<%BANNER%>

In Situ Lubrication of Boric Acid

Permanent Link: http://ufdc.ufl.edu/UFE0021824/00001

Material Information

Title: In Situ Lubrication of Boric Acid Powder Delivery of an Environmentally Benign Solid Lubricant
Physical Description: 1 online resource (43 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: acid, benign, boric, environmentally, friction, lubricant, manufacturing, solid, tribology, wear
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Mechanical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Due to growing social and environmental pressure on the manufacturing industry, the search is on to develop cost effective and environmentally benign alternatives to the currently used petroleum based lubricants. Boric acid is a non toxic, cost efficient, environmentally inert lamellar solid, similar in structure to solid lubricants graphite and molybdenum disulphide. The lubricant potential for boric acid was tested on the Pin on Disk Tribometer at the Tribology Lab at the University of Florida. Solid lubricant boric acid was delivered, via Nitrogen gas jet stream, to a loaded pin on disk interface with friction force and wear rates for the pin and disk being measured for 90 samples. Boric acid showed excellent friction reducing and wear protection capabilities. With further development, boric acid exhibited potential to supplant petrol based lubricants in certain manufacturing processes.
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: Sawyer, Wallace G.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021824:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021824/00001

Material Information

Title: In Situ Lubrication of Boric Acid Powder Delivery of an Environmentally Benign Solid Lubricant
Physical Description: 1 online resource (43 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: acid, benign, boric, environmentally, friction, lubricant, manufacturing, solid, tribology, wear
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Mechanical Engineering thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Due to growing social and environmental pressure on the manufacturing industry, the search is on to develop cost effective and environmentally benign alternatives to the currently used petroleum based lubricants. Boric acid is a non toxic, cost efficient, environmentally inert lamellar solid, similar in structure to solid lubricants graphite and molybdenum disulphide. The lubricant potential for boric acid was tested on the Pin on Disk Tribometer at the Tribology Lab at the University of Florida. Solid lubricant boric acid was delivered, via Nitrogen gas jet stream, to a loaded pin on disk interface with friction force and wear rates for the pin and disk being measured for 90 samples. Boric acid showed excellent friction reducing and wear protection capabilities. With further development, boric acid exhibited potential to supplant petrol based lubricants in certain manufacturing processes.
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: Sawyer, Wallace G.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021824:00001


This item has the following downloads:


Full Text





IN SITU LUBRICATION WITH BORIC ACID: POWDER DELIVERY OF AN
ENVIRONMENTALLY BENIGN SOLID LUBRICANT
























By

TIMOTHY PAUL BARTON


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 Timothy Paul Barton









ACKNOWLEDGMENTS

First and foremost, I would like to sincerely thank my family, my grandmother, Virginia

Smith, my mother, Michele Barton, my sisters Bonnie and Beth, and my girlfriend Merigan

Craig, for their unending support, motivation, and understanding throughout this entire academic

endeavor. It is with the love from this well of inspiration that I am able to finally complete this

journey.

A very important acknowledgement goes out to my graduate advisor, Dr W. Gregory

Sawyer, without whom I would have never found direction and certainly not completed my

graduate thesis. His support, academic, financial, motivational, helped to inspire me to complete

this degree. I would also like to acknowledge all members of my committee, both past and

present. Dr. John Schuller, Dr Nam Ho Kim, Dr John Zeigert, and Dr Tony Schmitz all played a

great part in my academic growth, and I sincerely appreciate all of their efforts.

I would like to thank Dr. Dan Dickrell, Pam Dickrell, and Ali Erdimer for their assistance

and lending their insight with their previous works with using boric acid as a solid lubricant. I

would like to note the members of the tribology lab, Dave Burris, Jason Steffans, and the rest and

thank them for their help in getting acclimated to the lab, and certainly for their understanding as

I made my way, often painfully for those around, through these experiments. I would also like to

acknowledge anyone else who was involved with the administrative, academic, and financial

efforts to complete this thesis.









TABLE OF CONTENTS

page

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

LIST OF TABLES ........... ... ...... ..................................................... 5

LIST O F FIG U RE S ................................................................. 6

ABSTRAC T ......................................................................... 7

CHAPTER

1 IN T R O D U C T IO N .................................................................................................... 8

1.2 Introduction to Petroleum Based Lubricants ............................................................8
1.2 E nvironm mental C concerns ................................................................................. 8
1.3 H health C on cern s ............................................................................... 9
1.4 C ost C concerns ................................................................. 9
1.5 M manufacturing C concerns ................................................................................. 10
1.6 L egal C concerns .................. ..................................... ... .......... ........ .. 11

2 LITERA TU RE REV IEW ........................................... .................. .................................. 12

2 .1 B o ric A cid ............... .... .................................................................................12
2.2 H health and Environm mental Im pact ........................................................................... 12
2.3 Boric A cid as a Solid Lubricant ............................................................. ....13

3 M A TH EM A TICA L M OD EL ................................. ....................................... .................... ....16

4 EX PER IM EN TA L SETU P ............................................................................ ......................20

4.1 Specimen Characterization and Preparation ............................................................ 20
4.2 P in on D isk T ribom eter........ .................................................................... ...............2 1
4.3 Pow der D delivery System ............................................................... ................24
4.4 E nvironm ent C ham ber ....................................................................... ........................27
4.5 Determination of Experimental M atrix ....................... ...........................................27

5 R E SU L T S .............. ... ................................................................ 30

6 DISCUSSION ........................ ......................................36

7 C O N C L U SIO N S ..........................................................................................................4 0

LIST OF REFEREN CE S ............................................................................................41

B IO G R A PH IC A L SK E T C H ............................................................................... .....................43



4









LIST OF TABLES

Table page

4-1 Surface Roughness Measurement results for prepared surface of 302 Stainless Steel
d isk s ......................................................................................... . .2 1

4-2 Experimental matrix (3*3*3*3) plus nine repeats at the midpoint and one
unlubricated control. .......................... ........................... .... ........ ......... 28

4-3 Randomized experimental testing sequence with all parameters per run..........................29

5-1 R results from the experim ental m atrix..................................................................... .....31









LIST OF FIGURES


Figure page

2-1 Lam ellar structure of boric acid............................................... .............................. 13

2-2 Experimental setup for Gearing et al. experiment to determine shear strength of boric
a c id ................... ...................1...................4..........

2-3 Experimental setup and results for Lovell et al. lubrication study using powder boric
acid as a Solid Lubricant in a loaded pin on disk experiment .......................................15

4-1 Actual boric acid powder used in this experiment taken using the Scanning Electron
M icroscope at the University of Florida................................ ......................... ........ 22

4-2 Rotating pin-on-disk tribometer used in this study ......................................................22

4-3 Rotating pin-on-disk tribometer used in this study ......................................................23

4-4 P article delivery scheme e .......................................................................... ....................24

4-5 Powder delivery calibration curves referenced from initial conditions.............................26

4-6 Incremental powder delivery calibration curves referenced from previous time step.......26

5-1 Pin volume lost due to wear during the experiment .................................. ...............30

5-2 Friction coefficient versus cycle number for the 9 repeat tests and the control ...............32

5-3 Friction coefficient versus cycle number for the experiments with terminal blockages
in the boric acid delivery during the experiment. ................................... ............... 33

5-4 Average friction coefficient versus normal load, flowrate, and sliding speed...................34

5-5 Average friction coefficients for all experiments plotted versus the boundary layer
thickn ess ......................................................................................................34

5-6 Friction coefficient versus cycle number for the two experiments that had the
shortest and longest transients to low friction coefficient ...........................................35

6-1 Wear marks on pin and disk after midpoint experimental conditions A) With boric
acid flow B) W without boric acid flow .......................................................... .... ........... 36









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

IN SITU LUBRICATION WITH BORIC ACID: POWDER DELIVERY OF AN
ENVIRONMENTALLY BENIGN SOLID LUBRICANT

By

Timothy Paul Barton

May 2008

Chair: W. Gregory Sawyer
Major: Mechanical Engineering

In-situ deposition of boric acid in dry powder form is investigated as a potential

environmentally benign solid lubricant for sliding metal contacts. Boric acid is widely used in

industrial processes and agriculture, is not classified as a pollutant by EPA, and produces no

serious illnesses or carcinogenic effects from exposure to solutions or aerosols. In this study,

boric acid powder is aerosolized and entrained in a low velocity jet of nitrogen gas which is

directed at a self-mated 302 SS sliding contact in a rotating pin-on-disk tribometer. The effects

of powder flow rate, sliding speed, normal load, and track diameter on coefficient of friction and

wear rate are investigated. Friction coefficients below [=0.1 can be consistently reached and

maintained as long as the powder flow continues. Wear rates are reduced over 2 orders of

magnitude.









CHAPTER 1
INTRODUCTION

1.2 Introduction to Petroleum Based Lubricants

Petroleum based lubricants are widely used in the manufacturing and industrial sectors, as

well as in automotive and many other mass market products. It is well recognized that the use of

these lubricants introduces significant quantities of used petroleum based substances into the

waste stream [1]. These lubricants impose significant negative impacts to the environment and

health both during their primary use and after disposal [2]. This study introduces a lubrication

concept aimed at reducing the need for petroleum based fluids in a wide range of industrial

processes and consumer applications by delivery of boric acid, an environmentally benign solid

lubricant, in powder form. The proposed lubricant and delivery method will avoid the waste

stream and environmental and health impacts associated with other lubricants used in many

industrial processes and products. This technique is not tied to any particular process or product,

but rather it has broad applicability; although customized delivery strategies will need to be

developed for the various applications.

1.2 Environmental Concerns

The environment has long been an afterthought regarding manufacturing concerns, with

the exception of clean up costs. As social awareness towards the environment rises to critical

levels, the impacts of industrial waste streams on the environment is subject to increased scrutiny

and requiring meaningful action. As of 1995 in the United States, 32% of all lubricants return to

the environment in a physically or visually altered state [1]. Their impacts on water and air

quality, wildlife, and human health have been found to be toxic. The behavior of the lubricant

when returning to the environment will govern the impairment to the environment. It follows









that there is significant push to develop lubricants with reduced toxicity and increased

biodegradability.

1.3 Health Concerns

Machining lubricants pose significant health hazards to human from both direct and

indirect contact. Some risks of lubricants include the development of nitrosamines in coolants,

skin disease from unprotected contact, carcinogens in used oils and lubricants, solvent containing

products, and heavy metal compounds in additives [2]. If not properly tended to, metalworking

fluid circuits can be subject to rancidity. Microorganisms such as bacteria, mold, and yeast cause

rancidity by multiplying in metalworking fluid circuits once exposed to air and water dilution.

Skin dermatitis has been linked to microorganism infected metalworking fluid mixtures [3].

Significant odor issues can also arise and respiratory irritation and infections have been linked to

rancidity of metalworking fluids.

1.4 Cost Concerns

The increasingly stricter environmental regulations and corresponding enforcement are

reducing the flexibility of metalworking fluids. As new metalworking formulations are

developed, they are missing Pb, S, or Cl compounds, which provide superior machinability

characteristics and are cost effective to produce. As noted above, the health and environmental

concerns of the metalworking fluids are cause for increased maintenance and disposal costs. The

labor and overhead costs in the U.S. in 1995 were estimated over $300 billion. These costs were

estimated to comprise approximately 5-6% of the total manufacturing costs in the U.S in 1995.

The costs associated with the use of cutting fluids, including purchase, maintenance, storage, and

disposal, is estimated to be about 16% of the manufacturing costs, many times more than the

labor and overhead costs [4].









1.5 Manufacturing Concerns

There have been many attempts to address the issue of reducing or eliminating petroleum

based lubricants from metal working processes. The impact of Minimum Quantity Lubricant

(MQL) volumes in machining was studied by Machado and Wallbank [5]. They compared the

use of dry cutting (no lubricant), an air-water mixture, and air-soluble oil lubricant (MQL) jet

streams versus overhead flood coolant in lathe cutting process. The results for cutting and feed

forces, surface finish, and tool life for MQL rivaled or exceeded those of standard flood cooling

at low cutting speeds. However, the effectiveness of MQL diminished as machining speeds

approached those commonly used in industry. MQL lubrication also introduces a significant

health risk by promoting a mist in the environment, requiring an extensive exhaust extraction

system. The results for air and water lubrication were promising, however, significant corrosion

problems were noted.

Kustas et al. [6] investigated the use of nanocoating on the cutting tools in a dry machining

operation. The attempt of the study is to prove the coatings will generate less heat during

machining and/or take away heat generated rapidly in dry machining by other means. One

hundred bilayer 13Angstrom B4C / 18 Angstrom W nanocoatings were deposited on cemented

WC-Co tools and HSS drills. Coated and uncoated dry machining tests were conducted. A 33%

reduction in torque requirement and a noticeable reduction of wear on the tool flank surfaces

were observed. No comparison between coated tool dry machining and non coated traditional

petroleum based lubricant machining was made. Significant efforts are being made to reduce or

eliminate the use of traditional petroleum based lubricants. Though none has yet risen to provide

the combination of surface quality, production cost, and tool life to supplant the standard

lubricants.









1.6 Legal Concerns

There are currently no US laws requiring the use of eco-friendly lubricants, but two

regulations, Executive Order 12873 (EO 12783) and the Great Lakes Water Quality Initiative

(GLWQI), have made significant impacts on the use and disposal of conventional lubricants. EO

12783 provides preferential treatment of government contracts for bidders who use recycled oils

and/or use environmentally compatible oils where possible. The Great Lakes Water Quality

Initiative puts stringent requirements on zinc limitations such that the use of zinc containing

lubricants is effectively banned. Government regulations on use of environmentally harmful

lubricants, though placing limits on human activity with regards to the environment, are far from

exhaustive and many problems exist in spite of them. Over 30% of all lubricants return to the

environment in a harmfully altered state. Outside the realm of regulation, there exists a need to

find an environmentally benign alternative to the current petroleum based lubricants without

sacrificing manufacturing quality or dramatically increasing costs.









CHAPTER 2
LITERATURE REVIEW

2.1 Boric Acid

This experiment explores the potential of boric acid as an environmentally viable

alternative to the petroleum based lubricants used today. Boric acid, whose chemical formula is

H3B03, is also known as orthoboric acid or boracic acid. Boric acid is a hydrate of boric oxide,

B203. When boric oxide comes into contact with water, it will readily hydrate and convert to

H3BO3. Boric acid is a weakly acidic white powder that is soluble in water, approximately 27%

by weight in boiling water and 6% at room temperature. It is soft, ductile, stable, free flowing,

and easily handled. Finely ground technical grade boric acid powder (>99% pure) is readily

available for under $2US per pound.

2.2 Health and Environmental Impact

The Environmental Protection Agency has established that boric acid is benign and it is not

classified as a pollutant under the Clean Water Act or the Hazardous Air Pollutants Act of 1990.

Material safety data sheets for boric acid show no serious illness or carcinogenic effects from

exposure to solutions or aerosols.

The US is the world's largest producer of boron compounds [12]. Boric acid is recovered

from brines at Searles Lake in California, with large domestic reserves of boron materials

residing in other lake sediments and brines. Large quantities of boron ore are also collected from

an open pit mine in California.

The consumption of boric acid and boric oxide in the united states is distributed among

glass making (78%), fire retardant (9%), agricultural fertilizer (4%), and industrial applications

such as metal plating and finishing, paints and pigments, electroplating, and cosmetics (9%) [12].

A dilute water solution of boric acid is also commonly used as a mild antiseptic and eyewash.









The use of boric acid as a food preservative apparently dates back to the ancient Greeks. The

earliest available scientific study of boric acid was conducted in 1902 and reported in Science

1904 [13]. In this study, boric acid (0.5g) was introduced into food and ingested by a group of

participants with each meal. This produced occasional occurrences of "fullness in the head",

nausea, and loss of appetite in a few subjects.

2.3 Boric Acid as a Solid Lubricant

In the early 1990s, the lubricity of boric acid, an overlooked but extremely available and

environmentally benign lamellar solid, was demonstrated by Erdimer et al. [13-16]. Figure 2-1

shows the lamellar molecular structure of boric acid.

Lamellar Solid .
boric acid weak interactions
H3BO 3 a -- between lamellae




7 Boron +

S Hydrogen
Hydrogen
Bonds



Figure 2-1. Lamellar structure of boric acid

Experiments were conducted by Gearing et al. to determine the shear strength of boric acid

powder [17]. In the experiments, high pressure thrust washers were compressed against a 6111

aluminum alloy plate coated with boric acid at pressures above 500 MPa. A twisting moment

was applied by the compression tool loading the washers against the plate and sheared the boric

acid coating on the plate. Figure 2-2 shows the set up used in the experiment. The results

determined the shear strength of boric acid to be 23 MPa.











0.10 compression
o twisting tool
0.08 Fn
\ T

S0.06- boric acid
S'o applied to
0.04 s surface
0.04
Al 6111-T4
0.02- B. P Gearing etal.
Int. Journal of Plasticity
17 (2001) pp.237-271

0 100 200 300 400 500 600 700 800 indenter with knurled
Contact Pressure MPa surface eminence

Figure 2-2. Experimental setup for Gearing et al. experiment to determine shear strength of boric
acid.

The shear stress of boric acid in Gearing's study is almost the same as the experimentally

determined shear stress of molybdenum disulphide, which was found to be 24 MPa by Singer et

al. [18]. Figure 2-2 shows how the friction coefficient lowers with increased contact pressure

and is well below 0.1 for average pressures above 100 MPa. At sub-atmospheric pressures, boric

acid dehydrates and reverts back to boric oxide above 1700C. In machining applications, cutting

and forming interface temperatures are expected to exceed this hydration temperature, but the

contact pressures are expected to be much greater than atmospheric pressure. No published data

exists, however, detailing the hydration characteristics of boric acid above 1700C at higher

contact pressures.

Initial work with boric acid as a solid lubricant involved creating a solid film on the

surface of the work piece. Boric acid was dissolved in either water or alcohol and the surface

was coated and dried. After the solvent evaporated, the dried boric acid remained on the surface

as a thin coating. This method of boric acid deposition is incompatible with many practical









applications or products in the manufacturing industry. This may be one reason that boric acid

has not been accepted as a commonly used solid lubricant by industry.

Our study examined the potential of delivering dry powder via air jet to deliver sufficient

lubricant to adhere to the work piece and achieve good lubrication in situ. The feasibility of

boric acid powders to sustain low friction when delivered as a powder was shown by Lovell et al.

[20]. In this study, a concentrated sliding contact between an aluminum pin and AISI-M50

bearing steel was loaded to an initial maximum central Hertzian contact pressure of 1.9 GPa.

Figure 2-3 details the experimental setup and results.

0.7

normal 0.6-
load
friction .: on
o-rce 0.4 o ooo

0.3 4
I
boric acid 0.2 i off
spooned 0.1 --
onto
surface pin-on-disk apparatus I I I I
0 500 1,000 1,500 2,000 2,500
time (seconds)

Figure 2-3. Experimental setup and results for Lovell et al. lubrication study using powder boric
acid as a Solid Lubricant in a loaded pin on disk experiment

The sliding speed was approximately 1 m/s. Boric acid powder was delivered by manually

sprinkling it onto the disk surface; the corresponding lubricious surface film lowered the friction

coefficient from t = 0.4 to approximately t = 0.15. The friction coefficient immediately

responded to the application of boric acid by dropping to low friction state. Further, the reduced

friction was sustained until the powder delivery was halted.









CHAPTER 3
MATHEMATICAL MODEL

We developed a predictive mathematical model for the effects of boric acid delivery on the

sliding contact between the stainless steel surfaces. Assume a rotating disk has a contact

interface with a stationary sphere supporting a load. If boric acid is delivered into the interface, a

fraction of the powder may stick to the disk face along the wear track. The amount of boric acid

on the face may be represented by a fraction of surface covered with the powder. Initial

fractional coverage is represented by 00, where

AH3 O
o00 (3-1)
Total

As powder is continuously delivered, more will be deposited on the area. An amount of

the deposited powder, however, will adhere over top already covered surface. The coverage

after a platelet of material AO has been deposited on a surface with fraction coverage given by

Equation 3-1 is

00+ = 00 +/AO- AO-00 (3-2)

This pattern of application of boric acid platelets continues during powder exposure for n

platelets

0=00

0= 00 +A0-A000

02= 0 + 2. A0- 2. A0-0 -A02 + A020 (3-3)

03 = +3.A0-3-.A.00 -3.A02 +3A0920 ,+ a 3 -A03 .00

After reorganization the above pattern, namely by factoring out the AO terms, the sequence of

Equation 3-3 can be equivalently expressed as








0=00
01 =00 +A0.(1-.00)

02 =00 + 2-AO(1-00)- A02 .(1- 00) (3-4)
03 =00 + 3AO (1-00)-3.A02 .(1-00)+A03 (1-00)

This pattern can be equivalently expressed by the following series expression

SO=(n n!A (1-AO).(-1)(a+l)
a=0 +1 (3-5)
a=O a!. (n -a)!

Equation 3-5 can then be expressed in the compact and closed form

S= 1- (1- 0)- (1- AO) (3-6)

Equation 3-6 assumes that n platelets of boric acid powder sprayed into the interface stick

to the surface during a single revolution of the disk. After contact with the pin a fraction of the

material is removed. X is now introduced as a loss term indicating the amount of powder

removed each cycle by contact pressure between the pin and disk. Loss factor, X, is assumed to

be a linear coefficient term between 0 and 1 relating the coverage fraction of the disk surface

covered by the pin leaving of the contact interface to the coverage fraction of the disk surface

entering the contact interface. Factors affecting X include the normal load applied to the surface

through the pin as well. The terms of this relationship are defined below

,, = Coverage fraction entering contact interface
out = Coverage fraction exiting contact interface
A = Loss Factor

The coverage fraction leaving the contact interface is then defined as

Oout = A in (3-7)









Substituting Equation 3-6 into Equation 3-7 yields the following expression to determine

the fractional coverage for any given rotation N. This captures the transient behavior of the

coupled film deposition and removal process.

O,, = 1-(1- A. ,, _). (1-A )" (3-8)

Where N is the current cycle and n is the number of platelets that are deposited each

revolution of the disk. The final solution to the coupled deposition and removal sequence is

An .((1- AO)n) A ((1- AO)n)(N+') -1+(1- A0)n
0 = A(3-9)
S-.((1- A0)")" -1

The exponent n is the number of platelets deposited on any fractional area of the surface

between contacts. Powder mass flow rate is assumed to have a directly proportional effect on n,

as higher flow would increase the coverage fraction per unit time. Surface speed of the pin on

the disk is assumed to have an inversely proportional impact on n, as a faster surface speed

would reduce the amount of time for powder to enter the contact interface. Given these

considerations, the exponent n is assumed to be predicted by the following expression


n oc (3-10)
c)-r

m = mass flow rate (g/sec)
where o) = spindle speed (rad/sec)
r = radial pin location (mm)

The term N is the cycle count and can be calculated by



09.t
N 0 (3-11)
2/-









where t = time (sec). The term X describes the powder removal through the pin on disk

interface as a ratio of coverage fraction exiting the interface to the coverage fraction entering it.

Greater contact pressure would displace more powder per revolution thus reducing X. Contact

pressure is determined by the normal load, Fn. It is assumed in this model that X is

logarithmically proportional to the applied normal load and is described below


SoC e F) (3-12)

As the system approaches steady state, N approaches infinity. Equation 3-9 predicts that

the XN term, with X being a fraction between zero and one, will reduce to zero at steady state.


(1 -AO)- (3-13)
S-(I1- AO)" 1

This model assumes the amount of frictional load is determined by the coverage fraction.

The total friction coefficient is predicted to equal the sum of the coverage fraction for each

surface condition (i.e., powder or steel) multiplied by the friction coefficient for each surface

versus the pin.

total = 8 -IH3BO3 + (1 ). Alsteel (3-14)

where [LH3BO3 is the friction coefficient between boric acid and pin material. Experimentally,

Equation 3-14 will be verified by setting up an experimental matrix that varies the four

parameters predicted to affect the coverage fraction. Those four parameters are normal load,

spindle speed, powder mass flow rate, and radial location of the pin on disk contact interface.









CHAPTER 4
EXPERIMENTAL SETUP

For the effectiveness of the lubricating properties of boric acid powder to be characterized,

experiments studying surface to surface contact, especially the friction and wear between the

surfaces, needed to be conducted. The following chapter depicts the variables and methods

chosen to conduct the study of boric acid as a solid lubricant in metal on metal sliding contact.

4.1 Specimen Characterization and Preparation

The Tribology Lab at the University of Florida has a pin on disk tribometer that creates

relative rotating motion of a flat disk surface that maintains contact with a loaded stationary pin

surface. The pin samples in this experiment were 302 stainless steel balls of 4.76 mm (0.1875")

diameter purchased from McMaster Carr (part number 9291K18). As received, they had a

reflective surface finish so no further processing of the ball surface was performed prior to

experiment. The surface of the ball was analyzed using the Wyco Scanning White Light

Interferometer at MAIC (Major Analytical Instrumentation Center) at the University of Florida.

The mean surface roughness of the balls was 150 nm. The rotating surfaces in the experiment

were from 302 stainless steel disks with a 50.4mm (2.0") diameter. The disks were

manufactured from round stock purchased from McMaster Carr. The round stock was cut into

0.250 in. disks on a machine shop band saw. The disk surfaces were prepped using 3 stages of

sanding and polishing, using 100 grit, 250 grit, and 600 grit polishing paper, until all

manufacturing marks were removed and a mirror finish was obtained. Surface roughness

measurements on the prepared disk surface were repeated on 15 samples using white light

interferometry. Table 4-1 displays the results of these surface measurements. Average surface

roughness of the disks was calculated to be Ra 172 nm, O = 3 nm. Prior to installation to the

apparatus, each sample, pin and disk, were washed and sonicated in methyl alcohol.









Table 4-1. Surface Roughness Measurement results for prepared surface of 302 Stainless Steel
disks. Measurements performed on the WYCO White Light Interferometer at the
Major Instrumentation Analytical Center (MAIC) at the University of Florida
Scan # Ra Rms 20 Point Peak to 2 Point Peak to
Scan #
(nm) (nm) Valley (mm) Valley (mm)
1 288.0 367.1 3.3 3.8
2 287.5 366.2 3.3 3.8
3 287.5 366.0 3.4 3.9
4 286.4 364.8 3.3 3.7
5 286.4 364.7 3.3 3.6
6 288.2 367.1 3.3 3.7
7 287.8 366.5 3.3 3.6
8 287.6 366.2 3.2 3.8
9 287.7 366.3 3.2 3.6
10 287.5 366.1 3.3 3.9
11 288.0 366.7 3.4 4.3
12 286.3 364.7 3.3 3.7
13 287.2 365.8 3.3 3.7
14 287.5 366.3 3.3 3.7
15 289.2 367.9 3.3 3.6

RaAvg RaStdDev
287.5 0.7

The powder used in this study was technical grade (99% pure) with particle sizes ranging

from a few micrometers up to 100+ micrometers. The boric acid was purchased from Fischer

Chemical, p/n A74-3. The particle size of the powder was measured under an SEM at Major

Analytical Instrumentation Center at UF. Figure 4.1 shows the SEM photos of the boric acid.

Figures 4-2 and 4-3 show the diagram of the Pin On Disk system and apparatus used in this

experiment.

4.2 Pin on Disk Tribometer

The pin on disk tribometer used in this experiment is shown in Figure 4-3 Measurements

from the pin on disk tribometer were made through a compression load cell.





































Figure 4-1. Actual boric acid powder used in this experiment taken using the Scanning Electron
Microscope at the University of Florida

powder
delivery compression clean air
nozzle load cell flow
5,000 RPM Fn \
spindle & motor boric acid
& nitrogen gas







plastic enclosure



boric acid boric acid
collection & nitrogen gas
flow









Pin-on-Disk Tribometer
Figure 4-2. Rotating pin-on-disk tribometer used in this study. Airborne particle management
handled with the shop vacuum and filter stack.


powders used in this study










pin sample compression
& load cell
holder powder delivery
nozzle counterweight
50mm diameter e gimbal c
counterface (y& z rotations) -
0 0\ veunicaiadjusiment
5.000 RPM I ,nlcr| o0er
spindle ar. iriyoor
radial atdjugtmenr
optical table M -" + I rI d adju."menr
grid -





dead weigt loads --
Figure 4-3. Rotating pin-on-disk tribometer used in this study.

The load cell measured force from the pin sample holding arm due to the frictional

loading of the pin from the rotating disk. The other end of the arm supported the static load and

fixed the stationary ball to the desired location on the disk surface. Disk samples were fixtured

to the spindle and the testing surface exposed to contact with the pin sample. The pin on disk

tribometer spindle and motor had the capacity to rotate up to 4700 rpm in either clockwise or

counter clockwise orientation. Clockwise rotation was chosen to ensure constant contact

between the sample holding arm and load cell. The arm pivoted on a gimble and was allowed

pitch and yaw rotations. A threaded dead weight located at the end of the arm and a vertical

adjustment micrometer allowed for adjustment to balance the pitch of the sample holding arm.

When the pitch was properly balanced, the arm was considered level and an unloaded pin barely

made contact with the disk surface. A radial adjustment micrometer adjusted the radial location

of the system. When the system was radially balanced, the pin sat directly on the center of the

disk. Radial balance was calibrated by running a pin against a disk and the radial micrometer

adjusted until the wear track became a single dot.









4.3 Powder Delivery System

A system needed to be developed to aerosolize and continually deliver boric acid into the

contact interface. Nitrogen gas was first compressed and cleaned through the filter stack. There

it underwent two stages of filtration and desiccation to ensure gaseous purity. The nitrogen gas

was then regulated to 40 PSI and again desiccated. The nitrogen passed through a flow meter,

which allowed for variable flow output measured by a floating ball and calibrated scale. This

flow setting was the regulation for controlling the mass flow rate of boric acid in the pin disk

interface. Out of the flow meter, the clean, dry, metered gas then passed through 14" stainless

steel pipe into the boric acid bottle, shown in Figure 4-4 below.



nitrogen flow


nitrogen with '
boric acid particles .* .




fluidized bed of
boric acid powder




annular diffuser




Figure 4-4. Particle delivery scheme. The container was mounted on a mechanical shaker during
operation. Dry nitrogen gas flowrates were controlled through a series of flowmeters.

The boric acid bottle was mounted on a shaker, which allowed for enough particle

agitation to fluidize the powder and minimize powder clumping. As an additional method to

eliminate clumping, the powder was triple sifted prior to loading the bottle. As shown in the









figure above, the nitrogen input stem had many holes drilled through it and was formed into an

annular profile to have a large contact area with the powder. When the bottle was assembled, the

input stem descended into the powder. The gas pressure created a fluidized bed, which entrained

boric acid particles in the output stream. From the boric acid bottle, the nitrogen-boric acid

mixture flowed through a 1/4" stainless steel pipe to the pin on disk interface.

The flow meter setting was calibrated for powder flow rate prior to conducting the

experiments. The control variable to set the powder delivery rate was the flow meter setting,

which determined the gas flow into the powder delivery system. The flow meter had convenient

flow setting demarcations of 10, 20, 30, and 40. The powder container was loaded with enough

boric acid to establish a system weight of 600 grams prior to each calibration run. The mass of

the bottle loaded with powder was weighed prior to running gas through the system and

incrementally in five minute time steps. The time steps were chosen to verify consistency of the

flow rate over the run times observed throughout the experiment.

Figures 4-5 and 4-6 show the results of the powder delivery calibration. Figure 4-5 shows

the total powder delivery rate where the bottle weight measurements are referenced to the initial

weight and time (mo = 600 grams ; to = 0 min). The equation defining the powder mass delivery

rate referenced to to and mo is

m= M -m (4-1)
t -t

where ti is the time at iteration i and mi is the powder mass in the bottle at time = ti. Figure 4-6

shows the powder delivery results incrementally. Each powder delivery rate measurement

referenced the previous time step. The equation defining this incremental powder mass delivery

rate, which shows the impact of time on delivery rate, is given in Equation 4-2.














---10 FM setting
-2- 20 FM setting
--3- 0 FM setting
-0 --40 FM Setting


















0

0 5 10 15 20 25 30
Time (min)

Figure 4-5. Powder delivery calibration curves referenced from initial conditions, t = 0 min; mass

of boric acid in container = 600g, powder deliver rate = (m mo) / (tx to)



---10 FM setting

-*-30 FM setting
S--40 FM Setting




















0 5 10 15 20 25 30
Time (min)


Figure 4-6. Incremental powder delivery calibration curves referenced from previous time step;

powder deliver rate = (mx mx-1) / (tx tx-)










m=m n m (4-2)


Figures 4-5 and 4-6 show the non-linear nature of the powder delivery system. This

variability is due to several factors. The most significant issue is that there is no compensation

for the constant gas flow rate yet depleting powder levels. As time progressed, lesser amounts of

boric acid were able to get entrained in the outflow stream. Though the powder delivery rate was

not constant, the flow meter setting did correspond with a higher powder delivery rate

throughout the experiment. The actual powder delivery rate was verified for each experimental

run, by weighing the bottle prior to and following each run. For the experiments, settings 10, 20,

and 30 were chosen. A flow meter setting of 40 for the 60 minute test runs would have required

a larger powder container.

4.4 Environment Chamber

An environment chamber was implemented to the pin on disk tribometer set up to contain

boric acid, which could be a slip and inhalation hazard. The panels of the environment chamber

are made of Lexan sheets, providing clear view of the experimental apparatus, yet provide a

lightweight assembly and removal. A 2 horsepower shop-vac pulled a vacuum to create flow

throughout the chamber and reduce powder build up on the hardware. An air-filter stack was

attached to the chamber opposite the shop-vac interface, with a sheet filter attached to the inlet.

This provided clean air flow across the pin-disk interface.

4.5 Determination of Experimental Matrix

A matrix of testing variables was created to test the effectiveness of a continuous stream of

solid boric acid into the contact surface between the pin and disk. Using the predictions of the

mathematical model, the control variables were selected. The experimental parameters studied

in the experiment were normal force, sliding speed, wear track diameter, and powder flow rate.









The normal force was modified by adding a dead weight to the arm directly above the pin.

Sliding speed was adjusted by changing the spindle rotational speed. The adjustment for the

wear track diameter was set with the radial adjustment micrometer. The powder flow rate

measurement and adjustment was described in the previous section. Table 4-2 summarizes the

conditions tested in the experiment.

Table 4-2. The experimental matrix (3*3*3*3) plus nine repeats at the midpoint and one
unlubricated control.
Track Flow Meter
Fn (N) Q (RPM) Diameter, D (mm) Setting
0.65 400 3.8 10
3.30 2,000 25.0 20
6.30 4,000 38.0 30

The experimental matrix resulted in 81 individual tests. The midpoint condition (3.3 N

Load, 2000 rpm, 25.4 mm track diameter, flow meter setting = 20 ) was tested 9 additional times

to exhibit repeatability of the friction measurement. One midpoint test with no boric acid

delivery was conducted to provide a baseline measurement for friction between bare metal on

metal contact. The lengths of the tests were set to maintain 24,000 rotations of the disk, where

24,000 disk revolutions were divided by the spindle speed. The combination of wear track

diameter and spindle speeds allow for a range of sliding speeds that varied by 100 times.

The sequence of experiments was randomized to prevent any time-related bias as shown in

Table 4-3. As described earlier, the data acquisition system collected loading measurements of

the gimble arm on the load cell, which correlated to friction loads and friction coefficient of the

ball on the disk. The selection of test conditions allows for significant flexibility in grouping

data for analysis. For example, 27 tests were performed at constant diameter while varying

sliding speed, contact force, Fn, and flow rate, f; and 18 tests were performed at constant sliding

speed while varying track diameter, D, and rotational speed, Q, Fn, and f The variability of

testing conditions allowed for an expansive, yet detailed, study of the parameters impacting the












effectiveness of a solid boric acid lubricant on the wear rates and frictional forces in pin on disk


contact. Laboratory ambient conditions were measured daily before testing. Relative humidity


varied from approximately 20 to 55%. The ambient temperatures ranged from 250C to 280C.


Table 4-3. Randomized experimental testing sequence with all parameters per run.
Run Flow Meter Spindle Dia (in) Normal Length IRun Flow Meter Spindle Dia (in) Normal Length


Number Setting


(sec) INumber


(grams)


10 400 1.50 66.3
20 400 0.15 336.6
20 400 1.00 66.3
20 4000 0.15 66.3
10 4000 1.50 336.6
10 4000 1.00 642.5
10 4000 0.15 642.5
10 4000 1.50 66.3
10 2000 1.50 642.5
20 2000 1.00 336.6
20 2000 1.00 66.3
20 400 1.50 642.5
30 400 1.00 336.6
10 2000 0.15 642.5
20 4000 1.00 66.3
30 400 1.50 336.6
20 2000 1.50 642.5
30 2000 0.15 642.5
30 2000 1.50 642.5
20 4000 0.15 336.6
30 2000 1.00 642.5
10 400 1.50 336.6
20 2000 0.15 66.3
20 4000 1.50 336.6
20 2000 1.00 336.6
30 4000 1.50 66.3
20 4000 1.00 336.6
20 2000 1.50 66.3
30 400 0.15 642.5
30 2000 1.50 66.3
30 400 1.00 642.5
10 400 0.15 336.6
10 400 1.00 336.6
10 4000 1.50 642.5
20 2000 1.00 336.6
30 4000 1.50 642.5
10 4000 0.15 66.3
20 400 1.00 642.5
30 2000 1.00 336.6
30 4000 0.15 336.6
10 400 0.15 642.5
30 2000 1.00 66.3
30 2000 1.50 336.6
10 2000 1.50 66.3
10 400 1.50 642.5


Setting


Speed
(RPM)


Load
(grams)


20 2000 1.00
30 4000 1.50
20 2000 1.00
10 4000 1.00
20 400 1.50
20 2000 1.50
30 400 1.50
10 400 1.00
30 4000 1.00
30 2000 0.15
30 4000 0.15
20 2000 1.00
30 400 0.15
30 400 1.00
10 2000 0.15
10 400 0.15
20 4000 1.50
30 4000 1.00
30 400 0.15
20 4000 1.00
20 4000 1.50
30 400 1.50
20 2000 1.00
30 4000 1.00
20 400 1.00
30 2000 0.15
20 400 0.15
20 2000 1.00
10 2000 0.15
20 4000 0.15
10 4000 0.15
10 2000 1.00
30 4000 0.15
10 400 1.00
20 2000 0.15
20 2000 0.15
10 2000 1.00
20 400 0.15
20 400 1.50
10 2000 1.50
10 2000 1.00
20 2000 1.00
10 4000 1.00
20 2000 1.00
20 2000 1.00


(sec)


642.5 720
336.6 360
336.6 720
66.3 360
336.6 3600
336.6 720
66.3 3600
66.3 3600
336.6 360
336.6 720
642.5 360
336.6 720
336.6 3600
66.3 3600
66.3 720
66.3 3600
642.5 360
642.5 360
66.3 3600
642.5 360
66.3 360
642.5 3600
336.6 720
66.3 360
336.6 3600
66.3 720
642.5 3600
336.6 720
336.6 720
642.5 360
336.6 360
66.3 720
66.3 360
642.5 3600
336.6 720
642.5 720
642.5 720
66.3 3600
66.3 3600
336.6 720
336.6 720
336.6 720
336.6 360
336.6 720
336.6 720


* shaded entries denote a midpoint run


Speed
(RPM)


_ __ __ ___ ____ __ ____ __ ___ ___









CHAPTER 5
RESULTS

The critical measurements taken in this study were the thickness of the wear track left on

the sample disk and the load on the load cell. The wear track diameter and thickness are the key

measurements to correlating the wear rate and ultimately the effectiveness of the boric acid

powder as a solid lubricant. To measure the thickness of the wear track, an optical microscope

was used. A calibrated measurement grid was attached to the scope, and the inside of the wear

track was located at 0.000". The measurement at the outside of the wear track then established

the wear track thickness. Measurements were made at each quadrant of the wear track to

establish consistency. Those measurements were averaged and the standard deviation taken.

The wear rate for the pin was used as the metric for wear in this study. To calculate the wear on

the pin, the volume of material lost was first determined. Equation 5-1 was used to determine

the pin volume lost due to wear. Figure 5-1 graphically shows the relationship between the wear

scar diameter and the lost volume.

h2 (3R h)
V= (5-1)
3

where V is the volume of the pin lost due to wear, R is the radius of the pin and h is the radial

height of the lost volume. Calculated wear rates for all runs are summarized in the table below.





Sh





Figure 5-1 Pin volume lost due to wear during the experiment










Table 5-1. Results from the experimental matrix. D is wear track diameter in millimeters, f is
commanded boric acid flowrate in grams/minute, K is wear-rate x 10-6 mm3/(Nm).
Q = 400 RPM Q = 2,000RPM Q = 4000 RPM
Fn = 0.65 N Fn = 0.65 N Fn = 0.65 N
D F K D F K D F u K
3.8 10 0.196 2.92 3.8 10 0.157 3.8 10 0.090
3.8 20 0.080 3.8 20 0.094 14.93 3.8 20 0.085
3.8 30 0.138 3.8 30 0.085 1.98 3.8 30 0.092 0.63
25.0 10 0.147 0.30 25.0 10 0.036 36.65 25.0 10 0.086 0.22
25.0 20 0.122 25.0 20 0.035 42.48 25.0 20 0.075
25.0 30 0.159 0.67 25.0 30 0.062 21.31 25.0 30 0.047 4.90
38.0 10 0.147 0.44 38.0 10 0.058 1.25 38.0 10 0.080 0.21
38.0 20 0.138 0.03 38.0 20 0.077 0.38 38.0 20 0.049 4.13
38.0 30 0.173 0.19 38.0 30 0.045 1.56 38.0 30 0.089 0.57

Fn= 3.3 N Fn= 3.3 N Fn = 3.3 N
D F K D F K D F u K
3.8 10 0.143 0.23 3.8 10 0.148 0.82 3.8 10 0.081 1.48
3.8 20 0.139 4.03 3.8 20 0.12 3.8 20 0.088 5.55
3.8 30 0.162 0.31 3.8 30 0.11 0.12 3.8 30 0.087 0.03
25.0 10 0.152 0.76 25.0 10 0.056 0.80 25.0 10 0.055 1.05
25.0 20 0.156 0.48 25.0 20 0.044* 1.54* 25.0 20 0.035 0.52
25.0 30 0.163 0.41 25.0 30 0.05 2.00 25.0 30 0.030 1.55
38.0 10 0.113 9.02 38.0 10 0.049 0.14 38.0 10 0.321 0.03
38.0 20 0.116 1.36 38.0 20 0.039 0.27 38.0 20 0.040 0.34
38.0 30 0.101 13.26 38.0 30 0.045 0.19 38.0 30 0.026 0.69

Fn = 6.3 N Fn = 6.3 N Fn = 6.3 N
D F K D F K D F u K
3.8 10 0.155 0.42 3.8 10 0.117 0.35 3.8 10 0.086 0.07
3.8 20 0.138 0.06 3.8 20 0.108 3.8 20 0.079 0.60
3.8 30 0.165 3.8 30 0.103 3.8 30 0.078 0.85
25.0 10 0.137 5.28 25.0 10 0.036 0.24 25.0 10 0.059 0.27
25.0 20 0.122 1.05 25.0 20 0.061 0.72 25.0 20 0.029 0.46
25.0 30 0.121 0.01 25.0 30 0.059 3.44 25.0 30 0.035 1.09
38.0 10 0.100 20.78 38.0 10 0.047 0.30 38.0 10 0.230 2.35
38.0 20 0.106 32.86 38.0 20 0.039 0.21 38.0 20 0.023 14.08
38.0 30 0.090 21.15 38.0 30 0.046 0.09 38.0 30 0.073 0.06
Midpoint condition uavg = 0.044, standard deviation, a 0. 009; Kvg = 1.54 a = 1.15










The other critical measurement taken was the frictional load on the load cell. The location

of the load cell is at a one to one distance from the arm pivot. This corresponds to a one to one

loading relationship between the load cell measurement and the radial loading of the arm relative

to the rotating disk. The coefficient of friction is then approximated using the Equation 5-2:


P = (5-2)
N

where N is the dead weight load applied to the arm. The friction coefficient measurements were

averaged over the duration of the low friction state and reported in Table 5-1 above. The nine

repeats of the midpoint test condition and the control (midpoint condition with no lubricant) are

plotted together in Figure 5-2.



0.8 no boric acid flow
J load cell saturation
0.7
o 0.6
S-- Run 10
'4-
S0.5 ---'n 25
S4 spikes related to momentary
u loss of powder delivery
03 \ ..----. ,,,n 57
0 Run 68
P0.2 iw7 73
0.12 If 4 -- Run 87
0.1 I Run 90

0
0 4 8 12 16 20 24
number of cycles (thousands)

Experimental conditions: Fn = 3.3 N, 2 = 2,000 RPM, D = 25.4 mm, f= 20 g/min.

Figure 5-2. Friction coefficient versus cycle number for the 9 repeat tests and the control, which
had no boric acid in the gas flow.

This plot displays the effect on friction coefficient of the boric acid as compared to

interfacial contact with no lubricant. The figure shows how the boric acid, once introduced into










the contact interface, quickly reduced the friction coefficient. The spikes show interruptions in

powder flow, and show the sensitivity of the system to the presence of the lubricant.

The powder delivery system experienced complete delivery failure on two experimental

runs. This powder delivery failure occurred due to the delivery tubes clogging with powder.

Figure 5-3 shows the effects of clogging on the friction coefficient.



0.5
powder delivery
failure
0.4
c Fn=6.3 N
Uo
S0.3

0.2 Fn=3.3 N
"o 0.2


0.1


0
0 I I l-2 lI6 i I
0 4 8 12 16 20 24
number of cycles (thousands)

Experimental conditions: D = 25 mm, surface speeds =5. 0 m/s, Fn identified in the
plot.

Figure 5-3. Friction coefficient versus cycle number for the experiments with terminal blockages
in the boric acid delivery during the experiment.

The data was parsed several ways to search for trends in wear rate and friction coefficient.

Since there were four variables in the experiment, a series of plots needed to be created. For

each wear track diameter, three plots were created to detail the relationship between average

friction coefficient and load, flow rate, and sliding speed, respectively. This series of plots is

shown in Figures 5-4. Efforts were made to correlate the experimental data to the Von Karman

mathematical model. Figure 5-5 shows the relationship between average friction coefficient and

boundary layer thickness.











wear track diameter 25 mm
surface speed 2.5 m/s


10grams/minute
20grams/minute
30grams/minute


0.20


0.15


0.10


0.05



06


6 7


0.20 wear track diameter 25 mm
[ surface speed 2.5 m/s


0.15


0.10


Fn= 0.65 N
Fn = 3.30N
Fn=6.30N


{


0 1 I I I
0 0.05 0.10 0.15 0.20
measured flowrate (grams/second)


wear track diameter 25 mm
normal load 3.3 N


10 grams/minute
20 grams/minute
30 grams/minute


0 1 2 3 4 5
sliding speed (m/s)


Figure 5-4. Average friction coefficient versus normal load, flow rate, and sliding speed.


0.20


0.15




0.10




0.05


8 normal load (N)

o 0.63
o 3.3
o 6.3

1 2 3 4
boundary layer thickness (mm)


Figure 5-5. Average friction coefficients for all experiments plotted versus the boundary layer
thickness. The data is separated for the various normal loads as indicated.


C


0
U
t-


0
4-

(0


1 2 3 4 5
normal load (N)


0.20




0

S0.10
U
u

r-
m .0
0)


I I I I I I I









Many experiments showed a high degree of variability in the time taken to get to low

friction. Though the exact reason for this variability was undetermined, it is hypothesized that

unreliability in the powder delivery apparatus played a major role. As described in Chapter 4,

many precautions were taken to ensure the powder would be clump free and the air would be

properly desiccated. The mechanism responsible for delivering the boric acid powder to the Pin

on Disk interface is unknown, but significant variability with the delivery was detected. Figure

5-6 shows two runs with different times to reach a low friction state. As is shown in the figure,

the measured wear rates of the pin differ by two orders of magnitude. The measured friction

coefficients are not significantly different.


0.5

Fn=0.65N
0.4 7 5 mm3
S K~ 3.7x105 Nm





c 0.2
0


0.1 Fn=3.3N
K 7.5x1 0-7 m


0 4 8 12 16 20 24
number of cycles (thousands)
Experimental Conditions: D = 25 mm, surface speed =2.5 m/s, normal loads are
identified in the figure.

Figure 5-6. Friction coefficient versus cycle number for the two experiments that had the shortest
and longest transients to low friction coefficient.









CHAPTER 6
DISCUSSION

When the boric acid was successfully flowing to the interface, friction and wear between

the pin and disk was significantly reduced. This trend is shown in Figure 5-2, where the control

condition is plotted against the midpoint runs. The only difference between the control and the

midpoints runs was the presence of boric acid lubricant. The average friction coefficient drops

from 0.641, non-lubricated sliding contact, to 9.3x10-4 when boric acid is introduced to the

system. That represents a two order of magnitude reduction in friction force for the same

operating conditions. The measured friction coefficient is rapidly impacted by the presence of

boric acid in the contact interface. Figure 5-2 details how a low friction state is achieved after

tens of revolutions.

The wear rate of the pin shows a similar trend to the friction reduction, reducing from 4.86

x 10-4 mm3/Nm to 1.55 x 10-6 mm3/Nm when boric acid is introduced into the sliding contact

interface. Photographs detailing the difference in wear between the lubricated and non lubricated

test conditions are shown in Figure 6-1.


Figure 6-1. Wear marks on pin and disk after midpoint experimental conditions A) with boric
acid flow and B) without boric acid flow









The measured friction coefficient is very sensitivity to interruptions in flow. Figure 5-3

shows several spikes in friction coefficient that corresponded to perturbations in flow. In Figure

5-3, which highlights the two runs where the powder delivery system failed prematurely, the

friction coefficient spikes at the point of failure and stays high for the duration of the run. As

was seen with the intermittent delivery failures, the transition from low to high friction

coefficient occurred within tens of revolutions. This transition from low to high friction suggests

that the rate of powder deposition in the contact interface is very close to the rate of removal.

In Figure 5-4, the plots detail the effect on friction coefficient by a single experimental

variable, while holding all other variables constant. Plot 5-4a show the effect of normal load on

friction coefficient, in groups of powder flow rate, while holding wear track and surface speed

constant. Across the data plots, there was no apparent significance between coefficient of

friction and normal load or sliding speed. In Figure 5-4c, the lower sliding speeds resulted in a

higher coefficient of friction than the higher sliding speed tests under the same conditions. Using

the competitive rate models for the deposition and removal of boric acid, it is hypothesized that

removal rate should increase with increasing sliding speed. This would lead to less boric acid

getting entrained in the contact zone and result in a higher coefficient of friction. The

experimental data suggested the opposite to be true. Several conclusions may be drawn from this

data, however. It may suggest that the removal rate is suppressed at higher speed, thereby

increasing the amount of boric acid in the interface and reducing friction. The data may also

suggest that at higher sliding speeds, the rate of film formation on the surface of the disk and/or

pin increases.

It is hypothesized that the direct injection of nitrogen and boric acid into the interface was

not the primary mechanism for boric acid transport to the pin contact. Using the model solved









by Von Karman [13,14], it is hypothesized that the fluid flow generated across the rotation of the

disk sample entrained the boric acid and delivered it to the pin contact. Figure 5-5 shows the

results of Van Karman's flow model. The rotation of the spinning disk pulls air above the disk

down at the center of the disk and discharges it radially outward. The air flow across the disk is

assumed to be laminar if the Reynold's number is less than 300,000. The Reynold's number is

given by Equation 6-1, where

Re= *R (6-1)


Re is the Reynolds number dimensionlesss), V is the peripheral speed (in m/s), R is the radius of

the disk (in mm), and v = 1.5x10 5 is used for the kinematic viscosity of the air. The highest

Reynold's number encountered in this experiment was 4000, signifying that the flow across the

disk was laminar. Rogers and Lance [24, 25] provided a numerical solution to this problem.

Their study demonstrated that the boundary layer thickness, 6, is relatively constant across the

disk and varies inversely with the square-root of the angular speed. The following equation

describes the results of their study


3= 5.4 V (6-2)


where a is the angular speed of the disk (in rad/s), and again v = 1.5x10 5 m2/s is used for the

kinematic viscosity of air. The entire data set is plotted versus the calculated boundary layer

thickness in Figure 5-5. The data trends in Figure 5-5 suggest that the thinner and higher speed

flows are more efficient at delivering boric acid to the contact.

Figure 5-2 details the measured friction coefficient for unlubricated pin on disk stainless

steel contact. The average wear rate was measured to be K = 5.0x10-4 mm3/Nm. All lubricated

experiments conducted showed significant decreases in pin wear rate, as shown in Table 5-1.









Eleven of the pin samples showed no detectable wear marks, these are indicated by dashes in the

wear rate column of Table 5-1. The lowest measured wear rate was K = 7.5 x10-4 mm3/Nm,

which occurred during the 25 mm diameter, 3.3 N load, 2.5 m/s sliding speed, and 0.078

grams/second boric acid flowrate conditions. The experiments exhibit over 500 times greater

improvement in wear resistance over the unlubricated contact measurements.

Substantial effort was made to correlate the friction coefficient and the wear rate with the

selected experimental variables. There were no strong indicators tying the experimental

variables to either friction coefficient or wear rate. The only qualitative explanation for the

variation in wear rate is the significant loss of material during start up transients. The tests were

initiated on nascent surfaces with the boric acid flow expected create to replenish a surface film

in situ. The startup transient varied widely between experimental runs, though all experiments

eventually reached a low friction state. It is suggested that the majority of material lost to wear

occurs during this period, and better understanding of the film formation must be developed.

Figure 5-6 shows the friction coefficient plots for experiments with the shortest and longest

initial transient conditions.

These experiments were both at a radial position of 2.5 mm and a sliding speed of 2.5 m/s,

although they were at two different normal loads. The experiment with the shortest transient had

a wear-rate of 7.5x10-7 mm3/(Nm) and the test with the longest transient had a wear-rate of

3.7x10-5 mm3/(Nm). The ratio of the volumes lost between the shortest and longest transient was

0.1 (i.e., the test with a shorter transient lost 10 percent of the material of the test with the longest

transient), while the ratio of the frictional energy dissipated during the transient region is about

1/3.









CHAPTER 7
CONCLUSIONS

* Conclusion 1: These experiments were the first reported demonstrations of continuous
powder delivery of boric acid as a solid lubricant.

* Conclusion 2: These experiments clearly indicate that powder delivery of boric acid is a
viable technique for providing in situ lubrication for concentrated metal contacts. Wear
rate reductions of over 100 times and friction coefficients of well under 0.1 were
demonstrated.

* Conclusion 3: Further development and refinement of the powder delivery system must
be explored to develop an accurate understanding of the impact of the variables on
friction coefficient and wear rate.

* Conclusion 4: A future study is suggested for exploring the impact of a pre-deposited
boric acid film on the surface of pin and/or disk to verify the impact of the initial
transient on friction coefficient and wear rate.









LIST OF REFERENCES


[1] P. Hamblin, Environmentally compatible lubricants: trends, standards, and terms. Proc.
Environmental Aspects in Production and Utilization of Lubricants, Sopron, (1995) 1-10

[2] W. Bartz, Lubricants and the environment, Tribology International 31 (1998) 35-47

[3] G. Johnson, Milacron Marketing Co, CIMCOOL technical report, (1999) 1-12

[4] A.R. Machado, J. Wallbank, The effect of extremely low lubricant volumes in machining,
Wear 210 (1997) 76-82

[5] M. Rahman, A. Senthil-Kumar, Salam M.U., Experimental evaluation on the effect of
minimal quantities of lubricant in milling, International Journal of Maching Tools and
Manufacture 42 (2002) 539-547

[6] F.M. Kustas, L. FehrehnbacheR, R. Komanduri, Nanocoatings in cutting tools for dry
machining, Annals of the CIRP (1997) 39-42

[7] L.W. Jelinski, T. Graedel, R. Laudise, D.W. McCall, C.K Patel, Industrial ecology:
concepts and approaches, Proceedings of the National Academy of Sciences of the United
States of America 89 (1992) 793-797

[8] L. Ward, Influence of boric acid and borax on digestion and health, Science 20 (1904) 26-
27

[9] G. Brown Jr, Remarks on industrial ecology, Proceedings from the National Academy of
Sciences 89 (1992) 876-878

[10] J. Sutherland, K. Gunter, D. Allen, D. Bauer, B. Bras, T. Gutowski, C. Murphy, T.
Piwonka, P. Sheng, D. Thurston, E. Wolff, A global perspective on the environmental
challenges facing the automotive industry: State-of-the-art and directions for the future, Int
J Vehicle Des 35 (2004) 86-110.

[11] D. Allen, D. Bauer, B. Bras, T. Gutowski, C. Murphy, T. Piwonka, P. Sheng, J. Sutherland,
D. Thurston, E. Wolff, Environmentally benign manufacturing: Trends in Europe, Japan,
and the Usa, J Manuf Scie E-T ASME 124 (2002) 908-920.

[12] P. Lyday, Mineral commodity summaries, US Geological Survey, (2003) 110-122.

[13] A. Erdemir, Tribological properties of boric-acid and boric-acid-forming surfaces. 1.
Crystal-chemistry and mechanism of self-lubrication of boric-acid, Lubr Eng 47 (1991)
168-173.

[14] A. Erdemir, R. Erck, and J. Robles, Relationship of hertzian contact pressure to friction
behavior of self-lubricating boric-acid films, Surf Coat Tech 49 (1991) 435-438.









[15] A. Erdemir, G. Fenske, and R. Erck, A study of the formation and self-lubrication
mechanisms of boric-acid films on boric oxide coatings, Surf Coat Tech 43-4 (1990) 588-
596.

[16] A. Erdemir, G. Fenske, R. Erck, F. Nichols, D. Busch, Tribological properties of boric-acid
and boric-acid-forming surfaces. 2. Mechanisms of formation and self-lubrication of
boric-acid films on boron-containing and boric oxide-containing surfaces, Lubr Eng 47
(1991) 179-184.

[17] B. Gearing, H. Moon, and L. Anand, A plasticity model for interface friction: Application
to sheet metal forming, International Journal of Plasticity 17 (2001) 237-271.

[18] I. Singer, R. Bolster, J. Wegand, S. Fayeulle, B. Stupp, Hertzian stress contribution to low
friction behavior of thin mos2 coatings, Appl Phys Lett 57 (1990) 995-997.

[19] M. Lovell, W.G. Sawyer, and A. Mobley, On the friction and wear performance of boric
acid lubricant combinations in extended duration processes, Tribol Tech 25 (2005) 73-81.

[20] N. Mccook, D. Burris, G. Bourne, J. Steffens, J. Hanrahan, W.G. Sawyer, Wear resistant
solid lubricant coating made from ptfe and epoxy, Tribo Lett 18 (2005) 119-124.

[21] T. Von Karman, Uber laminare und turbulente reibung, Z Angew Math Phys 1 (1921) 233-
252.

[22] M. Miklavcic, C. Wang, The flow due to a rough rotating disk, Z Angew Math Phys 55
(2004) 235-246.

[23] M. Rogers, G. Lance, The rotationally symmetric flow of a viscous fluid in the presence of
an infinite rotating disk, J Fluid Mech 7 (1960) 617-631.

[24] M. Rogers, G. Lance, Boundary layer on disc of finite radius in rotating fluid, Q J Mech
Appl Math 17 (1964) 319-25.









BIOGRAPHICAL SKETCH

The author was born in Oakland, California, in 1977. Timothy Barton was raised

in the San Francisco bay area for the duration of his childhood. In 1995, he moved to

Florida to attend the University of Florida to pursue a degree in mechanical engineering.

He graduated with his Bachelor of Science in mechanical and aerospace engineering in

2001. He obtained his Master of Science in mechanical and aerospace engineering from

the University of Florida in 2007.





PAGE 1

1 IN SITU LUBRICATION WITH BORIC ACID: POWDER DELIVERY OF AN ENVIRONMENTALLY BENI GN SOLID LUBRICANT By TIMOTHY PAUL BARTON 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 Timothy Paul Barton

PAGE 3

3 ACKNOWLEDGMENTS First and forem ost, I would like to sincerely thank my family, my gr andmother, Virginia Smith, my mother, Michele Barton, my sister s Bonnie and Beth, and my girlfriend Merigan Craig, for their unending support, motivation, and understanding throughout this entire academic endeavor. It is with the love fr om this well of inspiration that I am able to finally complete this journey. A very important acknowledgement goes out to my graduate adviso r, Dr W. Gregory Sawyer, without whom I would have never f ound direction and certainly not completed my graduate thesis. His support, academic, financial, motivational, helped to inspire me to complete this degree. I would also like to acknowledge all members of my committee, both past and present. Dr. John Schuller, Dr Nam Ho Kim, Dr John Zeigert, and Dr Tony Schmitz all played a great part in my academic growth, and I sincerely appreciate all of their efforts. I would like to thank Dr. Dan Dickrell, Pam Dickrell, and Ali Erdimer for their assistance and lending their insight with their previous works with using boric acid as a solid lubricant. I would like to note the members of the tribology lab, Dave Burris, Jason Steffans, and the rest and thank them for their help in ge tting acclimated to the lab, and certainly for their understanding as I made my way, often painfully fo r those around, through these experiments. I would also like to acknowledge anyone else who was involved with the administrative, academic, and financial efforts to complete this thesis.

PAGE 4

4 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................3 LIST OF TABLES................................................................................................................. ..........5 LIST OF FIGURES.........................................................................................................................6 ABSTRACT.....................................................................................................................................7 CHAP TER 1 INTRODUCTION....................................................................................................................8 1.2 Introduction to Petroleum Based Lubricants................................................................... 8 1.2 Environmental Concerns................................................................................................. 8 1.3 Health Concerns..............................................................................................................9 1.4 Cost Concerns.................................................................................................................9 1.5 Manufacturing Concerns...............................................................................................10 1.6 Legal Concerns..............................................................................................................11 2 LITERATURE REVIEW.......................................................................................................12 2.1 Boric Acid................................................................................................................ .....12 2.2 Health and Envi ronm ental Impact................................................................................ 12 2.3 Boric Acid as a Solid Lubricant....................................................................................13 3 MATHEMATICAL MODEL................................................................................................. 16 4 EXPERIMENTAL SETUP....................................................................................................20 4.1 Specimen Characterization and Preparation................................................................. 20 4.2 Pin on Disk Tribometer................................................................................................. 21 4.3 Powder Delivery System...............................................................................................24 4.4 Environment Chamber.................................................................................................. 27 4.5 Determination of Experimental Matrix......................................................................... 27 5 RESULTS...............................................................................................................................30 6 DISCUSSION.........................................................................................................................36 7 CONCLUSIONS.................................................................................................................... 40 LIST OF REFERENCES...............................................................................................................41 BIOGRAPHICAL SKETCH.........................................................................................................43

PAGE 5

5 LIST OF TABLES Table page 4-1 Surface Roughness Measurement results fo r prep ared surface of 302 Stainless Steel disks...................................................................................................................................21 4-2 Experimental matrix (3*3*3*3) plus nine repeats at the m idpoint and one unlubricated control........................................................................................................... 28 4-3 Randomized experimental testing se quence with all param eters per run.......................... 29 5-1 Results from the experimental matrix................................................................................ 31

PAGE 6

6 LIST OF FIGURES Figure page 2-1 Lamellar structure of boric acid......................................................................................... 13 2-2 Experimental setup for Gear ing et al. experim ent to dete rmine shear strength of boric acid.....................................................................................................................................14 2-3 Experimental setup and results for Lovell et al. lubrication stu dy using powder boric acid as a Solid Lubricant in a loaded pin on disk experim ent........................................... 15 4-1 Actual boric acid powder used in this experim ent taken using the Scanning Electron Microscope at the University of Florida............................................................................22 4-2 Rotating pin-on-disk tribom eter used in this study. ........................................................... 22 4-3 Rotating pin-on-disk tribom eter used in this study............................................................ 23 4-4 Particle delivery schem e.................................................................................................... 24 4-5 Powder delivery calibration curves referenced from initial conditions............................. 26 4-6 Increm ental powder delivery calibration curv es referenced from previous time step....... 26 5-1 Pin volume lost due to wear during the experiment.......................................................... 30 5-2 Friction coefficient versus cycle number for the 9 repeat tests and the control ................32 5-3 Friction coefficient versus cycle number for the experim ents with terminal blockages in the boric acid delivery during the experiment............................................................... 33 5-4 Average friction coefficient versus norm al load, flowrate, and sliding speed................... 34 5-5 Average friction coefficients for all experim ents plotted versus the boundary layer thickness.............................................................................................................................34 5-6 Friction coefficient versus cycle numb er for the two experim ents that had the shortest and longest transients to low friction coefficient................................................. 35 6-1 Wear marks on pin and disk after midpoint experimental conditions A) With boric acid flow, B) Without boric acid flow............................................................................... 36

PAGE 7

7 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 IN SITU LUBRICATION WITH BORIC ACID: POWDER DELIVERY OF AN ENVIRONMENTALLY BENI GN SOLID LUBRICANT By Timothy Paul Barton May 2008 Chair: W. Gregory Sawyer Major: Mechanical Engineering In-situ deposition of boric acid in dry pow der form is investigated as a potential environmentally benign solid lubricant for sliding metal contacts. Boric acid is widely used in industrial processes and agriculture, is not classified as a pollutant by EPA, and produces no serious illnesses or carcinogenic effects from exposure to solutions or aerosols. In this study, boric acid powder is aerosolized and entrained in a low velocity jet of nitrogen gas which is directed at a self-mated 302 SS sliding contact in a rotating pin-on-disk tr ibometer. The effects of powder flow rate, sliding speed, normal load, and track diameter on coe fficient of friction and wear rate are investigated. Friction coefficien ts below =0.1 can be consistently reached and maintained as long as the powder flow continue s. Wear rates are re duced over 2 orders of magnitude.

PAGE 8

8 CHAPTER 1 INTRODUCTION 1.2 Introduction to Petroleum Based Lubricants Petroleum based lubricants are widely used in the manufacturing and i ndustrial sectors, as well as in automotive and many other mass market pr oducts. It is well rec ognized that the use of these lubricants introduces signi ficant quantities of used petroleum based substances into the waste stream [1]. These lubri cants impose significant negative impacts to the environment and health both during their primary use and after disposal [2]. Th is study introduces a lubrication concept aimed at reducing the need for petroleum based fluids in a wide range of industrial processes and consumer applications by delivery of boric acid, an environmentally benign solid lubricant, in powder form. The proposed lubr icant and delivery method will avoid the waste stream and environmental and health impacts associated with other lubricants used in many industrial processes and products. This technique is not tied to a ny particular process or product, but rather it has broad applicab ility; although customized delivery strategies will need to be developed for the various applications. 1.2 Environmental Concerns The environm ent has long been an afterthought regarding manufacturing concerns, with the exception of clean up costs. As social awar eness towards the environment rises to critical levels, the impacts of industrial waste streams on the environment is subject to increased scrutiny and requiring meaningful action. As of 1995 in the United States, 32% of all lubricants return to the environment in a physically or visually alte red state [1]. Their impacts on water and air quality, wildlife, and human health have been found to be toxic. The behavior of the lubricant when returning to the environment will govern the impairment to the environment. It follows

PAGE 9

9 that there is significant push to develop lubricants with re duced toxicity and increased biodegradability. 1.3 Health Concerns Machining lubricants pose significant health hazards to hum an from both direct and indirect contact. Some risks of lubricants include the developmen t of nitrosamines in coolants, skin disease from unprotected cont act, carcinogens in used oils a nd lubricants, solvent containing products, and heavy metal compounds in additives [2]. If not properly tended to, metalworking fluid circuits can be subject to rancidity. Microorganisms such as bacteria, mold, and yeast cause rancidity by multiplying in metalworking fluid circ uits once exposed to air and water dilution. Skin dermatitis has been linked to microorganism infected metalworking fluid mixtures [3]. Significant odor issues can also ar ise and respiratory irr itation and infections have been linked to rancidity of metalworking fluids. 1.4 Cost Concerns The increasingly stricter environm ental regulations and corres ponding enforcement are reducing the flexibility of metalworking flui ds. As new metalwor king formulations are developed, they are missing Pb, S, or Cl com pounds, which provide superior machinability characteristics and are cost effective to produce. As noted above, the health and environmental concerns of the metalworking flui ds are cause for increased mainte nance and disposal costs. The labor and overhead costs in the U.S. in 1995 were estimated over $300 billion. These costs were estimated to comprise approximately 5-6% of the total manufacturing costs in the U.S in 1995. The costs associated with the use of cutting flui ds, including purchase, maintenance, storage, and disposal, is estimated to be about 16% of th e manufacturing costs, many times more than the labor and overhead costs [4].

PAGE 10

10 1.5 Manufacturing Concerns There have been m any attempts to address th e issue of reducing or eliminating petroleum based lubricants from metal working processes. The impact of Minimum Quantity Lubricant (MQL) volumes in machining was studied by Mach ado and Wallbank [5]. They compared the use of dry cutting (no lubricant), an air-water mixture, and air-s oluble oil lubricant (MQL) jet streams versus overhead flood coolant in lathe cut ting process. The results for cutting and feed forces, surface finish, and tool life for MQL riva led or exceeded those of standard flood cooling at low cutting speeds. However, the effectiv eness of MQL diminished as machining speeds approached those commonly used in industry. MQL lubrication also introduces a significant health risk by promoting a mist in the envir onment, requiring an extensive exhaus t extraction system. The results for air and water lubricat ion were promising, however, significant corrosion problems were noted. Kustas et al. [6] investigated the use of na nocoating on the cutting tools in a dry machining operation. The attempt of the study is to prov e the coatings will generate less heat during machining and/or take away heat generated ra pidly in dry machining by other means. One hundred bilayer 13Angstrom B4C / 18 Angstrom W nanocoatings were deposited on cemented WC-Co tools and HSS drills. Coated and uncoated dry machining tests were conducted. A 33% reduction in torque requirement and a noticeable reduction of wear on the tool flank surfaces were observed. No comparison between coated tool dry machining and non coated traditional petroleum based lubricant machining was made. Significant efforts are being made to reduce or eliminate the use of traditional petroleum based lubricants. Though none has yet risen to provide the combination of surface quality, production cost, and tool life to supplant the standard lubricants.

PAGE 11

11 1.6 Legal Concerns There are currently no US laws requiring th e use of eco-friendly lubricants, but two regulations, Executive O rder 12873 (EO 12783) and the Great Lakes Water Quality Initiative (GLWQI), have made significant im pacts on the use and disposal of conventional lubricants. EO 12783 provides preferential treatment of government contracts for bidders who use recycled oils and/or use environmentally compatible oils wh ere possible. The Great Lakes Water Quality Initiative puts stringent require ments on zinc limitations such th at the use of zinc containing lubricants is effectively banne d. Government regulations on us e of environmentally harmful lubricants, though placing limits on hu man activity with regards to th e environment, are far from exhaustive and many problems exist in spite of them. Over 30% of all lubricants return to the environment in a harmfully altered state. Outsid e the realm of regulation, there exists a need to find an environmentally benign alternative to th e current petroleum base d lubricants without sacrificing manufacturing quality or dramatically increasing costs.

PAGE 12

12 CHAPTER 2 LITERATURE REVIEW 2.1 Boric Acid This experim ent explores the potential of boric acid as an environmentally viable alternative to the petroleum base d lubricants used today. Boric acid, whose chemical formula is H3BO3, is also known as orthoboric acid or boracic acid. Boric acid is a hydr ate of boric oxide, B2O3. When boric oxide comes into contact with water, it will readily hydrate and convert to H3BO3. Boric acid is a weakly acidic white powder that is soluble in wa ter, approximately 27% by weight in boiling water and 6% at room temperature. It is soft, ductile, stable, free flowing, and easily handled. Finely ground technical gr ade boric acid powder (> 99% pure) is readily available for under $2US per pound. 2.2 Health and Environmental Impact The Environm ental Protection Agency has established that boric acid is benign and it is not classified as a pollutant under the Clean Water Act or the Hazardous Air Pollutants Act of 1990. Material safety data sheets fo r boric acid show no se rious illness or carc inogenic effects from exposure to solutions or aerosols. The US is the worlds largest producer of boron compounds [12]. Boric acid is recovered from brines at Searles Lake in California, with large domestic reserves of boron materials residing in other lake sediments and brines. Larg e quantities of boron ore are also collected from an open pit mine in California. The consumption of boric acid and boric oxide in the united states is distributed among glass making (78%), fire retardant (9%), agricultur al fertilizer (4%), and industrial applications such as metal plating and finishing, paints and pigments, electroplating, and cosmetics (9%) [12]. A dilute water solution of boric acid is also co mmonly used as a mild antiseptic and eyewash.

PAGE 13

13 The use of boric acid as a food preservative appa rently dates back to th e ancient Greeks. The earliest available scientific study of boric ac id was conducted in 1902 and reported in Science 1904 [13]. In this study, boric acid (0.5g) was introduced into food and ingested by a group of participants with each meal. This produced occa sional occurrences of fullness in the head, nausea, and loss of appetite in a few subjects. 2.3 Boric Acid as a Solid Lubricant In the early 1990s, the lubricity of boric ac id, an overlooked but extrem ely available and environmentally benign lamellar solid, was demons trated by Erdimer et al [13-16]. Figure 2-1 shows the lamellar molecular structure of boric acid. Figure 2-1. Lamellar structure of boric acid Experiments were conducted by Gearing et al. to determine the shear strength of boric acid powder [17]. In the experiments, high pressure thrust washers were compressed against a 6111 aluminum alloy plate coated with boric acid at pressures above 500 MPa. A twisting moment was applied by the compression tool loading the wa shers against the plate and sheared the boric acid coating on the plate. Figure 2-2 shows the set up used in the experiment. The results determined the shear strength of bor ic acid to be 23 MPa.

PAGE 14

14 Figure 2-2. Experimental setup for Gearing et al. experiment to de termine shear strength of boric acid. The shear stress of boric acid in Gearings study is almost the same as the experimentally determined shear stress of molybdenum disul phide, which was found to be 24 MPa by Singer et al. [18]. Figure 2-2 shows how the friction coeffi cient lowers with increased contact pressure and is well below 0.1 for average pressures above 100 MPa. At sub-atmospheric pressures, boric acid dehydrates and reverts back to boric oxide above 170C. In machining applic ations, cutting and forming interface temperatures are expected to exceed this hydration temperature, but the contact pressures are expected to be much greater than atmospheric pressure. No published data exists, however, detailing the hydration charac teristics of boric acid above 170C at higher contact pressures. Initial work with boric acid as a solid lubricant involved creati ng a solid film on the surface of the work piece. Boric acid was dissolv ed in either water or alcohol and the surface was coated and dried. After the solvent evaporated, th e dried boric acid remained on the surface as a thin coating. This method of boric acid deposition is incompatible with many practical

PAGE 15

15 applications or products in the manufacturing indu stry. This may be one reason that boric acid has not been accepted as a commonly used solid lubricant by industry. Our study examined the potential of delivering dry powder via air jet to deliver sufficient lubricant to adhere to the work piece and achieve good lubrication in situ. The feasibility of boric acid powders to sustain lo w friction when delivered as a powder was shown by Lovell et al. [20]. In this study, a concentrated sliding contact between an aluminum pin and AISI-M50 bearing steel was loaded to an initial maximum central Hertzian contact pressure of 1.9 GPa. Figure 2-3 details the experime ntal setup and results. Figure 2-3. Experimental setup and results for Lovell et al. lubric ation study using powder boric acid as a Solid Lubricant in a loaded pin on disk experiment The sliding speed was approximately 1 m/s. Boric acid powder was delivered by manually sprinkling it onto the disk surf ace; the corresponding lubricious surface film lowered the friction coefficient from = 0.4 to approximately = 0.15. The friction coefficient immediately responded to the application of boric acid by droppi ng to low friction state. Further, the reduced friction was sustained until the pow der delivery was halted.

PAGE 16

16 CHAPER 3 MATHEMATICAL MODEL We developed a predictive m athematical model fo r the effects of boric acid delivery on the sliding contact between the stainless steel surf aces. Assume a rotating disk has a contact interface with a stationary sphere su pporting a load. If boric acid is delivered into the interface, a fraction of the powder may stick to the disk face along the wear track. The amount of boric acid on the face may be represented by a fraction of surface covered with the powder. Initial fractional coverage is represented by 0 where 330 H BO totalA A (3-1) As powder is continuously delivered, more w ill be deposited on the area. An amount of the deposited powder, however, will adhere over top already covered surface. The coverage after a platelet of material has been deposited on a surface with fraction coverage given by Equation 3-1 is 010 0 (3-2) This pattern of application of boric acid platelets conti nues during powder exposure for n platelets 0 100 22 2000 2233 30 0 0 022 3333 (3-3) After reorganization the above patt ern, namely by factoring out the terms, the sequence of Equation 3-3 can be equivalently expressed as

PAGE 17

17 0 100 2 20 0 0 23 30 0 0 0(1) 2(1)(1) 3(1)3(1)(1) (3-4) This pattern can be equivalently expr essed by the following series expression (1) 0!(1)(1) 1 !()!aa n an ana (3-5) Equation 3-5 can then be expressed in the compact and closed form 01(1)(1)n (3-6) Equation 3-6 assumes that n platelets of boric acid powder sprayed into the interface stick to the surface during a single revo lution of the disk. After contact with the pin a fraction of the material is removed. is now introduced as a loss te rm indicating the amount of powder removed each cycle by contact pressure be tween the pin and disk. Loss factor, is assumed to be a linear coefficient term between 0 and 1 rela ting the coverage fraction of the disk surface covered by the pin leaving of the contact interface to the coverage fraction of the disk surface entering the contact interf ace. Factors affecting include the normal load applied to the surface through the pin as well. The terms of this relationship are defined below Coverage fraction entering contact interface Coverage fraction exiting contact interface Loss Factor in out The coverage fraction leaving the co ntact interface is then defined as outin (3-7)

PAGE 18

18 Substituting Equation 3-6 into Equation 3-7 yields the following expression to determine the fractional coverage for any given rotation N. This captures the transient behavior of the coupled film deposition and removal process. ,, 11(1)(1)n iN iN (3-8) Where N is the current cycle and n is the number of platelets that are deposited each revolution of the disk. The final solution to the coupled deposition a nd removal sequence is (1)((1))((1))1(1) ((1))1Nn NNn Nn nN (3-9) The exponent n is the number of platelets de posited on any fractional area of the surface between contacts. Powder mass flow rate is a ssumed to have a directly proportional effect on n, as higher flow would increase the coverage frac tion per unit time. Surface speed of the pin on the disk is assumed to have an inversely pr oportional impact on n, as a faster surface speed would reduce the amount of time for powder to enter the contact inte rface. Given these considerations, the exponent n is assumed to be predicted by the following expression m n r (3-10) where mass flow rate (g/sec) spindle speed (rad/sec) radial pin location (mm) m r The term N is the cycle count and can be calculated by 2 t N (3-11)

PAGE 19

19 where time (sec) t The term describes the powder remo val through the pin on disk interface as a ratio of coverage fraction exiting the interface to th e coverage fraction entering it. Greater contact pressure would displace more powder per revolution thus reducing Contact pressure is determined by the normal load, Fn. It is assumed in this model that is logarithmically proportional to the applie d normal load and is described below ()n F e (3-12) As the system approaches steady state, N appr oaches infinity. Equation 3-9 predicts that the N term, with being a fraction between zero and one, w ill reduce to zero at steady state. (1)1 (1)1n n (3-13) This model assumes the amount of frictional lo ad is determined by the coverage fraction. The total friction coefficient is predicted to equal the sum of the coverage fraction for each surface condition (i.e., powder or steel) multiplied by the friction coefficient for each surface versus the pin. 33(1)total HBO steel (3-14) where H3BO3 is the friction coefficient between boric acid and pin material. Experimentally, Equation 3-14 will be verified by setting up an e xperimental matrix that varies the four parameters predicted to affect the coverage fr action. Those four parameters are normal load, spindle speed, powder mass flow rate, and radial location of the pin on disk contact interface.

PAGE 20

20 CHAPTER 4 EXPERIMENTAL SETUP For the effectiveness of the l ubricating properties of boric acid powder to be characterized, experim ents studying surface to surface contact, especially the friction and wear between the surfaces, needed to be conducted. The following chapter depicts the variables and methods chosen to conduct the study of boric acid as a solid lubricant in metal on metal sliding contact. 4.1 Specimen Characterization and Preparation The Tribology Lab at the Univer sity of Florida has a pin on di sk tribom eter that creates relative rotating motion of a flat disk surface that maintains contac t with a loaded stationary pin surface. The pin samples in this experiment were 302 stainless steel balls of 4.76 mm (0.1875) diameter purchased from McMaster Carr (par t number 9291K18). As received, they had a reflective surface finish so no further processing of the ball surface was performed prior to experiment. The surface of the ball was an alyzed using the Wyco Scanning White Light Interferometer at MAIC (Major Analytical Instrumentation Center) at the Un iversity of Florida. The mean surface roughness of the balls was 150 nm The rotating surfaces in the experiment were from 302 stainless steel disks with a 50.4mm (2.0) diameter. The disks were manufactured from round stock purchased from Mc Master Carr. The rou nd stock was cut into 0.250 in. disks on a machine shop band saw. The disk surfaces were prepped using 3 stages of sanding and polishing, using 100 grit, 250 gr it, and 600 grit polishing paper, until all manufacturing marks were removed and a mirror finish was obtained. Surface roughness measurements on the prepared disk surface we re repeated on 15 samples using white light interferometry. Table 4-1 displays the results of these surface measur ements. Average surface roughness of the disks was calculated to be Ra 172 nm, = 3 nm. Prior to installation to the apparatus, each sample, pin and disk, were washed and sonicated in methyl alcohol.

PAGE 21

21 Table 4-1. Surface Roughness Measurement results for prepared surface of 302 Stainless Steel disks. Measurements performed on the WY CO White Light Interferometer at the Major Instrumentation Analyt ical Center (MAIC) at the University of Florida 1288.0367.13.3 3.8 2287.5366.23.3 3.8 3287.5366.03.4 3.9 4286.4364.83.3 3.7 5286.4364.73.3 3.6 6288.2367.13.3 3.7 7287.8366.53.3 3.6 8287.6366.23.2 3.8 9287.7366.33.2 3.6 10287.5366.13.3 3.9 11288.0366.73.4 4.3 12286.3364.73.3 3.7 13287.2365.83.3 3.7 14287.5366.33.3 3.7 15289.2367.93.3 3.6 Ra AvgRa Std Dev 287.50.7 2 Point Peak to Valley (mm) Scan # Ra (nm) Rms (nm) 20 Point Peak to Valley (mm) The powder used in this study wa s technical grade (99% pure) with particle sizes ranging from a few micrometers up to 100+ micrometers. The boric acid was purchased from Fischer Chemical, p/n A74-3. The particle size of the powder was measured under an SEM at Major Analytical Instrumentation Center at UF. Fi gure 4.1 shows the SEM photos of the boric acid. Figures 4-2 and 4-3 show the diagram of the Pin On Disk system and apparatus used in this experiment. 4.2 Pin on Disk Tribometer The pin on disk tribom eter used in this experi ment is shown in Figure 4-3 Measurements from the pin on disk tribometer were made through a compression load cell.

PAGE 22

22 Figure 4-1. Actual boric acid powde r used in this experiment ta ken using the Scanning Electron Microscope at the University of Florida Figure 4-2. Rotating pin-on-di sk tribometer used in this study. Airborne particle management handled with the shop vacuum and filter stack.

PAGE 23

23 Figure 4-3. Rotating pin-on-disk tribometer used in this study. The load cell measured force from the pin sample holding arm due to the frictional loading of the pin from the rotating disk. The ot her end of the arm supported the static load and fixed the stationary ball to the desired location on the disk surface. Disk samples were fixtured to the spindle and the testing su rface exposed to contact with th e pin sample. The pin on disk tribometer spindle and motor had the capacity to rotate up to 4700 rpm in either clockwise or counter clockwise orientation. Clockwise rotation was chosen to ensure constant contact between the sample holding arm and load cell. The arm pivoted on a gimble and was allowed pitch and yaw rotations. A threaded dead weight located at the end of the arm and a vertical adjustment micrometer allowed for adjustment to balance the pitch of the sample holding arm. When the pitch was properly balanced, the arm wa s considered level and an unloaded pin barely made contact with the disk surface. A radial ad justment micrometer adjusted the radial location of the system. When the system was radially ba lanced, the pin sat directly on the center of the disk. Radial balance was calibrated by running a pin against a disk and the radial micrometer adjusted until the wear tr ack became a single dot.

PAGE 24

24 4.3 Powder Delivery System A system needed to be developed to aerosoli ze and continually delive r boric acid into the contact interface. Nitrogen gas was first compre ssed and cleaned through the filter stack. There it underwent two stages of filtration and desiccation to ensure gaseous purity. The nitrogen gas was then regulated to 40 PSI and again desiccat ed. The nitrogen passe d through a flow meter, which allowed for variable flow output measured by a floating ball and calibrated scale. This flow setting was the regulation for controlling the mass flow rate of boric acid in the pin disk interface. Out of the flow meter, the clean, dry, metered gas then passed through stainless steel pipe into the boric acid bottle, shown in Figure 4-4 below. Figure 4-4 Particle delivery scheme. The container was mounted on a mechanical shaker during operation. Dry nitrogen gas flow rates were controlled throu gh a series of flowmeters. The boric acid bottle was mounted on a sh aker, which allowed for enough particle agitation to fluidize the powder and minimize powder clumping. As an additional method to eliminate clumping, the powder was triple sifted prior to loading the bo ttle. As shown in the

PAGE 25

25 figure above, the nitrogen input stem had many ho les drilled through it and was formed into an annular profile to have a large contact area with the powder. When the bottle was assembled, the input stem descended into the powder. The gas pressure created a fluidized bed, which entrained boric acid particles in the output stream. From the boric acid bottle, the nitrogenboric acid mixture flowed through a stainless steel pipe to the pin on disk interface. The flow meter setting was calibrated for powder flow rate prior to conducting the experiments. The control variable to set the powder delivery rate wa s the flow meter setting, which determined the gas flow into the powder delivery system. The flow meter had convenient flow setting demarcations of 10, 20, 30, and 40. The powder container was loaded with enough boric acid to establish a system weight of 600 gr ams prior to each calibration run. The mass of the bottle loaded with powder was weighed pr ior to running gas th rough the system and incrementally in five minute time steps. The time steps were chosen to verify consistency of the flow rate over the run times observed throughout the experiment. Figures 4-5 and 4-6 show the results of the powder delivery calibra tion. Figure 4-5 shows the total powder delivery rate where the bottle wei ght measurements are referenced to the initial weight and time (m0 = 600 grams ; t0 = 0 min). The equation defining the powder mass delivery rate referenced to t0 and m0 is 0 i iomm m tt (4-1) where ti is the time at iteration i and mi is the powder mass in the bottle at time = ti. Figure 4-6 shows the powder delivery results incrementa lly. Each powder delivery rate measurement referenced the previous time step. The equati on defining this incremental powder mass delivery rate, which shows the impact of time on delivery rate, is given in Equation 4-2.

PAGE 26

26 0 5 10 15 20 25 0 5 10 15 20 25 30 Time (min) 10 FM setting 20 FM setting 30 FM setting 40 FM SettingPowder Delivery Rate (gram/min) Figure 4-5. Powder delivery calibration curves re ferenced from initial conditions, t = 0 min; mass of boric acid in container = 600g, powder deliver rate = (mx m0) / (tx t0) 0 5 10 15 20 25 051015202530 Time (min) 10 FM setting 20 FM setting 30 FM setting 40 FM SettingPowder Delivery Rate (gram/min) Figure 4-6. Incremental powder delivery calibrati on curves referenced from previous time step; powder deliver rate = (mx mx-1) / (tx tx-1)

PAGE 27

27 1 1 ii iimm m tt (4-2) Figures 4-5 and 4-6 show the non-linear nature of the powder delivery system. This variability is due to several fact ors. The most significant issue is that there is no compensation for the constant gas flow rate yet depleting powder levels. As time progressed, lesser amounts of boric acid were able to get entrained in the ou tflow stream. Though the powder delivery rate was not constant, the flow meter setting did correspond with a higher powder delivery rate throughout the experiment. The actual powder deliver y rate was verified for each experimental run, by weighing the bottle prior to and following each run. For the experiments, settings 10, 20, and 30 were chosen. A flow meter setting of 40 for the 60 minute test runs would have required a larger powder container. 4.4 Environment Chamber An environ ment chamber was implemented to th e pin on disk tribometer set up to contain boric acid, which could be a slip and inhalation hazard. The panels of the environment chamber are made of Lexan sheets, providing clear view of the experimental apparatus, yet provide a lightweight assembly and removal. A 2 horsepower shop-vac pulled a vacuum to create flow throughout the chamber and reduce powder build up on the hardware. An air-filter stack was attached to the chamber opposite th e shop-vac interface, with a sheet filter attached to the inlet. This provided clean air flow across the pindisk interface. 4.5 Determination of Experimental Matrix A m atrix of testing variables was created to test the effectiveness of a continuous stream of solid boric acid into the contact surface between the pin and disk. Using the predictions of the mathematical model, the control variables were se lected. The experimental parameters studied in the experiment were normal force, sliding sp eed, wear track diameter, and powder flow rate.

PAGE 28

28 The normal force was modified by adding a dead weight to the arm dir ectly above the pin. Sliding speed was adjusted by changing the spin dle rotational speed. The adjustment for the wear track diameter was set with the radial ad justment micrometer. The powder flow rate measurement and adjustment was described in th e previous section. Table 4-2 summarizes the conditions tested in the experiment. Table 4-2. The experimental matrix (3*3*3*3) pl us nine repeats at the midpoint and one unlubricated control. Fn (N) (RPM) Track Diameter, D (mm) Flow Meter Setting 0.65 400 3.8 10 3.30 2,000 25.0 20 6.30 4,000 38.0 30 The experimental matrix resulted in 81 i ndividual tests. The midpoint condition (3.3 N Load, 2000 rpm, 25.4 mm track diameter, flow mete r setting = 20 ) was tested 9 additional times to exhibit repeatability of the friction measurem ent. One midpoint test with no boric acid delivery was conducted to provide a baseline measurement for friction between bare metal on metal contact. The lengths of the tests were se t to maintain 24 ,000 rotations of the disk, where 24,000 disk revolutions were divi ded by the spindle speed. The combination of wear track diameter and spindle speeds al low for a range of sliding sp eeds that varied by 100 times. The sequence of experiments was randomized to prevent any time-related bias as shown in Table 4-3. As described earlier, the data acquisition system colle cted loading measurements of the gimble arm on the load cell, which correlated to friction loads and friction coefficient of the ball on the disk. The selection of test conditions allows for significant flexibility in grouping data for analysis. For example, 27 tests were performed at constant diameter while varying sliding speed, contact force, Fn, a nd flow rate, f; and 18 tests were performed at constant sliding speed while varying track diam eter, D, and rotational speed, Fn, and f. The variability of testing conditions allowed for an expansive, yet detailed, study of the parameters impacting the

PAGE 29

29 effectiveness of a solid boric acid lubricant on the wear rates and frictional forces in pin on disk contact. Laboratory ambient cond itions were measured daily befo re testing. Relative humidity varied from approximately 20 to 55%. The ambi ent temperatures ranged from 25C to 28C. Table 4-3. Randomized experimental testi ng sequence with all parameters per run. 1 104001.5066.3360046 2020001.00642.5720 2 204000.15336.6360047 3040001.50336.6360 3 204001.0066.3360048 2020001.00336.6720 4 2040000.1566.336049 1040001.0066.3360 5 1040001.50336.636050 204001.50336.63600 6 1040001.00642.536051 2020001.50336.6720 7 1040000.15642.536052 304001.5066.33600 8 1040001.5066.336053 104001.0066.33600 9 1020001.50642.572054 3040001.00336.6360 10 2020001.00336.672055 3020000.15336.6720 11 2020001.0066.372056 3040000.15642.5360 12 204001.50642.5360057 2020001.00336.6720 13 304001.00336.6360058 304000.15336.63600 14 1020000.15642.572059 304001.0066.33600 15 2040001.0066.336060 1020000.1566.3720 16 304001.50336.6360061 104000.1566.33600 17 2020001.50642.572062 2040001.50642.5360 18 3020000.15642.572063 3040001.00642.5360 19 3020001.50642.572064 304000.1566.33600 20 2040000.15336.636065 2040001.00642.5360 21 3020001.00642.572066 2040001.5066.3360 22 104001.50336.6360067 304001.50642.53600 23 2020000.1566.372068 2020001.00336.6720 24 2040001.50336.636069 3040001.0066.3360 25 2020001.00336.672070 204001.00336.63600 26 3040001.5066.336071 3020000.1566.3720 27 2040001.00336.636072 204000.15642.53600 28 2020001.5066.372073 2020001.00336.6720 29 304000.15642.5360074 1020000.15336.6720 30 3020001.5066.372075 2040000.15642.5360 31 304001.00642.5360076 1040000.15336.6360 32 104000.15336.6360077 1020001.0066.3720 33 104001.00336.6360078 3040000.1566.3360 34 1040001.50642.536079 104001.00642.53600 35 2020001.00336.672080 2020000.15336.6720 36 3040001.50642.536081 2020000.15642.5720 37 1040000.1566.336082 1020001.00642.5720 38 204001.00642.5360083 204000.1566.33600 39 3020001.00336.672084 204001.5066.33600 40 3040000.15336.636085 1020001.50336.6720 41 104000.15642.5360086 1020001.00336.6720 42 3020001.0066.372087 2020001.00336.6720 43 3020001.50336.672088 1040001.00336.6360 44 1020001.5066.372089 2020001.00336.6720 45 104001.50642.5360090 2020001.00336.6720 Flow Meter Setting Spindle Speed (RPM) Dia (in)Normal Load (grams) Length (sec) Run Number Flow Meter Setting Spindle Speed (RPM) Dia (in)Normal Load (grams) Length (sec) Run Number shaded entries denote a midpoint run

PAGE 30

30 CHAPTER 5 RESULTS The critical m easurements taken in this study were the thickness of the wear track left on the sample disk and the load on the load cell. The wear track diameter and thickness are the key measurements to correlating the wear rate and ultimately the effectiveness of the boric acid powder as a solid lubricant. To measure the thic kness of the wear track, an optical microscope was used. A calibrated measurement grid was att ached to the scope, and the inside of the wear track was located at 0.000. The measurement at the outside of the wear track then established the wear track thickness. Measurements were made at each quadrant of the wear track to establish consistency. Those measurements were averaged and the standard deviation taken. The wear rate for the pin was used as the metric for wear in this study. To calculate the wear on the pin, the volume of material lo st was first determined. Equa tion 5-1 was used to determine the pin volume lost due to wear. Figure 5-1 gr aphically shows the relatio nship between the wear scar diameter and the lost volume. 2**3 3 hRh V (5-1) where V is the volume of the pin lost due to wear, R is the radius of the pin and h is the radial height of the lost volume. Calculated wear rates for all runs are summarized in the table below. h V R Figure 5-1 Pin volume lost due to wear during the experiment

PAGE 31

31 Table 5-1. Results from the experimental matrix. D is wear track diameter in millimeters, f is commanded boric acid flowrate in grams/minute, K is wear-rate x 10-6 mm3/(Nm). = 400 RPM = 2,000RPM = 4000 RPM Fn = 0.65 N Fn = 0.65 N Fn = 0.65 N D F K D F K D F K 3.8 10 0.196 2.92 3.8 10 0.157 3.8 10 0.090 3.8 20 0.080 3.8 20 0.094 14.93 3.8 20 0.085 3.8 30 0.138 3.8 30 0.085 1.98 3.8 30 0.092 0.63 25.0 10 0.147 0.30 25.0 10 0.036 36.6525.0 10 0.086 0.22 25.0 20 0.122 25.0 20 0.035 42.4825.0 20 0.075 25.0 30 0.159 0.67 25.0 30 0.062 21.3125.0 30 0.047 4.90 38.0 10 0.147 0.44 38.0 10 0.058 1.25 38.0 10 0.080 0.21 38.0 20 0.138 0.03 38.0 20 0.077 0.38 38.0 20 0.049 4.13 38.0 30 0.173 0.19 38.0 30 0.045 1.56 38.0 30 0.089 0.57 Fn = 3.3 N Fn = 3.3 N Fn = 3.3 N D F K D F K D F K 3.8 10 0.143 0.23 3.8 10 0.148 0.82 3.8 10 0.081 1.48 3.8 20 0.139 4.03 3.8 20 0.12 3.8 20 0.088 5.55 3.8 30 0.162 0.31 3.8 30 0.11 0.12 3.8 30 0.087 0.03 25.0 10 0.152 0.76 25.0 10 0.056 0.80 25.0 10 0.055 1.05 25.0 20 0.156 0.48 25.0 20 0.044*1.54*25.0 20 0.035 0.52 25.0 30 0.163 0.41 25.0 30 0.05 2.00 25.0 30 0.030 1.55 38.0 10 0.113 9.02 38.0 10 0.049 0.14 38.0 10 0.321 0.03 38.0 20 0.116 1.36 38.0 20 0.039 0.27 38.0 20 0.040 0.34 38.0 30 0.101 13.26 38.0 30 0.045 0.19 38.0 30 0.026 0.69 Fn = 6.3 N Fn = 6.3 N Fn = 6.3 N D F K D F K D F K 3.8 10 0.155 0.42 3.8 10 0.117 0.35 3.8 10 0.086 0.07 3.8 20 0.138 0.06 3.8 20 0.108 3.8 20 0.079 0.60 3.8 30 0.165 3.8 30 0.103 3.8 30 0.078 0.85 25.0 10 0.137 5.28 25.0 10 0.036 0.24 25.0 10 0.059 0.27 25.0 20 0.122 1.05 25.0 20 0.061 0.72 25.0 20 0.029 0.46 25.0 30 0.121 0.01 25.0 30 0.059 3.44 25.0 30 0.035 1.09 38.0 10 0.100 20.78 38.0 10 0.047 0.30 38.0 10 0.230 2.35 38.0 20 0.106 32.86 38.0 20 0.039 0.21 38.0 20 0.023 14.08 38.0 30 0.090 21.15 38.0 30 0.046 0.09 38.0 30 0.073 0.06 Midpoint condition avg = 0.044, standard deviation, = 0.009; Kavg = 1.54 = 1.15

PAGE 32

32 The other critical measurement taken was the fr ictional load on the load cell. The location of the load cell is at a one to one distance from the arm pivot. This corresponds to a one to one loading relationship between the load cell measurem ent and the radial loading of the arm relative to the rotating disk. The coefficient of friction is then approximated using the Equation 5-2: F N 5-2) where N is the dead weight load applied to the arm. The friction coefficient measurements were averaged over the duration of th e low friction state and reported in Table 5-1 above. The nine repeats of the midpoint test condition and the control (midpoint condition with no lubricant) are plotted together in Figure 5-2. Experimental conditions: Fn = 3.3 N, = 2,000 RPM, D = 25.4 mm, f = 20 g/min. Figure 5-2. Friction coefficient vers us cycle number for the 9 repeat tests and the control, which had no boric acid in the gas flow. This plot displays the effect on friction co efficient of the boric acid as compared to interfacial contact with no lubri cant. The figure shows how the boric acid, once introduced into no boric acid flow

PAGE 33

33 the contact interface, quickly redu ced the friction coefficient. The spikes show interruptions in powder flow, and show the sensi tivity of the system to the presence of the lubricant. The powder delivery system experienced comp lete delivery failure on two experimental runs. This powder delivery failure occurred due to the delivery tubes clogging with powder. Figure 5-3 shows the effects of cl ogging on the friction coefficient. Experimental conditions: D = 25 mm, surface speeds = 5.0 m/s, Fn identified in the plot. Figure 5-3. Friction coefficient vers us cycle number for the experiments with terminal blockages in the boric acid delivery during the experiment. The data was parsed several ways to search for trends in wear rate and friction coefficient. Since there were four variables in the experiment, a series of pl ots needed to be created. For each wear track diameter, three plots were crea ted to detail the relatio nship between average friction coefficient and load, flow rate, and sliding speed, respectiv ely. This series of plots is shown in Figures 5-4. Efforts were made to corre late the experimental data to the Von Karman mathematical model. Figure 5-5 shows the rela tionship between average friction coefficient and boundary layer thickness.

PAGE 34

34 Figure 5-4. Average friction coefficient versus normal load, flow rate, and sliding speed. Figure 5-5. Average friction coefficients for al l experiments plotted versus the boundary layer thickness. The data is separated for the various normal loads as indicated.

PAGE 35

35 Many experiments showed a high degree of vari ability in the time taken to get to low friction. Though the exact reason for this variab ility was undetermined, it is hypothesized that unreliability in the powder delivery apparatus play ed a major role. As described in Chapter 4, many precautions were taken to ensure the powde r would be clump free and the air would be properly desiccated. The mechanism responsible for delivering the boric acid powder to the Pin on Disk interface is unknown, but significant variabi lity with the delivery was detected. Figure 5-6 shows two runs with different times to reach a low friction state. As is shown in the figure, the measured wear rates of the pin differ by tw o orders of magnitude. The measured friction coefficients are not significantly different. Experimental Conditions: D = 25 mm, su rface speed = 2.5 m/s, normal loads are identified in the figure. Figure 5-6. Friction coefficient vers us cycle number for the two experiments that had the shortest and longest transients to lo w friction coefficient.

PAGE 36

36 CHAPTER 6 DISCUSSION When the boric acid was successfully flowing to the interface, friction and wear between the pin and disk was significantly reduced. This trend is shown in Figure 5-2, where the control condition is plotted against the midpoint runs. The only difference between the control and the midpoints runs was the presence of boric acid lu bricant. The average friction coefficient drops from 0.641, non-lubricated s liding contact, to 9.3x10-4 when boric acid is introduced to the system. That represents a tw o order of magnitude reduction in friction force for the same operating conditions. The measured friction coeffi cient is rapidly impacted by the presence of boric acid in the contact interf ace. Figure 5-2 details how a low friction state is achieved after tens of revolutions. The wear rate of the pin s hows a similar trend to the fr iction reduction, reducing from 4.86 x 10-4 mm3/Nm to 1.55 x 10-6 mm3/Nm when boric acid is introduced into the sliding contact interface. Photographs detailing the difference in wear between the lubricated and non lubricated test conditions are shown in Figure 6-1. A B Figure 6-1 Wear marks on pin and disk after midpoint experimental conditions A) with boric acid flow and B) without boric acid flow

PAGE 37

37 The measured friction coefficient is very sens itivity to interruptions in flow. Figure 5-3 shows several spikes in friction coefficient that corresponded to perturbations in flow. In Figure 5-3, which highlights the two runs where the po wder delivery system failed prematurely, the friction coefficient spikes at the point of failure and stays high for the duration of the run. As was seen with the intermittent delivery failures, the transition from low to high friction coefficient occurred within tens of revolutions. This transition from low to high friction suggests that the rate of powder de position in the contact interface is very close to the rate of removal. In Figure 5-4, the plots detail the effect on friction coeffi cient by a single experimental variable, while holding all other va riables constant. Plot 5-4a s how the effect of normal load on friction coefficient, in groups of powder flow rate, while holdi ng wear track and surface speed constant. Across the data plots, there was no apparent significance between coefficient of friction and normal load or slidi ng speed. In Figure 5-4c, the lo wer sliding speeds resulted in a higher coefficient of fric tion than the higher sliding speed tests under the same conditions. Using the competitive rate models for the deposition an d removal of boric acid, it is hypothesized that removal rate should increase with increasing sliding spee d. This would lead to less boric acid getting entrained in the contact zone and result in a higher coefficient of friction. The experimental data suggested the opp osite to be true. Several conc lusions may be drawn from this data, however. It may suggest that the remova l rate is suppressed at higher speed, thereby increasing the amount of boric acid in the interf ace and reducing friction. The data may also suggest that at higher sliding speeds, the rate of film formation on the surface of the disk and/or pin increases. It is hypothesized that the dire ct injection of nitrogen and boric acid into th e interface was not the primary mechanism for boric acid transport to the pin contact. Using the model solved

PAGE 38

38 by Von Karman [13,14], it is hypothe sized that the fluid flow gene rated across the rotation of the disk sample entrained the boric acid and delivered it to the pin contact. Figure 5-5 shows the results of Van Karmans flow m odel. The rotation of the spinning disk pulls air above the disk down at the center of the disk and discharges it radially outward. The air flow across the disk is assumed to be laminar if the Reynolds number is less than 300,000. The Reynolds number is given by Equation 6-1, where Re VR (6-1) Re is the Reynolds number (dimensionless), V is th e peripheral speed (in m/s), R is the radius of the disk (in mm), and 51.510 x is used for the kinematic viscosity of the air. The highest Reynolds number encountered in this experiment was 4000, signify ing that the flow across the disk was laminar. Rogers and Lance [24, 25] provided a numerical solution to this problem. Their study demonstrated that the boundary layer thickness, is relatively constant across the disk and varies inversely with the square-root of the angular speed. The following equation describes the results of their study 5.4 (6-2) where is the angular speed of th e disk (in rad/s), and again 51.510 x m2/s is used for the kinematic viscosity of air. The entire data se t is plotted versus the calculated boundary layer thickness in Figure 5-5. The data trends in Figure 5-5 suggest that the thinner and higher speed flows are more efficient at delive ring boric acid to the contact. Figure 5-2 details the measured friction coeffi cient for unlubricated pin on disk stainless steel contact. The average wear rate was measured to be K = 5.0x10-4 mm3/Nm. All lubricated experiments conducted showed significant decreases in pin wear rate, as shown in Table 5-1.

PAGE 39

39 Eleven of the pin samples showed no detectable w ear marks, these are indicated by dashes in the wear rate column of Tabl e 5-1. The lowest measured wear rate was K = 7.5 x10-4 mm3/Nm, which occurred during the 25 mm diameter, 3.3 N load, 2.5 m/s sliding speed, and 0.078 grams/second boric acid flowrate conditions. The experiments exhibit over 500 times greater improvement in wear resistance over the unlubricated contact measurements. Substantial effort was made to correlate the friction coefficient and the wear rate with the selected experimental variables. There were no strong indicators ty ing the experimental variables to either friction coefficient or wear rate. The only qualitative explanation for the variation in wear rate is the significant loss of ma terial during start up transients. The tests were initiated on nascent surfaces with the boric acid flow expected crea te to replenish a surface film in situ The startup transient vari ed widely between experiment al runs, though all experiments eventually reached a low friction state. It is sugg ested that the majority of material lost to wear occurs during this period, and be tter understanding of the film formation must be developed. Figure 5-6 shows the friction coe fficient plots for experiments w ith the shortest and longest initial transien t conditions. These experiments were both at a radial posit ion of 2.5 mm and a sliding speed of 2.5 m/s, although they were at two different normal loads. The experiment with the shortest transient had a wear-rate of 7.5x10-7 mm3/(Nm) and the test with the longe st transient had a wear-rate of 3.7x10-5 mm3/(Nm). The ratio of the volumes lost betw een the shortest and longest transient was 0.1 (i.e., the test with a s horter transient lost 10 percent of the ma terial of the test with the longest transient), while the ratio of the frictional ener gy dissipated during the tran sient region is about 1/3.

PAGE 40

40 CHAPTER 7 CONCLUSIONS Conclusion 1: These ex periments were the first re ported demonstrations of continuous powder delivery of boric acid as a solid lubricant. Conclusion 2: These experiments clearly indicate th at powder delivery of boric acid is a viable technique for providing in situ lubrication for concentrat ed metal contacts. Wear rate reductions of over 100 times and friction coefficients of well under 0.1 were demonstrated. Conclusion 3: Further development and refinement of the powder deliv ery system must be explored to develop an accurate unders tanding of the impact of the variables on friction coefficient and wear rate. Conclusion 4: A future study is suggested for expl oring the impact of a pre-deposited boric acid film on the surface of pin and/or disk to verify the impact of the initial transient on friction coefficient and wear rate.

PAGE 41

41 LIST OF REFERENCES [1] P. Hamblin, Environmentally compatible lubr icants: trends, standards, and terms. Proc. Environmental Aspects in Production and Util ization of Lubricants, Sopron, (1995) 1-10 [2] W. Bartz, Lubricants and the enviro nment, Tribology Intern ational 31 (1998) 35-47 [3] G. Johnson, Milacron Marketing Co, CIMCOOL technical report, (1999) 1-12 [4] A.R. Machado, J. Wallbank, The effect of extremely low lubricant volumes in machining, Wear 210 (1997) 76-82 [5] M. Rahman, A. Senthil-Kumar, Salam M.U., Experimental evaluation on the effect of minimal quantities of lubricant in milling, In ternational Journal of Maching Tools and Manufacture 42 (2002) 539-547 [6] F.M. Kustas, L. FehrehnbacheR, R. Koma nduri, Nanocoatings in cutting tools for dry machining, Annals of the CIRP (1997) 39-42 [7] L.W. Jelinski, T. Graedel, R. Laudise, D.W. McCall, C.K Patel, Industrial ecology: concepts and approaches, Proceedings of the National Academy of Sciences of the United States of America 89 (1992) 793-797 [8] L. Ward, Influence of boric acid and bor ax on digestion and health, Science 20 (1904) 2627 [9] G. Brown Jr, Remarks on industrial ecology, Proceedings from the National Academy of Sciences 89 (1992) 876-878 [10] J. Sutherland, K. Gunter, D. Allen, D. Bauer, B. Bras, T. Gutowski, C. Murphy, T. Piwonka, P. Sheng, D. Thurst on, E. Wolff, A global perspective on the environmental challenges facing the automotive industry: State-of -the-art and directions for the future, Int J Vehicle Des 35 (2004) 86-110. [11] D. Allen, D. Bauer, B. Bras, T. Gutowski, C. Murphy, T. Piwonka, P. Sheng, J. Sutherland, D. Thurston, E. Wolff, Environmentally beni gn manufacturing: Trends in Europe, Japan, and the Usa, J Manuf Scie E-T ASME 124 (2002) 908-920. [12] P. Lyday, Mineral commodity summaries, US Geological Survey, (2003) 110-122. [13] A. Erdemir, Tribological properties of boric-acid and boric-acid-forming surfaces.1. Crystal-chemistry and mechanism of self-lubr ication of boric-acid, Lubr Eng 47 (1991) 168-173. [14] A. Erdemir, R. Erck, and J. Robles, Relati onship of hertzian contact pressure to friction behavior of self-lubricati ng boric-acid films, Surf Coat Tech 49 (1991) 435-438.

PAGE 42

42 [15] A. Erdemir, G. Fenske, and R. Erck, A study of the formation and self-lubrication mechanisms of boric-acid films on boric oxid e coatings, Surf Coat Tech 43-4 (1990) 588596. [16] A. Erdemir, G. Fenske, R. Erck, F. Nichol s, D. Busch, Tribological properties of boric-acid and boric-acid-forming surfaces. 2. Mechanis ms of formation and self-lubrication of boric-acid films on boron-containing and boric oxide-containing surfaces, Lubr Eng 47 (1991) 179-184. [17] B. Gearing, H. Moon, and L. Anand, A plasti city model for interface friction: Application to sheet metal forming, International Journal of Plastic ity 17 (2001) 237-271. [18] I. Singer, R. Bolster, J. Wegand, S. Faye ulle, B. Stupp, Hertzian st ress contribution to low friction behavior of thin mos2 coa tings, Appl Phys Lett 57 (1990) 995-997. [19] M. Lovell, W.G. Sawyer, and A. Mobley, On the friction and wear performance of boric acid lubricant combinations in extended durat ion processes, Tribol Tech 25 (2005) 73-81. [20] N. Mccook, D. Burris, G. Bourne, J. Steffe ns, J. Hanrahan, W.G. Sawyer, Wear resistant solid lubricant coating made from ptfe and epoxy, Tribo Lett 18 (2005) 119-124. [21] T. Von Karman, Uber laminare und turbulente reibung, Z Angew Math Phys 1 (1921) 233252. [22] M. Miklavcic, C. Wang, Th e flow due to a rough rotating disk, Z Angew Math Phys 55 (2004) 235-246. [23] M. Rogers, G. Lance, The rotationally symmet ric flow of a viscous fl uid in the presence of an infinite rotating disk, J Fluid Mech 7 (1960) 617-631. [24] M. Rogers, G. Lance, Boundary layer on disc of finite radius in rotating fluid, Q J Mech Appl Math 17 (1964) 319-25.

PAGE 43

43 BIOGRAPHICAL SKETCH The author was born in Oakland, Californi a, in 1977. Tim othy Barton was raised in the San Francisco bay area for the durat ion of his childhood. In 1995, he moved to Florida to attend the University of Florida to pursue a degree in mechanical engineering. He graduated with his Bachelor of Science in mechanical and aerospace engineering in 2001. He obtained his Master of Science in mechanical a nd aerospace engineering from the University of Florida in 2007.


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 E20101203_AAAAFK INGEST_TIME 2010-12-04T04:34:56Z PACKAGE UFE0021824_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 808 DFID F20101203_AACKIJ ORIGIN DEPOSITOR PATH barton_t_Page_05.txt GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
1f4c6b64f5f396a2786bd4a31e58ad10
SHA-1
e5a3a803f45c2175cf97cba22ce9fab77d4b5d8d
25271604 F20101203_AACKDM barton_t_Page_08.tif
2137f5cb53b66f3cd818e99dbd2a28d1
8fb8e674309433865e1cff7378bb6db2a7507357
2054 F20101203_AACKIK barton_t_Page_08.txt
c883aeed3f6fabfecee848717a57be4e
2558ea0eb0b5db2e783c234ee7dc69cb6608d4ed
8615 F20101203_AACKDN barton_t_Page_43.QC.jpg
bec7bb246283c31edbc703c5ee8b5c81
c13471c21b5e65fb8cfaeaceb96ef77d5fc996e9
1206 F20101203_AACKIL barton_t_Page_11.txt
ec0ce807f4c536f103bd21014dc521b6
b17a9942a5ca797e66f6b54a8588952034d6870f
59008 F20101203_AACKDO barton_t_Page_41.pro
2d3fe2226023b5168fc417761d8a36cd
db12fc1f2bf9ae9b19c200d78d96b673e70d2293
2019 F20101203_AACKIM barton_t_Page_12.txt
a4f9e8646a056b13a60f33d95104eb03
e3da7521d88158947eb15e1a10d4471f2d34d310
20650 F20101203_AACKDP barton_t_Page_42.QC.jpg
21458be89f3e1265f11f5b81d3fe46fd
634e02a3e41823b84a5b433f28c8032a85540070
2175 F20101203_AACKIN barton_t_Page_14.txt
b78b51a1990aba62f39c850ccdf62074
b52eb75c0b63436d22966163aeb0f94e01e0d210
4432 F20101203_AACKDQ barton_t_Page_34thm.jpg
4c29d6314e60f5cdc001eee0c3fdb828
6b6e90508c52dad6e2bc20c43f970228da3b7654
2169 F20101203_AACKIO barton_t_Page_20.txt
4d733327601b09e3ca38773428ee0c96
a15b4b64043479283dfd66c9292c2e1ee604feec
35703 F20101203_AACKDR barton_t_Page_24.pro
f1191bf8bb2b700334577adbf59d0168
22a906074caae93f2526bdf0191997e7a3a8af4b
2213 F20101203_AACKIP barton_t_Page_21.txt
9f01eb840fa971a5ac781bd931024257
0b04ddfe2c5af79345aa4ce6a64c846daf0df7f6
964 F20101203_AACKDS barton_t_Page_02.pro
cededf17356bc1aaaddb7326a0567c6b
e84bd7bd645ea08fef72018d78edffa865fd39b4
1229 F20101203_AACKIQ barton_t_Page_22.txt
62c5a54f1ed26cdf94c29008676f6c16
e336f8f7629e96ef2e22fcdb9c93e33baa06dca7
21727 F20101203_AACKDT barton_t_Page_23.QC.jpg
5e81c16318b33e6ccdb381f8cd96042f
9c9788e2cdfc716e353f4e9902921f7d8f04aeae
2048 F20101203_AACKIR barton_t_Page_25.txt
27d10edc9f89adbf53ee0d0868f2793d
c3ebea99c82469347485d4f5e45bde13e1871270
F20101203_AACKDU barton_t_Page_17.tif
48560a49a169c5901b28ec06d948f2b9
04a3168a25d68a47758994c30fd07ef146ae75dd
1244 F20101203_AACKIS barton_t_Page_26.txt
c69a8440c3babdd622ad56a1f8a5a63b
81e66f18b1c71380547f72820c7fc734d9c64d03
F20101203_AACKDV barton_t_Page_18.tif
d04d564975ef1e34dc7a9def06358d4f
9fb6908bc9100a8935141fd849e9aae4df1d8e7a
3480 F20101203_AACKIT barton_t_Page_29.txt
a744901d78ce5cd662c3a819d438fc27
60b2600955cd1068d3f750e53c4a60994cbca7a2
31772 F20101203_AACKDW barton_t_Page_07.pro
2ec08b37441800dd9c830d44d0b27033
8a29ec9f63c3547a94fdd61d137159cbeaa07249
2753 F20101203_AACKIU barton_t_Page_31.txt
2cd923513acb55e9916f453ccc3a56b3
91ec189290b3433e354d914b0fbb4366ac8019be
5703 F20101203_AACKDX barton_t_Page_19thm.jpg
0590f1b4047e7d72f51f3d16ba80ee32
2cd6a8d61f6e5e5fced9ea8b5ddba52e44264b2e
F20101203_AACKIV barton_t_Page_37.txt
066145893cb9e410002e3465d0bb7dd8
46d22f74034506d6061ff7f762f6e998df6c4387
1053954 F20101203_AACKBA barton_t_Page_21.tif
4b46af8456d6fef528ed2689184f364f
3bc7e1cbabcbb82138f6b9fb4d34af8c460d8999
1051949 F20101203_AACKDY barton_t_Page_23.jp2
e2a988dc1c428ecb454fecf1d8be678b
b3e99b78946e27116b752b80c5dc96ba46f4f99e
1001 F20101203_AACKIW barton_t_Page_40.txt
a2444ed7b76b35251213e65d92635ef4
d47fe97e9c73cc07b79213709884c7a13596fa49
F20101203_AACJZA barton_t_Page_38.tif
8e01061480c34142d09ed42402ba7ba0
d46c802e7ffeea20303d68f62af61c9e991141b3
482 F20101203_AACKDZ barton_t_Page_01.txt
256266ee71a79bf72e8456522bb50b86
b963e5d4d2d5bf50b467c1fc913dbe943593e433
6714 F20101203_AACJZB barton_t_Page_08thm.jpg
7b0321318f690e2b3d6a71df9e5b9e92
10895b17c4e5522844b6512077e754665e8c24b2
36028 F20101203_AACKBB barton_t_Page_16.pro
fc889e455acb060000604658c8d28e01
f9995c44f86592abcdad50201dd84300a6e53576
546 F20101203_AACKIX barton_t_Page_43.txt
4dca1413d03ef84ab28639ca90da4504
476eb735c92962f6931529125f7c9a1cc8c6c3cd
61905 F20101203_AACJZC barton_t_Page_33.jpg
21d97ec3d994a038617583b93e27ec0c
04f639fe2132f382acd6e727ae64828c51dc4ab2
7000 F20101203_AACKBC barton_t_Page_27thm.jpg
c04bb09870ece4418033490ed70a46b2
c5410533dbeec2440637c941d849fd032c069f37
95040 F20101203_AACKGA barton_t_Page_29.jpg
cdcc2f98792d874caed7768c2d5660ef
f95f7125b7147bcf388bb176d047ae7017d25292
1194379 F20101203_AACKIY barton_t.pdf
856896e74aea0e7758d8ba395fa70158
74c23b2bb804ba94d254cf6f025ef1299d3de585
27156 F20101203_AACKBD barton_t_Page_05.jpg
ecc19d2b764333df3890791c860c6a5a
435f9f9072cff0ace1e073681a017493cb11b0f3
62453 F20101203_AACKGB barton_t_Page_30.jpg
4f08c978a729e571242aaa2e8d139ff5
7b92db3d9d69f31b81450ebf44bf6f9d774a0805
19700 F20101203_AACKIZ barton_t_Page_17.QC.jpg
f0fb84f82fd843ef9f49860c6765c83c
26654c93887cab875658991fc402c1e095df8cdd
6356 F20101203_AACJZD barton_t_Page_13thm.jpg
63a6782cb8650b877a04dd004dd1dbd4
f2d4d292f63cc74202a2779799c5ddcae297701f
1687 F20101203_AACKBE barton_t_Page_19.txt
ccfb4f22e08349d382cb5cf3c700accd
c785e8cff4f6d07f169e5bd95f2b6ceafd2836f1
83144 F20101203_AACKGC barton_t_Page_31.jpg
e00c1c716556c8c9fe4ef79379ff75c1
60ce75fb0593c3d6a299b145aca7ec2ae1c7fee6
6865 F20101203_AACJZE barton_t_Page_38thm.jpg
1ead89973b8d7a64bebd024a8d955b53
aa051cc1eb84cc95e83284af097fd2962bccbd55
F20101203_AACKBF barton_t_Page_11.tif
d29fcd4c19f5638f4f3272a5bd55cd2b
6716ff93b19bf94fe62cbbfea5c7ad59ef0ea2e8
65399 F20101203_AACKGD barton_t_Page_36.jpg
cc4e71cd7a80ff1488e0e9be64824ac5
45a86f332e5883f00ca08b848c38ed7b90e5f97a
45590 F20101203_AACJZF barton_t_Page_14.pro
1d5bb2636e362bb605501e81ce8b7548
d3e32c23f9cd3fb614fea7de46aafead394f4bdc
1051935 F20101203_AACKBG barton_t_Page_06.jp2
04e7831ef2e85dfaf5662c693143671d
f29a268700e8a7722a8c19c535308c7b46716637
39042 F20101203_AACKGE barton_t_Page_40.jpg
fc6b8b2ad6e6dbe6f5c2e07a44941f4e
702327f17a9366ed5ed3782ea5188d4fa7180867
2240 F20101203_AACJZG barton_t_Page_32.txt
d6e9a482166d50669d6e727a3f71b8f1
3ea6af2b738f4954fa0614fef5fcd6bb05eeb2a3
581615 F20101203_AACKBH barton_t_Page_34.jp2
7185fa037a95c9a667a3145579d7a2c7
a3bbab21277d698658b07872fa10bfe2720b1c79
95681 F20101203_AACKGF barton_t_Page_41.jpg
5b5d4dc90663c85aa7c435525b7e1c3c
5a4395760a2b05f57e370745d35880e6622cd6fc
1476 F20101203_AACJZH barton_t_Page_07.txt
0af01dfb0cf0aad6d43f4694be7127dc
353121acd76717c414f3421edc7868bdbfa8f693
20564 F20101203_AACKBI barton_t_Page_36.QC.jpg
90fe090567f7980d6f8bbe2158952459
f5707969922064aa6461c6d80c89b2062d499b0a
37304 F20101203_AACJZI barton_t_Page_30.pro
1276bad6e35e9c9bba190065553c5e21
cfe26de215e79ace5288064f318acd6e4f4d6169
F20101203_AACKBJ barton_t_Page_25.tif
3658212777ef6d2c4b8cd53cdec44e8c
161af22b4fa1564fd21273e46f9ee80e3e78d6bc
73969 F20101203_AACKGG barton_t_Page_42.jpg
fcf510d230fae2d71c1f51deae5bdfff
939f225d7d00a6bc534ea752f1ab90784bda31ee
F20101203_AACJZJ barton_t_Page_07.tif
8bf0d87d099fd47f50179c782a45ccc5
39d681d31667088d31de9b5ae81324d18461cffb
1051950 F20101203_AACKGH barton_t_Page_04.jp2
c43bb95c61d88521b6e43548e270be33
413f4fc44a211a0b713ff06c2d555625765b78b9
F20101203_AACJZK barton_t_Page_23.tif
37688a66c2790841ca34899210b16b99
2d47b923e59a495672fe6cb353bf8d11a166953c
5323 F20101203_AACKBK barton_t_Page_21thm.jpg
b26294b8e0aa4b9838354f9504e0a934
ae376d200bc5bfa31e352febb60f5a374ec06447
731992 F20101203_AACKGI barton_t_Page_07.jp2
d38bfbc8365633ad8dad40a78fa6b6cf
5307dd09a774c1001d551814212e59336f0a58bc
41426 F20101203_AACJZL barton_t_Page_21.pro
0086fa75581f3556045f29a52a12c98a
f710e7a42c1a67f9a952d31ae8b2e29ca6ae27f9
27270 F20101203_AACKBL barton_t_Page_34.pro
ad527b3085dfdf334c4f24a425d729be
9ca2be5c8860dbf2651337f61439c9a9cc1bc477
1051980 F20101203_AACKGJ barton_t_Page_08.jp2
0625c9a62192861a26276e7fc35850d1
7e666d82af27914e5e5129c4ed31ea9505075d42
20054 F20101203_AACJZM barton_t_Page_19.QC.jpg
1a666a2e1676d3abb1ff164e24705729
e6ea830a1730b1942029b08a4bf56a9d850a47b8
50762 F20101203_AACKBM barton_t_Page_39.pro
810d5cef4b88f7ffe170d89a8d414c97
9bb5a0be00168876c5fa3d0bdd140c965428b93d
1051927 F20101203_AACKGK barton_t_Page_10.jp2
98e2487dc4b27990593bcc60d8ed6577
7f25169f30e612503fbeeaa6f481a81c16cf3b65
2596 F20101203_AACJZN barton_t_Page_05thm.jpg
b4a20e3f381b4ac2c9308a4fccebfe9a
1b2924d6186037a410ed8d7b4f8666c4fac74d45
15322 F20101203_AACKBN barton_t_Page_26.pro
6bad10aa6da31fd24fb46a959b7c206c
a3b9fd8c47ce3ec8078c9669eddbd76942727a77
1051960 F20101203_AACKGL barton_t_Page_12.jp2
f0e47798a818c62371cfd629b2acb63b
4d3359244ac768fd0d617c4f51e28a8c47f6be01
55564 F20101203_AACJZO barton_t_Page_22.jpg
b9faae78963b4cd1726fdc4ed96beabb
119469dc7d720127625ee6403c31e0fa023363a5
27658 F20101203_AACKBO barton_t_Page_37.QC.jpg
200aec1b9dfc71125e97eb533a3aaac4
29acde26a79e718affa32013e014cf996c0508f7
911056 F20101203_AACKGM barton_t_Page_14.jp2
eb29f5a8e244ab2904b262228047b228
8a14ddc437b774c3857ddb3f02496996c2e6165b
F20101203_AACJZP barton_t_Page_28.tif
c1257f27e49d0156fb5686889630f0b3
e953d1b025050747048953163b5b6974a2b5739f
78112 F20101203_AACKBP barton_t_Page_25.jpg
0b881b344381debe4de56f07d06cb89c
0b7b587d6ca21a5289d984e06f3a276d1b4ff4c4
971031 F20101203_AACKGN barton_t_Page_15.jp2
8b3beb64837deae37eb373566187ec1f
595a88f118338c2efdd4883c2a238a20856958a3
76535 F20101203_AACJZQ barton_t_Page_28.jpg
b6358ae0ced73d27961ed84c6dc8cb0f
e2e83ca73912773e4e0051328f8c69869f4d967e
19058 F20101203_AACKBQ barton_t_Page_04.QC.jpg
7425636e559f191ae3241f701bc770e5
6ccf9723a87e1dce32ad2396ee23e67739bf071f
755337 F20101203_AACKGO barton_t_Page_18.jp2
40aea13fd72e3e0a990e1c8127f287d1
289ab8d529afa594f5dd116d1c139dc40f2e4763
76320 F20101203_AACJZR barton_t_Page_12.jpg
cad6560cf5b9efef5957ab44c88742b6
c3fc7024f937be3b5d145ca87b27a57dfc1025d1
44032 F20101203_AACKBR barton_t_Page_34.jpg
04c3fca52abe848bbddd75a5b4dd9871
ec757fe04c20411b4329ab5f7b2af023f2a11db9
81661 F20101203_AACKGP barton_t_Page_21.jp2
215cdb09819c9f4c5a53d4124b79c2a0
96078d4e8f455b02f07c05a10847776a5e2cab88
36079 F20101203_AACJZS barton_t_Page_18.pro
7454a0e5c36c31db82cb1db13702854a
50efc3d12df492189ae240efc6826dad3dd51e44
24718 F20101203_AACKBS barton_t_Page_12.QC.jpg
92e1e016e40f3e5c561ed2d8fa5618eb
efd7b5969ecd0433a406b2c06d3afd3c80ef8d59
936611 F20101203_AACKGQ barton_t_Page_22.jp2
49ee371367df67d77a3f314eb595c7f3
daaac95c69953827e8d4f4431878433a492c6fa5
15332 F20101203_AACJZT barton_t_Page_11.QC.jpg
8a0a4c8cb216dd60cb7608c31a858917
40802947fb10358848c90eb3db9db9206133a406
78990 F20101203_AACKBT barton_t_Page_08.jpg
32932013b249bfaf065e59ee4bf241d7
1fdcfc3fdd9ea4c0cd246ecbfbb0377aced3ba85
883796 F20101203_AACKGR barton_t_Page_24.jp2
0b01f131f894d403e22a135170ce5c5b
f2c7e9985653b46d0a0021da85f92c14ae891653
2016 F20101203_AACJZU barton_t_Page_38.txt
ca41f41139e392ce9940419a4e6be0c8
7a60e9f583bfb2ba6e03cf53c874be8b4b88f391
F20101203_AACKBU barton_t_Page_05.tif
9a01f5c692f70bd767dbf36f45ab99a0
838da1ff9f2510ed61d647cbc767069cb9ab7969
116280 F20101203_AACKGS barton_t_Page_28.jp2
2d39538148e55622d33f96742f4ffb22
8b41a77fccf7fb7e8b48f1b0b14663f882a85798
64679 F20101203_AACJZV barton_t_Page_06.pro
d9ab3363f9e5547b3c51bbef55717ea9
0f43f24cfb1703dbcac92f64d3388d3c67bbddb5
1051971 F20101203_AACKBV barton_t_Page_25.jp2
af24352745c65c95f8e47c35ce1a2003
b22e32433e5d255d735ff3e6a3067e2516ffd3c1
856135 F20101203_AACKGT barton_t_Page_30.jp2
6039ec34e5a54c4573bb332e7c028a73
d2d5ea5037fd5b6b13da33fa2b11fce747c501e7
19019 F20101203_AACJZW barton_t_Page_24.QC.jpg
e8c6c32a7b3b90bf1170a7a17f352541
41b9a1d447eead14bec1b1b2b76565b038433131
1051967 F20101203_AACKBW barton_t_Page_13.jp2
47f4e9de4ba7db9fb2f4fadc09d467f6
9fdb4d5ae7598e33b072ef77d9c21a6a5460c1e4
115468 F20101203_AACKGU barton_t_Page_31.jp2
ae87010fb055ead1ec6a85697863d3d6
04626d29c64425f03787c4531a692cabc7c46eae
1051891 F20101203_AACJZX barton_t_Page_37.jp2
eb0e72c09f3c2225377f4c40506b5e54
8c26f21c972b5377a787f7d32cb451bbe7fe9e6a
7925 F20101203_AACKBX barton_t_Page_05.QC.jpg
2f588227aea25b02f96cdb39dbacc9dc
49815d052badb585b460e42ad9c88f3d5ee68aac
1051929 F20101203_AACKGV barton_t_Page_32.jp2
8cfe8214e90b371a2dd935e2364e5ccc
ab80ba7f1b2993663c2b3ec025bf998d8cf3b762
1051942 F20101203_AACJZY barton_t_Page_27.jp2
306f2cc706a1dea55275fdc7bc28338a
aaf383d3c356a6080dbc30b938c28e6a1eda672e
7742 F20101203_AACKBY barton_t_Page_01.QC.jpg
d713be41cae48961d43f586b0b0be888
e6eb8ff6ae6d56e8fc38fde63c9d20fbcb3548cc
862223 F20101203_AACKGW barton_t_Page_33.jp2
7820d7f36713e5669d17a826e29c500e
cf87fc59520cbf1d3b39bfbea9db36eaac13e34a
80351 F20101203_AACJZZ barton_t_Page_29.pro
f19ddb633d2575d85698d784fb200582
6fecdb8ed92fc1819e47382b12f66a3628e617f0
803952 F20101203_AACKBZ barton_t_Page_17.jp2
9947a8bf242b47c3063e96bda17c9821
dad95ce979b90e35f0e38b3014d39da68c1bb719
765760 F20101203_AACKGX barton_t_Page_35.jp2
237615574fa9027746704bf2aaa6f28d
f9ba275c257017af3305e74318f3002020f0c213
70713 F20101203_AACKEA barton_t_Page_15.jpg
6a81415542dda7037ed5a2440eda1ed2
2e15039517b441ed335ca06a4353c6389db2d1a5
1051978 F20101203_AACKGY barton_t_Page_39.jp2
0b6003f7d37328cd7a65669aba6597be
317e5f2712fd968748fbfe6b8de18d8338b81587
2218 F20101203_AACKEB barton_t_Page_28.txt
d5ffc327abbc2c8d3d9ea12d28ffc91d
558fa1ab7b55575e8cbfcf1ca961f6de924ddd5b
1051986 F20101203_AACKGZ barton_t_Page_41.jp2
5b617ab6d25c4586da577eefb31fe99e
c4a3e44b95ab098122fb32503ef8203d79263ecf
78838 F20101203_AACKEC barton_t_Page_39.jpg
a7d9fa5236573cae221895642a9e2847
b657cef09364713013330d964bd61d8d65a0bcbb
50579 F20101203_AACKED barton_t_Page_10.pro
777ebe9b722d8108150ea51fefb47028
3f748ae35d0c2246921b96ca2fd5c8520baef4fd
3138 F20101203_AACKJA barton_t_Page_02.QC.jpg
be430537cd5a59a88dea31da9f83b41e
6815c06df20034bad4cb02a89a068a528c9c5c93
19213 F20101203_AACKJB barton_t_Page_18.QC.jpg
cd55b28e21b21ea391a92c5c72f93d49
da4e16c873964c3280e65ebd63dcb2bc8589602e
254302 F20101203_AACKEE barton_t_Page_01.jp2
3febf114d4b97aef26c67034c7e06a11
d1a53b6f2dea9044cf290d8b0f39c3b726866d3b
22239 F20101203_AACKJC barton_t_Page_13.QC.jpg
afcfcc98781d0dd24a6081dca4988ed4
cf6145f28d66d10b20407e6b9bb0ff412574142d
26519 F20101203_AACKEF barton_t_Page_20.QC.jpg
6857038c140ddd83d4f19b9131d517cc
02660a50ec71cc7d1b77a0357b5ca865a3c17160
7451 F20101203_AACKJD barton_t_Page_37thm.jpg
3ebf91c79ecce37a0695071f93c3ec46
356c35347dc5c9b2210e11f0e399e5f6ac9a41b3
28417 F20101203_AACKJE barton_t_Page_29.QC.jpg
59516e9462e83517498aa66f3af399bd
c45b5202c132ad32ed8b78eabd21eb27ba1d8890
16014 F20101203_AACKEG barton_t_Page_22.QC.jpg
deda4bad9dc2cd35f16799f6e2e4e22e
8ba6f54a2bb31018d4f96ec1525705774f3e2332
5761 F20101203_AACKJF barton_t_Page_33thm.jpg
055a04c26bf400966ed267a6f60df407
9f944a69ba93bfa7b0efe962f35f8e2b1f282cd0
F20101203_AACKEH barton_t_Page_43.tif
abfc2d5a0a0270404b8826cd9d27b937
8bb82bfa402f1223bf63bd8a40debf1a3dbf217e
21897 F20101203_AACKJG barton_t_Page_15.QC.jpg
dc91fec3430bf43bf2b34f8c2f2fd953
346bc82bd9ace81e162c7441edc5f7b6c70dd89d
25264 F20101203_AACKEI barton_t_Page_28.QC.jpg
21563d596a227b58b454866dceffb721
2d988497035b84dc10540a8f6e4811f6b2f9a757
6015 F20101203_AACKJH barton_t_Page_23thm.jpg
20616b8f226a698ce321e6d80410a5fa
b723ffce696de6a05fd38d30b7da545fa16aa1ff
18827 F20101203_AACKEJ barton_t_Page_21.QC.jpg
683d62ad123e03ea2ffcec787d0c44a8
a69d59657574bb97c1c15465cf3cddce0d5e24c0
24024 F20101203_AACKJI barton_t_Page_38.QC.jpg
198dd0f1ff877561d224a858c7163f71
4567ae1c341998e3741d48a93e8d2f6e6b6e96f4
23028 F20101203_AACKEK barton_t_Page_40.pro
7a7687984da2591e6b26f3f194142f62
9bbc462c610fa1c7447af7c6bc5569d1c3eeec76
4092 F20101203_AACKEL barton_t_Page_26thm.jpg
e7802ada482c5868343ea06fdca88567
f4cb0a8716d3e0db7626e30f850896d751b7e4cb
4526 F20101203_AACKJJ barton_t_Page_11thm.jpg
676ad42cbfc349ec3d33e42bc2fb6a15
5517f535016d95a62f4db9d895147f4c8eb1318b
787039 F20101203_AACKEM barton_t_Page_16.jp2
1c5aae75f1f620380d2e40d0f276fe32
8152c5016f0094a35e583270e4cb8c82c0cb4709
6369 F20101203_AACKJK barton_t_Page_15thm.jpg
7ef24133d822a41a20bdab3dc65ab517
99de819bac5d67a981db71dd6d189bb3c51dfee6
1969 F20101203_AACKEN barton_t_Page_09.txt
136c6ac63ac9c2edd24fbf6b2dbd065f
cb1f5ef2a28dc0590f20850257d38f4f78495d78
5826 F20101203_AACKJL barton_t_Page_16thm.jpg
8507a8df303ccb2b89451ed5cae7d39a
eb03443ca1f4dae6f60916c35ce2a63e9fed62bd
76001 F20101203_AACKEO barton_t_Page_38.jpg
6515eab0ecc178e544364886fd19e699
d713828640116329e0814a3661ae9ee91ca105f7
7631 F20101203_AACKJM barton_t_Page_29thm.jpg
8c34fe80c46db9f1252e18ed11314f7d
75bc43d77e0b18f55fe177eee082ecc06345c686
F20101203_AACKEP barton_t_Page_03.tif
9eb542eae517981ba1ced007c9308fe7
3969e7fb1baf6ec45453c6c4f99a6cce0af642a4
2706 F20101203_AACKJN barton_t_Page_43thm.jpg
83058604fa46850bd6d29f77e8a3794b
b280f63dcb259371287a50e8d001184a31869b59
1446 F20101203_AACKEQ barton_t_Page_35.txt
1e0e849081c10d66ed9cf88593c6ae52
6bf15d2a5434eca635eb6297306ea63717bbe9d4
3539 F20101203_AACKJO barton_t_Page_40thm.jpg
23699914563b1c2da8aa55bef0c87a30
0512e6b4ef1159f4240ab10a5befc0c1c9bc1c01
50652 F20101203_AACKER barton_t_Page_25.pro
0f6a1b292f58afa3e8e037faf3e7c524
89eba0e8dd1407e4d951a565b6f540927ac19b9d
67727 F20101203_AACKJP UFE0021824_00001.xml FULL
fb6f8217c1e7a0d5d132a1d5d20e0f97
dc6a4e398d0408a9f61fbb0cc8676cfec7b1826d
2036 F20101203_AACKES barton_t_Page_33.txt
0d1bcce1e78db1c0e536bdf2dc0ed418
1afb1ea8b3a444f223c668b04961b0ccbfffebf7
2336 F20101203_AACKJQ barton_t_Page_01thm.jpg
08df84e6c00e5086ab7ad83449c2f620
4faa2e3fe3b53f5beb1f9c156742b550d1df6ae8
F20101203_AACKET barton_t_Page_40.tif
9fb7b5e32358a249502d6757689c49a3
94b1dbb9713b53acaf4793f8bbd1a1871d380e06
20678 F20101203_AACKJR barton_t_Page_03.QC.jpg
3c76b6a1ef0e705fe62c21e8b5a24afe
e4f5b17633f710f421b8e91d02cce798ad7d9622
2062 F20101203_AACKEU barton_t_Page_23.txt
0af6149e459bcb6acca252560ef35c0e
5ab5052be763f29e886078c795e4372230b3688c
5797 F20101203_AACKJS barton_t_Page_03thm.jpg
c1a082fc41c950fcba989651fc3c043a
9133e1d1ac3626dafe3baea04c0515c6442ee89a
50114 F20101203_AACKEV barton_t_Page_11.jpg
a00836c931423028d5d27cb445304080
7a580917eb560a805c8adcabc5445f6b84f1ae38
4978 F20101203_AACKJT barton_t_Page_04thm.jpg
3dc2fad68e5fbff91d56934ff85d3bea
e7ef147b066cd110da8c6e4565a2a4cb02aef02b
F20101203_AACKEW barton_t_Page_29.tif
0538ed89db25e7e317528c246f360621
ecd6c649e9365fd983c4e14c430dc569ae464cbf
22284 F20101203_AACKJU barton_t_Page_06.QC.jpg
5e4991b3bd7ce7243a5160a318148a49
41ce0b1c089c23f22c99ea9785085dfb0a087ad6
2377 F20101203_AACKEX barton_t_Page_41.txt
39ff5522338702ca9c8edf63766dee66
a79782cd2456b1bd735179c92afcb5e7dffa0f70
24806 F20101203_AACKJV barton_t_Page_08.QC.jpg
f2c744ec02a86ec2836d9e1084a49b93
1542c98f48304498f76f72caf0d128fb3d40a031
2228 F20101203_AACKCA barton_t_Page_15.txt
77df0bd5fdece78b5a57a4ca0e43c35a
2483d22858b482d93c2044a5122308d475e1dec3
6649 F20101203_AACKEY barton_t_Page_09thm.jpg
ed6d8f699a741758106564ec69ea0831
2af865103106f1fdae9a71c4f3c01c4d6c4e29b7
6954 F20101203_AACKJW barton_t_Page_10thm.jpg
0987e46fade07a0270fee2bfd5aed362
3177709d4fe88ded97304691ff625d708927c070
59116 F20101203_AACKCB barton_t_Page_31.pro
20f7f03577bd84f64889379cd2734e57
1ca88b580c27dd5a2e3d2ea2971a8eaa8d3e5c99
18215 F20101203_AACKEZ barton_t_Page_05.pro
13e2a6a630700d5870f9ccc38ce960e6
727addd40a9f82070233e0965066ba24190e7ab0
22025 F20101203_AACKJX barton_t_Page_14.QC.jpg
f7d5ab699bda4c1efa4d2260b6d191fa
f3108915bda0b916e39488a644e49309b6b34296
303218 F20101203_AACKHA barton_t_Page_43.jp2
9e546152ac7ff504f3f6caf76c5aead0
84937767e45f9bcf209809054ab4daae1552f663
6382 F20101203_AACKJY barton_t_Page_14thm.jpg
7b99677ffea7c02906976ebcff87a673
67a2e44d74d74b6d5e00a8b09e1a877d022b70c9
27918 F20101203_AACKCC barton_t_Page_02.jp2
9af56fe1bd59a3f14be3a451783a64e9
b382d723bc25dac70d0f3e3960040942776059a1
F20101203_AACKHB barton_t_Page_01.tif
68b073619bc35f8b3e1f7eea3258decc
d040fe264dc340827591d4e5424929c48a513373
7351 F20101203_AACKJZ barton_t_Page_20thm.jpg
3ca89ebb9ea35720f9674dfb1d8fdb57
88e5c7f0b3613d25ed079b51da85af6c4eed5d87
50489 F20101203_AACKCD barton_t_Page_27.pro
d1b5ca5c4d55ad09571565ae110f620a
cb3bb2701212c6c1012473ce3c49e5185eafd707
F20101203_AACKHC barton_t_Page_02.tif
c69061f863ca3a35dd431cda2877d582
2bcf235c72c9b88cfe87f08b62d968bef2d18465
1051981 F20101203_AACKCE barton_t_Page_09.jp2
d4f0e28d6ef2aa923e519342aad7bb84
7e0efa5d5fb4cfe34280d7eaff7f826dd6661118
86633 F20101203_AACKCF barton_t_Page_37.jpg
e7db4118de12977927913047571b1db6
3b7a6756ba44a7029c084812f56c7cfa80adc11f
F20101203_AACKHD barton_t_Page_04.tif
dbc1446f0cb9860a887101e4513f740b
822360793c4a97e06c3690d5f6f90f5803233744
47535 F20101203_AACKCG barton_t_Page_12.pro
18346c3125373d51aecf229de4f51af7
5792f2248d3b5e35d6c27e749f820a9932ec65cb
F20101203_AACKHE barton_t_Page_06.tif
9a6b3cb5331ddd9ac879315cc52319a0
fc77fc297a64f8b8a48799f11312ca39c1a9bcc6
84242 F20101203_AACKCH barton_t_Page_04.pro
c9f0a810cf354753b861673e2fa9c7b1
b9d373eb667b9b3ec3c58f104d8c468720c64e88
F20101203_AACKHF barton_t_Page_09.tif
3e50bd5c0ba7511856268e26b487124b
09a23cb0e43fee17b38b30b60faf9a3a8c13561f
661856 F20101203_AACKCI barton_t_Page_11.jp2
cc1b51660cf5f0d65b0a007933d0c0df
b5cdacce97ba398bbf29fb71d5396fd6e07c6097
F20101203_AACKHG barton_t_Page_12.tif
5aab4a421043b0528e5f78f2d0513df8
6a847c2dd436532913758e85fced8f8c02588ecf
63892 F20101203_AACKCJ barton_t_Page_24.jpg
bc6ae9dcb0eda091223b26b66b861412
1c74de4a1f65b3964b6b743db284a8e7b26a0384
52909 F20101203_AACKCK barton_t_Page_20.pro
2843ec2a4e7eabaec27fc59518e4a315
dd305b16e811f68316a9bce8a518152e08d33a3e
F20101203_AACKHH barton_t_Page_13.tif
31f4e281b488beffb52c8bb4db9878d6
dc6e31f3e5c90e7015472dbc62defc707f5072c1
1591 F20101203_AACKCL barton_t_Page_24.txt
950d157a1f2087c8b9a533a425ddcb91
d5854c54577e0fc60d220d277ad7e4d348e0d59c
F20101203_AACKHI barton_t_Page_16.tif
e5e085c10eec8ed816c2d6c11c01744e
b631a1537ce90bb0d06630e8dc789c5ab6e5da9e
562699 F20101203_AACKCM barton_t_Page_05.jp2
9ca25d3df373e1aa85dc1922cc5a10b7
791fbebb1dbd2e4676f8058562a90e575ffc014a
F20101203_AACKHJ barton_t_Page_20.tif
8928c9d2b4750b8d723d8b6643a24f88
1e3792bd0cfbcf8d4bd01a224beeea7d14410df8
19923 F20101203_AACKCN barton_t_Page_32.QC.jpg
4a99a8e33da2bfdda4c4616b357cda52
1a1a53718c264e39eefa1528b32564025760b9b0
F20101203_AACKHK barton_t_Page_22.tif
89a688eec406d49027885c04c78fb643
3ed99a3ce6b20a17b5d57c148ea6fa25a8dd6bbe
17037 F20101203_AACKCO barton_t_Page_22.pro
b5194dfcd082d4adc881769c7d27df39
f0e4a2c51ac120e7c72f0952db67b48734f83c6b
F20101203_AACKHL barton_t_Page_26.tif
7fdf6d10b2a848c28a7e0704ca7fca81
391f52203314c71957305b751eef2b0a16eb370d
62013 F20101203_AACKCP barton_t_Page_19.jpg
2dbee65d65d1e44b2526d6b9e9288e67
ce0fe6d082234a4ea9c2421452935740282bf4ac
F20101203_AACKHM barton_t_Page_27.tif
7e0fcc736c3c68f2d2d7699532e0667c
c739117e482041f0931af1656b0afb0917fc5408
43976 F20101203_AACKCQ barton_t_Page_32.pro
428118ff03f199165e507caeba7cd3a0
ac106dec880c05567c62f55ef6d42d02a25693cc
F20101203_AACKHN barton_t_Page_30.tif
472b96f204c092debb9ad942ec63f17a
4da640a19af3d1ee0446ea00f7fd9ceae8d61da5
78246 F20101203_AACKCR barton_t_Page_06.jpg
a431443fde49b893ad3f2d2252fa32e2
0bf2b0bd473460da75c4424233519808b4a69406
F20101203_AACKHO barton_t_Page_31.tif
c149cef938338a16bfbb6e603bdee3c1
ae6f3628be155d5a4af53d12e7cb7391f58c3bdc
19211 F20101203_AACKCS barton_t_Page_16.QC.jpg
1901e6e15361ba13075b5ab48de34cdc
dd2a5d29d582bd3f04b6a8e1c4f5afeee4a61729
F20101203_AACKHP barton_t_Page_32.tif
fbfa28168df06fbb5380c741fcf625a3
9d1bc09d38a5bf2bcceb0896bbe5880bd11c3586
1051898 F20101203_AACKCT barton_t_Page_36.jp2
3dc99f887608b356e2da9b236ade58bb
8adb483bf3f4fb88d93ede0a49e75797d0df2ca0
F20101203_AACKHQ barton_t_Page_33.tif
9f67338125c54da8dab1211642d5b633
89fef66333739ae8a2280000037a118eab4f7208
1696 F20101203_AACKCU barton_t_Page_17.txt
d2b548d329623f1373e419d45d56e5b3
be1768595114b6d6f0d35c19945e99caaad50249
8423998 F20101203_AACKHR barton_t_Page_35.tif
24f37987e5ba6fab35d5893e18ad704c
5ad963bb8117b2af8ce9d1b7541c8a19cc6c9a7b
6906 F20101203_AACKCV barton_t_Page_28thm.jpg
2f088b4c07f950be2c5a726a413e3e3a
7e794557880efac9ee2f51bdcde389aecdfeed65
F20101203_AACKHS barton_t_Page_36.tif
2859239cf1670cd8a3fc6ac55d92e75b
6cf89e4627877ed7f2ee679d881374effff4a025
F20101203_AACKCW barton_t_Page_19.tif
c79728813dcf89bda618141c1bc2f089
f2ebf3eaf1e318d9c09f718caeaf4966b633913c
F20101203_AACKHT barton_t_Page_37.tif
b160d382cdd1630e0c53fd935e066998
eaf593e8858af9ae23703faa8b044e7536852e2a
1051985 F20101203_AACKCX barton_t_Page_20.jp2
4770f744c1236eef1e2fab314951a98f
efa4b852a874769f1b6f5cbe5f328d615b285a10
F20101203_AACKHU barton_t_Page_41.tif
bd4f434c218cca93e2444d1728d9ec47
28eb1246ec163d9c62ac42c455ea32ed2cde0fc6
6000 F20101203_AACKCY barton_t_Page_32thm.jpg
3cd1eb42c92e85a211d9e21b9e1b525c
6093588c5c5d39c4fdefc10d5656ee672445aec2
F20101203_AACKHV barton_t_Page_42.tif
799efc41ffe5af06f27f6b31a3f9f0bb
b0a58e7dd77b93869e7d94eb1df7ea137b94b779
40444 F20101203_AACKCZ barton_t_Page_26.jpg
42078a7849ac18cb9efa273dc0c5ca28
68d8a25b3f8adae25d642c33e7752123e6ca6b86
8562 F20101203_AACKHW barton_t_Page_01.pro
94a27eb0b4336a0b86f439e9dc945555
43eb03479a1fd293b34d2f567023262b3e83a9ea
1634 F20101203_AACKAA barton_t_Page_16.txt
628dfacc45b16edc3137fe04c7eabe63
8fadf9f93f45dbbf8185507a17972a8fa1483df8
40493 F20101203_AACKHX barton_t_Page_03.pro
864305fe286654b205458becf1889820
e860c8e6e7d2a470767e7399b3f4a0bf53b7625a
5930 F20101203_AACKAB barton_t_Page_30thm.jpg
9f0f5693c5fdbf52844634a828f8a360
da0a0efc3d7643a254de8cc6d47682108512d6b2
20102 F20101203_AACKFA barton_t_Page_33.QC.jpg
13e028d065eaced394c628c10177cf26
80a3a659548e5c68a773731e2a4e910a961d927d
48812 F20101203_AACKHY barton_t_Page_08.pro
69d8b5a769119e037e41e96b16650683
e2e7431f920589f629a78a29f41bba429ab693cd
57846 F20101203_AACKAC barton_t_Page_21.jpg
ee5663f5fce7602394a79325a9967a39
dd10835078c7a8685205ad66e834480bca4767da
25337 F20101203_AACKFB barton_t_Page_10.QC.jpg
61e5fcdf8943c15394bc19cbf0536ea3
3faa35ff4f6dda87a0581f74534d0686fa160ff6
29353 F20101203_AACKHZ barton_t_Page_11.pro
110177d46030d31f98ad0e7971e3cba8
a25213580d26aa9f9d5568005b550d76ebf19587
37328 F20101203_AACKAD barton_t_Page_17.pro
6fc0b800519be9e1d07ff062142a7bf2
8c1d6379c034e817dff6a9cc5231c2795d422a2a
17455 F20101203_AACKFC barton_t_Page_07.QC.jpg
c418a34b5616664003448a87e1c76cd0
5463cec696b126ece67be47d58afde583589ac70
56110 F20101203_AACKAE barton_t_Page_37.pro
80b6d4f2ee8b0cac0c77d91cc74cde24
09f66d694fa321c70da5fda9f3ad6ad668686848
57751 F20101203_AACKFD barton_t_Page_18.jpg
9d22fd1ffc815775e4f7bec92d49d802
120d4cbc14acd9315efa93ff3cd8781434dd923d
1040308 F20101203_AACKAF barton_t_Page_38.jp2
539e65250bb15ecd0d6236c6f838eed8
b07c3bf7a92b0d6c6829ff8a08277acd3af0aa48
4950 F20101203_AACKKA barton_t_Page_22thm.jpg
9347bc0253b8e52be874dc3680e516f0
320cd6d7246706324eb4a653d9f7e6e13cc08289
F20101203_AACKFE barton_t_Page_15.tif
e1f46b34fb050a21019dadff2c6f50e8
5c3212da0e693323383a62779a9b389721be5dc5
1716 F20101203_AACKAG barton_t_Page_13.txt
bec9e0a075d1f628f03bd5c7df07c3af
59dc91f506d749355279948ffc48f3eeb4a0a9dc
6864 F20101203_AACKKB barton_t_Page_25thm.jpg
0edb9f309f872fcf5e9bb7f73bf938ba
b74a2f5724250d7110f711f5224cd75722669871
56129 F20101203_AACKAH barton_t_Page_35.jpg
47ab5fe7c404e04d4fc1b43e06e124c5
9857db5e36de745459755a85fac57787143c9f61
12745 F20101203_AACKKC barton_t_Page_26.QC.jpg
44d26392922665ed8dd272c9c6f3a8ce
47cb5dc7874ffc1111909f3c190b75559c932ed5
65081 F20101203_AACKFF barton_t_Page_32.jpg
5bbfa497909edffcc5cb58f123e4cc4b
206a74f48233f2ae0e739b8b70a66da0c965c658
1757 F20101203_AACKAI barton_t_Page_42.txt
c3052e326a41d4c7aef6fdb803349945
353388074825764edd6e02bff96154a2ff353e65
24917 F20101203_AACKKD barton_t_Page_27.QC.jpg
9fb653691663efb1e8166ab0c4e588ce
f78e559f75180dc58ebfd0b21d620b1d06592ad6
32648 F20101203_AACKFG barton_t_Page_35.pro
c113f59599cfe9debe4b5c875fd1c43f
8f3d1f434e6bd05c27c2d2a85c013a2c07f60a00
7033 F20101203_AACKAJ barton_t_Page_41thm.jpg
20d7d5c11532671bc4a6b79617f09e14
3db48963306a82312ad27425a030077e1f9a7fd8
25014 F20101203_AACKKE barton_t_Page_31.QC.jpg
dcc094648c3058b6196d081f0fccec24
3b76cad5cf725ea5a3810100da44e3fd59282153
5029 F20101203_AACKFH barton_t_Page_07thm.jpg
792cf54edf4b40f1aa209ed07e50eca3
7a5a804fd56ea0b7e5de92cd3c7e86cda63b7ef9
45106 F20101203_AACKAK barton_t_Page_23.pro
6bf48867632409a86973e4befe9b7a8d
2523ddba9f822d67472557ce3819aac07d65767f
13756 F20101203_AACKKF barton_t_Page_34.QC.jpg
a41f63f164b6e00f0bdead725b97573a
d323e5d19e9c6f0ab15dfd4bc90b6153465cae1d
5931 F20101203_AACKFI barton_t_Page_06thm.jpg
e9f9051dc3e773e79ba7ab88c591ceeb
6dda3b6ef0550370b590efd941e9329f1423c211
24742 F20101203_AACKAL barton_t_Page_39.QC.jpg
de8504a3c6a1accd2dd77f1dd8ac35bf
d6c371684f60444fc9662db7b1a23bbae0c50048
16965 F20101203_AACKKG barton_t_Page_35.QC.jpg
df58de233d1e4b0de73a40739eee4540
898aedd7e606f4dd69c029963997261a2e11bf9d
1051982 F20101203_AACKFJ barton_t_Page_29.jp2
f73725e07af2299731c8f0c08606c566
8fbbf62407367435b7599ead39483da2e4dab2df
949888 F20101203_AACKAM barton_t_Page_42.jp2
8473bfc744d98de46e04ba4a56a2e0d7
67099dc6f6b0885a0fc46bd689664ca06643de2f
7063 F20101203_AACKKH barton_t_Page_39thm.jpg
17a1d2bc8ec977d53bd16575fe151b7c
5b803c5c83b97898b69be151fdcad3a506f8f48f
6702 F20101203_AACKFK barton_t_Page_12thm.jpg
fb2d8cb855e16eaef2b441950cacb404
445ddefbc4890d9449abfc72c7b9a7e5594293cf
827579 F20101203_AACKAN barton_t_Page_19.jp2
997560a18ee0f3ab1761c4961f21ccc2
a199cb23c510dbbbee21fdb8f1837353b138f6ac
27009 F20101203_AACKKI barton_t_Page_41.QC.jpg
28ab924ef180099e48e30215cd42bcf4
db07df88ffe59bb05c0945bc543e79958bd11603
914136 F20101203_AACKFL barton_t_Page_03.jp2
a1b8e57c79e98622d9d6ad12bf8f5ba8
21fb5745e5d2374f9ac054e7ad51dbb05673a35b
2006 F20101203_AACKAO barton_t_Page_39.txt
191530d0b513779282a550e30c6c3117
acedfba78332ae1e6ec273da26eaec9a2ce6e950
5526 F20101203_AACKKJ barton_t_Page_42thm.jpg
c3c6fdc3d323a6c65036cb1917eaa6db
799ead8e80097545e1e8e4efb28d37f094d3a67b
52614 F20101203_AACKFM UFE0021824_00001.mets
901858cca2fb9551cfd20f00a61cad86
c90c7b6607b4f6d48860a6ac4db192f683a233a8
5689 F20101203_AACKAP barton_t_Page_17thm.jpg
7a88e766a718d39e96c61bfe40a38bc8
6eab4e08243835a89aad509d7618d36a45a22a74
47984 F20101203_AACJYQ barton_t_Page_09.pro
74989dc7eb0d3105ca7d4e6a8cc24e28
9a4230aece66b54f500de52841e793fbc05d3412
6449 F20101203_AACKAQ barton_t_Page_31thm.jpg
c5770a7717dbc39aa9ecf0db0347fd43
91beadcf4a343f1a3614f26cac7c935e35455071
F20101203_AACJYR barton_t_Page_10.tif
cb273c10edc24720336e8f13829fd6ed
3ce7c49e0c1e273a95bfe71bbe6c3e38f0cb6cf0
505463 F20101203_AACKAR barton_t_Page_40.jp2
b44c5567769771d485a082098d6ef456
91da116840597c676ce9e4ab0453b613bdf8db10
25954 F20101203_AACKFP barton_t_Page_01.jpg
0b5a83ff55933fad8872a00390070d7e
ca40ad6daadc69515d8fb0af06bda25c81631601
69449 F20101203_AACJYS barton_t_Page_04.jpg
bd6dd0a8a6e0758d204f4b9c788f4d10
dbd566ffb04b0c2b49f00d81989150a532e1beb0
4652 F20101203_AACKAS barton_t_Page_35thm.jpg
80232d2a0229318b3583995e1c161536
f1c4cf2da526c29f29d3da7af605619ea45368b9
66733 F20101203_AACKFQ barton_t_Page_03.jpg
96fc49f4763286b8721c9607228e2c7d
a03d7ecd89b4b8138f9d9a1ea3a764f84ace7c72
27866 F20101203_AACJYT barton_t_Page_43.jpg
9bf62f32ac9618d210251641bb6f4c83
d3a80de2006c46e77c2f460abe3eb353d4fbbe52
F20101203_AACKAT barton_t_Page_24.tif
0b7a582cc7946eea21b8cae9ea663060
c3bf2f0398cdef2f217a92a51b03b8c474a602e1
56743 F20101203_AACKFR barton_t_Page_07.jpg
8285321b22d39d656221acebcabdb1f7
abb2505f2a10dc3872eb8d755959cf433adfc5fb
12830 F20101203_AACJYU barton_t_Page_43.pro
f2779da13f223426c9af9a5d032741f7
65abca05e3e7d199b338bc48ce847cbd9445dfe1
2026 F20101203_AACKAU barton_t_Page_10.txt
5baeed21213624a143b148bde44d7d7d
4296d5fe01e4c0405c35749467564c687a293ae7
80403 F20101203_AACKFS barton_t_Page_10.jpg
8da73b2533a38a03b79a0ebc36fc1eb6
0e6e8b437653c53d9536ee026e6f2ac448510f91
24444 F20101203_AACJYV barton_t_Page_09.QC.jpg
ed7f1c2e643e6aae2f7f59e921398f7a
371c8d65b995591e0cea4646687dd32d62027699
6011 F20101203_AACKAV barton_t_Page_36thm.jpg
65621dca35b39286331d9c2fefe5eb10
9628038f9efae27181c4ee289e0b7c9069a154d9
69393 F20101203_AACKFT barton_t_Page_13.jpg
cad92676a71d141242ec9b856258df3d
466d914f787ccbf1395d8691fa1786b96b3c7b71
F20101203_AACJYW barton_t_Page_14.tif
3c0a7b792168ea9abff7e2dcdebf3ae9
fefd65e3243695a214f87cb6518d372b63d4370e
1418 F20101203_AACKAW barton_t_Page_36.txt
31b350b32735686ccdcec56a851c6028
ad4d3d16a2ace958ce8fa9febc337c6127c2723c
67741 F20101203_AACKFU barton_t_Page_14.jpg
ad249b04a8800d48361b0d574e953d45
a8986fd02aab914356e1c0aabcb0e019cefc56a0
F20101203_AACJYX barton_t_Page_39.tif
0f12a468125c357a7250236414e5a85a
ff3a72e81a4b9c8fa0fadb3617804185d20fe0dc
2626 F20101203_AACKAX barton_t_Page_06.txt
494e99e15c4422f5371d1593359105dd
786b0e1403dcfdd6122039932fa276e46c0547ce
58573 F20101203_AACKFV barton_t_Page_16.jpg
45c0163479ff7d6990dd7c3115a7767c
3de51a091ff813fa31b45b5363d790acf2d2bc48
5757 F20101203_AACJYY barton_t_Page_24thm.jpg
fa1d3e64aa6307d6b687fc075e2dc0ba
448b425d996c93ee8a7321102670c7e53dd6f588
F20101203_AACKAY barton_t_Page_34.tif
78d47b4b7d2c3b5db8868d94a2d5e956
7147f331f959c383d1497963fad010747bfc46db
61258 F20101203_AACKFW barton_t_Page_17.jpg
79ff8b85580958123cbb8ae1478db4e6
792ab6b5e3fae1ffd656a146676e4b2cd8544acf
33408 F20101203_AACKAZ barton_t_Page_36.pro
c47a40f7907b82b3e7e01457ba7f04b5
b7dd788c54cfb288b0d4e11cb793880bb3773aa1
84382 F20101203_AACKFX barton_t_Page_20.jpg
5cdaad24a639fb8fd8b26d7b3d4c4a7a
e836ca3de31ce035386e65d057b5f888cdb44afe
19832 F20101203_AACJYZ barton_t_Page_30.QC.jpg
bbe5db1dd8111d3a7c1572ae5c37f390
6cccf67a283cd92a17e4a81b47b7878892e36470
71961 F20101203_AACKFY barton_t_Page_23.jpg
eeac87ab4599622d06ced51a447da8fd
90b322ff947f64c3f33bef97549ca2dbb981578b
43127 F20101203_AACKDA barton_t_Page_42.pro
75ff40f2603cb8937d1d567773d05a57
43eb8cf93b83ec6f643c9dea1362ef92b0455e0c
78440 F20101203_AACKFZ barton_t_Page_27.jpg
ada26d7ae2d2f9aceb945205662bb94f
d0e2fb13261475621763fc14f9d8411c5ce3ec16
10262 F20101203_AACKDB barton_t_Page_02.jpg
b05d373e17d954cd49f3e2204f565ae6
64a909fde829ae4a7ad016c6f273c46793aa7539
1699 F20101203_AACKDC barton_t_Page_18.txt
8e51a5f00a8f15ef189765135771c97e
cd3882a12757ee0d0aacebdb88ea5d2063aac84c
39149 F20101203_AACKIA barton_t_Page_13.pro
05ddb5ebd8b3122840ff42bb5c5790c1
245f81c868be4f5254481399d0f85aac35a9b751
42931 F20101203_AACKIB barton_t_Page_15.pro
c789a10bbae971cb22e0f61ba992adda
93ad484db2e7449ed764bd749d165f5e4f4eaf83
11215 F20101203_AACKDD barton_t_Page_40.QC.jpg
d6b3e0ad84bb7026a89d0ed8f84b86d8
a9bb57b038cfb879a78ee874a23965b2699a5166
39060 F20101203_AACKIC barton_t_Page_19.pro
e160774f4547a7f1b54a14283fbf7eca
baf7171394a2f269cb6f17828c5a550a8004b65f
5242 F20101203_AACKDE barton_t_Page_18thm.jpg
0d97c6e56d909f39575d5c54bbc9e807
21257be0ce6d1e19ed35826af093ac501574f819
54971 F20101203_AACKID barton_t_Page_28.pro
8ccc14514df37d831c4a5326383a574d
455a74e855e8991c7475e0eecd27fb122648130e
1348 F20101203_AACKDF barton_t_Page_02thm.jpg
2fb9587ccc20cd28f04f1911dcd0dbc2
9f99cd3d91f763494c2dcc850bc206bf6be15490
39317 F20101203_AACKIE barton_t_Page_33.pro
6e17b3ac6447a6a87040f0e4660df1d3
5af07f48f2bd29a870457564c1008f8fc46ff3f1
1416 F20101203_AACKDG barton_t_Page_34.txt
5a314cf1ead0ecbb24ee28c7c6fdda81
71cc53932f01111c01070cde6b823bda3d46407d
49151 F20101203_AACKIF barton_t_Page_38.pro
4d0005276e59fcad4f03ea2f13e84745
649ebf4cb44dfa61230e71ba6bd9e6b2d3b129b5
1694 F20101203_AACKDH barton_t_Page_30.txt
f1bf9fac00b107893537cc47e62dd305
62d4e943a72d5d86dac946e6049c94f828bff888
93 F20101203_AACKIG barton_t_Page_02.txt
ea310784c745fffb544e85ccc086b607
1bf96f127ec40fb5e386634ce2d34b2717c889db
2086 F20101203_AACKDI barton_t_Page_27.txt
97db54919ce33ed0fa6d0bfa4f774c21
78f59a6e70ce0aa3c4e56d6d3497ec59e58ad337
1658 F20101203_AACKIH barton_t_Page_03.txt
5d89549be606b7999d6867a61ce9c7ea
346941963b1d8350cd65bd53e125ca5a27c0d390
76503 F20101203_AACKDJ barton_t_Page_09.jpg
19c47777026e65f01c12ab1b1f643d94
8f0c553145b19c506b1658369246401f1c67ec40
25440 F20101203_AACKDK barton_t_Page_25.QC.jpg
cd1e331955dc454dfc6d2f2b63ca3a9c
12fbb33d3d6365e4d72e5b51dc317b2241a681e7
3470 F20101203_AACKII barton_t_Page_04.txt
940de0a8a06417058cdf4db3bcc242cb
de905c7e51e2bcbab453eb63681013e2af0afe3a
677556 F20101203_AACKDL barton_t_Page_26.jp2
231351671f51d8a2a8954e09a6731b2c
9cbf6c01c64480aa3cd59fd8745a193ca4a7884e