3e823ebfa68fc565250542dfc4a
38080d1ee674f9acaee9a1139ffcbda506339c2f
48737 F20101204_AAABZD chu_b_Page_16.pro
8ad45cf9a5be22eaea6547b5afd2e254
b5dd22059f9b4b7cf21779e1784c5c83d37775e1
6875 F20101204_AAACBF chu_b_Page_22thm.jpg
57594e5d4844bc5755ea91b6e957d90c
29edb8b944a9a53bd3d439657bc86c3164fd2328
F20101204_AAABUH chu_b_Page_03.tif
23f59f9bed0d53d5cc3c2447afa02f9e
f077075c796d78017d65da8d0685a0ea4454bc91
53832 F20101204_AAABZE chu_b_Page_17.pro
dcf9bc22acec187d9c4658ef23cda188
2a55aa9ede6af0527f14d7403aeb1e89deec396d
1564 F20101204_AAACBG chu_b_Page_46thm.jpg
fd1d18b79e1040cd2a036c0aa1a2beee
39cd270e8db0473fa793d9825553d41027642d14
20839 F20101204_AAABUI chu_b_Page_21.QC.jpg
8c9048a182a615670656425ea375b8b6
6bd84eb1cce368ad55b26ba5cc8a982f0b1c1a82
9635 F20101204_AAABZF chu_b_Page_19.pro
3745a8c7a67d083793ff700a6cc4067e
83d5b15826d12428500bbd24faa862164fe7755d
13863 F20101204_AAACBH chu_b_Page_26.QC.jpg
bbc7db73f0eba5ef83e26a431de2a473
bc10c2976e202c9e5ffebc7b98024cd09c06e6de
27155 F20101204_AAABUJ chu_b_Page_23.pro
f215571bbc51b55dfc288fc168458ab4
5889ab3b27e33849dc6f518c63bbcebd1fb8c188
31680 F20101204_AAABZG chu_b_Page_21.pro
6699e76b160a366126ee24a12b8f8342
58ae764b2b31a35ae9635491e46dd56595aa232a
F20101204_AAABUK chu_b_Page_13.tif
9527a21bd58ccf7797e006c9725b6e30
ac5390da1db865423d44c36dff6fa5a29703dca7
33764 F20101204_AAABZH chu_b_Page_24.pro
e0dd50c50efac200d753cdbeecb39fd4
d3239f72baa0920e7f93e9d5ec90629becbcb340
22763 F20101204_AAACBI chu_b_Page_28.QC.jpg
969e2cb3478b4cddcfa0d1aa03f5d77c
00d5850bf9ac6276ec0bfe154e5023967f504801
F20101204_AAABUL chu_b_Page_09.tif
842bd696a07f5bbea9b29d2c69149b3c
12165b05e393af84bd5bc91cacb788644d3d14f0
55783 F20101204_AAABZI chu_b_Page_25.pro
489440cdca9b9b30d1afc68e4f3ae8f7
37977d3720e8cf415a3ff638ec5c4ae4288ef774
6205 F20101204_AAACBJ chu_b_Page_34thm.jpg
818a6ea6a934afb797877673a8c88f84
2e54a7c74be98abae197d773b936dba41eb73808
24691 F20101204_AAABUM chu_b_Page_47.jpg
e934fafcf7131bbd8f3033e4fb0a3c84
9817a36b6ed39e98159cfdbcd3421a81f8d868c5
25445 F20101204_AAABZJ chu_b_Page_26.pro
a0d72fcbf2db231ba4417b0d08b2d587
953aeb18b0a0aeb9e7e4a4b21efe32b7afcbe191
5852 F20101204_AAABUN chu_b_Page_43thm.jpg
22d98956c761a4adbffde28e51e4b162
6d087e2755e2aec564996d94d338ddc04e7a1b32
38267 F20101204_AAABZK chu_b_Page_28.pro
dc36715de60572be7018f59df9342555
62116bcdc8aa5b6b076f7bd49eac4ff78e1d3c69
3102 F20101204_AAACBK chu_b_Page_02.QC.jpg
6ccccf80c5eaf229cb7ffa5d128f30da
599c6a4f1b4c27122b54281dcf60a208a9be2257
25911 F20101204_AAABUO chu_b_Page_22.pro
31b3412a63f651bc17f32c9e13ca9cf7
0af2c2fd796044b2f79327dcb4d5706ea24258b7
40695 F20101204_AAABZL chu_b_Page_29.pro
9f84ccd907bda4c5d237b582fb97a0f5
3199a2c25e90f7e09beba96b8f8b536fea15f4e2
7510 F20101204_AAACBL chu_b_Page_25thm.jpg
dad800874942134938c3512ced8c2765
7413ff83bc65d53a8e14739520fcba5a268eecd4
3895 F20101204_AAABUP chu_b_Page_04thm.jpg
9361a19a59976d764900bebcd4851945
ee078f6f0a6006dfd37c1a97012f6853cffea7f3
16790 F20101204_AAABZM chu_b_Page_30.pro
6c6083b30f7909c2ba9e918b122f583c
ca3a1c6111e2d505b88a25cbca55d64e2bbaea2c
20986 F20101204_AAACBM chu_b_Page_34.QC.jpg
0ffeeaa75bdc2a807520ca5faf747a24
bfe78bab6cca8e6d17fc4e23dee9862ad7e55a7e
2373 F20101204_AAABUQ chu_b_Page_27.txt
cc7ae284138ccfbeb1d700f4628ba256
5adc2e776fc4ddfeb6690a5bc217367fffdd3e86
38995 F20101204_AAABZN chu_b_Page_31.pro
bce2aae01ad010ea6b63ad3d5e7bd4af
641d04a7ef340795c76b70858efde344db90772d
5532 F20101204_AAACBN chu_b_Page_37thm.jpg
8b8cea06a8e24b29aab455640c9bce4a
03dbcb075e8ebcfa5e29686c7075092e322b8f34
87482 F20101204_AAABUR chu_b_Page_25.jpg
bfec42a5e97b598520e5740400036e79
a872fb8aa102988a6f555e8e66c4b762ebed8807
37455 F20101204_AAABZO chu_b_Page_34.pro
48f438fe59f904b571ccc35e95244581
d7da34956170439c258d41d4d47a43b8f5dcdd20
2128 F20101204_AAACBO chu_b_Page_07thm.jpg
5d31433a0cd091b379d0f5d61740b92c
9523c9ea0be62ffed1adc697f08bdf57a364a659
45031 F20101204_AAABUS chu_b_Page_11.pro
08ee3b5568208f08ba7bedd29565d058
16eb217672d1804b07818ffd535a6727dd1b184c
19447 F20101204_AAABZP chu_b_Page_35.pro
cb032b2da2d668f7417510f7e7fc0a2f
75622fc8bcfd69895d0c8fcc0c06b31de3191179
13673 F20101204_AAACBP chu_b_Page_04.QC.jpg
63c469421986f1868c51c047b8f1085e
8a1ade32e20d6fc235cf8db934ec4fcc50718883
17001 F20101204_AAABUT chu_b_Page_13.pro
fed777a9eff4b2adce45733b5521782a
72713f777765fbb5a691a2766634843677caa677
21923 F20101204_AAABZQ chu_b_Page_37.pro
e46776df4bd4f470981335017841de00
6a1cbdb5c40d4cb15e59ab46f00fcc5de6306da8
8807 F20101204_AAACBQ chu_b_Page_06.QC.jpg
79d34a3947ca161fff01776fd06b68c1
17e1b1798689466ff0044b62bbc50ebcdb7de7bf
25940 F20101204_AAABUU chu_b_Page_17.QC.jpg
fc8e60f6ac3579a9d31b991c1501803b
0fc580afe373299527cc18cd790985c6ba5f979b
37788 F20101204_AAABZR chu_b_Page_38.pro
fc3e8b502c46de807601713ce08e786f
c3073cc923148cd522df6b1d68ee6e4c5df358fa
17475 F20101204_AAACBR chu_b_Page_37.QC.jpg
1d3929f7e15846102e88404304f86a6f
28119ff4914ad64122c59814d1ebaaeabc6e24d4
7166 F20101204_AAABUV chu_b_Page_17thm.jpg
00c5969443ae4cfb0186864cb98493c7
84d198644d82191784645688ec9ceb1ce9bc72fb
43111 F20101204_AAABZS chu_b_Page_40.pro
e27b15e33182e23776f6320fb174aff7
c3834451b913898945368dffc5ee41f0f59c1749
6012 F20101204_AAACBS chu_b_Page_38thm.jpg
557d08e40bdb146bce36682be27aef97
79b248491248865696672c4297da10a175e1606e
25866 F20101204_AAABUW chu_b_Page_18.QC.jpg
aaf8dc1118f3ffee832cf8b15c7941a1
f7bb49c9dec108b69730ac188c2629b0f7c38e99
56616 F20101204_AAABZT chu_b_Page_42.pro
0c74e4480bfd88d6185693f47ed8fe92
0d7faad1db69bbe4ef44a830aef154bd2a27f39e
6183 F20101204_AAACBT chu_b_Page_11thm.jpg
097d4314331f053c12355a19b610d406
2dd66fd233cac2ca584a6a2710d43714c04eefb2
66669 F20101204_AAABUX chu_b_Page_21.jpg
b31d6da60b39b304faf3f5044c072d63
ffd69f7b5e7b5a8babded8dd9b0f98ed09bc1d34
41291 F20101204_AAABZU chu_b_Page_43.pro
7ad330aa9e37b567b313655312e201f7
274d601740335e42c6a8439514e864185bcde1c2
21821 F20101204_AAACBU chu_b_Page_32.QC.jpg
474e84715fa106053e714684736d3e56
1cf6b8d18ac3d989c8aaf3f255aedb8cb080b34f
24352 F20101204_AAABUY chu_b_Page_39.pro
eb3bc8623a84b0cf1438a0bfb5ac3e02
6bb414976e07d88dd4c8ab9c0e1056a897d68c3b
54640 F20101204_AAABZV chu_b_Page_44.pro
f62540e1498945413a9cc5c6848d5f56
85270e8d9d1eedc11d410df7373706c484f4ec63
21141 F20101204_AAACBV chu_b_Page_43.QC.jpg
8a432e1abd3adfc77eb3e8d738476345
5da7c1763166219d77a16b8c0a76912c3b217219
56429 F20101204_AAABUZ chu_b_Page_33.jpg
768cffd199ca1c0a7ae92875d579febd
b18c2b43df3da776d4ac234c0eef40c8ecebe811
63303 F20101204_AAABZW chu_b_Page_45.pro
4115b641c7243aa9792f7e5843f9cb26
53859411da924bcb9554279b63d6540fc88d5fd7
6571 F20101204_AAACBW chu_b_Page_31thm.jpg
efafac93fb04ad1a46c23ca9cc132ee3
8e95e285410babbac5dbfd6ea9ea0f7cbc1c0c51
F20101204_AAABSA chu_b_Page_27.tif
15c5ccf2f6f72fddc964f63b9fc369fb
0a3303c07d4f794c0ca45969e94277c64045a15a
471 F20101204_AAABZX chu_b_Page_01.txt
f15d52b9869f6aedbe9b0705d033ab17
fba0569e4718329fad4872b7ce029b6222ba168a
4824 F20101204_AAACBX chu_b_Page_23thm.jpg
aa6e47cd22cdf571a0776c9f5b19e1e6
a9e2f9fc007d97d88ea782e7d1854167fbb37dda
10384 F20101204_AAABSB chu_b_Page_07.pro
e85a81e4ca5b66d01ff0c57d5b6dc1cd
fa1995eaaafc285c01124ebb30dcd19e51f35d52
3634 F20101204_AAABZY chu_b_Page_05.txt
7940b0a2163c87f5585b7652dc27ecaf
5fcd0bcb5a5bf20196a62a33fd0ba81579f1b23b
4809 F20101204_AAACBY chu_b_Page_30thm.jpg
2dcdbe88e035f1bac59b3d1128802282
02cda4739c03bbf9578e7b22c09a94f86848da4d
4233 F20101204_AAABSC chu_b_Page_46.QC.jpg
f06baf10dfea8e17fbae56da30735f09
6ae094f605f8e61a5ddb381adc67adf2ceeda332
1051983 F20101204_AAABXA chu_b_Page_17.jp2
6c07b7c068b5db169b7d77fbfe4c880f
4453eb3cc643859dd6af2225a7d36ac240c834bd
1353 F20101204_AAABZZ chu_b_Page_06.txt
080bb78ee55789c71b025428619db607
090b3b76a5ea79283a35eeb7b48eeace2a288537
7513 F20101204_AAACBZ chu_b_Page_42thm.jpg
9e618d79fb6258cf18634ea6ecc13024
d15ef85810a4163613c03c359a11bf32bc3781ed
2272 F20101204_AAABSD chu_b_Page_10thm.jpg
8ad7bc2ba03c0f7550d6f3ba32032817
7b773404d5a3abbf1cd48414cb4e79ce76c88bb6
1051975 F20101204_AAABXB chu_b_Page_18.jp2
895e592f598e386d0c08569019e31833
eafe048d7cffb0468dac455da5bb66beada00fd6
F20101204_AAABSE chu_b_Page_21.jp2
970af51c413d12fecab99d4dde27eef1
83d5ad9e8f849fba6da0903b3f599fdadb680c47
F20101204_AAABXC chu_b_Page_22.jp2
e8a49cabe60b636514ef042bd74ff326
176274cf54b76885ed4169df2a6e77cab4186fe2
6268 F20101204_AAABSF chu_b_Page_21thm.jpg
fa84cb4efd2016b4323adb8e7d7e4876
f79ba2b3deef596c8c7e95016de10c01ef41bb36
662791 F20101204_AAABXD chu_b_Page_23.jp2
be4fb771e07cf9e6ff9edf5bf9293d4b
4a67c9100dc5485c0172708a1899a32f4ba7d045
7134 F20101204_AAABSG chu_b_Page_18thm.jpg
594d51d59b2514295591abd2a7c06d60
0f57434817f0b444f0c1ef61f547a6b3b436f00d
1051973 F20101204_AAABXE chu_b_Page_24.jp2
ac1d956245ccae0e30862e13f9a2b022
4f4c683ba2aefd76f0eb9bbec53236ee58874d31
22154 F20101204_AAABSH chu_b_Page_27.QC.jpg
12129f28e87f28d9604446cf201f9a42
2e616f95df2f216963d05784778881c9a512dbef
1051965 F20101204_AAABXF chu_b_Page_25.jp2
bda845225dc1ccf521fdcfa6920a3fc5
90d84c6487ebd7a8c6a198f864e793cebaefb6d1
F20101204_AAABSI chu_b_Page_10.tif
548fbb89479d7ebd38dc92713bfb2eb6
a715bc3340051ef7ec6c7dfee7ce7fb34a8cab4a
569740 F20101204_AAABXG chu_b_Page_26.jp2
13d2aae7f91323321b388b7a6abcfda5
45f84eb75299956858a8402649e7a76b66c11785
1126 F20101204_AAABSJ chu_b_Page_22.txt
2305e00ce6a15727a678e5604abf5e9e
2b452074963c825eddbc6bbdcf2bd1e2629a2329
883118 F20101204_AAABXH chu_b_Page_27.jp2
ccc6e0416ae0041fe5b556d7a8e7a02e
e84aa969817776001a3f3a3b38e6c8caa4a74e3e
F20101204_AAABSK chu_b_Page_45.tif
f01cbbcf0bb64ec912169c6a8910c868
c2b1d9c103303de25b2a84118eea739c169dc8cf
886043 F20101204_AAABXI chu_b_Page_28.jp2
a0b94ba2366d52913f5d0fce48ab119a
a961600b30e9d701af248aa91b8c3fb5a217e5e1
1002720 F20101204_AAABSL chu_b_Page_32.jp2
b8ce923149ac451062b800bf28fe0544
836806d189c67f7d053c00dd49d0864f4ebcf120
1050091 F20101204_AAABXJ chu_b_Page_29.jp2
0cbe48eff569b51e92ad883efdf386ba
20c08bec5168019ebd3d60c0c1034524301799c9
25339 F20101204_AAABSM chu_b_Page_12.QC.jpg
6515380e05caf8e66ae6a9f5df72342d
7f0bb7048afe419000b3aac3bd1d85f8f22ff3f3
583386 F20101204_AAABXK chu_b_Page_30.jp2
64e4db965826c9e3c39487fee15ea060
d966a80455bbc3beb1cf3524cb3bd86e19b17fdb
52584 F20101204_AAABSN chu_b_Page_18.pro
0cb2ba2aeb23727793ca4b24c93112f8
1f67b7503a8b3311c597749b53c1253b3b2e5766
1051981 F20101204_AAABXL chu_b_Page_31.jp2
1f795d375ae83e8e2fdac022fbc402d5
f366fd506944d51c18d724e5beb9ae7ca60b1564
34414 F20101204_AAABSO chu_b_Page_36.pro
e386c7a99de020757ca4cf07bf4c4701
f63cf8e5e02c2d45687abd38fcbf5f1b053b4725
726365 F20101204_AAABXM chu_b_Page_35.jp2
4d1b5fa4109e8f085cd6a4a007e01371
0e7287ba69bad6616022c71bb5338a45de7859f6
1318 F20101204_AAABSP chu_b_Page_23.txt
d37f03ec3c2571ae8149cadf7ad90766
5a740e0109c8be34928eda68858493a57a6aa3d8
742986 F20101204_AAABXN chu_b_Page_37.jp2
759165a0607844e29258fad09b94e2b9
96f78193248a10cc4e8e8a6ebe17f1d3b7bd043d
26041 F20101204_AAABSQ chu_b_Page_15.QC.jpg
53073ce7b3a1a83b047a04cd2b7a832d
7b357791aa367a0657a831355d94b6e51db138ba
934551 F20101204_AAABXO chu_b_Page_38.jp2
712966ce290230f3a7bd7204cad8003d
4949aa9c0f0e89e050f439d89a18f018b99bbe58
F20101204_AAABSR chu_b_Page_25.tif
eb9c46f2fc36d1945313df56040d5b87
715fbf306a0ec48ef1f25d88e28c51031e9b00a7
816971 F20101204_AAABXP chu_b_Page_39.jp2
6e14a69dc687d6e81cd0e9264594c5b4
3a2e904b5428e549b1d47c0b42160e91516cc7f2
6946 F20101204_AAABSS chu_b_Page_08thm.jpg
fab002ad591f77c7b59c6516442372eb
b44812c8b36a72ba8439a7b770d4fe333949f84f
944789 F20101204_AAABXQ chu_b_Page_40.jp2
5265934b30a8ba8c63778b622e172c6c
21da3606a393c3506d0bf070624a2a8891e1fb3e
19751 F20101204_AAABST chu_b_Page_41.pro
e226f27bd4f3112d686d53c65b6f5037
bdaf01044c5e85d6e8c567480c6725e6261a13d9
1051986 F20101204_AAABXR chu_b_Page_42.jp2
1254712c07386ee175889aca597349e7
527cd6d0ff14d538c59900f0d51ff1f05f150af8
1051915 F20101204_AAABSU chu_b_Page_45.jp2
f461badb6afaf53be2ec14e3612703df
9eebae0f2705578eadc6a2311660426201244c43
916784 F20101204_AAABXS chu_b_Page_43.jp2
41e84e279c081d50762151f312e75f54
68c1931bfd05edfc364b0bd9e30157d6ff003e6c
830 F20101204_AAABSV chu_b_Page_02.pro
1932963499e21e0ce3d700c903398c1d
551f93a2b0730b2b95be26272d119d61bb0c3250
1051943 F20101204_AAABXT chu_b_Page_44.jp2
a2717e280f78bf4333e54cf90a41adfe
8ed22fe8922b52a40faf180c1d3763889fe71367
F20101204_AAABSW chu_b_Page_32.tif
b62addf56c31c2dc293ab841f1366af8
df11f98eaeee5ad30f1615a4277044978205ad7e
261759 F20101204_AAABXU chu_b_Page_47.jp2
ab7798277710dc472b5a3b03c4afa3bb
82bf863feb3f3a018659ed3bdd7e92a0dd7cb781
766783 F20101204_AAABSX chu_b_Page_06.jp2
6c3c41a92dfa93fe7d56992802a0d8c5
8c4b87e13819faaffbac5acdc1fa5d6ed652ff10
F20101204_AAABXV chu_b_Page_02.tif
021a4140afb71c41c2170f97cc8b0b27
8e99989bd468e74a595d670b262a47237ee8a008
F20101204_AAABSY chu_b_Page_44.tif
ecf1e12e65fb70fc323a1524132de641
de054ada2216caa25f698faeb9fd47f16bc604e2
F20101204_AAABXW chu_b_Page_05.tif
1bd2dc7d12821098315d6be075c25ea8
05270ee1b5134e029607b7032d1ff75251985d17
3183 F20101204_AAABSZ chu_b_Page_46.pro
2f776c7f41abfc4f136e275f108bc4ab
b8eff69f477b01ce82262d8edb68ceb215e971bf
F20101204_AAABXX chu_b_Page_06.tif
0348fb5ee9d9ff83d81f40fe9f72f5c8
decb53a55d1879b1495599ff64b95484a3dcc8a9
F20101204_AAABXY chu_b_Page_07.tif
3ddb5ab14a67e7017b9e53dd61753042
3cb59ee39ea69662d37f68d4f9180f2ff782cad2
F20101204_AAABXZ chu_b_Page_08.tif
e1f1c7fcee34c8882dcc5f01bee23a35
f12cb56168fb84647ec6630c5a40d19bcbbad9eb
6840 F20101204_AAABVA chu_b_Page_14thm.jpg
ebb20cb29c2e2abf2d57f3c36e7fa792
3db579ffa2c4dc025657763af7058bbfc09d9da7
76340 F20101204_AAABVB chu_b_Page_16.jpg
8a2f03f27b56089a19284efb6f4e29b9
2d86b0f06f342a096974ae51e3018e3b117776d3
F20101204_AAABQF chu_b_Page_33.tif
f4fd2a29a2592975e55b722fa7a839c6
365c2711fe1f2965aea2528a6691066aca1317bc
7712 F20101204_AAACCA chu_b_Page_47.QC.jpg
9c3a7a90fbc9f45268c097088f76ffdd
b670293d160b53f2d6f3fe0743a2301e93b82588
918720 F20101204_AAABVC chu_b_Page_36.jp2
bb5fda188502042239863957c52d0d95
6a5fb77c63f71a9f5c7d4d85b79fd0c57b1d060a
6936 F20101204_AAABQG chu_b_Page_12thm.jpg
d0b39b5a2c91b9791f83a39b18e9e175
be5ba81fcf46ab32bb8f8d380a8e09b8b4f8210b
21541 F20101204_AAACCB chu_b_Page_36.QC.jpg
5b90ca3f9eac119ea2613b26dc04664a
f9e8680659890268316e5d53c83de571c4649165
2568 F20101204_AAABVD chu_b_Page_06thm.jpg
9ee761b77e9e7577b82960536bf06c3b
29c15f684821a461cf81aead837bedca75b3f045
383274 F20101204_AAABQH chu_b_Page_13.jp2
b805f2173d541fbd9dda810391edc0ca
7b9f0ef686e95ba6e5cad615f71ed3c12d8537cb
2416 F20101204_AAACCC chu_b_Page_19thm.jpg
2f2210fc71f22e25adcd1e2a02eb2af3
95a8cc9769f9cfdad132d8be5d3a5a316591303b
53126 F20101204_AAABVE chu_b_Page_35.jpg
16056af8d292be9be2b8bff379b06386
eb8fcf044f61bf845439cd11b048bfdee64ae1d9
7720 F20101204_AAABQI chu_b_Page_01.QC.jpg
32d402034c980e13c7221c2e54ad33ab
6de59500482fba41ae962508ac7ccf496fa19647
27748 F20101204_AAACCD chu_b_Page_25.QC.jpg
10c59f0d9a9dcc212b4016a8e468325d
042bac8b661aa16b835d726b5c2ce290a89f774c
5945 F20101204_AAABVF chu_b_Page_40thm.jpg
ca8c4a29f1f77fd6b2cd52deb92506c8
d15a66ea1d321bdd13208170888aea70b8681b0f
4097 F20101204_AAABQJ chu_b_Page_41thm.jpg
b3579924d7ad178b77ab2aebd5b91d69
dd106557a2d40636fa7e6c4534a49b4ba8b851e0
5932 F20101204_AAACCE chu_b_Page_07.QC.jpg
74dc90f0aa52146fc3ffbc29f1361cfc
7ff61dedad147cec8174b35c5d01ed61b6b226ce
221815 F20101204_AAABVG chu_b_Page_19.jp2
ba2122a3d6c1884b092148b0300cb110
34db5f0ad9756813d42269965aa7699c4adb114e
76602 F20101204_AAABQK chu_b_Page_08.pro
79453661421e1df04730966d81fb279c
71f73382c1edcf24f66ef196bac473821962d274
23936 F20101204_AAACCF chu_b_Page_44.QC.jpg
08cc1969b1cbf27899fe17e5f88d7968
2c89d36c9368ae7b876e63af9efa63d7aba19151
875848 F20101204_AAABVH chu_b_Page_33.jp2
c97c0e15d6efa20e0c663ec032ac7986
459c6dbf6d803b91a2c1562bae301b5e36235ec9
7438 F20101204_AAABQL chu_b_Page_15thm.jpg
2810889b0b2f0a05637d5e14cc7b6a1f
cd25810cc0613b95dcaeafdf41616b5ef56a5190
6782 F20101204_AAACCG chu_b_Page_29thm.jpg
c26ff8d3e6b4136652fc0028eee49d83
6240fad538ed2b12c46f883305bed80656efc931
87 F20101204_AAABVI chu_b_Page_03.txt
04eb3bf5c1d68be19d466793be97874c
656c776d7fa399fee25945735c19f6b1442743ea
46231 F20101204_AAABQM chu_b_Page_20.pro
f3ab2977b6207ecf120a51114af41c1f
711d0a8fc9f3551b60de9239f5e50bd79a37f51a
1334 F20101204_AAACCH chu_b_Page_03thm.jpg
a81a122c8689d206e2bc50eb76f0dea7
8db385ee5891066d53c293cb035824a4621db9b7
474 F20101204_AAABVJ chu_b_Page_47.txt
580484df5228440ee4bfd9e6b22c1d69
bff2e4ad6338f8965471f275549c2fa228541dab
1141 F20101204_AAABQN chu_b_Page_39.txt
5102a9a527cf6b87e288a5b8063055c5
77f505d26faf8727a2dbb8a9704d4f1dd3d3e264
6192 F20101204_AAACCI chu_b_Page_27thm.jpg
a03118d4440ba7df4ab704f9f5df1ede
d8bdef56638abb4fcb21896a82d8288e15ccddd0
51733 F20101204_AAABVK chu_b_Page_12.pro
0465ae64ed10ac7ae49fd9cd587c105a
91137d2c68ac5ce424bfa9dec6166b77ff819814
F20101204_AAABQO chu_b_Page_21.tif
93f27e19ffc2277560197b744eb86a33
d35b895b7ca7a13051c17d053283dab56c4ad23f
56264 F20101204_AAABVL UFE0021894_00001.mets FULL
5a30d3641c1fd5d1a768e3c14bd6a896
445dcee8b34e0d2cda005519a2e944180615f8a9
28027 F20101204_AAACCJ chu_b_Page_45.QC.jpg
2cd19241a8b103f2f7379af67555dcb7
78204365ac3954581c788092401eac51ad01c306
2465 F20101204_AAABQP chu_b_Page_47thm.jpg
d5a7c4cd1311c0310ef617938734968f
9da2c2d92527d941ff6f6b749f0c6fd3be7c5291
5340 F20101204_AAACCK chu_b_Page_35thm.jpg
e9816a8544725a08ec407944ed4b9eb9
7258473f7c5b4217e219a1a3efd5f3600e6e26e6
985 F20101204_AAABQQ chu_b_Page_04.txt
1cf45d4beb1085d2bba96136dcd4bcb1
82da9558c20f43859b14c55446c0edeb17ba591c
72422 F20101204_AAACCL UFE0021894_00001.xml
28a3cabc2e9abb74b5f82e53ecbd5c2a
e314461756c5f0084312a402981f110c5e6e97a7
23089 F20101204_AAABVO chu_b_Page_01.jpg
072a4306e8f7359bce083fadd3a0b268
dd61a6e86a36af05a843f906def2ada1a5f47f25
40233 F20101204_AAABQR chu_b_Page_27.pro
62c7bdc750891a6fa5d5a7c30f5d00be
a7c2eda6ba58561a1c5253f5e6c412856d378e44
5076 F20101204_AAACCM chu_b_Page_05thm.jpg
6f2e9821bbbdac55014cf7f929ecba9f
5b45d468218b502594556fe2aa26c2fe17ff2d25
10131 F20101204_AAABVP chu_b_Page_02.jpg
6acbf1b98539583a91bb8b888e3f71f7
735f4c59371c468750eb25595ec83fb9b1489f5a
68223 F20101204_AAABQS chu_b_Page_22.jpg
571f1c79f7cb4ea909327005efa6702d
c7670fcff0d0491a4f28a919513c85244ef25dee
26392 F20101204_AAACCN chu_b_Page_08.QC.jpg
4159986cd01d0cb01f2da16043752097
9f456519cd5a2b3c3c2812673a7cd8e9d83c9f61
43189 F20101204_AAABVQ chu_b_Page_04.jpg
a23ce8f147d63391053e5c5f367aee06
cdfad46aa70ffc434d0e501be6d59f54ed5cd194
135 F20101204_AAABQT chu_b_Page_46.txt
e7b626a20319c53d62b0b5b894b1c70d
ce1a4a383aa9525c62bc42e78d830be5b1dcb779
12568 F20101204_AAACCO chu_b_Page_09.QC.jpg
852ee9666620d8eee635493ea436e8e6
f7949ca4830e1fc114dae6b0cd0478038ee7a9eb
33284 F20101204_AAABVR chu_b_Page_06.jpg
fef03e355f1ad88bcf84058578623319
64e961273d35cf70da1991f60dd836b1c67d269c
52354 F20101204_AAABQU chu_b_Page_23.jpg
4323aef8fff68250f7e9cfeaf7f495f1
3b3cde7024058f81971c09bec0d622a80d63db21
3565 F20101204_AAACCP chu_b_Page_09thm.jpg
8afc96662594a4ba4d777accd16e56c1
d397431484bd2f5470a6f43e3947479267af03d7
19396 F20101204_AAABVS chu_b_Page_07.jpg
b1ff03cde0b62a38071a4e10e7ba5bbb
4b37837ba48c8ec68ab9a7aadd59ead3c7bdff99
1575 F20101204_AAABQV chu_b_Page_36.txt
9b13ca6bd9f12664ce2a04747b77d3bc
207b1fc157799d11746b648b078025d107faf4c9
22627 F20101204_AAACCQ chu_b_Page_11.QC.jpg
36505f6ef1976a1ee8ae95653447897c
73525673cf3adda0af0181bdb925fe825b9ac2bc
93874 F20101204_AAABVT chu_b_Page_08.jpg
f25b8156ae06ddcc8649943f03d5a5c1
1579f7d46a3705d4a24de073f4663b01a579f667
F20101204_AAABQW chu_b_Page_46.tif
ae84e5cd4cfaa253678a25e7ac115317
62d68fb94c225665fd2df8856189193a4f3749c3
10570 F20101204_AAACCR chu_b_Page_13.QC.jpg
96b790d43dcc48f42d3a3f2b61c4c91a
d6696d988d119983387cd2479444024f929fa9e1
41996 F20101204_AAABVU chu_b_Page_09.jpg
fd9057478f055302609ea918d91a097c
eccc1172c4183532c6eb61f2876b7114aba315c5
22945 F20101204_AAABQX chu_b_Page_24.QC.jpg
d2749aa4db5167d16ac8368f8eaca890
670096ec645c97b2d8a46738b10af3e5d46cf5fb
6762 F20101204_AAACCS chu_b_Page_20thm.jpg
93803302ecc11a7117936c7b45230448
b54b45e14e4feb2e091506f0b1dc15b945a22c4e
73683 F20101204_AAABVV chu_b_Page_11.jpg
f7aa3e2c769fda0e3f8f0cde08753428
efbcf87bdc409f2b4af907ce38f5950a26755e6c
6532 F20101204_AAACCT chu_b_Page_28thm.jpg
ce1cadac8c5e8cf690bafa3180b7433b
128d58888427c672990b542b0018f55531e70cad
81744 F20101204_AAABVW chu_b_Page_12.jpg
093e29633d3f78d2b7138d52a00be15e
85659acefdeb40765e3ca8d0b4676d60cc0e4a07
86446 F20101204_AAABQY chu_b_Page_42.jpg
19cdcde7426828b137a9e902f1ab605d
4cbf5791c9704b6202c005cfd377ca94d67c5125
14913 F20101204_AAACCU chu_b_Page_30.QC.jpg
93e104897d9d39e5362e0ddc7ec4b921
ba2f45c05304dfa26bdace2c3404510d236cc53e
31435 F20101204_AAABVX chu_b_Page_13.jpg
f77d6d04c84869b0c2e26d93a78c8285
357412daea16c928b2806473cf16e42d3bf51265
F20101204_AAABQZ chu_b_Page_38.tif
0410133f7a18dec3d9c9b896c4e13b55
68f79114c6ebe80fb18979bdda76385c5befaff0
5514 F20101204_AAACCV chu_b_Page_33thm.jpg
d1d19d5fe769f79d4a13b92f4c189767
cf2a23b5677edffe1e72944a202bc9c2c2ad2ab1
78153 F20101204_AAABVY chu_b_Page_14.jpg
459f296d314db8eb3c6a85d453460ef5
0ec49501f5353a7bf07ed20a28d8334a25d37666
20721 F20101204_AAACCW chu_b_Page_38.QC.jpg
bad3493b6c25cbc1dbd2282d9df461f4
61c588d3e6187ec7998f041792afcf0c0fd1caa5
22442 F20101204_AAABTA chu_b_Page_33.pro
31f32a19e2f074d43921fa632711b557
cce0ac0befc8d073af1aa51b2927b08d5012b5c8
85008 F20101204_AAABVZ chu_b_Page_17.jpg
b1e3136acb821064775b05e4c4e02c8c
5b72b00d661db40dcf84314daab8ac46ee38e9d6
18893 F20101204_AAACCX chu_b_Page_39.QC.jpg
00afd6915390209e8044d47b6eefd684
adcb9d4751a2fadf369a3925257a811fa4e79143
72374 F20101204_AAABTB chu_b_Page_05.jpg
99955b15b83afee0ae960885b824dc8f
365774f6566264fdf59b8f39e22876d2fe9bb5c2
5963 F20101204_AAACCY chu_b_Page_39thm.jpg
d3a807db952e1610edc5cc3e36f202cb
233f4c2f7e3558794f1b72744bd59a64c8d99140
533781 F20101204_AAABTC chu_b_Page_41.jp2
cea6bd2a6c1d7d5ce0212b4e26b0854b
d987da8b7e5b456bb9540e0594b5932e3ed1e8fc
453 F20101204_AAACAA chu_b_Page_07.txt
762e6c0cb6570fd6e8d2a719ca201e33
26d6cb830c8cf2f229fafcc1dea9580a104c5add
21670 F20101204_AAACCZ chu_b_Page_40.QC.jpg
0bacb08d95823cc48cc0b5f650593e67
60a0e4f892b1e858fc53ec2711499458696096f0
68238 F20101204_AAABTD chu_b_Page_34.jpg
d3e1ff0a1a329c22e56b861082c54308
00c16ffc79d5eedf0d5d111aa933add1c8cf65a4
F20101204_AAABYA chu_b_Page_11.tif
eff5b9445608b6322513a0f3212628ea
856e76f7e66e43e7d1c69281944be448fdf4bff5
3111 F20101204_AAACAB chu_b_Page_08.txt
29ab54ac85b7860b6707d52b3a0f1ff0
6df24beca6980112a7bc7aeff96f481cc4ff1245
3998 F20101204_AAABTE chu_b_Page_26thm.jpg
bbf033f14acdc829432470c5e9167fed
d8b511685dc5b838058654e41ad4ed322d69a763
F20101204_AAABYB chu_b_Page_12.tif
4cee800f6e9840b478ad70c87ffbebc0
dd4e670931af49c9169ed50950179c54b8b65e9f
363 F20101204_AAACAC chu_b_Page_10.txt
f0ee496f2ead6b1720986daf05eb04ef
25c2b5b8b00cddf12f582982ccabd67b609090ed
3405 F20101204_AAABTF chu_b_Page_13thm.jpg
8b606690d25b3655040275124ca945c7
4049bdccbb4f0bbe32273273ededa1d1cb9ba3b4
F20101204_AAABYC chu_b_Page_14.tif
06121883d34d139576fb1a0534535ad4
6fdb06a87b3e231b52dc0dda4bf830f8fb49cb0b
2123 F20101204_AAACAD chu_b_Page_12.txt
5625129ec3d98256f73e705db0a56a63
ad5bcd90643ba06ed565fcceb5388c42931da6be
F20101204_AAABTG chu_b_Page_16.tif
d86a11b7cea66ab6d622eef8eed9a276
d05b1a8a020b2c57fb8e98a348128ebe4219e100
F20101204_AAABYD chu_b_Page_17.tif
3335198a5b9bab9b72aaa5f12904f834
8f18866e1a72785bba1ee158ec2d183728dd69d5
2188 F20101204_AAACAE chu_b_Page_15.txt
04a142abfffadbb1a884e54483b131e8
52b63ec52dc82a556d44ed9fa79644cf744d1200
28895 F20101204_AAABTH chu_b_Page_09.pro
7de870a4a16e02e87d658100b21c9788
0f55de6ebfe92b5138dcdfc5ba4bdcae6c0558f0
F20101204_AAABYE chu_b_Page_20.tif
6061349004e494dab9551625ad9c2bfe
ca129920180fcaa85244163644d571e163af1698
2000 F20101204_AAACAF chu_b_Page_16.txt
39d1ddc1c335396b278b6c966cfa15ed
e7e59bc5fd00345a330614f4d1f47ee5d5e2fa7c
22820 F20101204_AAABTI chu_b_Page_22.QC.jpg
6d31b94a06577e0ace576e223cd8ad1a
f919273970854782130f5813a0f1535931c72a7f
F20101204_AAABYF chu_b_Page_22.tif
0551029468b8553c92a24bb30f75253d
2ed45a91c5c1bae1ef810486c431d4f6fa036d8e
2115 F20101204_AAACAG chu_b_Page_18.txt
8bbf77021199c3b3799540ee37396c87
fd0aaf76685d81b06f8cb17f5c86a533d29ef582
F20101204_AAABTJ chu_b_Page_04.tif
7db970c88d0bb21e2ed5b2c7965ebb16
3ae281ce73dca869a1e7c974ea5cf4638daefaae
F20101204_AAABYG chu_b_Page_23.tif
a6e396ef8c679b685139684c9b40c5c7
e5dc69052698627de1f4e7a9dced09086ba09b0b
81539 F20101204_AAABTK chu_b_Page_15.jpg
7d81d81ffba70f3fee8ff9ced1ecc084
f77b4bd350148a018732361995e047ea7ab5189f
F20101204_AAABYH chu_b_Page_24.tif
b1a374b762433bf5431427ceca39d8a3
8956157e0d6cd9ec0a2b5cd72a03556e651dcc81
390 F20101204_AAACAH chu_b_Page_19.txt
c88f4997894412db96ead7d87c23a41f
3bf1e0ca89b14704df7e7e4983cbc66cf43f8647
F20101204_AAABYI chu_b_Page_26.tif
5509e0e084596c3af808ce2387870a66
0dece764f35ffa2c08c274bd7ddafac2ec22a17b
1946 F20101204_AAACAI chu_b_Page_20.txt
751092b6bd93dbe3cfe63a6f92554238
2c0da768a843189cd33ff59ba293f4e1a3460de9
22936 F20101204_AAABTL chu_b_Page_31.QC.jpg
a6fa9076d3a1d4098f083b833f5ffa20
81ff9fe25f880c9dfc846ceac597d86a05dfdd2e
F20101204_AAABYJ chu_b_Page_28.tif
0d65ae3f69a5464f667c53cf4a7b53a8
f003aafbfb08cd1096ca094d3d72e8ee56a9c984
1658 F20101204_AAACAJ chu_b_Page_21.txt
40fd6ceb9f29c12e590658a53e0e69f3
33e5ad649dba489b5d3711cab541dd7ad977be35
1051962 F20101204_AAABTM chu_b_Page_14.jp2
1e4e0071784d8b3c72be45510f806e40
a20b0b29f36e9d0dcd0144bb37a46e24116cd6cf
F20101204_AAABYK chu_b_Page_29.tif
3c74822a72fa62b468d08c125d9a6def
d92edbf575b37e4350dd32c70b5ecbb98baa6464
1414 F20101204_AAACAK chu_b_Page_24.txt
2734e558e92605ff419168fa9fb8f0ca
b697f63c2176f61273b8b35f25271d025b76a2f2
6810 F20101204_AAABTN chu_b_Page_24thm.jpg
09d045fa5ff2e3a0ee848f3fb9cf5bd7
3caef2cd77712ff63e5946ce094b744ec7e612d3
F20101204_AAABYL chu_b_Page_31.tif
d2e7a9e069bb9e90d8653cf756ec7b10
c381385b15025b9b3deb381556d93f96436f999e
1040 F20101204_AAACAL chu_b_Page_26.txt
89981a5f617ee8ff352091d8f1166424
273e740115529305773c1993800549047ddcbf5a
24042 F20101204_AAABTO chu_b_Page_16.QC.jpg
0b47523feb03dc55f968f7827839d92a
7ed19d2a656bf11926afd0b0e6a8a916e0c1cc83
F20101204_AAABYM chu_b_Page_34.tif
553fa56ef9ef4f101540fd9bb4393fbb
b1839319ff724ef56356dc9f5aefcb2dd38cfa7c
1606 F20101204_AAACAM chu_b_Page_28.txt
5d4eb57c46a0bce1c831a0cafa7d1d66
5d8a1a167f877f9052d377e0009212bcd5cfe6b1
43905 F20101204_AAABTP chu_b_Page_26.jpg
fb5c11acd1094d4197231341e8cb8e85
3bd5948a581b8c9de693996b67fab9856662b702
F20101204_AAABYN chu_b_Page_35.tif
eb7a8568503795cffbab7cf59d660131
6c128c2c16e6a4e68498a43eb077f66485d67f48
1958 F20101204_AAACAN chu_b_Page_29.txt
e7b7c8da0bd8a21299bc8a3f987f00a3
c8398a77318ddfc9c9c79718969151045b9d13be
6420 F20101204_AAABTQ chu_b_Page_32thm.jpg
ffe37f3ed645f03362af1003c7c2c295
5504d5b2a5bcb038497d9d05d2b5385bf314b641
F20101204_AAABYO chu_b_Page_36.tif
d973cd37f224c315fbc60aafa3778ccb
31efa7c44c098bcc44c6c18ee9ada64052050bf5
933 F20101204_AAACAO chu_b_Page_30.txt
20f5206a0274a995c56e65814f65607f
3de6d9805470152e4625f165bc1870aa0e8acb76
1008263 F20101204_AAABTR chu_b_Page_11.jp2
d3a34395e2c69f089da7ebeab784979b
02a3d9c1e9ca311ad42b306c4499bf7c6e12d31e