Physical, mineralogical and interfacial bonding properties of carbonate and silicate mineral aggregates used in portland cement concrete in Florida

PHYSICAL, MINERALOGICAL AND INTERFACIAL BONDING PROPERTIES
OF CARBONATE AND SILICATE MINERAL AGGREGATES
USED IN PORTLAND CEMENT CONCRETE IN FLORIDA

By

ROBIN ERIC GRAVES

A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1991

ACKNOWLEDGEMENTS

This research paper was sponsored through grants from

the Florida Department of Transportation. The project was

under the general administration of Mr. L.L. Smith, State

Materials Engineer; Mr. R.E. Daniels, Deputy State Materials

Engineer; and Dr. J.M. Armaghani, Pavement and Materials

Engineer. Additional funding was provided by the Chemstar

Corporation, under the administration of Mr. D.D. Walker,

Vice President of Technical Services. The project directors

were Dr. J.L. Eades and Dr. G.H. McClellan of the Department

of Geology, University of Florida.

I extend my sincere gratitude to the many people who

have given me their help and support. This project would

not have been possible without the support and guidance of

Dr. G.H. McClellan and Dr. J.L. Eades who served as chairman

and co-chairman of my committee. Our many discussions have

been invaluable in carrying out this work and have broadened

my education to include many other aspects of industrial

mineralogy. I thank them for being my teachers and my

friends. I would like to thank Dr. F.N. Blanchard and Dr.

A.F. Randazzo of the Department of Geology, and Dr. B.E.

Ruth of the Department of Civil Engineering for their

efforts in serving as additional members of my supervisory

committee. A special thank you goes to Dr. C.C. Mathewson

of the Department of Geology, Texas A&M University for the

extra effort required in serving as an outside member of my

committee. I would also like to thank Dr. D.M. Patrick of

the Department of Geology, University of Southern

Mississippi for his enthusiasm for teaching, which led to my

great interest in engineering geology.

I would like to acknowledge the personnel of the

Florida Department of Transportation, State Materials Office

for their assistance in aggregate and concrete testing, and

the staff of the Major Analytical Instrumentation Center at

the University of Florida for their assistance in conducting

spectroscopic analyses. I also thank the graduate and

undergraduate students in the Department of Geology who

assisted me in the project and were very helpful in sample

acquisition, preparation, and analysis.

I am greatly indebted to my wife Jennifer for her love,

patience, understanding, and support, which have been a big

part in achieving this educational goal. Her word

processing skills were invaluable in the preparation of this

manuscript and many others. My thanks also go to my many

friends because they truly make life more enjoyable.

Finally, I would like to thank my parents for their

love, support, and encouragement to excel in my educational

pursuits.

iii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS . . . . .

ABSTRACT . . . . .

INTRODUCTION . . . . .

Purpose and Scope . . . .

Methods . . . . .

ECONOMIC AND GEOLOGIC ASPECTS OF FLORIDA AGGREGATES

Coarse Aggregate Sources . . .

Taylor County Region . . .

Hernando County Region . . .

Lee and Collier Counties Region .

Dade and Broward Counties Region .

Levy and Citrus Counties Region .

Fine Aggregate Sources . . .

AGGREGATE PROPERTIES . . . .

Physical Properties . . . .

Mineralogical Properties . . .

Sample No. 08-005 (Brooksville, Florida)

Sample No. 87-145 (Miami, Florida) .

Sample No. 87-090 (Miami, Florida) .

Sample No. 12-008 (Ft. Myers, Florida) .

. 17

. 17

. 18

. 19

. 20

. 20

. 21

. 22

. 22

. 41

. 42

. 45

. 47

. 47

page

* ii

* vi

. 1

S 2

* 4

. 7

Sample No. 86-062 (Pembroke Pines, Florida)

Sample No. 34-106 (Gulf Hammock, Florida)

Sample No. AL-149 (Calera, Alabama) .

Sample No. GA-177 (Tyrone, Georgia) .

Sample No. 50-120 (Chattahoochie, Florida) .

Shape Characterization . . . .

CEMENT PASTE-AGGREGATE INTERFACIAL BONDING . .

Literature Review . . . .

Textural Characteristics of

the Interfacial Region . . .

Bonding of Carbonate and Silicate Aggregates

Surface Chemical Considerations . .

Microtextural Studies . . . .

Aggregate Treatment Investigations . .

Improvement of the Cement Paste-Silicate

Aggregate Bond . .

Concrete Materials and Design

Aggregate Treatment Methods

Strength Testing . .

Microscopic Examinations .

DISCUSSION . . . .

CONCLUSIONS . . . .

REFERENCES . . . .

BIOGRAPHICAL SKETCH . . .

. 114

. . 116

. . 120

. . 130

. . 134

. . 150

. . 163

. 169

. 176

. 49

. 50

S 52

S 52

S 54

S 56

S 76

. 79

S 79

80

S 81

S 90

S112

. .

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

PHYSICAL, MINERALOGICAL AND INTERFACIAL BONDING PROPERTIES
OF CARBONATE AND SILICATE MINERAL AGGREGATES
USED IN PORTLAND CEMENT CONCRETE IN FLORIDA

By

ROBIN ERIC GRAVES

August 1991

Chairman: Dr. Guerry H. McClellan
Major Department: Geology

Physical, mineralogical, and interfacial bonding

properties of carbonate and silicate mineral aggregates,

which are being used in a Florida Department of

Transportation (FDOT) comprehensive concrete study, were

investigated to develop basic data to aid in evaluation of

concrete properties. Compilation of published information

on geologic and economic factors important to the Florida

concrete aggregate industry was an integral part of the

investigation.

Physical properties reported for the aggregates include

size gradation, specific gravity, unit weight, water

absorption, abrasion, strength, and particle shape.

Mineralogical compositions and textural properties were

investigated using X-ray diffraction and petrographic

microscope methods. Interfacial bonding properties of the

aggregates in portland cement concrete specimens also were

characterized by microscopic methods.

Physical data indicate that the aggregates meet general

requirements for use in concrete and reflect typical values

based on historical FDOT information. Mineralogical

analyses show that the aggregates have limited variability

in compositions and contain few deleterious constituents.

Physical properties can be related to textural

characteristics which are controlled by geologic history of

the materials.

Microscopic studies of concrete specimens containing

crushed limestone and granite aggregates suggest that a good

mechanical component of cement paste-aggregate interfacial

bonding is provided because of rough surface textures.

Smooth surface textures of uncrushed siliceous river gravel

aggregates would not provide good mechanical bonding.

Limestone aggregates also appear to provide a chemical

component of interfacial bonding, resulting in good

attachment of cement hydration products to the aggregate

surfaces. Extensive separations were evident at cement

paste-silicate aggregate interfaces, suggesting a poor

chemical bond. This is believed to result from the

different surface chemical properties of carbonate and

silicate minerals.

A granite aggregate treatment method, using a calcium

hydroxide solution followed by drying, resulted in higher

concrete compressive strength and improved interfacial

vii

bonding with cement paste. Spectroscopic analyses indicate

that this may be because of stronger adsorption of cations

on the silicate mineral surfaces.

Rapid population growth and development has caused

concerns regarding aggregate resources in Florida.

Increasing demands, fixed resources, expanding urban

development, and increasing environmental regulation are

factors that may necessitate the use of alternative

aggregate sources to meet future concrete production

requirements.

viii

INTRODUCTION

The State of Florida has experienced rapid population

growth and development in recent years, and this trend is

expected to continue into the future. This growth has led

to tremendous needs for expansion, repair, and maintenance

of transportation systems. These needs have placed great

demand on both the financial and material resources of the

State. Design, construction, and maintenance of the

transportation infrastructure presents challenging problems

to planners, engineers, and scientists responsible for its

development. Continued success in providing quality

transportation systems is highly dependent on the material

resources available for construction.

Asphaltic and portland cement concretes are the primary

materials utilized for construction of pavements and

structures composing highway systems. Mineral aggregates

are the major component of these materials. Rapid growth, a

limited resource base, and increasing environmental

restrictions on mining and processing activities are causing

concern over the future availability of quality construction

aggregates for concrete production within the State. As the

raw materials that have traditionally been used for concrete

production are depleted, new sources may have to be found to

accommodate the State's transportation needs. The Florida

Department of Transportation (FDOT) has considerable

experience with concrete construction using these

traditional raw materials. However, the inevitable future

need to use alternative resources requires that existing

materials be thoroughly investigated so that aggregates from

non-traditional sources can be compared with these materials

to assure the satisfactory performance of future structures.

Purpose and Scope

This study is one component of a comprehensive concrete

research program organized and managed by the State

Materials Office of the Florida Department of

Transportation. This program is designed to evaluate

systematically the behavior of portland cement concretes

made from Florida materials. Extensive testing of fresh and

hardened concrete containing aggregates from a wide range of

sources and prepared using various mix designs and

admixtures is being conducted for evaluation of strength and

durability characteristics at both the macroscopic and

microscopic levels.

At the macroscopic level, the concrete mixtures are

being tested for strength, elastic properties, permeability,

sulfate resistance, and corrosion. At the microscopic

level, petrographic, electron microscopic, and chemical

techniques are being used to evaluate microtextural and

microstructural properties that may effect strength and

durability of concrete.

One objective of this study was to characterize

concrete aggregates from major Florida source areas which

are currently being utilized in the comprehensive FDOT

study. The data collected will be used in the overall

interpretation of results from concrete testing using

various materials and mix designs and will serve as a source

of basic information on fundamental characteristics of

Florida concrete aggregates for future reference and

comparisons with alternative source materials. Economic and

geologic aspects of the Florida aggregate industry also were

surveyed to provide an overview of aggregate resources and

associated current and future problems.

Portland cement concrete is a composite material and

its behavior is dependent upon many variables. One

important aspect of composite concrete behavior is the

interfacial bond between the cement paste and aggregates,

which has been the subject of an increasing number of

investigations in recent years. The cement paste-aggregate

interfacial bond is usually considered to be the weakest

component of composite concrete strength and may also be an

important factor for permeability and durability (Mindess

and Young, 1981). Therefore, it is important to better

understand the physical and chemical properties affecting

the bond between aggregates and cement hydration products

formed in the cement paste.

A second objective of the research project was to

compare the interfacial bonding properties of carbonate and

silicate mineral aggregates in portland cement concrete and

to better understand the effects of surface properties of

aggregates on interfacial bonding. In addition, a method

for improving the bond between cement paste and silicate

aggregates was investigated based on surface chemical

considerations.

Methods

General geology and data on Florida aggregate resources

were compiled from the literature and discussions with

industry analysts. This information was used to provide an

overview of current resources, extraction methods, and

associated technical, economic, environmental, and social

factors important to the mining industry.

Limestone coarse aggregates from selected Florida

sources, which are commonly utilized in concrete production,

were tested and analyzed to determine their physical,

mineralogical, and textural characteristics. These are

properties that are important for concrete design,

production, and performance. In addition to the Florida

limestone aggregates, limestone aggregates from an Alabama

source, granite aggregates from a Georgia source, and

siliceous river gravel aggregates from the Florida Panhandle

area, which have been utilized in some State projects and

are included in the FDOT comprehensive study, were

evaluated.

Physical properties of the various aggregates that were

investigated included gradation, specific gravity, unit

weight, water absorption, strength, and particle shape.

Mineralogical investigations included petrographic

descriptions, X-ray diffraction analyses, and surface

textural studies. Historical FDOT information and

literature reports were used for evaluation of physical and

mineralogical data obtained for the various aggregates.

A literature review was conducted on previous studies

of cement paste-aggregate bonding and surface properties of

aggregates, which may be important in controlling

interfacial bonding characteristics. Microscopic

examinations were performed on field and laboratory concrete

specimens to compare interfacial bonding properties of the

limestone, granite, and siliceous river gravel aggregates.

Based on the published data and microscopic

examinations, a concrete testing program was implemented to

further investigate interfacial bonding characteristics.

Concrete mixtures were prepared with crushed aggregates from

a commonly used Florida limestone source and a Georgia

granite source. The concrete mixtures were cured, tested,

and examined with microscopic techniques. In addition,

concrete mixtures were prepared with granite aggregates

that were subjected to pretreatment methods based on the

literature review and analytical laboratory investigations.

These concrete mixtures were cured, tested, and examined for

comparison with concrete prepared with untreated granite

aggregates.

6

The information collected in this study is intended to

serve as a basic source of information on concrete aggregate

resources commonly utilized within the State and their

associated physical and mineralogical properties that may be

important for concrete production and performance.

ECONOMIC AND GEOLOGIC ASPECTS OF FLORIDA AGGREGATES

Surface and near-surface carbonate deposits in Florida

are mined for many uses, primarily for portland cement

manufacture and as aggregate for road base and concrete

production. However, many other chemical, industrial, and

environmental applications also occur (Schmidt et al.,

1979). The geographic distribution of these deposits is

controlled by the geologic history of the State.

Florida is underlain by more than 4000 feet of

Cretaceous to Holocene sedimentary rocks (primarily

carbonates) which overlie a basement of older sedimentary,

metamorphic, and igneous rocks. Miocene to Holocene surface

sands, silts, and clays, which may reach several hundred

feet in thickness, cover subsurface carbonate units in many

areas (Schmidt et al., 1979; Cooke, 1945). Eocene to

Miocene surficial carbonate units occur over a large portion

of the western one-half of the central peninsula and over a

small area of northwest Florida (Figure 1). These materials

were deposited in a broad carbonate bank overlying the

Florida Platform that composes the Florida peninsula, the

adjacent continental shelves off the east and west coasts,

and the Great Bahama Bank (Chen, 1965).

Deposition of Eocene and Oligocene carbonate rocks was

followed by a period of post-Oligocene erosion and

SERIES FORMATION .i .
CITRUS LAKE
Anastasia Fm.
ARN ORANGE
PLEISTOCENE Miami Oolite Ao- ----
OSCEOLA
Key Largo Limestone i <, POiLK

PLIO-MIOCENE Tamiami Fm. / 1--' -
MANATEEI HARDEE
ST. tUCI
^ Hawthorn Fm. L._--j DE O TOIGLOT ANO
MIOCENE SARASOTA, MARTI
St. Marks Fm. and ,icARLOTTEi GLADES .-
Chattahoochie Fm. AND
LL; H PAIM BEACH
OIGOCENE Suwannee Limestone and i
Marianna Limestone "ii iii

SOcala Limestone
Eoc(E
Avon Park Limestone ,

Figure 1. Geologic map of surface and near-surface carbonate
deposits in Florida (after Schmidt et al., 1979).

structural uplift ("Ocala Uplift") in the western peninsula

(Vernon, 1951). Miocene sediments subsequently covered much

of the peninsula. Extensive erosion has occurred on the

structural high, resulting in exposure of the Middle Eocene

Avon Park Formation, roughly parallel to the northwest-

southeast trending crest of the structure. The surface

exposure of the Avon Park Formation is surrounded in a

roughly oval pattern by the Upper Eocene Ocala Limestone,

which unconformably overlies the Avon Park and dips in all

directions away from the elongate structural high. The

Oligocene Suwannee Limestone unconformably overlies the

Ocala Limestone and, because of erosion or non-deposition,

is only exposed at the northwestern and southeastern ends of

the Ocala outcrop area (Schmidt et al., 1979).

Lithologically, Lower to Middle Eocene deposits are

predominantly fossiliferous limestones and fine to medium

crystalline dolomites, along with small amounts of

carbonaceous material and gypsum. Overlying Upper Eocene

and Oligocene deposits are generally high calcium limestones

with sparse to abundant fossil contents (Chen, 1965; Vernon,

1951; Cooke, 1945). However, these may be dolomitized to

various degrees in northwest Florida (Schmidt et al., 1979).

Detailed lithology of these carbonate sequences can be

highly variable, resulting from changing depositional

environments and periods of non-deposition or erosion

resulting from sea level fluctuations, and uplift and

downwarping of the Florida Platform (Chen, 1965; Vernon,

1951).

The Pleistocene Anastasia Formation, a sandy lithified

coquina that represents an ancient beach facies, extends

along the Atlantic coast and forms the Atlantic Coastal

Ridge from St. Johns County to Palm Beach County (Figure 1).

Southward, the Anastasia interfingers with and grades into

the Pleistocene Miami Oolite and Key Largo Limestone. The

Miami Oolite is a locally cross-bedded, sandy to pure

limestone with a variable oolite content and abundant

solution features. The Key Largo Limestone is a highly

fossiliferous deposit from a coral reef environment. The

Pleistocene Ft. Thompson Formation underlies these units and

is characterized by alternating marine shell beds and

freshwater limestones and marls (Schmidt et al., 1979;

Cooke, 1945).

Surficial carbonates of Southwest Florida are from the

Plio-Miocene Tamiami Formation. These are sandy, shallow

water carbonate deposits with abundant fossil contents and

dissolution features, and often have a recrystallized matrix

(Schmidt et al., 1979; Cooke, 1945).

Concrete coarse aggregates (>4.75 mm) in Florida are

supplied primarily by crushed limestone operations.

Production of coarse aggregates is primarily limited to five

geographic regions of the State (see Figure 1):

1 the Taylor County region of northwest Florida

2 the Hernando County region of central Florida

3 the Levy and Citrus Counties region of central

Florida

4 the Lee and Collier Counties region of southwest

Florida

5 the Dade and Broward Counties region of southeast

Florida

In addition to crushed limestone, some gravel deposits are

utilized as concrete aggregate in the Florida panhandle.

Many other areas have surface or near-surface limestone

deposits, but these generally do not meet physical,

mineralogical or chemical requirements needed for use as

concrete aggregate. The Ocala Limestone deposits are good

examples. Although they are widespread in surficial

exposure (Figure 1), they are generally too soft and friable

for use as concrete aggregate. Local post-depositional

diagenetic modification can result in small quantities that

may be suitable for this use, but this is not common.

Concrete fine aggregates (<4.75 mm) in Florida are

derived from two principal sources. Quartz sands, which are

mined in west Florida and the Ridge Section of north and

central Florida from plastic Pleistocene deposits, and

manufactured sands, which are produced from crushed

limestone operations in south Florida, are the fine

aggregates commonly used.

In 1989, Florida ranked second nationally in total

crushed stone production and fourth nationally in production

of all types of construction aggregates (U.S. Bureau of

Mines, 1990a). The rapid growth and development of the

state has caused a steadily increasing demand for aggregates

for the period from 1960 to 1989 (Figures 2 and 3). In

1989, the crushed stone industry in Florida produced a total

output of approximately 82,000,000 short tons with a value

of approximately $382,000,000 (U.S. Bureau of Mines, 1990b).

A large portion of these aggregate materials are used

for construction of the state's roads, bridges and general

infrastructure. During the fiscal year of 1989, the FDOT

let approximately $500,000,000 worth of road and bridge

construction contracts (Spencer, 1990). The large increase

in demand and production during recent years has led to

concerns over future availability of quality aggregate

resources to meet the state's construction needs. The

Hernando County region, which has been an aggregate source

for concrete production in much of north and central

peninsular Florida, is rapidly being depleted and industry

analysts estimate only five to ten years of supplies

remaining in this area. The Taylor County region of

northwest Florida is believed to have abundant aggregate

resources remaining. However, the marginal quality of these

aggregates and a lack of rail transportation may

economically limit the use of these resources in the large,

expanding markets in other areas of the State. South

Florida deposits also are believed to have relatively

abundant resources and reserves remaining. However,

extensive urban development and strict environmental

90000

80000

o 70000
I-

o 60000
U)
'3 50000

40000
0

30000

20000
1960 1970 1980 1990

Years

Figure 2. Production of crushed stone in Florida (compiled
from U.S. Bureau of Mines statistics).

400000

C 300000
0

o
0 200000

0 100000

0 --I--------- i -- --

1960 1970 1980 1990

Years

Figure 3. Value of crushed stone in Florida (compiled from
U.S. Bureau of Mines statistics).

regulations are making extraction of these materials

increasingly difficult.

These factors have led many planners and industry

analysts to conclude that importation of aggregate materials

from out-of-state sources may be necessary to supplement

existing resources to meet future concrete production

demands. Limestone aggregates already are being imported to

the Tampa and Jacksonville market areas from the Bahamas and

granite aggregates are being imported to these market areas

from Canada. Granite aggregates from Georgia and Canada

have been used in some State projects. Rapidly increasing

demands, declining quality and quantity of resources, urban

development, and environmental restrictions are all factors

that may make increased importation of aggregate materials

necessary in the future. This is not a situation that is

unique to Florida. At the national level, limestone

aggregate materials are being imported to gulf coast markets

from Mexico, and granite aggregates are being imported to

east coast markets from Canada (McConville, 1990).

Geologic factors that control the mining potential of

aggregate resources are lithology, structure, stratigraphy,

and geomorphology of the deposits. Lithology controls the

physical and mineralogical properties that determine whether

the material meets specific requirements for its intended

use, and the percentage of mined material that can be used

after mining and processing procedures (yield). Structure,

stratigraphy, and geomorphology control the thickness and

areal extent of the deposit and also the amount, type, and

thickness of overburden that must be removed during mining.

These are important factors in determining the feasibility

for mining of the deposit. In addition, groundwater

conditions also may impose limits on mining activities in

some areas.

Economic factors may be even more important than these

technical considerations. Transportation is the principal

factor that determines price and market for low value, bulk

materials such as limestone aggregates. The high cost of

truck transportation (approximately 35 cents per ton for the

first mile and 7 cents per ton per mile thereafter) normally

limits haul distances to about 50 miles. Longer distances

require less expensive rail, barge, or ship transport.

Barge shipment rates may only be one or two cents per ton

per mile. Rail costs are slightly more expensive (Langer,

1988). Therefore, ready access to local markets and/or

access to low-cost rail or sea transportation is essential

for an economically viable limestone aggregate industry.

The current uncertainty on future fuel prices makes the

transportation costs of aggregates difficult to forecast.

High transportation costs often cause mining operations

to be in close proximity to urban markets. This can result

in a variety of social and political problems concerning air

and water pollution, noise, and traffic. Land-use conflicts

also are common in urban areas. Quality materials often are

not available for mining due to overlying residential,

retail, or industrial development. These are common

problems in many urban areas nationwide. Zoning to protect

potential mining areas and sequential land-use planning are

steps being taken by some cities to mitigate this problem

(Harben and Bates, 1990; Langer, 1988).

Exploration and development of mining deposits also are

subject to many market and financial parameters. Many site-

specific issues such as quantity and quality of reserves,

proximity to markets, land values, and royalty requirements

determine the long-term cash flow feasibility of these large

capital investments. Examples of valuation methods applied

to limestone deposits can be found in Lewis and Moran

(1990).

Coarse Aqqregate Sources

The following sections present a summary of particular

geologic and economic factors important to the production of

aggregates from major source areas in Florida. The

information reported was compiled from several literature

sources including Anonymous (1990), Spencer (1990), Yon et

al. (1989), Campbell (1986), Miller (1986), Schmidt et al.,

(1979), Edgarton (1974), Yon and Hendry (1972), White

(1970), and Puri and Vernon (1964), Vernon (1951), and Cooke

(1945).

Taylor County Region

The mining region of Taylor County is located in the

Gulf Coastal Lowlands physiographic province with elevations

retail, or industrial development. These are common

problems in many urban areas nationwide. Zoning to protect

potential mining areas and sequential land-use planning are

steps being taken by some cities to mitigate this problem

(Harben and Bates, 1990; Langer, 1988).

Exploration and development of mining deposits also are

subject to many market and financial parameters. Many site-

specific issues such as quantity and quality of reserves,

proximity to markets, land values, and royalty requirements

determine the long-term cash flow feasibility of these large

capital investments. Examples of valuation methods applied

to limestone deposits can be found in Lewis and Moran

(1990).

Coarse Aqqregate Sources

The following sections present a summary of particular

geologic and economic factors important to the production of

aggregates from major source areas in Florida. The

information reported was compiled from several literature

sources including Anonymous (1990), Spencer (1990), Yon et

al. (1989), Campbell (1986), Miller (1986), Schmidt et al.,

(1979), Edgarton (1974), Yon and Hendry (1972), White

(1970), and Puri and Vernon (1964), Vernon (1951), and Cooke

(1945).

Taylor County Region

The mining region of Taylor County is located in the

Gulf Coastal Lowlands physiographic province with elevations

ranging from 10 to 20 feet above mean sea level (MSL).

Aggregate materials have been produced in this region from

dolomitic limestones of the Oligocene Suwannee Limestone

(Figure 1) since the early 1970s. The Suwannee Limestone in

this area is highly variable with regard to degree of

induration (i.e., hardness). Therefore, the material must

often be selectively mined before crushing and sizing

according to its intended use and associated specifications.

Mining operations continue to a depth of approximately 45

feet, with draglines utilized for mining activities below

the water table. The primary market for these aggregate

materials is in the Tallahassee-Leon County area with truck

operations being the primary method of transportation.

Hernando County Region

Mining activity in this region is commonly along the

north-south trending Brooksville Ridge, a sub-unit of the

Central Highlands geomorphic province, with elevations

ranging from 75 to greater than 250 feet above MSL.

Aggregate materials in this region also are produced from

the Oligocene Suwannee Limestone. This is a high calcium

limestone deposit that has been mined since the 1920s. The

limestone deposits in this area may be variable in degree of

induration and interbedded with chert, sands, silts and

clays. The deposit has an irregular karstic surface, but

may be in excess of 100 feet in thickness in some areas. It

is overlain by younger, siliciclastic sediments. Mining

activity is concentrated along topographic highs and

regional elevations often permit dry mining techniques, but

local mine dewatering may be required. Mined depths range

from approximately 50 to 100 feet below land surface.

Overburden materials, which must be removed initially, may

reach up to 100 feet in thickness. The high cost of

removing large amounts of overburden currently prohibits

economic feasibility of mining in some areas that may have

good aggregate materials. Blasting methods are utilized to

fracture the limestone which is then sent to crushers for

processing and sizing. Primary markets for these aggregate

materials include the Tampa, Orlando, and Gainesville-Ocala

regions along with numerous other areas of central Florida.

Truck operations are the primary method of transportation.

Lee and Collier Counties Region

Mining activity in this region occurs along the

Southwestern Slope geomorphic subdivision of the Southern

Peninsular zone. Elevations along the Southwestern Slope

range from 0 to 25 feet above MSL. Limestone aggregate

materials have been mined from the Plio-Miocene Tamiami and

Pleistocene Ft. Thompson Formations (Figure 1) in this

region since the mid-1950s. Overburden materials of sand,

clay and shell reach thicknesses of 15 feet. Wet mining

techniques are utilized after overburden removal due to

shallow water table conditions. The limestone deposits in

this region are variable in composition and often are

interbedded with sands. Therefore, selective mining

techniques are sometimes used to extract limestone materials

River (Citrus County) regions in the western part of

northern peninsular Florida (Figure 1). These materials are

variable in induration and physical properties and generally

produce low yields. Relative deep deposits are extracted

using wet mining techniques. These materials are primarily

used in local north Florida markets. However, the Crystal

River materials are used in other markets because of sea

transport availability.

Fine Aggregate Sources

Fine aggregate materials in north and central Florida

normally consist of quartz sands mined from Pleistocene

terrace deposits. These materials may be produced using

either wet or dry mining techniques depending on local water

table conditions. Suction dredges are often employed for

wet mining the deposits. The sands are usually wet screened

for size grading purposes.

South Florida concrete producers commonly use

manufactured limestone sands as fine aggregate materials.

These are often produced as processing by-products from

coarse limestone aggregate operations discussed previously.

possessing the desired properties. Thickness of mineable

deposits is typically 50 feet. Extensive development in the

Ft. Myers and Naples areas provides an excellent local

market and other truck transport markets range as far north

as Tampa. In addition, there is rail service to the Orlando

market.

Dade and Broward Counties Region

This region lies within the Everglades Trough

geomorphic subdivision of the Southern peninsular zone with

elevations averaging about 10 feet above MSL. Aggregate

materials in this region are produced from the Pleistocene

Miami Oolite and the underlying Ft. Thompson Formation

(Figure 1). These deposits have been extensively mined in

this area since the early 1900s. Mining operations in this

area typically involve removal of 10 feet or less of marl

and peat overburden. Wet mining techniques are normally

used due to very shallow water table conditions in this

area. The tremendous urbanization of Florida's southeastern

coastal areas provides an excellent local market for these

materials. Products also are shipped by rail to other north

Florida markets, such as Orlando, Tampa, and Jacksonville.

Levy and Citrus Counties Region

This area of Florida has recently been opened to mining

operations for coarse aggregate production. These materials

are dolomitic limestones mined from the Eocene Avon Park

Formation in the Gulf Hammock (Levy County) and Crystal

possessing the desired properties. Thickness of mineable

deposits is typically 50 feet. Extensive development in the

Ft. Myers and Naples areas provides an excellent local

market and other truck transport markets range as far north

as Tampa. In addition, there is rail service to the Orlando

market.

Dade and Broward Counties Region

This region lies within the Everglades Trough

geomorphic subdivision of the Southern peninsular zone with

elevations averaging about 10 feet above MSL. Aggregate

materials in this region are produced from the Pleistocene

Miami Oolite and the underlying Ft. Thompson Formation

(Figure 1). These deposits have been extensively mined in

this area since the early 1900s. Mining operations in this

area typically involve removal of 10 feet or less of marl

and peat overburden. Wet mining techniques are normally

used due to very shallow water table conditions in this

area. The tremendous urbanization of Florida's southeastern

coastal areas provides an excellent local market for these

materials. Products also are shipped by rail to other north

Florida markets, such as Orlando, Tampa, and Jacksonville.

Levy and Citrus Counties Region

This area of Florida has recently been opened to mining

operations for coarse aggregate production. These materials

are dolomitic limestones mined from the Eocene Avon Park

Formation in the Gulf Hammock (Levy County) and Crystal

AGGREGATE PROPERTIES

Studies were conducted to determine physical and

mineralogical properties of limestone aggregates from six

Florida sources, limestone aggregates from an Alabama

source, siliceous river gravel aggregates from a Florida

source, and granite aggregates from a Georgia source.

Physical properties are critical in proper concrete mix

design and preparation and also influence both fresh and

hardened concrete properties (Mindess and Young, 1981).

Petrographic studies were conducted to determine

mineralogical and textural properties that often control

physical and chemical properties of aggregates and their

related performance in concrete.

Computerized image analysis techniques also were

applied to characterize aggregate shapes from the various

mining sources. Aggregate shape may be an important factor

influencing economy of mix design and properties of fresh

and hardened concrete.

Physical Properties

The limestone, granite, and siliceous river gravel

coarse aggregates and quartz fine aggregates were sampled

from stockpiles at the FDOT State Materials Office in

accordance with ASTM D 75, Standard Practice for Sampling

AGGREGATE PROPERTIES

Studies were conducted to determine physical and

mineralogical properties of limestone aggregates from six

Florida sources, limestone aggregates from an Alabama

source, siliceous river gravel aggregates from a Florida

source, and granite aggregates from a Georgia source.

Physical properties are critical in proper concrete mix

design and preparation and also influence both fresh and

hardened concrete properties (Mindess and Young, 1981).

Petrographic studies were conducted to determine

mineralogical and textural properties that often control

physical and chemical properties of aggregates and their

related performance in concrete.

Computerized image analysis techniques also were

applied to characterize aggregate shapes from the various

mining sources. Aggregate shape may be an important factor

influencing economy of mix design and properties of fresh

and hardened concrete.

Physical Properties

The limestone, granite, and siliceous river gravel

coarse aggregates and quartz fine aggregates were sampled

from stockpiles at the FDOT State Materials Office in

accordance with ASTM D 75, Standard Practice for Sampling

Aggregates (Annual Book of ASTM Standards, 1988). Sample

quantities obtained for each aggregate source consisted of

one bag containing approximately fifty pounds. Sample

identifications along with corresponding source locations

and mined geologic units are given in Table 1.

The representative samples collected were analyzed in

accordance with ASTM C 136, Standard Method for Sieve

Analysis of Fine and Coarse Aggregate (Annual Book of ASTM

Standards, 1988). Results of the gradation analyses are

summarized in Table 2 along with recommended FDOT

specifications for No. 67 aggregate, which was the specified

size for the comprehensive concrete study. Gradational

properties for several of the samples analyzed failed to

meet FDOT specifications. Additional shipments of

aggregates were obtained for two source area materials (08-

005 and GA-177).

Gradational analyses for two shipments of fine

aggregates are given in Table 3 along with recommended FDOT

specifications. Fineness modulus values for the two samples

were slightly different. However, both values fall within

acceptable limits and the slight difference should not be

significant for concrete testing purposes.

Other physical properties of the aggregates that are

important for concrete mix design, preparation, and

properties were determined at the FDOT laboratories.

Properties determined included specific gravity, unit

weight, and water absorption for each aggregate sample

Table 1. Aggregate sample identifications.

FDOT Pit
Identification No.

08-005

87-145

87-090

12-008

86-062

34-106

AL-149

GA-177

50-120

71-132

Location

Brooksville, FL
(Hernando County)

Miami, FL
(Dade County)

Miami, FL
(Dade County)

Ft. Myers, FL
(Lee County)

Pembroke Pines, FL
(Broward County)

Gulf Hammock, FL
(Levy County)

Calera, AL
(Shelby County)

Tyrone, GA
(Fayette County)

Chattahoochie, FL
(Gadsden County)

Keystone Heights, FL
(Putnam County)

Geologic Unit

Suwannee Ls.
(Oligocene)

Miami Oolite
and Ft.
Thompson Fm.
(Pleistocene)

Miami Oolite
Fm.
(Pleistocene)

Tamiami
(Plio-Miocene)
and Ft.
Thompson
(Pleistocene)
Fms.

Miami Oolite
and Ft.
Thompson Fm.
(Pleistocene)

Avon Park Fm.
(Eocene)

Newala Ls.
(Ordovician)

Precambrian
granite

Pleistocene
gravels

Pleistocene
sands

Table 2. Coarse aggregate gradations (% passing).

Sieve FDOT
Size Specificationsa

1 in. 100

3/4 in. 90-100

3/8 in. 20-55

No. 4 0-10

No. 8 0-5

Sample Identification

08-005A 08-005B 87-145 87-090 12-008 86-062

100 100 100 100 100 100

86 97 84b 97 98 95

36 38 19b 52 60b 38

4 4 3 4 6 7

3 2 2 2 3 4
u1

aStandard Specifications for Road and Bridge Construction (1986)

bDoes not conform to FDOT specifications

Table 2.--continued.

Sieve FDOT
Size Specificationsa

1 in. 100

3/4 in. 90-100

3/8 in. 20-55

No. 4 0-10

No. 8 0-5

34-106

100

98

43

3

1

Sample Identification

AL-149 GA-177A GA-177B

100 100 100

96 93 91

27 54 12b

1 16b 1

1 9b 1

50-120

100

99

47

7

3
0^

aStandard Specifications for Road and Bridge Construction (1986)

bDoes not conform to FDOT specifications

Table 3. Fine aggregate gradations (% passing).

Sieve Size Spec

No. 4

No. 8

No. 16

No. 30

No. 50

No. 100

No. 200

Fineness modulusb

FDOT
ifications8

95-100

85-100

65-97

25-70

5-35

0-7

0-4

Sample Identification

71-132A 71-132B

100 100

99 98

88

56

18

2

0

2.38

85

50

17

2

0

2.48

aStandard Specifications for Road and Bridge Construction
(1986)

bASTM C 136, Method for Sieve Analysis of Fine and Coarse
Aggregate (Annual Book of ASTM Standards, 1988)

(Table 4). Los Angeles abrasion values (Table 4) meet the

FDOT requirement of 45 percent maximum loss. Results of

these tests reflect typical values reported for Florida

aggregate properties compiled from FDOT records and Ho and

Hendricks (1991) (Table 5), and other aggregate studies

(Anonymous, 1990; Spencer, 1990). Gradational,

mineralogical, and physical properties of aggregates may

vary over time as source locations, mining techniques, and

processing operations change. Therefore, quality control

and quality assurance programs are important for monitoring

mining activities to insure procurement of acceptable

materials.

Strengths of individual aggregate particles were

determined by loading each between two horizontal plates in

a hydraulic press and measuring the force required for

fracturing to occur. One hundred particles (+3/8 in.) from

each aggregate source were tested. Results of the strength

testing indicate that mean values (Table 6) for the Florida

limestone aggregates are generally in the same strength

range (200-300 lbs.). Mean values for the granite and

siliceous river gravel aggregates were slightly higher

(Table 6). Highest strengths were produced by the Alabama

limestone aggregate (Table 6). Distribution of strength

values (Table 6 and Figures 4 through 12) indicate that a

wide range of strength values occur for each aggregate

sample, with some samples possessing bimodal distributions.

Table 4. Physical properties of aggregate samples.

Sample No.

08-005

87-145

87-090

12-008

86-062

34-106

AL-149

GA-177

50-120

71-132

Specific
Gravity
(SSD)

2.45

2.39

2.42

2.43

2.54

2.48

2.75

2.67

2.63

Unit Wt.b
(lbs/ft.3)

92.12

86.00

86.04

89.00

90.08

92.72

105.80

93.62

104.80

2.63

Los Angeles
Abrasionc
% Absorptiona (% loss)

4.86

5.00

4.15

3.60

2.27

4.19

0.66

0.41

0.75

0.50

aASTM C 127, Standard Test Method for Specific Gravity and
Absorption of Coarse Aggregates (Annual Book of ASTM
Standards, 1988)

bASTM C 29, Standard Test Method for Unit Weight and Voids
in Aggregate (Annual Book of ASTM Standards, 1988)

CASTM C 131, Standard Test Method for Resistance to
Degradation of Small-Size Coarse Aggregate by Abrasion and
Impact in the Los Angeles Machine (Annual Book of ASTM
Standards, 1988)

Table 5. Summary of physical properties of aggregates compiled from historical FDOT data.

Aggregate ID

08-005
(Brooksville)

87-145
(Miami)

87-090
(Miami)

12-008
(Ft. Myers)

86-062
(Pembroke Pines)

34-106
(Gulf Hammock)

AL-149
(Calera)

GA-177
(Tyrone)

50-120
(Chattahoochie)

Specific
Gravity (SSD)
Mean Range

2.43 2.38-2.47

2.41 2.34-2.45

2.41 2.37-2.43

2.45 2.41-2.49

2.51 2.49-2.54

2.44 2.37-2.50

2.76 2.73-2.80

2.69 2.66-2.71

2.62 2.60-2.64

Absorption
(%)
Mean Range

3.87 2.28-5.80

4.48 3.38-7.16

4.25 2.75-6.04

3.62 3.05-4.50

3.42 0.85-5.50

5.22 3.56-7.68

0.53 0.44-0.56

1.52 0.76-2.28

1.05 0.60-1.49

Los Angeles
Abrasion
(% loss)
Mean Range

37 30-52

32 20-41

33 21-42

34 27-41

32 29-39

38 4-52

24 20-27

49 44-55

41 19-46

31

Table 6. Statistical distribution of coarse aggregate
strength data (values given in pounds).

Aggregate ID

08-005
(Brooksville)

87-145
(Miami)

87-090
(Miami)

12-008
(Ft. Myers)

86-062
(Pembroke Pines)

34-106
(Gulf Hammock)

AL-149
(Calera)

GA-177
(Tyrone)

50-120
(Chattahoochie)

Minimum

50

90

50

90

60

70

250

110

140

Maximum

740

780

710

710

790

870

>1000

>1000

900

Mean

214

295

285

286

237

280

703

367

454

Standard
Deviation

123

148

142

126

135

158

228

169

174

22.5--
20

15

0
10

5

0

0 100 200 300 400 500 600 700 800 900 1000
POUNDS

Figure 4. Distribution of strength values for aggregate
sample no. 08-005 (Brooksville).

8.- -~~

100 200 300 400 500 600

POUNDS

EI-----

700 800

900 1000

Figure 5.

Distribution of strength values for aggregate
sample no. 87-145 (Miami).

T

m

20-
18
16
14
12
z
o 10
8
6
4

0 100 200 300 400 500 600 700 800 900 1000
POUNDS

Figure 6. Distribution of strength values for aggregate
sample no. 87-090 (Miami).

22.5--

20

15
z
0
o 10

5

0 100 200 300 400 500 600 700 800 900 1000
POUNDS

Figure 7. Distribution of strength values for aggregate
sample no. 12-008 (Ft. Myers).

20

1 15

0
10

5

01

100 200 300 400 500 600 700 800 900 1000

POUNDS

Figure 8. Distribution of strength values for aggregate
sample no. 86-062 (Pembroke Pines).

J lI

100 200 300 400 500 600
POUNDS

Figure 9.

700 800 900 1000

Distribution of strength values for aggregate
sample no. 34-106 (Gulf Hammock).

E-'-

Ir-
0

30

25

20

o 15

10

5

0 100 200 300 400 500 600 700 800 900 1000
POUNDS

Figure 10. Distribution of strength values for aggregate
sample no. AL-149 (Calera).

20--
18
16
14
I 12
z
O 10

6
4
2

0 100 200 300 400 500 600 700 800 900 1000
POUNDS

Figure 11. Distribution of strength values for aggregate
sample no. GA-177 (Tyrone).

18.

16

14
12

z 10
0

6

4

0 100 200 300 400 500 600 700 800 900 1000
POUNDS

Figure 12. Distribution of strength values for aggregate
sample no. 50-120 (Chattahoochie).

Mineralogical Properties

Petrographic and X-ray diffraction analyses were

used to determine mineralogic compositions and textures of

the coarse aggregate samples. Petrographic analyses were

performed by hand specimen descriptions and use of polarized

light microscopy on thin sections prepared from

representative samples of each aggregate source. X-ray

diffraction (XRD) analyses were conducted on the limestone

coarse aggregates for qualitative and quantitative

mineralogical determinations. Two size fractions of the

coarse aggregates were ground to minus No. 200 mesh (74 Am)

size fraction for XRD analyses.

The matrix-flushing method (Chung, 1974), using a Linde

C alpha alumina corundumm) standard, was utilized for

quantitative mineralogical determinations. Reference

intensity ratios were initially determined for the major

minerals (calcite, dolomite, and quartz) occurring in the

limestone aggregates using a 50/50 mixture of each mineral

phase and corundum. These values were used in calculation

of quantitative amounts of these minerals in the aggregate

samples according to the following equation:

P = (Xs/Kx)(Ix/Is)

where,
P = weight fraction of phase x in sample
Xs = weight fraction of corundum in sample
Kx = reference intensity ratio of phase x
Ix = measured intensity of strongest line of
phase x
Is = measured intensity of strongest line of
corundum

A summary of the petrographic and X-ray diffraction analyses

for the aggregate samples, along with corresponding

representative photomicrographs are presented below. Refer

to Table 1 for source locations and geologic units

corresponding to the sample numbers provided.

Sample No. 08-005 (Brooksville. Florida)

The aggregate sample consisted of crushed limestone

particles which were generally off-white to light tan in

color (Table 7). The particles were hard, firm and dense,

with angular to subrounded shapes and irregular surfaces

containing some crusher fines. A few moderately soft

aggregate particles were detected in hand samples. These

particles, and similar ones occurring in other samples

described below, probably represent particles occurring at

the lower end (<100 lbs.) of the strength value

distributions (Figures 4 through 12) previously discussed.

A few chert particles, which were very hard and dense with a

tan to dark brown color, also were noted .

Microscopic analyses revealed sparse to abundant fossil

fragments in a microspar to sparry calcite matrix. Pore

space was fairly sparse and isolated with a vuggy

appearance. Minor amounts of fine, subangular quartz grains

were seen within the aggregate. A typical photomicrograph

(Figure 13) shows sparse fossil fragments (F) in a microspar

to sparry calcite matrix (M), with a few quartz particles

(Q) and small, isolated pore spaces (P). XRD analysis

Table 7. Munsell color designations for coarse aggregate
samples.

Sample No.

08-005

87-145

87-090

12-008

86-062

34-106

AL-149

GA-177

50-120

Munsella Designations

2.5Y 8.25/2, 10YR 8/4, 10YR 5/2

10YR 7.5/2

10YR 7.5/1, 10YR 6/2, 5Y 8.75/1

10YR 7/1, 5YR 8/2, 10YR 8/1.5

5Y 8.5/1, 5YR 7.5/2, N6.25/

10YR 5/4, 10YR 6.5/4, 10YR 7.5/2,
N 6.25/

N 6/, N 8/

N 9/, 5 G 4/1

N 5.5/, 2.5y 8/2, 5Y 8.5/1, 5YR 7/4

aMunsell Book of Color (1976)

.. '" :- *"-'** ... .. t ",. 7 o'
S, "" "i.. -
..' ., .N-
.tr. '.P . I "

-"1*$t.M 2 A'<* A . ... !f ,.- "* ,'
a--

i i, 1 *. ^ '.1 .* t,, -< -1 .

r , ; *'i "i ; c .l.te I : "
de'4

I '*0 "^'~''C ''. p- *. .
,'k ; r. .. ;T lC ..."

P .oti -of a. a sape.0.

(Brooks.ville) Ag g ... c o of s ..,

.' w t W 1 A- .' ,.4;*t4

+"<_ ,jr ... :. ^ ^..
V ..

& ,."..' ,, '., _. '-- '"
-t ... .. ...
., 4'., 7A.* 4. t5 m-,

S --& ..... -

Figure 14. Photomicrograph of aggregate sample no. 87-145
(Miami). Aggregate is composed of sparry calcite filled
fossil fragments (F) and abundant quartz grains (Q) in a
microspar matrix. Crossed nicols.
(Mai Agregate r s cp dai

microspar mti Crossed .ic

Sample No. 87-090 (Miami. Florida)

The aggregate sample consisted of crushed limestone

particles that were generally off-white to light tan in

color (Table 7). The aggregates generally were hard and

firm, with occasional moderately soft and friable particles.

Particle shapes were angular to subrounded with some crusher

fines occurring on particle surfaces. Some fossil fragments

and fairly abundant porosity were noted in hand specimens.

Microscopically the particles contained sparse to

moderate amounts of fossil fragments in a microspar matrix.

Sparry calcite commonly occurred as fossil fillings. Pore

space was sparse to abundant and fairly isolated with sparry

calcite rims. Subrounded quartz grains were moderately

abundant and sometimes occurred as channel fillings.

Feldspar grains also were seen in minor amounts. A

photomicrograph (Figure 15) of the aggregate shows sparse

fossil fragments (F) and veins (V) filled with sparry

calcite occurring in a micritic matrix (M). Quartz grains

(Q) and small isolated pore spaces (P) also are evident.

XRD analysis (Table 8) showed the sample was composed

primarily of calcite, with a moderate quartz content.

Sample No. 12-008 (Ft. Myers, Florida)

The aggregate sample consisted of crushed limestone

particles which were light tan to gray in color (Table 7).

Particles shapes were angular to subrounded, with crusher

fines occurring on particle surfaces. The aggregate

particles generally were hard and dense, with some soft and

Sample No. 87-090 (Miami. Florida)

The aggregate sample consisted of crushed limestone

particles that were generally off-white to light tan in

color (Table 7). The aggregates generally were hard and

firm, with occasional moderately soft and friable particles.

Particle shapes were angular to subrounded with some crusher

fines occurring on particle surfaces. Some fossil fragments

and fairly abundant porosity were noted in hand specimens.

Microscopically the particles contained sparse to

moderate amounts of fossil fragments in a microspar matrix.

Sparry calcite commonly occurred as fossil fillings. Pore

space was sparse to abundant and fairly isolated with sparry

calcite rims. Subrounded quartz grains were moderately

abundant and sometimes occurred as channel fillings.

Feldspar grains also were seen in minor amounts. A

photomicrograph (Figure 15) of the aggregate shows sparse

fossil fragments (F) and veins (V) filled with sparry

calcite occurring in a micritic matrix (M). Quartz grains

(Q) and small isolated pore spaces (P) also are evident.

XRD analysis (Table 8) showed the sample was composed

primarily of calcite, with a moderate quartz content.

Sample No. 12-008 (Ft. Myers, Florida)

The aggregate sample consisted of crushed limestone

particles which were light tan to gray in color (Table 7).

Particles shapes were angular to subrounded, with crusher

fines occurring on particle surfaces. The aggregate

particles generally were hard and dense, with some soft and

48

D,.. ^i2..p, .^.,- :., i.. -. -^- ,p." ,. -,,,-. ,-- r C-. .

aif 4'1
SfeS^: 2?, -'^ $ -? i:`;,.,

(Q), in a micritic matrix (M). Crossed nicols.
4."
., v ,-,i ..,' i.- .., ,. .Q .
' .:% "*

I

Figure 15. Photomicrograph of aggregate sample no. 1287-090
(Ft. MyersMiami). Aggregate is composed of sparry calcite fossilled
fossil fragments (F) andlcite veins (V) along quartz grains (Q) in
(Q),a microsparitic matrix (M). Crossed nicols.
nr*" r, *i* .-

4**?;- ^. -m4 c..

Figure 16. Photomicrograph of aggregate sample no. 12-008
(Ft. Myers). Aggregate is composed of sparse fossil
fragments, sparry calcite veins (V) and quartz grains (Q) in
a microspar matrix (M). Crossed nicols.

friable pieces. Vuggy particles and fossil fragments were

abundant in hand specimens.

Microscopically the particles contained sparse to

moderate amounts of fossil fragments in a predominantly

micritic and microspar cement matrix. Sparry calcite was

common as fossil fillings and pore linings. Vuggy porosity

was fairly abundant, but usually isolated. Quartz particles

were abundant, with subrounded to subangular shapes.

Feldspar grains were noted in trace amounts. A

photomicrograph (Figure 16) shows sparse fossil fragments

(F) with sparry calcite fillings in a microspar (M) cement

matrix. Quartz grains (Q) are abundant and pore space (P)

is moderately abundant, but isolated. XRD analysis (Table

8) showed the sample consisted of calcite, with significant

amounts of quartz.

Sample No. 86-062 (Pembroke Pines, Florida)

The aggregate sample consisted of crushed limestone

particles which were off-white to tan in color (Table 7).

The aggregate particles were hard and dense, with a few soft

and friable pieces. Particle shapes were angular to

subrounded with crusher fines occurring on particle

surfaces. Fossil fragments and vuggy porosity were abundant

in hand specimens.

Microscopically the particles contained moderate to

abundant fossil fragments in a microspar to sparry calcite

matrix. Subangular to subrounded quartz grains were

abundant throughout the aggregate particles. Moldic and

vuggy porosity was fairly abundant, with sparry calcite

linings common. Feldspar grains were noted in trace

amounts. A photomicrograph (Figure 17) shows sparse fossil

fragments (F) with sparry calcite fillings in a micritic

matrix (M). Quartz grains (Q) are abundant and pore spaces

(P) are small and isolated. XRD analysis (Table 8) showed

the sample consisted primarily of calcite, along with

significant amounts of quartz.

Sample No. 34-106 (Gulf Hammock. Florida)

The aggregate sample consisted of crushed limestone

particles with distinct colors of off-white, light brown,

and gray (Table 7). The particles were angular to

subrounded, and generally hard and dense with some friable

pieces. A few particles appeared weathered and iron

stained. Fossil fragments were sparse, and porosity sparse

to moderate in hand specimens.

Microscopically the particles consisted of sparse to

moderately abundant fossil fragments in a microspar to

sparry calcite matrix. Channel fillings and veins of sparry

calcite were common. Quartz particles were sparse in the

aggregate. Porosity also was sparse and was very small and

isolated. A photomicrograph (Figure 18) shows a mixture of

microspar (M) with patches and veins of sparry calcite (S).

Note the absence of quartz grains and the very small,

isolated pore spaces (P). XRD analysis (Table 8) showed

that the aggregate consisted of roughly equal amounts of

calcite and dolomite. Pyrite has been known to occur in

Figure 17. Photomicrograph of aggregate sample no. 86-062
(Pembroke Pines). Aggregate is composed of sparse sparry
calcite filled fossil fragments (F) and quartz grains (Q) in
a micritic matrix (M). Crossed nicols.

<', -- T* ^*^.*1 -''"V If..

k N
4 A -)

*40.25mmr
.*: -^^-^..-'-.
'. .. -A

Figure 18. Photomicrograph of aggregate sample no. 34-106
(Gulf Hammock). Aggregate is composed of a mixture of
microspar (M) and sparry calcite (S). Crossed nicols.

(Table 8) confirmed that the aggregate is composed primarily

of calcite, with minor amounts of quartz.

Sample No. 87-145 (Miami. Florida)

The aggregate sample consisted of crushed limestone

particles which were off-white to light tan in color (Table

7). The aggregates generally were hard and firm with

occasional friable pieces. Particle shapes were angular to

subrounded and crusher fines were common on particle

surfaces. Fossil fragments and vuggy porosity were evident

in hand specimens.

Microscopically the particles contained sparse to

abundant fossil fragments, along with a few peloids in a

predominantly microspar matrix. Sparry calcite commonly

occurred on and within fossil fragments, and also as

infillings of pores and channels. Porosity was fairly

sparse and isolated. Subangular to subrounded quartz grains

were abundant in some aggregates. Trace amounts of feldspar

and opaque heavy minerals also were noted. A typical

photomicrograph (Figure 14) shows sparry calcite filled

fossil fragments (F) in a microspar matrix. Quartz grains

(Q) are abundant and pore spaces (P) are small and isolated.

XRD analysis (Table 8) confirmed petrographic observations,

showing that the aggregate is primarily calcite, but

contains a sizeable quartz component. Trace amounts of

feldspar and heavy minerals observed in microscopic analysis

did not occur in detectable amounts.

Table 8. Quantitative X-ray diffraction analyses of
limestone aggregates.

Calculated Percentages

Sample No. Size Fraction Calcite Dolomite Ouartz

08-005 +3/8 in. 98 2
-3/8 in. + No. 4 94 6

87-145 +3/8 in. 79 21
-3/8 in. + No. 4 87 13

87-090 +3/8 in. 91 9
-3/8 in. + No. 4 89 11

12-008 +3/8 in. 75 25
-3/8 in. + No. 4 83 17

86-062 +3/8 in. 82 18
-3/8 in. + No. 4 78 22

34-106 +3/8 in. 49 51 -
-3/8 in. + No. 4 57 43 -

AL-149 +3/8 in. 77 23 -
-3/8 in. + No. 4 57 41 2

50-120 +3/8 in. 100
-3/8 in. + No. 4 -- 100

generally were hard and dense, with occasional friable

schistose pieces.

Microscopically the particles contained anhedral

quartz, and subhedral to anhedral albite (sodium feldspar),

orthoclase and microcline (potassium feldspars) as the

primary minerals, and subhedral to euhedral biotite as a

minor constituent. Several accessory minerals were noted in

trace amounts including magnetite, zircon, apatite,

tourmaline, and garnet. The major mineral components

exhibited some undulatory extinction and fracturing, with

sericitic alteration occurring along feldspar fractures.

Some iron staining along fractures and alteration of biotite

to chlorite also was noted. A photomicrograph (Figure 21)

shows the coarsely crystalline and interlocking nature of

the mineral constituents of the aggregate. Quartz (Q),

biotite (B), and microcline (M), with its characteristic

twinning, can be seen in the photomicrograph. Some

fracturing also occurs in the aggregate (Figure 22).

Sample No. 50-120 (Chattahoochie. Florida)

The aggregate sample consisted of uncrushed siliceous

river gravel particles which were light brown to tan in

color (Table 7). The particles were hard and dense, with

subrounded to rounded shapes. Particle surfaces were

generally smooth textured with some dusty material. Some

particles contained visible fractures.

Microscopically the particles consisted of

polycrystalline quartz grains with some iron-stained

Figure 21. Photomicrograph of aggregate sample no. GA-177
(Tyrone). Aggregate is composed of coarsely crystalline
quartz (Q), alkali feldspar (M) and biotite (B). Crossed
nicols.

Figure 22. Photomicrograph of aggregate sample no. GA-177
(Tyrone). Aggregate is composed of coarsely crystalline
mineral grains with abundant fracturing. Crossed nicols.

these materials and is suspected to be the cause of the iron

staining. However, it was not detected in XRD analysis.

Sample No. AL-149 (Calera. Alabama)

The aggregate sample consisted of crushed limestone

particles which were light to medium gray in color (Table

7). Particle shapes were angular to subrounded with some

crusher fines occurring on particle surfaces. The

aggregates were hard and dense, with little visible

porosity.

Microscopically the aggregates ranged from those with

interlocking coarsely crystalline carbonate particles of

moderate interference color (Figure 19) to ones having a

matrix of fine grained particles (M), with abundant coarser,

sparry veins and infillings (S) (Figure 20). Very little

porosity was observed. Fossil fragments were absent, but

some recrystallized oolitic structures were evident in some

particles. XRD analysis (Table 8) showed that the particles

consisted of calcite and dolomite.

Sample No. GA-177 (Tyrone. Georgia)

The aggregate sample consisted primarily of crushed

granite particles, along with a few pieces of biotite gneiss

and schist. The particles were generally light to medium

gray, except for the very dark gneiss and schist particles

(Table 7), medium to coarse grained, and of granular

texture. Particle shapes were angular to subrounded, with

abundant crusher fines on particle surfaces. The aggregates

these materials and is suspected to be the cause of the iron

staining. However, it was not detected in XRD analysis.

Sample No. AL-149 (Calera. Alabama)

The aggregate sample consisted of crushed limestone

particles which were light to medium gray in color (Table

7). Particle shapes were angular to subrounded with some

crusher fines occurring on particle surfaces. The

aggregates were hard and dense, with little visible

porosity.

Microscopically the aggregates ranged from those with

interlocking coarsely crystalline carbonate particles of

moderate interference color (Figure 19) to ones having a

matrix of fine grained particles (M), with abundant coarser,

sparry veins and infillings (S) (Figure 20). Very little

porosity was observed. Fossil fragments were absent, but

some recrystallized oolitic structures were evident in some

particles. XRD analysis (Table 8) showed that the particles

consisted of calcite and dolomite.

Sample No. GA-177 (Tyrone. Georgia)

The aggregate sample consisted primarily of crushed

granite particles, along with a few pieces of biotite gneiss

and schist. The particles were generally light to medium

gray, except for the very dark gneiss and schist particles

(Table 7), medium to coarse grained, and of granular

texture. Particle shapes were angular to subrounded, with

abundant crusher fines on particle surfaces. The aggregates

bp~

2 0O.25mm

Figure 19. Photomicrograph of aggregate sample no. AL-149
(Calera). Aggregate is composed of coarsely crystalline
carbonate particles. Crossed nicols.

I' A- '.
-.' ^ s, ,. .. .-....
b .*,

JA^k>. j.r^^t^^s^
Figure 20. Photomicrograph of aggregate sample no. AL-149
(Calera). Aggregate is composed of coarse sparry calcite in
a fine-grained matrix (M). Crossed nicols.

fracturing (F) (Figure 23). Some aggregates consisted of

elongate polycrystalline quartz grains with highly sutured

contacts typical of metamorphic rocks (Blatt, 1982) (Figure

24). Undulatory extinction of quartz grains was common.

Quartz was the only mineral detected by XRD analysis of the

aggregates (Table 8).

Shape Characterization

Aggregate shape can be an important factor in concrete

mix design, properties and performance. In the past,

characterization of aggregate shapes has been limited to

simple visual descriptions or to indirect correlation with

various physical tests on bulk aggregate samples (Meier and

Elnicky, 1989). In this study, computerized image analysis

techniques were utilized for direct measurement of

individual aggregate particle shape and geometry. The

aggregates were cast in epoxy resins and, after hardening of

the epoxy, sawed or polished to their medial planes to give

a maximum projection area. One hundred particles of two

size fractions for each coarse aggregate sample were

analyzed using computerized imaging procedures for areas,

perimeters, Feret diameters, and shape functions.

Area to perimeter ratios were calculated for each

particle to provide an indicator of surface irregularity.

As particles become more irregular, perimeter values become

longer for a given area. Therefore, the area to perimeter

ratio becomes smaller as particle surface irregularity

becomes greater. Higher surface irregularity of concrete

Figure 23. Photomicrograph of aggregate sample no. 50-120
(Chattahoochie). Aggregate is composed of polycrystalline
quartz grains with iron stained fractures (F). Crossed
nicols.

Figure 24. Photomicrograph of aggregate sample no. 50-120
(Chattahoochie). Aggregate is composed of elongate
polycrystalline quartz grains with highly sutured contacts.
Crossed nicols.

aggregates should provide greater interlocking or mechanical

bond with cement paste. However, excessive irregularity of

particles may affect economy of mix designs by causing

higher cement requirements.

Shape functions (aspect ratios) were determined by

taking the minimum to maximum ratio of thirty-two Feret

diameters measured for each particle. The smaller the

aspect ratio, the more elongate the particle shape. In

general, high aspect ratios (more equant grains) are more

desirable for concrete aggregates.

Circularity shape factors are a measure of how closely

particle shape resembles the shape of a circle. Particles

that closely approximate a circle would be assigned a value

close to one with lower numbers assigned as the particle

shape deviates from a circle in elongation or surface

roughness. As in the case of area to perimeter ratios, more

irregular particles (low shape factor numbers) would

generally be more desirable for concrete aggregates due to

greater interlocking or mechanical bond with the cement

paste. Extremely low values may indicate very elongate

particles, which are not desirable for concrete.

Average values for area to perimeter ratios, aspect

ratios and circularity shape factors from the one hundred

particle measurements of two size fractions for each

aggregate sample were determined (Table 9). These data

indicate that no large differences exist between average

values of the parameters for the various aggregates

Table 9. Average values from image analysis dataa.

Aqqregate ID

08-005
(Brooksville)

87-145
(Miami)

87-090
(Miami)

12-008
(Ft. Myers)

86-062
(Pembroke Pines)

34-106
(Gulf Hammock)

GA-177
(Tyrone, GA)

Size Fraction

+3/8 in.
-3/8 in. + No

+3/8 in.
-3/8 in. + No

+3/8 in.
-3/8 in. + No

+3/8 in.
-3/8 in. + No

+3/8 in.
-3/8 in. + No

+3/8 in.
-3/8 in. + No

+3/8 in.
-3/8 in. + No

Area/
Perimeter

0.1051
0.0827

0.1205
0.0801

0.1095
0.0733

0.1008
0.0755

0.1155
0.0782

0.1172
0.0791

0.1170
0.0769

S4

S4

. 4

S4

. 4

. 4

Shape
Function

0.7413
0.7177

0.7491
0.6957

0.7569
0.6953

0.7194
0.7327

0.7360
0.7274

0.7302
0.7106

0.7379
0.7284

Circularity
Shape Factor

0.7030
0.7123

0.6060
0.6349

0.5958
0.6427

0.6563
0.6637

0.6497
0.6763

0.6541
0.6837

0.6032
0.6285

aData are based on measurements of at least 100 grains
to their medial plane.

of each aggregate sample, polished

measured. Comparison of values for the two size fractions

indicate that the smaller particles have more irregular

shapes and are less equant than the larger particles.

Although average values of the shape parameters are in close

agreement, the distribution of values for the aggregate

particles (Figures 25 through 38) differs between samples.

Standard deviation and variance values (Table 10) also

indicate slight variations in distribution of particle

values.

Because aggregate samples AL-149 (Calera) and 50-120

(Chattahoochie) were late additions to the project, these

materials were not analyzed for shape parameters. Sample

AL-149 would be expected to possess similar shape properties

as the other crushed materials. However, analysis of sample

50-120 would likely yield different shape parameter data

because it is an uncrushed aggregate, possessing smooth

particle textures.

0.06 0.1

0.14 0.18

AREA/PERIM

1 ,Mm-
0 0.125 0.375 0.625 0.875 1.125
SHAPE

0 0.125 0.375 0.625 0.875

Figure 25.

CIRCLE
Distributions of shape parameter values for
aggregate sample no. 08-005 (Brooksville),
plus 3/8 in. size fraction.

0 0.02

0 0.02

0.06 0.1
AREA/PERIM

ili
0 0.125 0.375 0.625
SHAPE

Figure 26.

0.14 0.18

0.875 1.125

0 0.125 0.375 0.625 0.875
CIRCLE
Distributions of shape parameter values for
aggregate sample no. 08-005 (Brooksville),
minus 3/8 in. plus No. 4 size fraction.

0 0.02 0.06 0.1 0.14 0.18

AREA/PERIM

0 0.125 0.375 0.625 0.875 1.125
SHAPE

Figure 27.

0 0.125 0.375 0.625 0.875
CIRCLE
Distributions of shape parameter values for
aggregate sample no. 87-145 (Miami),
plus 3/8 in. size fraction.

0 0.02

I

0.06 0.1 0.14 0.18

AREA/PERIM

0 0.125 0.375 0.625 0.875 1.125
SHAPE

0 0.125 0.375 0.625

0.875

Figure 28.

CIRCLE
Distributions of shape parameter values for
aggregate sample no. 87-145 (Miami),
minus 3/8 in. plus No. 4 size fraction.

0 0.02 0.06 0.1 0.14 0.18
AREA/PERIM

0 0.125 0.375 0.625 0.875 1.125

SHAPE

0 0.125 0.375

Figure 29.

0.625 0.875

CIRCLE
Distributions of shape parameter values for
aggregate sample no. 87-090 (Miami),
plus 3/8 in. size fraction.

0.14 0.18

AREA/PERIM

0 0.125 0.375 0.625
SHAPE

0.875 1.125

0 0.125 0.375 0.625 0.875
CIRCLE
Figure 30.. Distributions of shape parameter values for
aggregate sample no. 87-090 (Miami),
minus 3/8 in. plus No. 4 size fraction.

0 0.02

0.06 0.1

0.06 0.1

AREA/PERIM

0.14 0.18

*1~

0 0.125 0.375 0.625 0.875 1.125

SHAPE

0

Figure 31.

0 0.125 0.375 0.625 0.875
CIRCLE
Distributions of shape parameter values for
aggregate sample no. 12-008 (Ft. Myers),
plus 3/8 in. size fraction.

0 0.02

0 0.02 0.06 0.1
AREA/PERIM

0.14 0.18

0 0.125 0.375 0.625 0.875 1.125
SHAPE

Figure 32.

0 0.125 0.375 0.625 0.875
CIRCLE
Distributions of shape parameter values for
aggregate sample no. 12-008 (Ft. Myers),
minus 3/8 in. plus No. 4 size fraction.

0 0.02 0.06 0.1
AREA/PERIM

0.14 0.18

*1

0 0.125 0.375 0.625 0.875 1.125
SHAPE

Figure 33.

0 0.125 0.375 0.625 0.875
CIRCLE
Distributions of shape parameter values for
aggregate sample no. 86-062 (Pembroke
Pines), plus 3/8 in. size fraction.

0 0.02 0.06 0.1 0.14 0.18
AREA/PERIM

100

80

60

40

20

0 0.125 0.375 0.625 0.875 1.125
SHAPE

S 0.125
0 0.125 0.375 0.625 0.875

CIRCLE
Figure 34. Distributions of shape parameter values for
aggregate sample no. 86-062 (Pembroke Pines),
minus 3/8 in. plus No. 4 size fraction.

-

-

-

0 0.02 0.06 0.1
AREA/PERIM

0.14 0.18

0 0.125 0.375 0.625 0.875 1.125
SHAPE

Figure 35.

0 0.125 0.375 0.625 0.875
CIRCLE
Distributions of shape parameter values for
aggregate sample no. 34-106 (Gulf Hammock),
plus 3/8 in. size fraction.

0.06 0.1 0.14 0.18

AREA/PERIM

*1

0 0.125

0.375 0.625
SHAPE

0.875 1.125

Figure 36.

0 0.125 0.375 0.625 0.875
CIRCLE
Distributions of shape parameter values for
aggregate sample no. 34-106 (Gulf Hammock),
minus 3/8 in. plus No. 4 size fraction.

0 0.02

0 0.02 0.06 0.1 0.14 0.18
AREA/PERIM

0 0.125 0.375 0.625 0.875 1.125
SHAPE

0 0.125 0.375

0.625 0.875

CIRCLE
Figure 37. Distributions of shape parameter values for
aggregate sample no. GA-177 (Tyrone),
plus 3/8 in. size fraction.

50 "

0 I
0 0.02 0.06 0.1
AREA/PERIM

50

0.14 0.18

S0. 5 .
0 0.125 0.375 0.625 0.875 1.125

SHAPE

Figure 38..

0 0.125 0.375 0.625 0.875
CIRCLE
Distributions of shape parameter values for
aggregate sample no. GA-177 (Tyrone),
minus 3/8 in. plus No. 4 size fraction.

Table 10. Statistical distribution of shape parameter data.

Aggregate ID

08-005
(Brooksville)

87-145
(Miami)

87-090
(Miami)

12-008
(Ft. Myers)

Fraction

+3/8
-3/8

+3/8
-3/8

+3/8
-3/8

+3/8
-3/8

86-062 +3/8
(Pembroke Pines) -3/8

34-106
(Gulf Hammock)

GA-177
(Tyrone, GA)

+3/8
-3/8

+3/8
-3/8

Area/Perimeter

Std.
Deviation Variance

+ No. 4

+ No. 4

+ No. 4

+ No. 4

+ No. 4

+ No. 4

+ No. 4

0.0218
0.0174

0.0151
0.0154

0.0190
0.0159

0.0205
0.0135

0.0191
0.0162

0.0189
0.0145

0.0222
0.0153

0.0005
0.0003

0.0002
0.0002

0.0004
0.0003

0.0004
0.0002

0.0004
0.0003

0.0004
0.0002

0.0005
0.0002

ShaDe Function

Std.
Deviation Variance

0.1361
0.1471

0.1276
0.1387

0.1207
0.1402

0.1446
0.1332

0.1164
0.1399

0.1245
0.1440

0.1388
0.1459

0.0185
0.0216

0.0163
0.1092

0.0146
0.0196

0.0209
0.0177

0.0136
0.0196

0.0155
0.0207

0.0193
0.0213

Circularity

Std.
Deviation

0.0758
0.0714

Variance

0.0057
0.0051

0.2066 0.0427
0.0988 0.0098

0.1104
0.1017

0.0122
0.0103

0.0830 0.0069
0.0756 0.0057

0.0865 0.0075
0.0834 0.0070

0.0916 0.0084
0.0829 0.0069

0.1003
0.0837

0.0101
0.0080

CEMENT PASTE-AGGREGATE INTERFACIAL BONDING

Portland cement concrete is a composite material

consisting of a mixture of portland cement, aggregates, air,

and water. Various admixtures also may be used for

achieving specific properties such as increased workability,

entrained air content, and set retardation. Hydration

reactions between water and portland cement result in

formation of cementitious compounds which bind the system

together to form hardened concrete. The most abundant

compounds formed from hydration reactions are calcium

silicate hydrates and calcium hydroxide. Minor amounts of

other compounds are formed, including calcium aluminate

hydrates and calcium sulfoaluminates (Mindess and Young,

1981).

Composite strength of portland cement concrete may be

divided into three general components;

1 strength of the cement paste

2 strength of the aggregates

3 strength of the interface between the cement paste

and aggregates.

The water to cement ratio is one of the most important

properties affecting the strength of the cement paste. A

ratio of approximately 0.2 is required to hydrate the

cement. However, this will not produce a workable mix and

therefore, larger amounts of water are required. This

produces greater amounts of voids in the paste which reduces

its strength. Hydration of the cement paste is a gradual

process with most of the strength developed after 28 days.

Strength of the aggregates is dependent on

mineralogical composition, grain size, texture, and

alteration. These factors are determined by their geologic

mode of formation and subsequent geologic history.

The cement paste-aggregate interfacial bond is usually

considered to be the weakest component of composite concrete

strength (Mindess and Young, 1981) and has been the subject

of an increasing number of investigations in recent years.

The strength of the cement paste-aggregate interfacial bond

can be divided into mechanical and chemical components.

Mechanical bonds result from physical interlocking of cement

paste and aggregates as provided by physical properties of

the aggregates such as size, shape, and surface texture.

Chemical bonds result from chemical interactions between the

hydrating cement compounds and aggregate surfaces. These

interactions may be highly dependent upon the mineralogical

composition of the aggregate which controls its surface

properties. Surface properties of aggregates such as

chemical composition, atomic arrangement, and associated

surface charges are likely to determine cement paste-

aggregate interactions.

It is generally accepted that differences exist in the

bonding of cement paste with different aggregate types.

Carbonate aggregates are believed to react chemically with

cement hydration products, forming a good interfacial bond.

In contrast, silicate aggregates do not seem to bond well

with cement paste, as evidenced by more abundant separations

occurring at the cement paste-silicate aggregate boundary.

These interfacial separations may provide pathways for fluid

migration and therefore, may affect permeability and

durability of concrete (Mindess and Young, 1981).

Microtextural studies were conducted on concrete

specimens containing limestone aggregates from major Florida

sources, limestone aggregates from an Alabama source,

granite aggregates from a Georgia source, and siliceous

river gravel aggregates from a Florida source, that have

been utilized in State concrete projects, to characterize

their interfacial bonding properties.

While the interfacial region has been studied

extensively with regard to detailed textural

characteristics, relatively little work has been done to

improve the cement paste-silicate aggregate bond. A review

of some previous work on cement paste-aggregate bonding is

presented, followed by considerations of mineral surface

properties thought to be important for observed bonding

characteristics and their modification to improve bonding

with the cement paste. Aggregate treatment methods for

improving the cement paste-silicate aggregate bond were

investigated by analytical laboratory techniques, concrete

testing, and microscopic examinations.

Literature Review

Textural Characteristics of the Interfacial Region

Several studies have been conducted in recent years to

characterize the textural details of the interfacial region

between cement paste and aggregate. Barnes et al. (1978a)

conducted SEM investigations of early hydration

characteristics for interfacial regions between glass slides

and cement pastes. They observed a thin (1 um) "duplex"

film of calcium hydroxide (Ca(OH)2) crystals adjacent to the

glass slide with elongated hydrated calcium silicate (C-S-

H) particles covering the film and projecting toward the

body of the paste. Beyond the C-S-H particle occurred a

zone of hydrating cement grains. With increasing hydration,

the C-S-H particles shortened or formed a reticulated

network pattern and the zone beyond filled with abundant

Ca(OH)2 crystals. Subsequent work by the same authors

(1978b) showed similar features occurring in the interfacial

regions of mortars prepared with quartz sand.

Textural studies of the interfacial regions between

cement paste and carbonate aggregates by Barnes et al.

(1978) proved to be more difficult due to separation through

the aggregate rather than at the interface. However, some

textural studies of the cement paste-carbonate aggregate

interface have been carried out. X-ray diffraction studies

by Carles-Giberques et al. (1982) concluded that the Ca(OH)2

crystals at the interface of carbonate aggregates had much

less parallel orientation to the aggregate surface than

Literature Review

Textural Characteristics of the Interfacial Region

Several studies have been conducted in recent years to

characterize the textural details of the interfacial region

between cement paste and aggregate. Barnes et al. (1978a)

conducted SEM investigations of early hydration

characteristics for interfacial regions between glass slides

and cement pastes. They observed a thin (1 um) "duplex"

film of calcium hydroxide (Ca(OH)2) crystals adjacent to the

glass slide with elongated hydrated calcium silicate (C-S-

H) particles covering the film and projecting toward the

body of the paste. Beyond the C-S-H particle occurred a

zone of hydrating cement grains. With increasing hydration,

the C-S-H particles shortened or formed a reticulated

network pattern and the zone beyond filled with abundant

Ca(OH)2 crystals. Subsequent work by the same authors

(1978b) showed similar features occurring in the interfacial

regions of mortars prepared with quartz sand.

Textural studies of the interfacial regions between

cement paste and carbonate aggregates by Barnes et al.

(1978) proved to be more difficult due to separation through

the aggregate rather than at the interface. However, some

textural studies of the cement paste-carbonate aggregate

interface have been carried out. X-ray diffraction studies

by Carles-Giberques et al. (1982) concluded that the Ca(OH)2

crystals at the interface of carbonate aggregates had much

less parallel orientation to the aggregate surface than

those at silicate aggregate interfaces. They attribute the

decreased orientation to etching of the carbonate aggregate

surface due to reactions with the hydrating cement

compounds. These reaction mechanisms will be discussed in

more detail later.

Bonding of Carbonate and Silicate Aggregates

It is widely reported in the literature that the

carbonate aggregate-cement paste system provides superior

bonding properties relative to the silicate aggregate-cement

paste system. Bond strength measurements between hydrated

cement paste and various aggregate types generally show that

carbonate aggregates produce higher bond strength than

silicate aggregates (Farran, 1956; Hsu and Slate, 1963).

This is consistent with reports that highest concrete

compressive strength is obtained with limestone aggregates

(Bedard et al., 1984; Carrasquillo, 1987).

Some studies using polished aggregates have indicated

that extrusive siliceous aggregates may develop high bond

strengths (Alexander et al., 1965). This is likely due to

the amorphous nature of some extrusive aggregate components,

producing a pozzolanic type reaction with the cement paste

because of silica dissolution in the highly alkaline

environment. It is doubtful that this mechanism would occur

for the more commonly used non-extrusive, siliceous

aggregates of a highly crystalline nature, such as quartz

gravels and crushed granites, because of much lower silica

solubility.

SEM observations of cement paste-aggregate interfaces

have supported bond strength studies. Cracks are often

observed to exist at cement paste-silicate aggregate

boundaries (Christensen et al., 1981; Shah and Slate, 1965;

Regourd, 1987), even before any loading (Shah and Slate,

1965). Regourd (1987) states that carbonate aggregates

exhibit good interfacial bonding characteristics in contrast

to silicate aggregates on which the cement hydration

products are weakly bound to and from which they are easily

torn.

Surface Chemical Considerations

The contrasting bonding characteristics of carbonate

and silicate aggregates probably are a result of differences

in surface chemical properties which are affected by

mineralogical composition and crystallographic structure

(Figures 39 and 40). The higher bond strengths provided by

carbonate aggregates likely are due to chemical reactions

with the cement paste. These reactions may occur due to

favorable atomic composition and arrangement of calcite

mineral surfaces which are composed of Ca2+ and CO3" charge

centers (Figure 41) (Sherwood, 1967; Thompson and Pownall,

1989).

The two commonly proposed models of reaction between

limestone aggregates and cement paste are:

1- bonding of calcium hydroxide formed at the

interface to the calcite surface, either by

Carbon

*Oxygen

Charge not neutralized
on each oxygen = -

Carbon e
Calcium 0
Oxygen 0

Figure 39. Mineralogical properties of calcite, CaCO3.
(a) Bond strength and residual charges for
carbonate anionic complex. (b) Structure
of calcite (after Klein and Hurlbut, 1985).

^ J 5 ^ Charge not neutralized
on each oxygen =-1

Figure 40. Mineralogical properties of quartz, SiO2.
(a) Bond strength and residual charges for
the tetrahedral SiO4 group. (b)
Tectosilicate structure of quartz (after
Klein and Hurlbut, 1985).

Charge Surface
Centers /

+ + + +

Ca CO3 Ca C3 -Ca CO3 Ca CO3
I I I I I I I I
CO3 Ca CO -- Ca CO3 Ca CO3- Ca

CO Ca CO Ca CO Ca O Ca-- CO
3 3 3 3

Figure 41. Charges on calcite surface (after Sherwood,
1967).

epitaxial overgrowth (Farran, 1956) or non-

homogenous nucleation (Yuan and Guo, 1987), and

2- formation of the carboaluminate

3CaO.A1203.CaCO3.11H20 (Carles-Gibergues et al.,

1982; Yuan and Guo, 1987).

Some studies have concluded that the formation of

carboaluminates is the most beneficial of these reactions

for providing a strong cement paste-aggregate bond (Yuan and

Guo, 1987; Yuan and Odler, 1987).

The interactions between carbonate mineral surfaces and

cement hydration products likely are related to the highly

reactive nature of the carbonate mineral surfaces. Complex

and dynamic interactions occur at the interface of carbonate

mineral surfaces and aqueous solutions (Davis and Kent,

1990; Morse 1987). Studies by Davis et al. (1987) and Xu et

al. (1990) have shown that some cations initially adsorb

onto the hydrated calcite surface, followed by absorption

into the mineral by exchange with calcium, resulting in

solid solution formation. However, the ease of such

reactions may be influenced by many factors including ionic

radius and hydration properties, and solution chemistry

(Davis and Kent, 1990).

The surface properties of common silicate minerals,

such as quartz and feldspar, are different than those of the

carbonate minerals. Quartz is composed of Si and O atoms

linked in tetrahedral coordination (Klein and Hurlbut,

1985), and yields a negatively charged fracture surface

(Figure 42) (Sherwood, 1967). Dissociation of hydrogen

atoms results in negative charges at the quartz surface and

positive charges in the adjacent water medium.

The structure of oxides such as quartz causes them to

exhibit a strong hydrophilic (water attractive) character

(Iler, 1979). The water on the mineral surface can be

separated into weakly (physisorbed) and strongly

chemisorbedd) held layers (Figure 43). Electrical

imbalances at the mineral surface are satisfied by

chemisorption of water to form surface hydroxyl groups.

Hydrogen bonding between these groups and additional

adsorption of water forms a physisorbed water layer (Davis

and Kent, 1990; Parks, 1990). Adsorption of ions at oxide

mineral surfaces may occur by either chemisorption or

physisorption processes. Chemisorption occurs by formation

of inner-sphere complexes and involves loss of hydration

water and direct, short-range electrostatic or covalent

bonding of ions to surface oxygens (Figure 44).

Physisorption occurs by formation of outer-sphere complexes,

with ionic complexes retaining waters of hydration and

either hydrogen bonded to the surface or attracted by long-

range coulombic forces (Figure 44). The difference in the

strength between these long-range and short-range forces

causes outer-sphere adsorption complexes to be less strongly

bound to the mineral surface than inner-sphere adsorption

complexes (Brown, 1990).

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
1439	sobekcm_page_globals.constructor	Navigation Object created from URI query string
1439	sobekcm_database.verify_item_lookup_object
1442	sobekcm_page_globals.display_item	Retrieving item or group information
1442	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
1442	sobekcm_assistant.get_entire_collection_hierarchy
1442	cached_data_manager.retrieve_item_aggregation
1442	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
1442	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
1442	system.web.ui.page.page_load (ufdc.page_load)
1442	sobekcm_page_globals.constructor.on_page_load
1442	html_echo_mainwriter.add_style_references	Adding style references to HTML
1442	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
1493	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file