Citation
Beneficial use of recycled materials in concrete mixtures

Material Information

Title:
Beneficial use of recycled materials in concrete mixtures
Creator:
Maier, Patrick Lance
Publication Date:
Language:
English
Physical Description:
xv, 195 leaves : illustrations ; 28 cm

Thesis/Dissertation Information

Degree:
Master's ( Master of Science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Structural Engineering, CU Denver
Degree Disciplines:
Structural engineering

Subjects

Subjects / Keywords:
Concrete -- Additives ( lcsh )
Pavements, Concrete -- Recycling ( lcsh )
Glass waste -- Recycling ( lcsh )
Concrete -- Additives ( fast )
Glass waste -- Recycling ( fast )
Pavements, Concrete -- Recycling ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 164-168).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Patrick Lance Maier.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
747715186 ( OCLC )
ocn747715186
Classification:
LD1193.E53 2011m M34 ( lcc )

Full Text
R
BENEFICIAL USE OF RECYCLED MATERIALS
rN CONCRETE MIX TURES
by
Patrick Lance Maier
B.S. Civil Engineering, University of Colorado Denver, 2008
A thesis submitted to the
University of Colorado Denver
In partial fulfillment
of the requirements for the degree of
Master of Science, Structural Engineering
Civil Engineering
2011


This thesis for the Master of Science
degree by
Patrick L. Maier
has been approved
by
Dr. Chengyu Li
Date


Maier, Patrick Lance (M.S. Structural, Civil Engineering Department)
Beneficial Use of Recycled Materials in Concrete Mixtures
Thesis directed by Dr. Stephan A. Durham
ABSTRACT
The need to produce concrete mixtures with recycled materials is becoming more
important than ever before. Not only does using recycled materials in concrete
mixtures create landfill avoidance, but it decreases the depletion of virgin raw-
materials. The basis for this research was to investigate the effects of using recycled
materials, in varying amounts, on the fresh and hardened concrete properties. This
research includes the design of concrete mixtures composed of varying amounts of
recycled material replacements. The recycled materials in this study consisted of
ground granulated blast furnace slag (GGBFS), recycled concrete (crushed hardened
concrete) and crushed waste glass. The GGBFS was used as a replacement for the
cement. The recycled concrete and waste glass were used to replace the coarse and
fine aggregates, respectively. The concrete mixtures designed ranged from a twenty
five percent replacement to one hundred percent replacement with recycled materials.
These mixtures were compared against a standard concrete mixture using cement and
virgin aggregates. For comparison purposes, all mixtures were held constant in
regards to water to cementitious ratio. The fresh concrete properties examined
included slump, air content and unit weight. The hardened properties examined
included compressive strength, rate of strength gain, freeze-thaw durability,
permeability, and alkali-silica reactivity potential.
A concrete mixture composed entirely of recycled materials was developed.
This concrete mixture developed substantial strength and durability and is comparable
to a normal strength concrete mixture in several aspects. This concrete made from


100% recycled materials was a very low permeable concrete with a compressive
strength of 4300 psi (29.6 MPa). A concrete composed of 50% and 75% recycled
materials that achieved strengths of nearly 7000 psi (48 MPa) and 6300 psi (43.4
MPa) respectively were also developed. The beneficial and negative effects of using
recycled aggregates and GGBFS in a concrete mixture were determined. The
deleterious expansions caused by the waste glass reacting with alkalis in the
cementitious paste (ASR) were also determined. It was found that GGBFS, when
used at replacement levels of 50%, eliminated these concerns when waste glass is
used at even 100% aggregate replacement levels.
The point at which replacement with recycled materials becomes detrimental
to the concrete mixture, in regards to strength, durability and workability was
determined. A replacement of 50% recycled materials was determined to be an
optimum replacement amount for concrete. The use of recycled materials was
determined to be a benefit with regards to strength and durability up to 50% when
compared with a normal concrete made from virgin materials. However, it was shown
that even a concrete with recycled materials in excess of 50% cm be very beneficial
and comparable to a normal, regular strength concrete. Although freeze-thaw
durabilitys decreased for concretes made with recycled contents in excess of 75%,
the permeabilitys of these mixtures are extremely low and when coupled with
substantial strength, these concretes would be suitable for use in many applications.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Signed
Dr. Stephan A. Durham


DEDICATION PAGE
I dedicate this thesis to my family and friends for their never ending support. To my
mother for her continued support and showing me that giving up is not an option. To
my brothers for their inspiration to become an engineer. To the many relationships
that have been severed and the friendships that were lost over the many years during
my undergraduate and graduate studies.


ACKNOWLEDGMENT
I would like to thank my academic advisor, Dr. Stephan A. Durham. His continued
support throughout the years has encouraged me and many others. 1 would like to
thank my long time undergraduate advisor Dr. Kevin Rens for his patience, support
and humor over the last decade. I would like to acknowledge Dr. Jonathan Wu, for his
inspiration and always having his door open for a wondering mind. I would like to
thank Dr. Li for his support and participating on my thesis committee. I w'ould like to
thank Dr. N.Y. Chang for his guidance, support and kindness through the years. 1
would also like to thank Rui Liu, Brian Volmer, Driss Majdoub and Adam Kardos for
their support and much appreciated help during my research
I would like to thank Katie Bartojay and Bill Kepler for their support during
this research. Without their experience, knowledge and contacts in the concrete
industry, this research would not have been successful. I would like to thank Morgan
Johnson and LEHIGH Cement for their generous contribution of ground granulated
blast furnace slag for this research. 1 would like to thank Tony Able and Rocky
Mountain Bottle Recycling for their contributions of waste glass. I would also like to
thank John Kent and Oxford Recycling for their donations of recycled concrete. I
would like to especially thank Bud Werner and his remarkable staff at CTL
Thompson for their generous help testing for potential ASR.
Additionally, I would like to thank the faculty and staff of the University of
Colorado Denver, Civil Engineering Department for their support and guidance
throughout my educational career at UCD.


TABLE OF CONTENTS
Figures....................................................................xi
Tables.....................................................................xiii
Chapter
1. Introduction..................................................... 1
2. Literature Review................................................ 6
2.1 Preface.......................................................... 6
2.2 Ground Granulated Blast Furnace Slag (GGBFS)..................... 6
2.2.1 Production....................................................... 6
2.2.2 Physical, Chemical and Reactive Properties....................... 7
2.2.3 The Effects of GGBFS on Fresh Concrete Properties.................. 10
2.2.3.1 Slump.............................................................. 10
2.2.3.2 Air Content........................................................ 11
2.2.3.3 Time of Set........................................................ 12
2.2.3.4 Temperature........................................................ 13
2.2.4 The Effects of GGBFS on Hardened Concrete Properties............... 14
2.2.4.1 Strength........................................................... 14
2.2.4.2 Permeability....................................................... 21
2.2.4.3 Freeze-Thaw Durability............................................. 24
2.2.4.4 Resistance to Sulfate Attack....................................... 25
2.2.4.5 Alkali-Silica Reactivity........................................... 29
2.2.5 Summary............................................................ 33
2.3 Waste Glass as Aggregate........................................... 34
vii


I
2.3.1 Production......................................................... 35
2.3.2 Physical and Chemical Properties................................... 36
2.3.3 The Effects of Waste Glass on Fresh Concrete Properties............ 38
2.3.3.1 Slump.............................................................. 38
2.3.3.2 Air Content........................................................ 40
2.3.4 The Effects of Waste Glass on Hardened Concrete Properties....... 41
2.3.4.1 Strength........................................................... 41
2.3.4.2 Permeability....................................................... 44
2.3.4.3 Freeze-Thaw Durability............................................. 46
2.3.4.4 Alkali-Silica Reactivity (ASR)..................................... 47
2.3.5 Summary............................................................ 50
2.4 Recycled Concrete as Aggregate (RCA)............................... 52
2.4.1 Production......................................................... 52
2.4.2 Physical and Chemical Properties................................... 53
2.4.3 The Effects of RCA on Fresh Concrete Properties.................... 54
2.4.3.1 Slump.............................................................. 54
2.4.3.2 Air Content........................................................ 57
2.4.4 The Effects of RCA on Hardened Concrete Properties................. 58
2.4.4.1 Strength........................................................... 58
2.4.4.2 Permeability....................................................... 67
2.4.4.3 Freeze-Thaw Durability............................................. 71
2.4.4.4 Alkali-Silica Reactivity (ASR)..................................... 74
2.4.5 Summary............................................................ 74
2.5 Expectations....................................................... 76
3. Problem Statement.................................................. 78
3.1 Statement.......................................................... 78
viii


I
4. Experimental Plan............................................. 80
4.1 Design Summary................................................ 80
4.2 Material Properties........................................... 81
4.2.1 Ground Granulated Blast Furnace Slag (GGBFS).................. 81
4.2.2 Portland Cement............................................... 82
4.2.3 (Virgin) Coarse and Fine Aggregates........................... 84
4.2.4 (Recycled) Coarse and Fine Aggregates......................... 84
4.2.4.1 Waste Glass as Fine Aggregate................................. 85
4.2.4.2 Recycled Concrete as Coarse Aggregate (RCA)................... 88
4.2.5 Chemical Admixtures........................................... 91
4.2.5.1 Air Entraining Admixture (AEA)................................ 91
4.2.5.2 High Range Water Reducing Admixture (HRWRA)................... 91
4.3 Mixture Designs............................................... 92
4.3.1 Mixture Batching.............................................. 96
4.3.2 Curing........................................................ 97
4.4 Concrete Testing.............................................. 97
5. Experimental Results.......................................... 99
5.1 General....................................................... 99
5.2 Problems with this Study...................................... 99
5.2.1 Re-Batch of Mixture #2 (100-RA-C) and Mixture #3 (100-RA-BF)
Due to Consolidation Concerns................................ 99
5.2.2 Re-Batch of Mixture #2 (100-RA-C) Due to Lack of Specimen
Quantities..................................................... 101
5.2.3 Freeze-Thaw Chamber............................................. 101
5.3 Fresh Concrete Properties....................................... 103
5.3.1 Slump........................................................... 103
IX


5.3.2 Air Content................................................... 107
5.3.3 Unit Weight................................................... 108
5.4 Hardened Concrete Properties................................. 111
5.4.1 Compressive Strength.......................................... Ill
5.4.2 Permeability.................................................. 124
5.4.3 Freeze-Thaw Durability........................................ 130
5.4.4 Alkali-Silica Reactivity (ASR)................................ 149
6. Conclusions and Recommendations............................... 157
6.1 Summary....................................................... 157
6.2 Recommendations for Future Studies............................ 161
Appendix
A. Concrete Mixture Designs...................................... 169
B. Material Product Data Sheets.................................. 176
Bibliography.......................................................... 164
x


FIGURES
Figure
2.1 Average Compressive Strengths of Concretes Containing GGBFS
(Averaged from four cement brands and two aggregate types),
(Sippel, 2005).................................................. 16
2.2 Average Compressive Strengths of Concrete Mixtures Containing
70% GGBFS, With and Without HRWRA, (Richardson, 2006)......... 20
2.3 Compressive Strengths of RCA Concrete Mixtures Designed with the
EMV Method and AC1 Method, assumed 28-Day age, Fathifazl et al
(2009).......................................................... 61
2.4 Average Compressive Strength of RCA Concrete and Control
Concrete, Mirjana Malesev et al (2010).......................... 62
4.1 Average Waste Glass & UCD Sand Gradations..................... 87
4.2 Average RCA and UCD Rock Gradations........................... 90
5.1 Mixture Slump and HRWRA Type Used.............................. 106
5.2 Photographs of Failed Compressive Test Specimens, 100-RA-BF
(Mixture #3) & 100-RA-C (Mixture #2), Respectively............ 113
5.3 Compressive Strength Curves, 100-RA-BF (Mixture #3)........... 114
5.4 Compressive Strength Curves, 100-RA-C (Mixture #2)............. 115
5.5 Compressive Strength Curves, 75-RA-BF (Mixture #6)............ 115
5.6 Compressive Strength Curves, 50-RA-BF (Mixture #5)............. 116
5.7 Compressive Strength Curves, 25-RA-BF (Mixture #4)............ 116
5.8 Average Normalized Compressive Strengths, All Mixtures......... 117
5.9 28-Day Compressive Strengths and CDOT Class-D Requirements.... 121
5.10 Photograph of RCPT Test Setup and Apparatus.................... 126
XI


t
i
I
j
I
I
i
5.11 Results of RCPT Testing Performed at Various for All Concrete
Mixtures......................................................... 128
5.12 Photograph of Freeze-Thaw Test Specimens, 100-RA-BF (Mixture
#3) & 100-RA-C (Mixture #2), Respectively....................... 131
5.13 Photograph of Static Frequency Testing Apparatus................ 133
5.14 Photograph of Dynamic Frequency Test Apparatus.................. 133
5.15 Average Mass Losses vs. Number of Cycles for All Mixtures...... 145
5.16 Photograph of Preparing ASR l est Specimens, (100% GGBFS
Mixture)........................................................ 151
5.17 Measuring Length of an ASR Test Specimen........................ 152
5.18 Mixture #1 (100% PC) 14-Day Test Results, ASTM C 1567........... 153
5.19 Mixture #2 (50% PC & GGBFS) 14-Day Test Results,
ASTM C 1567...................................................... 154
5.19 Mixture #3 (100% GGBFS) 14-Day Test Results, ASTM C 1567...... 154
xu


TABLES
l
Table
2.1 Mass Percentage of Compounds in GGBFS (Mindess, 2003)......... 8
2.2 Average Compressive Strengths for Concretes with Varying GGBFS
Contents and Varying Curing Temperatures, (Flale, 2001)....... 17
2.3 28-Day Chloride-Ion Penetration Results for Concretes Containing
GGBFS vs. Control Mixtures with 100% PC, (Sivasundaram, 1992).. 23
2.4 Visual Results of GGBFS Beams Tested in Sulfate Rich Soil After 5
Years Exposure (Stark, 1989)..................................... 28
2.5 Results of ASR Expansion Tests on GGBFS Specimens Tested
According to ASTM-C1260 and ASTM-C1293, (Detwiler, 2003)......... 31
2.6 Typical Chemical Compositions Percentages of Soda-Lime
Container Glass, (Weihua Jin, 2000).............................. 37
2.7 Mixture Properties and Slump Results of Concretes Containing
Waste Glass as Fine Aggregate, (Shayan et al, 2005).............. 39
2.8 Mixture Properties and Compressive Strengths of Concretes
Containing Waste Glass as Fine Aggregate, (Shayan et al, 2005)... 43
2.9 Mixture Properties and RCPT Result of Concretes Containing Waste
Glass as Fine Aggregate at 380 Days of Age, (Shayan et al, 2005). 45
2.10 Slump Test Results of RCA Concrete Mixtures, (Malesev, 2010)..... 56
2.11 Compressive Strengths of Source Concretes and Concretes Made
With RCA From Same Source, (ACI 555R, 2001)...................... 59
2.12 Table 2.12 Compressive Strengths of Concretes Made with Varying
Amounts of CCA, Karthik Obla et al (2007)........................ 64
XIII


2.13 RCPT Test Results of Concretes Made with Varying Amounts of
CCA. Karthik Obla et al (2007)................................... 69
2.14 Average Results of Rapid Freeze-Thaw Testing on Concretes Made
with CCA, Karthik Obla et al (2007).............................. 72
4.1 Colorado Department of Transportation (CDOT) Class-D concrete
specifications and field requirements............................ 81
4.2 Lehigh Grade 120 GGBFS Physical and Chemical Properties.......... 82
4.3 Holcium Type-l-II Cement Physical and Chemical Properties........ 83
4.4 Testing of Recycled Materials.................................... 84
4.5 Fine Aggregate Properties of Waste Glass and UCD Sand............ 85
4.6 Coarse Aggregate Properties of RCA and UCD Rock.................. 88
4.7 Concrete Mixture Design Matrix................................... 92
4.8 Fresh and Hardened Concrete Properties Tested.................... 98
5.1 Fresh Concrete Properties.......................................... 103
5.2 Measured Unit Weights, Theoretical Unit Weights and Air
Contents......................................................... 109
5.3 Measured Unit Weights & Air Adjusted Theoretical Unit Weights.... 110
5.4 Average Compressive Strengths...................................... 112
5.5 Average Compressive Strengths Normalized for 6.5% Air Content... 113
5.6 ASTM C 1202 RCPT Permeability Classifications...................... 125
5.7 Results of RCPT Testing Performed at Various Ages for All
Concrete Mixtures.................................................. 127
5.8 Mixture #1, CC (Control), Transverse Frequencies & Pc Values..... 135
5.9 Mixture #2, 100-RA-C, Transverse Frequencies & Pc Values......... 135
5.10 Mixture #2 (Re-Batch), 100-RA-C-2, Transverse Frequencies & Pc
Values........................................................... 136
xiv


5.11 Mixture #3, 100-RA-BF, Transverse Frequencies & Pc Values...... 136
5.12 Mixture #4, 25-RA-BF, Transverse Frequencies & Pc Values...... 137
5.13 Mixture #5, 50-RA-BF, Transverse Frequencies & Pc Values...... 137
5.14 Mixture #6, 75-RA-BF, Transverse Frequencies & Pc Values...... 138
5.15 Mixture #1, CC (Control), Average Dynamic Frequencies &
Corresponding Pc Values from Both Specimens.................... 139
5.16 Mixture #2, 100-RA-C, Average Dynamic Frequencies &
Corresponding Pc Values from Both Specimens.................... 139
5.17 Mixture #2 (Re-Batch), 100-RA-C, Average Dynamic Frequencies
& Corresponding Pc Values for Both Specimens................... 140
5.18 Mixture #3, 100-RA-BF, Average Dynamic Frequencies &
Corresponding Pc Values for Both Specimens..................... 140
5.19 Mixture #4, 25-RA-BF, Average Dynamic Frequencies &
Corresponding Pc Values for Both Specimens..................... 141
5.20 Mixture #5, 50-RA-BF, Average Dynamic Frequencies &
Corresponding Pc Values for Both Specimens..................... 141
5.21 Mixture #6, 75-RA-BF, Average Dynamic Frequencies &
Corresponding Pc Values for Both Specimens..................... 142
5.22 Durability Factors (DF) for Each Specimen, All Mixtures........ 143
5.23 Average Durability Factors (DF) for Each Mixture............... 144
5.24 Mixture Proportions for Testing ASR Potential Based on ASTM C
1567 Procedures................................................ 150
xv


1.
Introduction
Concrete is the most commonly used building material in construction today. Over 2
billion tons of concrete are produced every year and there is a consistent 5% increase
each year (ASDSO, 2011). Anything and everything related to construction is open
for scrutiny in todays eco-conscience society. People today are more in tune and
informed about the negative effects we leave behind for future generations. Green is
the new buzz word, and every facet of todays industry is attempting to reduce their
carbon footprint. Builders today are under constant pressure to incorporate more and
more recycled materials into their products and to become more earth friendly. The
potential use of recycled materials in concrete is a growing interest. Although the use
of recycled materials in concrete is not a new advancement, typical replacement
values have commonly been on a small order.
Concrete has undergone significant advancements over the years, primarily
with regards to cement. Cement is the most important ingredient in a concrete
mixture. Cement is the glue of a concrete mixture, binding the aggregates together to
form a solid matrix and give the concrete its strength. The cement used in todays
concrete is referred to as a portland cement. The production of portland cement
(PC) requires significant amounts of energy. First, the raw materials used to make
cement must be acquired and processed. These raw materials must first be quarried
from the earth before being crushed and blended to an acceptable level. The blended
materials must then be put into a high temperature kiln that transforms these raw
materials into clinker. Clinkering is a term used to indicate the stage between
sintering and fusion. Sintering indicates no melting takes place, and fusion takes place
when 100% of the material is in a molten state. It is estimated that at any given time
in the kiln only one quarter of the material is in a molten state. After leaving the kiln
this clinker is then processed even further (ground down and mixed with gypsum)
before becoming portland cement. The production of portland cement accounts for
1


5% of the total global CO2 emissions caused by humans and it is estimated that for
every ton (metric ton) of PC produced, an equal ton of carbon dioxide gas (CO2) is
emitted into the atmosphere, (World Buisness Council for Sustainable Development,
2002). Carbon dioxide is a green house gas, and is believed to be a main contributor
to global warming. Most of the CO2 produced comes from the high temperature kilns
used in PC production plants. Of all the raw materials used in concrete today, PC is
the largest contributor to green house gases.
The use of recycled materials to replace cement content is a common practice
and has been for many years. It has been demonstrated that concretes strength,
durability and workability can be increased from the use of certain recycled materials.
Supplementary cementitious materials (SCM) have been used to replace cement in
concrete for thousands of years. Common cement replacements used today are fly
ash, silica fume and ground granulated blast furnace slag (GGBFS). Fly ash and silica
fume are byproducts of the power and silicon metal industry respectively. GGBFS is
a byproduct of the iron manufacturing industry. Fly ash and silica fume are known as
pozzolanic materials. GGBFS is a hydraulic cementitious material that also has
pozzolanic characteristics. The term pozzolan is used to describe any reactive
aluminosilicate material. Pozzolans react w ith by-products of the cement hydration
process in order to develop strength characteristics in concrete. Pozzolans will not
typically produce strength alone when mixed with water and therefore require cement
within the mixture. For this reason the common thought in the concrete industry is
that cement is needed, in at least a moderate amount, to produce strength. GGBFS is
also a cementitious material much like cement, and when mixed with water will
hydrate and produce strength alone.
Aside from cement, concrete contains several other ingredients. Aggregates
have almost always been used in concretes of the past and are always used in todays
concretes. Even todays grouts and mortars contain aggregates. Although important in
2


varying degrees to a concrete mixture, aggregates by and large are a filler material. In
fact, it is the bond between the aggregate and cement that is the weakest link in the
concrete matrix. Typically it is only in high strength concretes where aggregate
strength becomes a contributing factor. Quality aggregate sources are becoming more
difficult to find. Many aggregate sources used in the past have been depleted and
concrete batch plants are forced to use lesser quality aggregates. To acquire
aggregates from the earth, considerable energy must be used to quarry and refine the
rock before being suitable for use in concrete. Mining operations are always at the
forefront of environmental debate not only from the destructive aspect, but also from
an aesthetic standpoint. For these reasons, aggregates are of primary interest with
regards to potential replacement with recycled materials.
Many forms of aggregate replacement have been used in the past, from
recycled automotive tires and waste metal to pure trash. A less common coarse
aggregate replacement that is gaining more interest is recycled concrete aggregate
(RCA). Recycled concrete comes from the demolition of buildings, sidewalks, streets,
etc. Increasing concern over the potential harmful effects that crushed concrete can
have on the environment have caused growing concerns on how to dispose of the
materials (due to leaching of chemicals into the watershed), not to mention the
increased costs associated with concrete disposal. The diminishing landfill space is a
growing concern throughout the world. Reusing crushed concrete is of considerable
interest for these reasons.
Another less common aggregate replacement that is gaining more attention is
the use of recycled glass as a fine aggregate replacement. Glass containers are
typically recycled to make more glass containers, but the potential beneficial use in
concrete is gaining interest. Recycling of glass containers is difficult and may be
costly if separating the glass into colors is required. Removal of contaminants is also
difficult and much of the glass produced today ends up in landfills. In 2007
3


approximately 13.6 million tons (12.3 metric tons) of waste glass were generated in
the United States, and 76% of this glass was disposed of in landfills (Verdugo, 2009).
Many states and municipalities have attempted to avoid this difficulty by mandating
taxes on glass containers to help push manufacturers into re-using instead of
recycling. Whether glass containers are crushed and recycled or re-used, it is
estimated that for every ton of glass recycled, 1000 pounds (454 kg) of CO2 gas is
saved from being emitted into the atmosphere, (en.wikipedia.org, 2011).
This research investigates the effects that these recycled materials will have
on the fresh and hardened properties of concrete. Several concrete mixtures
containing recycled materials were designed and batched for this research. The
amounts of recycled materials used in each concrete mixture were varied, and their
fresh and hardened properties compared to a control mixture composed of natural
(virgin) aggregates and cement. The mixtures developed for this research project had
a percentage of aggregates and cement replaced with recycled materials. The natural
coarse aggregates were replaced with recycled concrete aggregate (RCA), the natural
fine aggregates with waste glass and the cement was replaced with GGBFS. To fully
investigate the effects these recycled materials have on the concrete, six mixtures
were developed, batched, and tested for structural and durability performance.
The recycled concrete aggregate and waste glasses were fully tested prior to
batching. Multiple gradations for both aggregates were completed as well as fineness
modulus, specific gravity, absorption capacities and unit weights.
A concrete mixture containing 100% recycled materials (RCA, waste glass
and GGBFS) was designed, batched and tested for this research. Three additional
mixtures with recycled material contents of 25, 50 and 75% were batched and tested.
One mixture was batched that contained 100% RCA and 100% cement as well as a
control mixture composed of 100% natural aggregates and 100% cement. All
4


mixtures were tested for fresh and hardened concrete properties. The fresh concrete
properties tested included slump, unit weight and air content. The hardened concrete
properties examined were compressive strength, rate of strength gain, permeability,
freeze-thaw resistance and alkali-silica reactivity (ASR).
A literature review was performed to investigate any past research completed
regarding GGBFS, recycled concrete, and waste glass on the effects each has on
concretes properties. No documented research could be found on the combined
effects that all three of these components together have on concrete. This research is
therefore considered to be a first of its kind. All testing conformed to ASTM testing
standards and when deviated from, notation was made. All data results, details and
conclusion of findings of this research are included within this thesis.
5


2.
Literature Review
2.i Preface
This literature review will not focus on the research that has been done on ordinary
concrete containing portland cement and normal aggregates. This review will only
focus on the recycled materials used to replace the cement and aggregates. The use of
recycled concrete, waste glass and GGBFS will be reviewed. However, no known
documented research on the combined effects of these three materials used together in
a concrete mixture could be found. Therefore, each material will be reviewed
individually in regards to any past research performed.
2.2 Ground Granulated Blast Furnace Slag (GGBFS)
2.2.1 Production
Ground granulated blast furnace slag (GGBFS) is a byproduct of the iron
manufacturing industry. It was first commercially produced in Germany in 1853 and
slag cements were used to build the underground metro station in Paris in 1889,
(PC A, 2005). Slags are residues that come from blast furnace production of various
metals and steel, as well as from the production of iron from ore. The slags that come
from the steel industry and other metals are not suitable for use in concrete unless
they are beneficiated (Mindess, 2003). Typically, the slags that are used in concrete
primarily come from the production of iron. During the production of iron, inorganic
lime based fluxes are used to remove impurities from the iron. These fluxes rise to the
surface of the molten iron and from there, they are removed and cooled. The slags
must be cooled quickly (i.e. not air cooled) in order to become a suitable form of
hydraulically active calcium aluminosilicate glass. If allowed to dry in air the slag
will form inert (non-reactive) calcium magnesium silicates (Mindess, 2003). The
6


process of quenching (using water to cool) is used to cool the slags. The method of
cooling varies from one plant to another and the amount of water used will have a
direct impact on the quality and strength of the slag (Bureau of Reclamation, 1988).
Granulation and pelletization are two ways in which the slag can be quenched. The
granulation process is used to produce GGBFS. During the granulation process the
molten slag is broken up by jets of water before being immersed in a water bath. This
process produces small 4 mm-sized granules which are then ground down to cement
fineness. Pelletization produces larger granules which are used in concrete as light
weight aggregates.
2.2.2 Physical, Chemical and Reactive Properties
There are three ASTM specified grades of GGBFS produced today, Grade 80, Grade
100 and Grade 120. This classification is called the slag-activity index and applies
to the GGBFS ability to improve the compressive strength of mortar cubes mixed
with 50% cement and 50% GGBFS when compared with reference cubes containing
100% cement (ASTM C 989). The 28 day compressive strength of Grade 80, Grade
100 and Grade 120 cubes must be at least 75, 95 and 115% of the strength of
reference cubes when compared respectively. GGBFS is usually ground down to a
fineness that exceeds normal portland cement to increase reactivity. Blaine fineness
ranges are anywhere from 3500 cm2/g to 8000 cm2/g. The fineness of the slag varies
with grade and will therefore also affect the fresh concrete properties as well. The
Blaine fineness of Grade 120 slag is higher than Grade 100, which in turn is higher
than Grade 80 slag. The higher the fineness the higher the water demand. Ultimately,
a concrete made with Grade 120 slag will have a lower slump value than a similar
mixture made with Grade 80 slag.
7


The chemical composition of GGBFS used in concrete today varies little from
plant to plant due to the fixed process of steel making and is not a significant concern
(Hooten, 2000). GGBFS is rich in lime, silica and alumina. The typical mass
compound percentages are shown below in Table 2.1. The amounts of sulfur are
limited to 2.5% as sulfide and 4.0% as sulfate per ASTM C 989. Residual iron is also
present as ferric oxide.
Table 2.1 Mass Percentage of Compounds in GGBFS (Mindess, 2003)
Compound Percentage Compound Percentage
CaO 35-45 MgO 5 15
Si02 32-38 Fe203 <2
AI2O3 8- 16 Sulfur 1 -2
When hydraulic cement (PC or GGBFS) is mixed with water, a chemical
reaction takes place known as hydration. Equations 2.1 and 2.2 below show the
generalized hydration process of tri-calcium silicate (C3S) and di-calcium silicate
(C2S) in cement (ceramic notation), (Mindess, 2003).
C3S + H20 C3S2H3 + CH (2.1)
C2S + H20 - C3S2H3 + CH (2.2)
Where: C3S2H3 is referred to as Calcium Silicate Hydrate (C-S-H)
CH is referred to as Calcium Hydroxide
The calcium silicate hydrates are the primary strength producing compounds
in a concrete mixture. Tri-calcium silicates typically form early in the hydration
8


process and give the concrete early age strength. The di-calcium silicates primarily
form much later and will contribute more to later age strengths. The hydration of slag
is much slower than a typical PC hydration and may take several months depending
on grade to reach equivalent 28 day strengths when compared with concretes made
with pure cement as the binder. The cause of this slow hydration is believed to be due
to impervious coatings of amorphous silica and alumina that form around the slag
particles during the first stages of the hydration process (Mindess, 2003). It is
believed that GGBFS hydration primarily produces C2S as opposed to a PC hydration
which primarily produces C3S (ASTM C-989). The hydration of GGBFS is also alkali
activated and the hydration rate will depend on the alkali content of the cement and
the alkali content of the slag.
The calcium hydroxide (CH) produced during the hydration process is not
considered a main strength producing compound, may actually cause durability
problems in a hardened concrete. CH in a hardened state is water soluble and
crystalline in structure. Leaching of CH from brick mortars can be readily seen as
well as from concrete (chalky white substance). This leaching process is called
efflorescence. Calcium hydroxide is also a component in alkali-silica reactions as
well as sulfate attack. Although CH is not a main strength producing component of a
cement hydration process, it is a primary ingredient for a pozzolanic reaction. ASTM
C 618 describes a pozzolanic material as a siliceous or siliceous and aluminous
material which in itself possesses little or no cementitious value but will, in finely
divided form and in the presence of moisture, chemically react with calcium
hydroxide at ordinary temperatures to form compounds possessing cementitious
properties. GGBFS contains amorphous silica and will react with calcium hydroxide
to form calcium silicate hydrates. Equation 2.3 below shows the generalized
pozzolanic calcium hydroxide and silica hydration process (ceramic notation),
(Mindess, 2003).
9


CH + S + H20 - C3S2H3
(2.3)
Due to this conversion of calcium hydroxide to calcium silicate hydrate, a
concretes strength and durability can be greatly increased by the use of pozzolanic
materials.
2.2.3 The Effects of GGBFS on Fresh Concrete Properties
2.2.3.1 Slump
A majority of past research indicates the effects of GGBFS on concretes workability
are favorable (i.e. increased workability). According to a report issued by ACI
regarding the use of GGBFS in concrete mixtures, typically the addition of GGBFS in
a concrete mixture will improve workability and increase slump, (ACI 233R, 2000).
This is believed to be caused by the texture and density of finely ground blast furnace
slag particles, which have smooth glassy surfaces and have a denser outer surface
than cement particles. GGBFS particles will therefore require less water immediately
after mixing than cement particles which absorb water rapidly (ACI 233R, 2000).
According to Wang Ting et al (2005), the increase in workability observed with
concretes containing GGBFS may be a result of an increased repulsion potential
between the cement and GGBFS particles. They also attributed this increase in
workability to the larger specific area of GGBFS particles.
Osborne (1989) investigated the effects of GGBFS on slump, compaction
factor and Vebe for concretes containing 0, 40 and 70 percent replacements with slag.
He found that as the percentage of GGBFS increased, the ratio of water to
cementitious materials needed to be reduced in order to maintain workability
properties similar to the control mixture having only cement as the binder. Research
conducted by the Virginia Department of Transportation (1999) showed that it is
10


possible to reduce the w/c (water to cement ratio) of a concrete mixture containing
GGBFS to achieve the same slump as a mixture containing only cement due to the
decreased water demand. Contrary to the previous work cited, the Missouri
Department of Transportation (2006) found that average slump values for a 70%
GGBFS mixture (Grade 120) were 2 inches (51 mm) less than that of control
mixtures having only PC as the binder and similar w/c. Similar results were reported
by Hale (2001) who investigated two GGBFS mixtures with 25 and 50%
replacements. Both mixtures had significantly less slump than the control mixture
with only cement. This was attributed to the higher fineness of the Grade 120 slag
when compared to the fineness of the cement.
2.2.3.2 Air Content
Information regarding the effects of GGBFS on air content was limited. Much
research was found regarding fresh concrete properties on concretes mixed with
GGBFS and air entraining admixtures however air content was rarely discussed.
According to the Portland Cement Association (2005), ground slags will have a
varying effect on the required dosage rates of air-entraining admixtures. ACI-233R
(2000) indicates that typically the required amount of air-entraining admixture will
increase if the GGBFS is finer than the cement.
Sivasundaram and Malhotra (1992) studied the effects of high volume
GGBFS concrete with target air contents of 5%. When studying the effects of air
entraining admixtures (AEA), they found that as GGBFS content increased from 60
to 75%, the amount of AEA needed almost doubled. When comparing GGBFS
mixtures to the control mixture with 100% cement, in some cases the AEA dosage
required more than quadrupled. They also found that if slag content was held constant
and cement content adjusted the effects on AEA dosage remained constant, which
11


would indicate the slag content is the governing factor. Richardson (2006)
investigated the air void system characteristics of several concrete mixtures composed
of 70% GGBFS replacement and found that the results were similar to control
mixtures made with only PC as the binder.
2.2.3.3 Time of Set
In general, research has shown that initial and final set times will be effected by the
use of GGBFS. According to ACI-233R (2000), the time of set for concretes
incorporating GGBFS will be increased and dependent on the initial temperature,
replacement amount, cement characteristics and w/c. The time to initial set for fresh
concrete temperatures around 73 F (23 C) will typically be one half to one hour
longer and for temperatures above 85 F (30 C), no significant changes were
observed when compared to a mixture with 100% cement. Significant retardations
have been observed at lower temperatures but these effects can be reduced by the use
of conventional accelerators such as calcium chloride.
Hogen and Meusel (1981) found that at higher temperatures (above 85 F) the
initial and final set times were similar to those of normal concrete with 100% cement.
Results were similar for slag fineness ranges between 4500 cm /g and 6000 cm /g
(2197 ft2/lb and 2930 ft2/lb) which give an indication that slag fineness is not related
to set time. Sakai et al (1992) found that only the final set times increased when
comparing normal concrete and GGBFS concretes. The initial set times were not
affected. They also used varying slag finenesses between 3000 cm /g to 6000 cm /g
(1465 ft2/lb and 2930 ft2/lb) which strengthens the belief that slag fineness is not a
factor in set time.
12


2.2.3.4 Temperature
The temperature rise in concrete is caused by the hydration process and is of
considerable interest in mass concrete placements. This rise in temperature can take
place over many days, months or even years depending on the size of placement.
Excessive temperatures can lead to thermal induced cracking, especially in mass
placements. The use of mineral admixtures such as fly ash and GGBFS in mass
concrete have essentially eliminated the need for production of Type IV cement (low
heat of hydration) in the United States. Since the reaction rate of GGBFS is much
slower, the heat is released gradually over longer periods, and the temperature of the
concrete is reduced (Mindess, 2003). According to ACPA, the decrease in
temperature is inversely proportional to the fineness, i.e. the finer the GGBFS the less
of a decrease in temperature due to the higher reactivity of finer slags
(www.ndconcrete.com).
The Bureau of Reclamation (2010) performed adiabatic temperature studies
on several concrete mixtures containing various pozzolans and Type II low heat of
hydration cement. The pozzolans used were a Class F fly ash and a Grade 120
GGBFS. Three mixtures containing 30% Class F fly ash and two mixtures containing
70% GGBFS were batched and tested. The total cementitious contents for the fly ash
and GGBFS mixtures were approximately 500 lb/yd3 and 375 lb/yd3 (297 kg/m3 and
222 kg/m3) respectively. The temperature rise for the GGBFS mixtures was
approximately 55 F (13 C) over the course of 56 days. The average temperature rise
for the fly ash mixtures was 73 F (23 C) over the course of 56 days (high of 80.1 F
(26.7 C) and low of 68.8 F (20.4 C)). This indicates a significant temperature
decrease between 14 F and 25 F (-10 C and -4 C) when comparing the GGBFS
mixtures to the fly ash mixtures. This decrease in temperature rise may also be due to
the decreased cementitious content of the GGBFS mixtures.
13


Sivasundaram and Malhotra (1992) investigated the autogenous temperature
rise and peak temperatures for concretes containing 50 to 70% GGBFS. They found
that high volumes of slag content would suppress the autogenous temperature rise of
concretes. In addition, they found that with increasing slag content the peak
temperature rise decreased as well.
2.2.4 The Effects of GGBFS on Hardened Concrete Properties
2.2.4.1 Strength
Studies on the effects of GGBFS on strength are varied. Typically, the early age
strengths of concretes containing GGBFS will be lower and later age strengths higher
when compared with normal PC concretes. The rate of strength gain appears to be
dependent on the grade of GGBS used in the mixture as well as the replacement
percentage. When compared to a normal cement concrete, mixtures with Grade 120
slag show decreased early age strengths and increased later age strengths (beyond 7-
days). If Grade 100 slag is used the general trend is lower early and mid age strengths
(1-21 days), and similar or greater strengths at later ages. Grade 80 slag will typically
lower all strengths at both early and later ages (ACI 233R, 2000). Other factors also
affected the strengths, such as the use of water reducing admixtures and curing
regimens (particularly temperature). There also appears to be ideal replacement
percentages were the greatest strength gains are realized.
Hogan and Meusel (1981) investigated the compressive strengths of GGBFS
concrete for a variety of replacement percentages. The GGBFS contents examined
were 40, 50, and 65 percent. These mixtures were compared to a control mixture
consisting of cement as the only binder. Two w/c ranges were used for air-entrained
mixtures and non-air-entrained mixtures. The w/c was 0.40 and 0.55 for air-entrained
14


mixtures and 0.38 and 0.55 for non-air-entrained mixtures. Ail specimens were moist
cured. The fineness of the slag used ranged from 4500 cm2/g to 6000 cm2/g (2197
y 'y
ft /lb to 2930 ft /lb). The results of the compressive strength tests show that the
GGBFS concretes gained strength at a slower rate during the early ages (1-7 days)
than the control mixture made with 100% cement as the binder. Between 7 and 10
days the strength gains of GGBFS concretes were greater than the control mixture.
When comparing 28-day strengths for all mixtures the concrete made with 40%
GGBFS, w/c of 0.38 and no air-entrainment had the highest strength of 8220 psi (56.7
MPa) compared with the control having a 28-day strength of 7240 psi (50.0 MPa).
For all mixtures, regardless of air-entrainment or w/c, the 28-day strengths decreased
as slag content increased from 40% to 65%. For non-air-entrained mixtures,
increasing the slag content from 40% to 50% had only a minimal effect on the
compressive strength at 28-days. When the slag content was further increased from
50% to 65%, the 28-day strengths decreased significantly by as much as 24%. All
mixtures incorporating GGBFS except the mixture with w/c of 55% and non-air-
entrained, exceeded the 28-day strengths of the similar control mixtures
Cramer and Sippel (2005) investigated concretes incorporating 30 and 50%
replacements with GGBFS and two different aggregate sources and four different
cement brands. The aggregates used were a limestone aggregate and an igneous
aggregate. Type I cement and Grade 100 GGBFS were used for all mixtures. All
mixtures were air-entrained and had air contents of 6%. They found that regardless of
replacement amount, most GGBFS mixtures had lower strengths than control
mixtures at all ages (with a few exceptions). Figure 2.1 shows the average
compressive strengths for the mixtures based on GGBFS replacement.
15


Control (100% Cement)------------30% GGBFS
50% GGBFS
Concrete Age (Days)
Figure 2.1 Average Compressive Strengths of Concretes Containing GGBFS
(Averaged from four cement brands and two aggregate types), (Sippel, 2005).
Although not shown on Figure 2.1, the 365-day strength trend was similar and the
control mixture on average had the highest strength. This is contrary to typical trends
were Grade 100 GGBFS concretes exhibit higher later age strengths (beyond 21 days)
than normal concrete (AC1, 2000).
Hale (2001) investigated the effects of GGBFS and curing regimens on
concrete. The cement used was a Type I and the GGBFS was a Grade 120. The total
cementitious content for all mixtures was 658 lb/yd3 (390 kg/m3). Normal aggregates
were used as well as a w/c of 0.39 for all mixtures. An air-entraining admixture was
used for all mixtures to achieve a target air content of 7%. Three different curing
16


regimens were followed; standard conditions of 73 F (23 C) and 100% humidity,
cold weather conditions of 50 F (10 C) and 100% humidity, and hot weather
conditions of 83 F (28 C) and 100% humidity. A replacement percentage of 50%
GGBFS was examined with standard and hot curing conditions, and 25% for cold,
standard and hot curing conditions. All mixtures were compared to a control mixture
consisting of 100% cement. Table 2.2 shows the results from this study.
Table 2.2 Average Compressive Strengths for Concretes with Varying CGBFS
Contents and Varying Curing Temperatures, (Hale, 2001).
Mix ID Curing Temp. 3-Day 7-Day 28-Day 56-Day 90-Day
Control 70 F 2820 3500 4510 4700 5090
25% GGBFS 70 F 3100 3940 5820 6640 7040
50% GGBFS 70 F 3050 4310 6390 7150 7670
Control 50 F 1200 2520 3700 3980 4560
25% GGBFS 50 F 1500 2900 5140 5840 6100
Control 83 F 3760 4380 5460 6070 6380
25% GGBFS 83 F 3670 4990 6610 7110 7200
50% GGBFS 83 F 3470 5500 6780 7570 8660
What is interesting in this study is that both the 25 and 50% replacement
mixtures exceeded the control mixtures strength at nearly all ages and curing
regimens. The only exception to this trend was for the hot cured specimens at 3-days
in which the control specimen had a strength of 3760 psi (25.9 MPa) and the 25 and
50% specimens had strengths of 3670 psi and 3470 psi (25.3 MPa and 23.9 MPa)
17


respectively. Yet after the 3-day mark, both GGBFS specimens rebounded and
overtook the control specimens during the hot cured regimen. As expected, the hot
curing regimen increased compressive strengths for all mixtures and was attributed to
the evaporation of water in the fresh concrete state which can lead to a reduced w/c.
As expected, the cold curing regimen decreased both specimens strength throughout
the range when compared to the standard curing temperature regimen. As stated in the
report, the air contents for the 50% GGBFS mixture were one to two percent less than
the other mixtures. Another interesting point is that the air content for the control
mixture cured under the cold regimen was 9.1%, which was 3.0% higher than the
25% GGBFS mixture which had an air content of 6.1%. These results do not indicate
whether or not the compressive strengths were normalized for air content and
therefore it assumed they were not.
Sivasundaram and Malhotra (1992) investigated the effects of GGBFS on
concretes using high replacement amounts between 50 and 75%. The w/c ranged
from 0.27 to 0.45. All mixtures contained air-entraining admixture and had target air
contents of 5.0%. A high range water reducing admixture (HRWRA) was also used
on all mixtures. All specimens were cured at 100% humidity and 68 F (20 C). The
cement used was a Type I, however the GGBFS grade nor the fineness was specified.
The total cementitious contents varied across the board (from 242 to 428 lb/yd' (144
to 254 kg/m3)) and when coupled with the varying w/c, the results are difficult to
interpret. They found that in all mixtures with 169 lb/yd3 of total cementitious, the
strengths for all ages increased when slag content increased from 50 to 70 and 75%.
They also determined that increasing the slag content from 70 to 75% does not yield
any significant benefits to strength. They found that for cementitious contents of 169,
211 and 253 lb/yd3 (100, 125 and 150 kg/m3) the optimum slag replacements were 70,
65 and 55% respectively. This indicates that as total cementitious contents increase,
the benefits of slag content decrease. In general, they determined that slag concretes
18


typically attain most of their strength by 28 days when compared to control mixtures
with only PC. All slag concretes were equal to or had higher strengths than control
mixtures beyond 7-days.
Richardson (2006) investigated two concrete mixtures with GGBFS contents
of 70% of the total cementitious materials. A HRWRA was used on only one GGBFS
mixture to study the effects of using a HRWRA on a slag concrete. A control mixture
was batched with a Type I cement while a Type II cement was used for the two
GGBFS mixtures. The type of slag used for both mixtures was a Grade 120. Normal
aggregates were used and average w/c values ranged from .399 to .419. Air entraining
admixtures (AEA) were used to achieve target air contents of 6.0% on all mixtures.
The average results of compressive tests performed on three mixtures each for the
control, 25 and 50 GGBFS mixtures are shown in Figure 2.2.
The results of this study showed that both 70% GGBFS mixtures had
significantly less strengths for all ages when compared to the control mixture. The
strength gain for early age (up to 7-days) was lowest for the 70%GGBFS when
compared to the 70%GGBFS-HRWRA and control mixture. The strength gain for the
HRWRA mixture was greatest between 7 and 28-days. The effects of using the
HRWRA on GGBFS mixtures can readily be seen. Beyond 7-days the HRWRA
mixture had greater strengths than the GGBFS mixture without HRWRA. This is no
surprise however since the use of HRWRA creates a superior microstructure within
the concrete matrix. The control mixture only had slightly higher air contents (around
1.0% higher) than both the GGBFS mixtures and therefore air content was probably
not a cause of the higher compressive strengths for the control mixture. A major
cause for the slower strength gains for both GGBFS mixtures is probably due to the
use of Type II cement. The use of Type II cement will typically lead to slower
strength gains caused by the lower heat of hydration. Although not shown in Figure
2.2, the 365 day strengths show similar trends with the highest strength of 6970 psi
19


(48.1 MPa) coming from the control, and strengths of 5645 and 5010 psi (38.9 and
34.5 MPa) from the 70%GGBFS-HRWRA and 70%GGBFS mixtures respectively. If
similar cement types were used in all mixtures, the results would most certainly be
different and the GGBFS mixtures may have been on par with the Control mixture,
especially at later ages.
Control (100% Cement) -70% GGBFS--70% GGBFS-HRWRA
Figure 2.2 Average Compressive Strengths of Concrete Mixtures Containing
70% GGBFS, With and Without HR WRA, (Richardson, 2006).
20


2.2.4.2 Permeability
Most studies of permeability on concretes containing GGBFS have shown that
permeability decreases with increasing slag content (ACI, 2000). A concretes
permeability depends on the interconnectivity of the pores within a hardened
concrete. The less connected these pores are, the more difficult it is for water and
chemicals to penetrate. These pores can vary in size from the microscopic to those
that can be seen by the naked eye. The use of GGBFS not only eliminates pores but is
also believed to reduce the size of pores as well. This is due in part to the
transformation of CH within a pore to C-S-H during a pozzolanic reaction. In
essence, this process decreases the solubility of the concrete (Mindess, 2003).
Another important factor on a concretes permeability is the w/c. The lower the w/c
the less permeable the concrete will be.
There are three categories of measurement used to determine a concretes
permeability. Two of these are traditional methods which involve the flow or
movement of water through a concrete, while a third indirect method involves the
flow or movement of an electrical current. The indirect method is the most commonly
used due to the difficulty in calculating the flow of water through a concrete medium.
The indirect methods are known as rapid chloride-ion permeability tests (RCPT). In
these tests a voltage is applied to the sides of a concrete specimen (typically a
cylinder 4 inch (102 mm) diameter by 2 inch (51 mm) height with solutions of
sodium chloride and sodium hydroxide on opposite ends. The amount of charge that
has passed through the specimen in 6 hours is used as a measurement of permeability.
Regardless of method, permeability measurements show significant variability with
coefficients of variation between 30 and 50% (Mindess, 2003). This has caused some
researchers to believe that the RCPT method only measures concretes conductivity
and not the permeability.
21


According to Shi, et al, (1998), who investigated the effects of GGBFS on the
permeability of concrete, the addition of mineral admixtures such as fly ash, silica
fume and GGBFS may affect the RCPT values due to the altering of the conductivity.
This change in conductivity is caused by the change in pore fluid chemistry'. A
hardened concretes pore fluid typically consists of sodium (Na), potassium (K) and
hydroxide (OH) ions. The report indicated that the introduction of GGBFS decreased
the concentration of the K and OH ions while increasing the Na ion concentration.
With these changes in pore chemistry comes a change in a concretes specific
conductivity. When comparing the specific conductivity of a concrete made from
100% PC to a mixture made with 50% GGBFS, the difference in specific
conductivity at 28 and 730 days of age was 3.25 and 24% respectively. They
concluded that RCPT tests should not be used to evaluate the permeability of
concretes containing mineral admixtures.
Sivasundaram and Malhotra (1992) studied the effects of GGBFS on a
concretes permeability using high replacement amounts between 50 and 75%. The
w/c ranged from 0.27 to 0.45. The permeability was determined based on the
specimens resistance to chloride-ion penetration measured according to AASHTO
T277-83. All mixtures contained air-entraining admixture and had target air contents
of 5.0%. A high range water reducing admixture (HRWRA) was also used on all
mixtures. All specimens were cured at 100% humidity and 68 F (20 C). The cement
used was a Type I; however neither the GGBFS grade nor the fineness was specified.
The chloride-ion penetration tests were performed at 28 days of age. The results of
this study are shown in 'fable 2.3.
22


Table 2.3 28-Day Chloride-Ion Penetration Results for Concretes Containing
GGBFS vs. Control Mixtures with 100% PC, (Sivasundaram, 1992).
Mixture Series w/c GGBFS Replacement Chloride-Ion Penetration
(#) ( ratio ) <%) ( Coulomb)
3 0.29 75 174
2 0.34 70 230
6 0.28 70 213
9 0.27 65 321
1 0.45 60 829
5 0.36 60 325
8 0.30 60 276
4 0.45 50 1159
7 0.38 50 383
Control-1 0.39 0 2984
Control-2 0.31 0 1283
Control-3 0.27 0 1305
The result show that the resistance of the concretes to chloride-ion penetration
increased with increasing slag content. Mixture #6 and #8 had chloride-ion
penetrations of 213 and 276 coulombs respectively. Compared with similar w/c
control mixtures #3 and #2 which had a chloride-ion penetration of 1305 and 1283
coulombs, the decrease in penetration is roughly between 85 and 80 percent. What is
also easily verified from this research is that as w/c decreases, the chloride-ion
penetration resistance increases substantially. When comparing the control mixtures
and GGBFS mixtures with two or more w/c values, as w/c decreased the penetration
decreased.
23


2.2.4.3 Freeze-Thaw Durability
The effect of GGBFS on a concretes resistance to freeze-thaw cycles is scattered.
Some studies have found little to no benefit from the use of GGBFS, while others
have found detrimental effects from the use of slag. Several factors determine how
well a concrete will withstand freeze-thaw cycles. A concretes resistance to freezing
and thawing depends on the permeability, the degree of saturation of the paste, the
amount of freezable water, the rate of freezing, the average distance from any point in
the paste to a free surface where ice can safely form and strength (Mindess, 2003).
Air-entrained concrete is commonly used today in areas where freeze-thaw durability
is needed. These air-entrained concretes have small evenly dispersed air pockets
which allow ice crystals to form (within the free surface) in the paste without causing
excessive tensile forces. Flowever, even air-entrained concrete will not be able to
withstand repeated freeze-thaw cycles without strength. Strength and w/c go hand in
hand and according to theory a fully hydrated concrete with w/c below 0.36 will not
even need air-entrainment. This is due to the lack of freezable water available within
the paste. It has also been noted that current test methods to determine a concretes
resistance to freezing and thawing such as ASTM C666, are not clear indicators to
actual field conditions and the results should be used with caution (Mindess, 2003).
Laboratory conditions caused by ASTM C666 are extremely harsh and unrealistic for
most applications, especially the high rate of freezing (5 F/h (2.8 C/h)).
Hogan and Meusel (1981) investigated the durability of concretes containing
50% GGBFS using ASTM C666 procedure A. Concrete beams were cast for a
mixture containing 50% GGBFS as well as concrete made from 100% PC as a
control. The total number of freeze-thaw cycles these beams were subjected to was
301. The results indicated that both concretes achieved similar results. The 50%
GGBFS had a durability factor of 91 and the control mixture had a durability factor of
98. They also concluded that the weight losses and expansion differences between the
24


mixtures determined during testing were negligible and both mixtures produced
sound concretes.
Lane and Ozyildirim (1999) performed freeze-thaw testing on several
concrete mixtures containing 25, 35, 50 and 60% GGBFS replacements. The GGBFS
was Grade 120 and all mixtures had w/c of 0.45. The testing method used was ASTM
C666 procedure A, where the specimens are subjected to 300 cycles of freezing and
thawing while being immersed in a 2% sodium chloride by mass solution. This
procedure is quite severe and all specimens performed very well with durability
factors greater than 100 for all specimens including the control beam made with
100% PC. The 25, 35 and 50% GGBFS beams had percentage of weight losses below
3% while the 60% GGBFS beam had a loss of 7.7% which exceeds the maximum
permissible limit of 7.0%. The scaling observed on all GGBFS specimens was only
on the surface and it is believed this would not be detrimental to use on roadways (a
primary basis for this study) since the loss of the top surface lends to increased
traction for tires.
2.2.4.4 Resistance to Sulfate Attack
Sulfate attack is a complex process which may involve all hydration products within
the cement paste. However, a clear correlation has been established between the tri-
calcium aluminate (C3A) content and a concretes susceptibility to sulfate attack.
When C3A hydrates with Gypsum (added to ground clinker during cement
production) and water, the product formed is called ettringite. Ettringite is stable as
long as gypsum is present. However, when the gypsum is used up, ettringite and any
remaining non-hydrated C3A will react a second time to form mono-sulfoaluminate. It
is the mono-sulfoaluminate that is very unstable in the presence of sulfate. Sulfates
are present in many ground waters, soils and clays and also in sea water. If sulfate
25


comes into contact with mono-sulfoaluminate, a reaction will take place and ettringite
will reform causing excessive stresses within the cement paste leading to rapid
deterioration. Reactions can also take place between sulfates and the CH and C-S-H
within a hardened cement paste. Reducing a concretes susceptibility to sulfate attack
involves decreasing the C3A content, lowering the w/c which increases the strength
and decreases the permeability.
It is well documented that concretes containing GGBFS replacements of 50%
or more have shown significantly greater resistances to sulfate attack. A minimum
replacement of 50% GGBFS is recommended when used with Type I cements that
have C3A contents up to 12% (AC1 233R. 2000). It is also recommended that the
alumina content of the GGBFS be no greater than 11% and in certain studies it was
found that when slags are used that have less than 11% alumina, increases in sulfate
resistance were found regardless of C3A content in the cement. Flogan and Meusel
(1981) investigated the resistance to sulfate attack with concretes having between 40
to 65% replacements with GGBFS. The method of testing used was the Wolochow
Method (lean mortar bar method). They found that mortar bars with 65% GGBFS had
expansions of only 0.07% at 70 weeks of age compared to 0.12% for similar bars
made with 100% PC.
Frearson and Higgins (1992) conducted research on mortar prisms to
determine the effects of GGBFS on resistance to sulfate attack. The prisms were
made with Type I cement and varying amounts of GGBFS up to 70% replacement.
Neither the grade of slag nor the fineness was specified in the report. The specimens
were cured in water for a period of 14 days. After curing the specimens were placed
and kept in a 0.31 molar solution of sodium sulfate (Na2S04) for 4 years. It was
determined from this research that as the GGBFS content increased the expansive
damage caused by sulfate attack decreased. The prism made with 70% GGBFS had a
26


total expansion less than 0.01%, compared to the control prism made of 100% PC
which had completely disintegrated after only 4 months of exposure.
Stark (1989) performed an extensive study into the effects of sulfate rich soils
on concrete. A total of 48 different concrete mixtures were tested during this research
and included three types of cement (I, II and V) as well as various percentages of
supplementary cementitious materials including GGBFS. Two different GGBFS
types were used, however the grade was not specified. The fineness of these two slags
were 5485 and 4415 cm /g (2678 and 2156 ft /lb). Three separate total cementitious
batches with contents of 376, 517 and 658 lb/yd3 (223,307 and 390 kg/m3) for each
mixture were investigated. Three beams measuring 6 inch x 6 inch x 30 inch (152 mm
x 152 mm x 762 mm) were made for each mixture and cured for 27 days in 100%
humidity and 73 F (23 C) and thereafter for the remainder of one year at 50%
humidity and 73 F (23 C) before testing began. Cylinders were also cast for testing
the 28-day strengths of the mixtures. After one year of curing the specimens were
placed outside three inches deep in sulfate rich soil for a total of 5 years and inspected
on an annual basis. Specimens were rated on a scale of 1 to 5 based on amount of
deterioration, were 5 is the worst rating and 1 indicates no deterioration. A summary
of the results including the control specimens and specimens made with GGBFS are
shown in Table 2.4.
In general the results show a significant reduction in sulfate resistance as w/c
increases. This is most probably due to the decrease in permeability as w/c increases.
All GGBFS beams with cementitious contents of 376 lb/yd (223 kg/m ) performed
poorly with the best rating of 4.6. Results indicate that conditions only slightly
improved for the GGBFS mixtures as cementitious contents increased to 517 and 658
lb/yd3 (307 and 390 kg/m3), however the w/c also decreased substantially as well.
Regardless of GGBFS replacement, all control beams made with 100% PC
(regardless of cement type) outperformed the GGBFS beams in nearly all the tests.
27


When comparing the 40% replacement to the 65% replacement GGBFS beams, the
increase in slag content decreased the resistance to sulfate attack. What is also
interesting is that when comparing the control beams made with 100% Type II PC to
the GGBFS beams made with Type II cement, the GGBFS had a negative effect. The
alumina content of both these slags was 9.40 and 6.1% which is below the maximum
recommended amount (11%) specified by ACI-233R to achieve adequate sulfate
resistance.
Tabic 2.4 Visual Results of GGBFS Beams Tested in Sulfate Rich Soil After 5
Years Exposure (Stark, 1989).
Cement Type Mix ID SLAG Fineness 376 lb/yd3 517 Ib/yd3 658 lb/yd3
w/c Strength Rating w/c Strength Rating 3 S' Strength Rating
(cm2/g) m (psi) (1-5) m (psi) (1-5) (psi) (1-5)
1 Control N/A 0.71 4030 5.0 0.46 6380 4.3 0.39 7440 1.5
11 Control N/A 0.71 3980 3.9 0.49 5640 2.9 0.38 7120 1.7
V Control N/A 0.65 4140 4.4 0.45 5420 1.8 0.37 6620 1.3
II 40% GGBFS 5485 0.67 4900 5.0 0.47 6750 3.4 0.39 7400 2.0
II 65% GGBFS 5485 0.66 4230 5.0 0.47 6180 4.0 0.37 6900 3.0
II 40% GGBFS 4415 0.71 3690 4.6 0.50 6080 3.3 0.41 7620 3.1
II 65% GGBFS 4415 0.72 3060 4.8 0.50 5290 3.7 0.40 7000 3.4
28


The author pointed out that these results are contrary to common laboratory
results regarding the sulfate resistance of GGBFS concretes, and this was attributed to
the harsher conditions of this test, albeit more realistic. Specimens were outside and
underwent constant wetting and drying cycles. This is opposite of laboratory
conditions were specimens are submerged in a sodium sulfate solution for the
duration of the test. It was also pointed out that almost all specimens were only
deteriorated above grade, and the bottoms of most beams were mostly undamaged.
T his was probably due to the more consistently moist environment below the soil (i.e.
less wetting and drying cycles). This strengthens the belief that laboratory conditions
due not clearly represent actual exposure conditions where concrete is in contact with
sulfate rich soils.
2.2.4.S Alkali-Silica Reactivity (ASR)
Studies have shown that the incorporation of GGBFS in concrete can reduce the
potential for ASR when reactive aggregates are used. In general, the causes of alkali-
silica reactions are due to the alkali content in cement and the silica in the aggregate.
T he reaction between a reactive aggregate and the alkalis from the cement can cause
an expansive gel to form around the aggregate particle. This expansion can ultimately
destroy a concrete. ASR is a complicated reaction which involves many different
factors. The general factors that can be used to control the effects of ASR in a
concrete are: control of the alkali concentrations (primarily from cement), control of
the amount of reactive silica (aggregate), control moisture penetration (permeability),
control the pH in the pore solution and alteration of the alkali-silica gel that forms
during the reaction (delay deterioration), (Mindess, 2003). GGBFS uses alkali to
hydrate and thus helps reduce the alkali content available to react with aggregates.
GGBFS also helps reduce permeability which helps reduce the ASR potential.
29


Simply finding aggregates that are not reactive is becoming more and more
difficult due to the availability of quality aggregate sources. As previously stated in
the introduction, the availability of quality aggregate sources has diminished over
time and concrete batch plants are being forced to use lesser quality aggregates.
Therefore, other techniques are being used today to eliminate the potential for ASR.
The use of pozzolans is a common method employed to reduce the potential for ASR
in concrete. According to ASTM-C989, concretes made with GGBFS replacements
greater than 40% showed reduced expansions due to alkali-silica reaction when used
with cements having alkali contents up to 1.0%. According to ACI-233R,
replacements of 50% or more with GGBFS have been effective in reducing the
potential for alkali-silica reactions when used with high alkali cement and reactive
aggregates.
Detwiler (2003) studied the effectiveness of using GGBFS in mitigating the
expansive reactions caused by ASR. The purpose of the research was to investigate
the effectiveness that pozzolans have on three particular factors that are believed to be
connected with use of pozzolans in controlling ASR. These factors included; dilution
of the alkalis in the cementitious paste, reduction of permeability and binding of the
alkalis. A moderately reactive aggregate was used as well as two different Type I
cements having alkali contents of 0.46 and 0.92%. Two GGBFS specimens were
made for each test and consisted of 35 and 50% replacements. Multiple specimens
were also made with fly ashes having varying CaO contents. Control specimens were
also made with 100% PC. Neither the grade nor the fineness of the slag was specified.
Mortar bars were prepared and tested according to ASTM-C1260 for 14 day
expansions. In the ASTM-C1260 test, an abundant supply of alkalis is available from
the solution in which specimens are submerged. Concrete prisms were also fabricated
and tested according to ASTM-C1293 for 2 year expansions. In the ASTM-C1293
test, only the alkalis from the cement paste are available for the duration of the test
30


(i.e. not submerged in alkali rich solution). The cements with alkali contents of 0.46
and 0.98% were used for the Cl293 and Cl260 tests respectively. The results are
shown in Table 2.5.
Table 2.5 Results of ASR Expansion Tests on GGBFS Specimens Tested
According to ASTM-C1260 and ASTM-C1293, (Detwiler, 2003).
ASTM-C1260 Expansion at 14-days, % ASTM-C1293 Expansion At 2-years, %
Mixture ID Mixture ID
Failure Criteria > 0.10% Failure Criteria > 0.04%
Control (0.98) 0.25 Control (0.46) 0.14
35% GGBFS 0.17 35% GGBFS 0.06
50% GGBFS 0.06 50% GGBFS 0.03
15% Low-CaO F.A. 0.09 15% Low-CaO F.A. 0.06
25% Low-CaO F.A. 0.03 25% Low-CaO F.A. 0.04
15% Med-CaO F.A. 0.17 15% Med-CaO F.A. 0.12
25% Med-CaO F.A. 0.16 25% Med-CaO F.A. 0.07
15% High-CaO F.A. 0.20 15% High-CaO F.A. 0.20
25% High-CaO F.A. 0.18 25% High-CaO F.A. 0.15
As can be seen by observing the results, GGBFS replacements of 35%
decreased the expansions during both procedures; however it was not enough to
achieve passing scores. For both the tests performed, the 35% GGBFS specimen did
not meet the passing criteria. However, only two specimens passed the ASTM-C1260
31


which included the 50% GGBFS specimen. The 50% GGBFS specimen
outperformed all specimens for the ASTM-C1293 test with an expansion of only
0.03%. Although not achieving a passing score, it should be noted that the specimen
with only 35% GGBFS performed better than all specimens except the 25% Tow-
CaO F.A. during the ASTMC1293 test.
Higgens and McLellan (2009) performed extensive studies over a period of
ten years to study the effectiveness of GGBFS in reducing ASR expansion. The
concrete mixtures were made such that the total reactive alkali contents varied from
0.31,0.37, 0.44 and 0.50 lb/ft3 (4.96, 5.92 and 7.05 kg/m3) using three types of
cement and two types of GGBFS. Several hundred concrete prisms were made with
varying amounts of GGBFS from 0-70% replacements. Natural aggregates were used
that were known to have caused ASR expansion in structures. The cement types used
were designated low, medium and high for alkali contents of 1.15, 0.87 and 0.54%
respectively. The testing was performed in the UK and methods as well as standards
varied from ASTM Standards and Methods. Whether or not the UK has a standard
method of rating GGBFS is unknown, however neither the grade nor the fineness of
the slag used was specified. The alkali contents for the two types of slag designated
low and high were 0.58 and 0.83% respectively. Two different temperatures were
used during testing, 68 F and 100 F (20 C and 37.8 C). The higher temperature
was chosen to investigate whether accelerated expansions would take place and more
importantly, whether the results from both tests would correlate. The specimens were
tested for a total of ten years according to British Standard Institution Procedure DD
218:1995.
The results if this study showed that the accelerated curing temperatures did
increase the expansion rates by as much as 3.0 %. The results for the accelerated tests
also correlated well with those from the standard test method. The author found that
in general as the amount of alkali increased, the expansions increased and as GGBFS
32


content increased, the expansions decreased. What is also interesting is that none of
the specimens made with 70% GGBFS expanded significantly, even with total alkali
-i
contents up to 0.6 lb/ft For the specimens made with 50% GGBFS, expansions
greater than 10% were only examined when the total alkali contents were far in
excess of what would be found in the field. They also determined that half of the
alkali content is contributed by the slag itself only when used in replacement
percentages up to 25%, and beyond this the alkali contents of the slag no longer
become a factor.
2.2.5 Summary
T his literature review showed that there can be both positive and negative effects
from the use of GGBFS in a concrete mixture. There was evidence that higher
dosages of air-entraining admixture and HRWRA may be needed if GGBFS is used.
This dosage increase appears to be related to the fineness of the GGBFS and with
increasing fineness, one can expect to see an increasing demand on admixture dosage.
Temperatures are expected to decrease in a concrete mixture and the degree will
depend on the amount of GGBFS used as replacement. The workability of concretes
made with GGBFS is expected to increase with increasing slag content. This increase
in workability will also be related to the grade and fineness of slag used and w ill
decrease as these attributes increase respectively. Concretes made with GGBFS will
typically show higher later age strengths than earlier age when compared with control
mixtures made of 100% PC. The higher grading of slag used will tend to reflect
higher strengths and strength gains due to reactivity. The optimum replacement levels
for the greatest strengths tend to be around 50% for GGBFS. Typically, strengths
have shown an increase up to 50% and decreasing or similar strengths thereafter
when compared to control mixtures.
33


The permeability of concrete will decrease with increasing slag content. This
decrease in permeability is associated with several other increases in durability for
GGBFS concretes. The use of GGBFS will typically increase sulfate resistance and
may or may not have an effect on the freeze-thaw durability. The increase in sulfate
resistance will increase with increasing slag content but may exhibit conflicting
results in actual field conditions where sulfate rich soils are encountered. The
resistance to freeze-thaw will typically decrease with replacement levels in excess of
50%. The resistance to ASR expansion will most probably increase with increasing
slag content when compared with concretes made of 100% PC and similar
aggregates. Significant increases in resistance to ASR expansion have been shown at
typically higher replacement levels of 50% or more.
2.3 Waste Glass as Aggregate
The majority of crushed glass (approximately 80%) that exists in the recycling stream
comes from bottles. These bottles will primarily be from beer and soda but may also
come from other food sources. The recycling of glass is difficult and costly due to the
usual need of separating the glass colors before recycling. This will depend on end
use though as the glass will maintain its color after recycling. Not all glass can be
recycled due to impurities and contaminants which are difficult to remove and
because frequently the bottles are broken or damaged which makes re-use impossible.
Difficulties in achieving the cullet (crushed waste glass) specifications required by
many bottle manufacturers have resulted in a low market value for crushed glass and
the bulk of crushed glass ends up in landfills. In 2007 an approximated 13.6 million
tons of waste glass was generated and only 24% of this was recycled (Verdugo,
2009). The excessive shipping costs for bulk glass to manufacturing plants are also
prohibitive due to the high density of glass.
34


The use of glass in concrete is gaining interest because of the potential bulk
use as a fine aggregate. However, many earlier studies on the use of glass in concrete
have shown that poor quality concrete is a result due to alkali-silica reactivity
between the glass particles and the cement paste. This trend is changing as more ways
to mitigate the ASR potential are becoming known. Due to the taboo of using glass in
concrete, the research is limited and allot of information on the subject aside from
ASR studies is minimal. Some newer research into glass as aggregate has found that
pulverized glass (powdered glass) will actually mitigate potential ASR and may also
act as a pozzolanic material.
2.3.1 Production
Waste glass comes from glass recycling plants which collect the glass from
households and commercial facilities. The glass is stockpiled at the recycling facility
and may or may not be separated based on color. The separation of glass with regards
to color is done because glass will maintain its color after recycling and depending on
end use this separation may be required. The color of the glass will also dictate the
chemical composition because of the chemicals added during original production to
alter the color. This separation of glass at the recycling plant may also be required
because of this chemical incompatibility of different colors. The primary colors that
make up the majority of the glass in the recycling mainstream are brown (amber),
green and clear (transparent).
The typical process of recycling bottles involves multiple crushers which
decrease the size of the glass particles more after each crusher. The glass moves along
a conveyer belt from one crusher to the next while vacuums and magnets remove any
labels or other foreign objects. After all crushing has taken place the glass drops into
a hopper and is ready for the furnace where it will be melted down. At this stage the
35


crushed glass is called cullet. Washing of the cullet may or may not take place
prior to entering the furnace and depends on the plant operation.
2.3.2 Physical and Chemical Properties
Since most of the glass that makes up the recycling stream is composed of containers
such as beer and other carbonated beverages, the chemical and physical composition
of these glasses will be discussed. Glass is an amorphous (non-crystalline) solid
structured material. The type of glass used to make common beverage containers is
called soda-lime glass and is primarily made from quartz with the addition of other
materials to help the fusing process. The term soda-lime comes from the addition of
lime (CaO) during the manufacturing process. Soda lime glass has a smooth surface
and is non-porous in nature (non-water soluble). This smooth surface raises concern
with regards to strength since the weakest link in the concrete matrix is the bond
between aggregate and cementitious paste. Replacing a natural sand aggregate which
has a rough surface with a smooth surfaced glass may result in decreased strength at
the interfacial transition zone (ITZ).
Ordinary soda-lime glass will be colorless without the addition of chemical
compounds to change the tint. FeO and O2O3 are added to the glass to make the
green colored bottles. Sulfur, carbon and iron salts are added to the glass if amber or
brown colored bottles are made (www.glassproperties.com, 2011). As will be
discussed later, the color of the glass is believed to have significant impacts on the
ASR potential due to the variation in chemical composition. Table 2.6 shows the
typical molecular compounds and percentages that make up soda-lime glass.
36


Table 2.6 Typical Chemical Compositions Percentages of Soda-Lime Container
Glass, (Weihua Jin, 2000)
Compound Clear Glass Amber Glass Green Glass
(%) (%) (%)
Si02 73.2-73.5 71.9-72.4 71.27
A1203 1.7- 1.9 1 1 bo 2.22
CaO + MgO 10.7-10.8 11.6 12.17
Na20 + K20 13.6-14.1 15.8-14.4 13.06
S03 0.20-0.24 0.12-0.14 0.052
Fe202 0.04-0.05 0.30 0.599
C^Ot 0.01 0.43
Aside from the properties associated with the glass manufacturing itself,
recycled waste glass will also contain other impurities on the surface that are
associated with the contents the glass container once held. Sugars and other organic
contaminants may be present. Washing of the waste glass will typically remove most
of these contaminants but this process can be costly. Other contaminants such as
remnants of the paper labels and various metal objects (i.e. aluminum, bottle caps,
etc.) not picked up by the magnets and vacuums during the crushing process can also
be found. If aluminum remaining from labels or other sources is not removed the
reaction between the alkalis and the aluminum will produce hydrogen bubbles. For
this research project the glass was used as received from the recycling plant
(unwashed).
The average specific gravity for soda-lime glass is 2.52, (en.wikipedia.org,
2011). The absorption capacity of the glass itself should be relatively 0.00%.
However, due to the contaminants within the waste glass stream, especially small
paper label fragments, will actually increase the absorption capacity slightly from
37


zero (C. Meyer, 2001). The fineness modulus, particle size and gradations of waste
glass will ultimately depend on the recycling plants crushing technique and will vary.
2.3.3 The Effects of Waste Glass on Fresh Concrete Properties
2.3.3.1 Slump
Polley (1996) performed extensive studies to determine the effects of waste glass on
fresh properties of concretes made with varying amounts of waste glass as aggregate.
The study included various replacement amounts of both coarse and fine aggregate
with waste glass. Waste glass replacement levels ranged from 0 (control) to 90% of
total aggregate volume. They investigated the required amount of mixing water
necessary to achieve a target slump of 50 mm (2.0 inch). Therefore the w/c was
varied throughout this research project. These concrete mixtures also had varying
amounts of Class F fly ash used as supplementary cementitious material ranging from
20 to 30%.
The results showed a linear trend of increasing water content as total glass
content increased. What was also observed and noted in the report was the
segregation of the waste glass with the cement paste. This was most notable with the
coarse glass aggregate. A poor bond between the glass and cement paste existed and
the angularity of the glass also contributed to the poor workability. The concretes
were difficult to work with and finish properly (actual sidewalk slabs were placed for
this research). They also core drilled these slabs at a later date and noted that the
consolidation at the bottoms was typically poor.
Ahmad Shayan and Aimin Xu (2005) examined the fresh concrete properties
of several concrete mixtures composed of varying amounts of waste glass as fine
aggregate. Replacement levels of 40, 50 and 75% were investigated for this research.
38


A constant w/c value as well as total cementitious content was used. Glass powder
was used as a supplementary cementitious material on several mixtures. Table 2.7
below summarizes the mixture properties and fresh properties for several selected
mixtures. The CGS in the mixture identification signifies crushed glass sand, or waste
glass used as fine aggregate and the replacement level used. It was indicated that
HRWRA was used on some mixtures but the documentation was vague regarding
which mixtures, as well as dosage rate.
Table 2.7 Mixture Properties and Slump Results of Concretes Containing Waste
Glass as Fine Aggregate, (Shayan et al, 2005).
Mixture II) Cement (Ih/yd3) Glass Powder (Ih/yd*) w/c Slump (inch)
#1 Control 640 0 0.49 2.75
#7 40% CGS 448 192 0.49 2.00
m 50% CGS-1 512 128 0.49 3.15
#9 50% CGS-2 640 0 0.49 2.75
#8 75% CGS 448 192 0.49 2.00
These results are interesting because it does not appear that the glass had any
effect on the workability of these mixtures. For instance, when comparing Mixture #1
(control) with Mixture #9 the slumps are identical. Both mixtures had the same
cementitious and only varied in waste glass replacement which was 50% for mixture
#9. The same similarities can be observed when comparing Mixture #7 with Mixture
#8. The replacement of fine aggregate with waste glass between these mixtures is
39


35%, but the slumps were identical. A relationship can be seen between glass powder
and increase in slump when observing Mixture #6 and Mixture #9. As mentioned
previously, the usage of HRWRA was specified but the information was vague and it
is assumed that all mixtures contained some amount of HRWRA. The author did
point out that the workability of the waste glass mixtures was harsh and that the
harshness increased as glass content increased.
2.3.3.2 Air Content
The air contents of concretes made with crushed glass as aggregate do not show any
significant differences when compared with natural aggregate concretes based on the
studies investigated. Polley (1996) investigated multiple replacement levels of natural
coarse and fine aggregates with crushed waste glass and found only a small difference
in AEA dosage requirements. However, the small additional AEA dosage that was
required on mixtures that had replacements of aggregate with waste glass also
contained fly ash and the increase was associated mainly with the fly ash, not the
waste glass. There was a difference in required AEA dosage for the glass powder
mixtures once the replacement level exceeded 15%. However, powdered glass is not
considered an aggregate as the fineness can be well in excess of even cement (8000
cm2/g (3906 ft2/lb.
40


2.3.4 The Effects of Waste Glass on Hardened Concrete Properties
2.3.4.! Strength
Polley (1996) performed extensive studies to determine the effects of waste glass on
the strength of concretes made with varying amounts of waste glass as aggregate. The
glass used was a combination of various colors directly from the recycling plant. The
study included various replacement amounts of both coarse and fine aggregate with
waste glass. The w/c values ranged considerably throughout this study due to the
increased water demand as glass content increased. A target slump of 50 mm (2.0
inch) was specified and the additional water added to achieve this workability
sometimes caused a useless mixture. These concrete mixtures also had varying
amounts of Class F fly ash used as supplementary cementitious material ranging from
20 to 30%. A low alkali cement and a moderate alkali cement were used.
The results of these studies are too extensive to tabulate and therefore only the
key findings will be discussed. The results of this study showed that as glass content
increased the strengths decreased. This relationship was most pronounced with the
coarse glass aggregate concretes and was attributed to the plane of weakness
associated with the larger glass particles. The glass particles are also not as strong as
natural aggregates and are susceptible to fracture which may cause lower strengths as
well. The effects of reduced strength decreased as the fineness of the glass increased.
The low alkali cement mixtures exhibited greater strengths than the moderate alkali
mixtures which suggest that ASR was a contributing factor to strengths. A reduction
in strength loss from 20% to only 5.0% when using low alkali cement as opposed to
moderate alkali cement was realized. They concluded that an optimum replacement of
20 to 24% and limiting the glass to only fine aggregates would limit the decrease in
strengths to only 5.0% when compared with a similar control mixture. Another
substantial observation made was the differences in strength when comparing the
41


unwashed glass to the washed glass concrete mixtures. They investigated similar
mixtures that used the same replacement amounts of glass and either used the glass
as received or washed the glass prior to usage. For mixtures containing 40% coarse
glass aggregates the difference in strengths is 20%. For mixtures containing 25%
glass as fine aggregate the difference in strengths are 40% at 28-days of age and 44%
at 56-days of age. This reduction in strength was attributed to the presence of sugars
and other contaminants remaining on the glass.
M. Mageswari et al (2010) investigated the strengths of concretes containing
various replacement amounts of natural sand fine aggregate with waste glass fine
aggregate. This waste glass was composed entirely of sheet glass scraps collected
from local shops near the area were the research was conducted. The sheet glass was
crushed and graded similar to the fine aggregate it was to replace. The glass also
consisted of multiple colors and was not separated prior to usage. To investigate the
effects on concrete strength, replacement levels of 10, 20... and up to 100% of the fine
aggregate was used. Their results indicated that the optimum replacement levels were
between 10 to 20% of the fine aggregate. All mixtures showed an increase in strength
when compared to the control mixture for the earlier age groups (up to 90-days).
However, at 180 days of age all mixtures showed decreased strengths when compared
to the control specimens. Although the concretes were not tested for this reaction, the
strength reductions were attributed to ASR.
Ahmad Shayan and Aimin Xu (2005) examined the effects of glass aggregate
on concrete strengths for various replacement ranges of fine aggregate. They also
investigated the effects of glass powder (fineness of 8000 cm2/g (3906 fT/lb)) on the
strengths of concrete. Both the powdered glass and glass fine aggregate came from
soda-lime waste glass. The mixtures were designed and later compared to a 5800 psi
(40 MPa) 28-day structural concrete mixture. Specimens were tested at 7, 28, 90 and
404-days of age. Table 2.8 shows several key mixture designs as well as 28 and 404-
42


day strengths. The CGS in the mixture identification refers to crushed glass sand as
replacement of natural sand. Some mixtures had glass powder as a supplementary
cementitious as well. It should be noted that the compressive strengths were
approximated from bar chart as no tabulation of values was provided. Therefore,
minor discrepancies in data may exist.
Table 2.8 Mixture Properties and Compressive Strengths of Concretes
Containing Waste Glass as Fine Aggregate, (Shayan et al, 2005).
Mixture ID Cement (Ib/yd3) Glass Powder (Ib/yd*) w/c Compressive Strength (lb/in2)
28-Day 404-Day
#1 Control 640 0 0.49 8412 11,168
#7 40% CGS 448 192 0.49 5076 8557
m 50% CGS-1 512 128 0.49 5221 8412
#9 50% CGS-2 640 0 0.49 6237 8992
#8 75% CGS 448 192 0.49 4641 8267
When observing the compressive strengths of the waste glass concretes, only
Mixture #9 achieved the required strength of 5800 psi (40 MPa) at 28-days of age. All
mixtures achieved the required strength by 404 days. The slowest strength gain came
from the 75% CGS mixture but this mixture also had a 30% cementitious replacement
with glass powder. Although all mixtures achieved significant results, the decrease in
strength due to waste glass is evident when compared with the control mixture. When
comparing the control mixture (Mixture #1) with Mixture #9 which both had 100%
43


PC as cementitious, the reduction in strength is over 2000 psi (13.8 MPa) for both age
groups. The inclusion of the glass powder also had significant effects on strengths,
but the reduction in PC used may well be worth this reduction. Unfortunately
information regarding the waste glass was vague at best. No information was given as
to whether or not the glass was processed further before mixing or whether or not any
washing took place.
2.3.4.2 Permeability
Ahmad Shayan and Aimin Xu (2005) investigated the effects of crushed glass
aggregate and powdered glass cementitious replacements on the permeability of
hardened concrete at 380-days of age. The testing was performed according to ASTM
C 1202 procedures. The replacements of fine aggregate with waste glass were 40, 50
and 75% of the total fine aggregate content. Some mixtures also incorporated
powdered glass as supplementary cementitious. The w/c was held constant as well as
the total cementitious content. This study involved actual field testing in which slabs
were placed and finished outside at a test facility. Cylinders were cored from the slab
at the specified time of testing. When the slabs were constructed separate cylinders
were also cast from the same batch and cured in the laboratory for comparison with
field specimens. Table 2.9 summarizes the results from several selected tests of both
the lab cured specimens as well as the field cured specimens for comparison. These
results were scaled from bar graphs and therefore small discrepancies may exist.
44


Tabic 2.9 Mixture Properties and RCPT Result of Concretes Containing Waste
Glass as Fine Aggregate at 380 Days of Age, (Shayan et al, 2005).
Mixture ID Glass Powder (Ih/yd?) Field Charge Passed (Coulombs) Permeability Class Laboratory Charge Passed (Coulombs) Permeability Class
#1 Control 0 2400 Moderate 2450 Moderate
#7 40% CGS 192 1200 Low 800 Very' Low
#6 50% CGS-1 128 850 Very Low 1000 Very Low
#9 50% CGS-2 0 3400 Moderate 2600 Moderate
#8 75% CGS 192 650 Very Low 600 Very Low
The results show an increase in permeability as waste glass increases. When
comparing field specimens for Mixture #1 with Mixture #9 the difference is 1000
Coulombs. Mixture #9 had 50% waste glass as fine aggregate and did not have
powdered glass as cementitious. The laboratory specimens for these two mixtures
achieved conflicting results for Mixture #6 with a difference of 800 Coulombs. The
exact opposite of this trend was observed for mixtures #7 and #8. Mixture #7 had
40% glass content while Mixture #8 had 75% glass as aggregate. Mixture #7 with the
lower amount of glass aggregate had significantly higher laboratory results when
compared with Mixture #8. However, the field cured specimens both achieved similar
results and a decrease in results was observed. Both of these mixtures as well as
Mixture #6 did have powdered glass as cementitious added and this may have been
the cause of the lower permeabilitys. When comparing Mixture #6 and Mixture #9
the effects of powdered glass on permeability can readily be seen.
45


2.3.4.3 Freeze-Thaw Durability
Policy (1996) investigated several different glass aggregate concrete mixtures for
free-thaw durability. Most studies on glass aggregate are related to strength and ASR
and therefore this was the only study that could be found on this subject. Several
beams were fabricated with different glass contents and tested for 600 cycles of
freezing and thawing. For this experiment, several mixtures with coarse aggregate
replacements and fine aggregate replacements were investigated. The replacement
amounts ranged from 12% to 36% for the coarse plus fine and 20% to 24% for the
fine only. The degradation in durability factors for the coarse glass aggregate beams
is more rapid than the fine glass aggregate beams. There is a correlation between
increasing degradation rate and increasing glass content for the coarse plus fine glass
replacement beams. This correlation does not exist for the fine glass only beams. All
beams that had only fine glass replacements had durability factors greater than 85%.
The coarse plus fine glass beams did not favor as well and several beams fell below
80%. The author pointed out that all beams that failed exhibited strong performance
up to about 70 cycles before an alarming drop in durability factor occurred. Another
interesting point made is that all beams (even the control) showed a decrease in
durability factor for the first 10 cycles by around 6 to 7%. The mass loss observed for
the coarse plus fine glass mixtures showed very similar results between 2.5 and 3.5%
loss at 350 cycles.
The mass loss for the fine glass replacement beams showed more variations
and were not as tightly grouped. These specimens performed worse than the
combination specimens with mass losses between 2.5 and 6.5%. The severity of the
mass loss for both the fine and fine plus coarse glass replacement specimens was also
correlated with the amount of replacements. As glass content increased the mass loss
increased.
46


2.3.4.4 Alkali-Silica Reactivity (ASR)
The alkali-silica reactivity between glass aggregate and cement is well documented.
Many studies have shown that when crushed glass is used as an aggregate
replacement, ASR is almost certain. However, more recent studies have also shown
there are ways to successfully use glass in a concrete if proper techniques are
employed. Mitigation techniques such as the use of pozzolans and even the type or
color of glass has proven to be successful in reducing expansions caused by ASR to
negligible levels.
In general, the causes of alkali-silica reactions are due to the alkali content in
cement and the silica in the aggregate. The reaction between crushed glass and the
alkalis from the cement can cause an expansive gel to form around the glass particle.
This reaction can cause excessive internal forces and may ultimately destroy a
concrete. ASR is a complicated reaction which involves many different factors. The
general factors that can be used to control the effects of ASR in a concrete are:
control of the alkali concentrations (primarily from cement), control of the amount of
reactive silica (aggregate), control moisture penetration (permeability), control the pH
in the pore solution and alteration of the alkali-silica gel that forms during the
reaction (delay deterioration), (Mindess, 2003).
Most studies found that tested the potential for ASR in glass aggregate
concretes followed the ASTM C 1260 procedures. This procedure is convenient
because it lasts for only 2 weeks. However, this test as opposed to ASTM C 227,
involves submergence of the specimens in a highly alkaline NaOH solution at
elevated temperatures and is not representative of actual field conditions. This test has
been shown to be highly reliable however and is the most common used (Weihua Jin,
2000). Two reports on ASR testing conducted with concretes containing glass will be
discussed later. Most research on glass aggregates and potential ASR have come to
47


the following similar conclusions. Where reactive glass was used, the particle size
appears to have a direct influence on the ASR. This is because the ASR process
involves the reaction between a solid (aggregate) and a solution (cement paste) and
the surface area of the aggregate will play a major role in reaction rate. There is also a
pessimum size that will cause the most significant expansions. Most past studies that
found this pessimum effect theorized that it was related to the variation in alkali to
silica ratio. They argued that the pessimum behavior was associated with a pessimum
gel ratio. However, these tests were based on ASTM C 227 which has a fixed supply
of alkalis during the procedure.
There are two processes involved with ASR, gel formation and gel
permeation. These two processes are both related to aggregate size. Gel formation
will build up internal stresses within the concrete matrix, while gel permeation will
tend to relieve these pressures. This is somewhat similar to the freeze-thaw process.
Ice crystal formation causes excessive stresses in confinement, but where air voids
are present the ice crystals will permeate to these areas and relieve stress. Gel
formation predominates the coarser aggregate domain while gel permeation
predominates the finer domain. This pessimum point is when there exists a balance
between these two and the maximum expansion will take place. It has also been
shown that decreasing the size of the glass particles to below a standard U.S. #50
sieve (i.e. passing) will eliminate most ASR potential.
Weihua Jin et al (2000) performed various studies into the effects of ASR on
glass aggregate concretes. They first studied the effects of particle size on reactivity.
They used a clear soda-lime glass for this portion of study and separated the glass into
representative sizes corresponding to U.S. sieve sizes. They replaced 10% of the
natural sand in the concrete with only one size for each mixture. Mortar bars tested
according to ASTM Cl260 indicated that the maximum, or pessimum size was the
#16 sieve size (1.18 to 2.36 mm). What was also interesting is that for mortar bars
48


made with size #50 glass the expansions were identical to the reference bars made
from only cement and natural aggregates. Mixtures made with glass particles equal to
and smaller than #100 sieve actually showed less expansions than reference bars.
The second phase of testing involved the study of glass content on expansion
rates. They made various mixtures with glass contents ranging from 0 to 100% of the
fine aggregate. These mixtures were made with glass matching the gradations of the
control specimens. A clear correlation was made between expansion rates and glass
content. As glass content increased, so did the expansion rates. The relationship was
rather linear as well.
The third phase of this study involved the investigation of glass color on ASR
potential. Three separate groups of mortar bars were prepared for this investigation.
Each group consisted of 10% replacement amounts of only one color of glass. The
three groups consisted of green, brown (amber) and clear glass (previously
investigated during phase I). The results showed that clear glass has the greatest
expansions. Amber glass was shown to be considerably less reactive followed by
green glass which showed non-reactive behavior and when compared with reference
bars, showed less expansions. The pessimum size for the amber glass was found to be
#8 size and although not reactive comparatively with the reference bars, #16 for the
green glass.
They concluded that glass color and particle size has an effect on reactivity.
They confirmed that a pessimum size exists and that this size depends on color and
glass type but will shift toward smaller particles the more reactive it is. They
concluded that green glass caused no expansions to speak of and that finely ground
green glass can be a relatively low cost alternative to ASR mitigation in concretes
containing reactive aggregates. They indicated this was caused by the O2O3 content
in the green glass.
49


Amhad Shay an (2002) studied the effects of crushed glass and powdered glass
(8000 cm2/g (3906 ft /lb)) on the ASR potential of concretes made with crushed glass
as fine and coarse aggregate replacements. The coarse aggregate particle size range
was 12 mm to 4.75 mm (1/2 inch to 3/16 inch). The fine aggregate particle size range
was 4.75 mm to 0.15 mm (1/2 inch to 1/200 inch). The waste glass used was a mixed
colored soda-lime glass cullet from a local bottle recycling plant. The powdered glass
was made from the cullet by further grinding. The Blaine fineness of the powdered
glass was 8000 cm2/g (3906 ft2/lb). The testing procedure followed was ASTM C
1260 for 28-days.
They found similar results as Jin with regards to glass content and expansion.
As the percentage of glass content increased the expansions increased linearly. They
also found that replacements up to 30% of total aggregate were non-detrimental and
expansions were insignificant when combined with low alkali cement. When
investigating the gradation effects on expansion, they found direct correlations
between particle size and expansion. The results show that expansions are negligible
for particle sizes below 0.30 mm and that the coarse aggregates (greater than 0.60
mm) showed the greatest expansion rates. Furthermore, it was concluded that as the
particle size decreases the ASR potential not only decreases but the fine glass
particles will actually help suppress ASR when combined with the larger coarse glass
particles. Powdered glass was also found to be a good suppressant of ASR expansions
when used at cement replacement amounts of 20 to 30%.
2.3.5 Summary
Most studies on the effects of waste glass on concrete properties were limited to ASR
research. This topic is the most widely investigated because of the almost certain
problem that arises when using a known highly reactive aggregate. The use of waste
50


glass as aggregate is becoming more excepted if techniques are employed to suppress
the ASR. It was shown that particle size does have a pronounced effect as well as
glass color on the ASR expansion rate. However, most waste glass is not sorted and
the use of only green glass (which is non-reactive) is not a practical approach.
Reducing the size of the waste glass to below a #50 sieve appears to eliminate the
ASR potential, but at the same time is no longer an aggregate but more of a
cementitious material or powder. Incorporating glass fines does tend to decrease the
ASR potential wrhen used in combination with fine aggregate waste glass. The
existence of a pessimum size of glass particles appears to be well founded and hovers
around the #16 sieve size. The pessimum effect does not appear to be present when
relating to content. As glass replacements increased the potential for ASR expansion
also increased when mixed colored glass was used.
The absorption of the glass will be relatively zero, but the effects on
workability will increase as content increases. Typically most mixtures made with
glass as aggregate exhibited harsh characteristics and these increased as content
increased. The effects of AEA will probably remain the same or slightly increase for
concretes made with waste glass aggregate. The strength of the concrete will probably
decrease with increasing waste glass content. This is due to the relatively low strength
of glass when compared to natural sand aggregate. The bond between the glass
particle and cement paste will not be as strong because of the glassy smooth structure.
Whether or not the glass has been washed will also have major effects on concretes
strength. A reduction in strength of up to 40% was observed when comparing washed
to unwashed glass.
The permeability and freeze-thaw durability of concretes made with glass is
not well understood or studied. Few studies could be found on these subjects. Of
those found the general results showed that freeze-thaw durabilitys may decrease
with increasing content. This decrease will also be more pronounced with increasing
51


glass size (i.e. coarse as opposed to fine). Once again, this can probably be related to
the decreased strength and bond between glass particles and cement paste. The
permeability of concretes using waste glass as aggregate will probably increase as
replacement increases. This increase will depend on replacement amount and also on
the fineness of the glass. The use of powdered glass actually decreases the
permeability of concrete but this is more of a cementitious material and not an
aggregate.
2.4 Recycled Concrete as Aggregate (RCA)
2.4.1 Production
The recycled concrete used as aggregates in concrete comes from buildings, roads,
bridges and other civil structures that have been demolished. Only non-contaminated
concrete is accepted at plants and cannot contain any trash, wood or other materials.
In the past, most recycling plants would not accept concrete with reinforcing however
nowadays these concretes are being accepted more and more. Any concrete that is
suspected of having alkali-silica reactivity or other deleterious mechanisms will also
not be accepted.
Typically the concrete is crushed to a reasonable size at the construction or
demolition site before being transported to the recycling plant. There are portable
crushers that can crush the concrete down into gravel at location and these are
becoming more common as opposed to hauling large chunks to the recycling plant. At
the recycling plant the concrete will be crushed further by primary and secondary
crushers and filtered to remove contaminants and reinforcing steel. The reinforcing is
removed by large magnets. The crushed concrete is then graded and washed. Most
recycling plants will separate the crushed concrete into different piles which
52


correspond to a nominal maximum aggregate size (NMAS). Any contaminants
removed including steel from reinforcing will then be sent to other recycling plants or
landfills.
RCA usage has increased over the years due to increased savings in mixture
design as well as an increase in studies and data supporting the use of RCA as a
respectable aggregate. Most commonly RCA usage is primarily in roadways but
usage in buildings is growing (PCA, 2002). Another new approach to recycling
concrete and reusing as aggregate will also be examined. This newer approach
involves reusing the concrete that gets returned to batch plants after a job has
finished. This concrete is crushed and stockpiled similar to RCA but is called crushed
concrete aggregate (CCA). The differences between these two aggregates are greater
control of properties due to the known properties of the source concrete, and also in
the lack of potential contaminants in CCA. CCA has never been used in the field and
chemical contamination from chlorides will also not be a concern.
2.4.2 Physical and Chemical Properties
The physical and chemical properties of recycled concrete aggregate (RCA) will vary
greatly depending on the source of the demolished concrete. These properties will be
directly related to the original aggregate used and also on the cementitious materials.
Crushed concrete will not only contain the coarse aggregate but also chunks of motor
containing the fine aggregate and cementitious paste. This paste will also surround
portions of the coarse aggregate in varying amounts. Usually the NMAS will be
consistent with the recycling plants designation when tested separately. Even if the
concrete being recycled does not show signs of distress from alkali-silica reactivity, it
should be tested because the aggregate may have been reactive but low alkali cement
or other measures may have been used. Chloride content should also be tested
53


because of the high volumes of recycled concrete that come from roadways and the
chlorides used to melt snow may be present. ACI 555R (2001) has developed
requirements pertaining to the amounts of deleterious materials acceptable in coarse
and fine RCA depending on type of deleterious materials.
The unit weights and specific gravity will also vary greatly depending on the
source but are typically lower than a normal virgin aggregate of similar properties.
This is caused by the hydrated cement mortar still attached to the aggregates as the
mortar will typically have lower density than the source aggregate (ACI 555R, 2001).
The absorption capacity of recycled concrete aggregate will usually always be higher
due to the increased porosity of the mortar chunks and cementitious paste surrounding
the aggregate. Absorption values will typically range between 3 and 10% and will
increase as NMAS decreases (PCA, 2002). The shape of RCA will typically be
angular and resemble crushed rock. This will also cause workability concerns as
opposed to round natural stone.
2.4.3 The Effects of RCA on Fresh Concrete Properties
2.4.3.1 Slump
Gholamreza Fathifazl et al (2009) examined the effects of RCA on the fresh
properties of concrete. For this particular study a new mixture design was developed
to incorporate the existing mortar content within the RCA. As previously stated, the
RCA will contain chunks of mortar as well as cementitious paste surrounding the
aggregates. This new method accounts for this preexisting mortar in the design
process and subtracts this amount from the new cementitious quantities. Several
mixtures were developed from varying RCA sources and compared with a similar
control mixture composed of normal virgin aggregates and PC. A total of six RCA
54


mixtures were developed and tested. Five of these six mixtures contained varying
amounts of water reducing admixtures. All RCA mixtures had significantly less
slumps than the control mixture, even though the control mixture did not have
HRWRA. This was attributed to the reduction in new mortar caused by this new
design method.
The U.S. Department of Transportation, Federal Highway Administration
(FHWA) performed an extensive survey of all state transportation agencies to
investigate the usage of RCA in roadway construction (U.S. Department of
Transportation, FHWA, 2004) They found that although 41 states currently recycle
aggregate from demolished roadways, only 11 of these states reuse the RCA in new
roadways. They recommend that RCA must be treated as a lightweight aggregate and
RCA requires special attention in order to achieve the same workability as a normal
PC concrete. This is due in part to the high absorption capacity and also because of
the coarseness of the RCA. They estimate that concretes made with RCA will require
on average 5% more water and if the percentage of fines increases as much as 15%
more water may be needed. They also recommend that when RCA is used it must be
in saturated surface dry (SSD) condition prior to batching and that RCA stockpiles be
constantly wetted with water. They mentioned that several state agencies have
reported decreased workability in concretes made with RCA but this was attributed to
the lack of preparation and increased fines content. Most state agencies that use RCA
have limited the amount of fines to 20% due to concerns with workability.
Mirjana Malesev et al (2010) investigated the fresh concrete properties of
concrete made with RCA. The RCA replacement was investigated at 50 and 100% of
the total coarse aggregate content. Natural sand was used for the fine aggregate on
both these mixtures and for the 50% RCA mixture, the remaining 50% of coarse
aggregate was a natural aggregate. A control mixture was also made with the same
natural sand as the RCA mixtures and the same natural coarse aggregate as the 50%
55


RCA mixture. A Type-II PC was used for all mixtures and the w/c was started at
0.514 and varied depending on mixture.
It is unclear which design method was used when designing these mixtures. It
is assumed these mixtures were designed based on SSD and the additional water is
due to the moisture contents and absorption capacity of the RCA. Therefore it is
assumed this extra water reported is actually required to achieve the same w/c when
comparing the control mixture with the RCA mixtures. The average absorption
capacity of the RCA aggregate was 4.0% and the percentage of fines was less than
1.0%. To study the effects on workability two methods were employed, the first was
to test the concrete mixtures immediately after batching and the second involved
testing slump after 30 minutes of mixing. This second method was employed because
of a belief that the workability of the concrete should be tested at a time similar to
what would be seen in the field. Table 2.10 summarizes the results of these three
mixtures.
Table 2.10 Slump Test Results of RCA Concrete Mixtures (Malesev, 2010)
Concrete Mixture ID W/C (ratio) Immediate Slump, (cm) 30 min Slump, (cm)
Control (Natural Agg.) 0.514 16 10
50% RAC 0.568 14.5 8.5
100% RAC 0.620 11 9
56


The results indicated that immediately after batching the RAC mixtures
showed decreased slump values, and these values decreased as RAC content
increased. However, after 30 minutes of mixing the slumps for both RAC mixtures
decreased but not as substantially as the control mixture. After 30 minutes of mixing
the slumps of all three mixtures were quite similar. This indicates that mixing time is
important and may need to be considered when using RCA in concrete. The cause of
this was attributed to the additional mixing time and also to the additional water
required by the RAC mixtures. However, as previously stated it is unclear whether or
not this extra water was indeed extra or simply that which would be required based
on conditions of the aggregate, (i.e. absorption capacity and moisture content).
2.4.3.2 Air Content
According to the U.S. Department of Transportation, Federal Highway
Administration (2004), the use of air entraining admixtures in concrete made with
RCA should follow the same guidelines as any other natural aggregate and testing
should be conducted to determine the proper dosage. They also indicated that if the
source concrete from which the RCA came from performed poorly in regards to
durability (i.e. freeze-thaw resistance) then the RCA should not be used in new
concrete that will be susceptible to freezing and thawing cycles. The source of RCA
is sometimes difficult to determine and selecting an RCA from a known source may
not be an option. They also indicated that AEA dosage in general may be slightly
lower due to increased fines and the angularity of the RCA aggregate.
According to ACI 555R (2001) the air contents of concretes made with RCA
will vary more than similar mixtures made with natural aggregates, but the general
trend was that air contents would be slightly higher. Studies performed by Karthik
Obla et al (2007) support these findings. They investigated the effects of crushed
57


concrete aggregate (CCA) on fresh and hardened concrete properties. They studied
various concrete mixtures with different replacement levels of CCA ranging from 10
to 100% and found that the air contents of concretes made with CCA tended to be
only slightly higher than the control mixture made with natural aggregate. This trend
became more noticeable as the amount of CCA fines increased. The air contents
measured were entrapped air only because AEA was not used for these mixtures.
2.4.4 The Effects of RCA on Hardened Concrete Properties
2.4.4.1 Strength
According to ACI 555R (2001), the strengths of concretes made with RCA will
depend greatly on the source concrete. If concrete is made from varying sources of
RCA then the strength trends will also vary. Based on previous studies of concrete
containing RCA, if a single source for the RCA is used the variation in strength can
be minimized. Table 2.11 summarizes results of compressive strengths for concretes
with known sources and strengths with concretes made with RCA from the same
sources. Although the w/c varied considerably in the source concretes, the new RCA
concrete w/c was held constant.
When observing the Table 2.6, there doesnt appear to be a good correlation
between source concrete strength and RCA concrete strength. There also is not a good
correlation between source concrete w/c and the RCA concrete strength as well. The
highest source w/c values do appear at the lower end of the strength spectrum, but this
was not always the case. When comparing source strength to RCA strength, there is
not much of a correlation at all. The highest strength RCA mixtures did come from
the higher strength source concretes but this was not always the case as two of the
highest strength source mixtures with accompanied low w/c values were also some of
58


the lowest RCA strength contributors. The only information provided for these
mixtures was the w/c but cementitious content as well as other factors may have
played a role in the scattering of data.
Table 2.11 Compressive Strengths of Source Concretes and Concretes Made
With RCA From Same Source, (ACI 555R, 2001).
Source Source Concrete RCA RCA Concrete
Concrete 15-year Compressive Concrete 28-Day Compressive
w/c Strength, (lb/in2) w/c Strength, (lb/in2)
0.53 10,900 0.57 7120
0.50 10,600 0.57 6870
0.59 9050 0.57 6280
0.65 8600 0.57 6250
0.65 9850 0.57 6040
0.67 7470 0.57 5840
0.50 8980 0.57 5770
0.80 5640 0.57 5510
0.50 12,300 0.57 5340
0.50 9290 0.57 5100
0.81 6100 0.57 4640
0.53 10,600 0.57 4380
Gholamreza Fathifazl et al (2009) examined the effects of RCA on the
hardened properties of concrete. For this particular study a new mixture design was
59


developed to incorporate the existing mortar content within the RCA. This new
method accounts for this preexisting mortar in the design process and subtracts this
amount from the new cementitious quantities. This new method of design is called the
Equivalent Mortar Volume method (EMV). Several mixtures were developed from
varying RCA sources and compared with a similar control mixture composed of
normal virgin aggregates and PC. A total of six RCA mixtures were developed and
tested. To illustrate the need for a new mixture design when RCA is used, two
methods of design were used when designing each RCA mixture. The new proposed
design method was used on mixtures and designated with an E- at the beginning.
The conventional AC1 method of design was also used for comparison and designated
with a C- at the beginning. Depending on what type of supplementary cementitious
material was incorporated in these new concretes, if any, was also determined and
designated as B- for GGBFS, F- for fly ash and C- for no supplementary used
at the end of the mixture ID. For example, the mixture designated as ERAC-B
indicates this mixture was designed with the new method and the cementitious
contained GGBFS. The mixture CRAC-C was designed by the traditional ACI
method and the cementitious content of the concrete only contained cement. The
results of compressive testing performed on the RCA mixtures as well as the control
mixture designated as NAC-C are shown in Figure 2.3. It should be noted that these
values were scaled off a bar chart as no table reporting the exact values was available.
Therefore, minor discrepancies may exist. It was also assumed these specimens were
tested at 28 days of age as this information was also not given.
The results of the compressive tests show that this new method of design
produces comparable strengths with the ACI method. It should be noted that on
average the mixtures designed by this new method had 22% less new cement due to
the existing mortar being accounted for in the mixture design. This could lead to
substantial cost savings. What is also interesting is that GGBFS and Cement RAC
60


mixtures obtained greater strengths than the control mixture. The author attributes this
greater strength to the stronger interfacial transition zone (ITZ) between the RCA and
cementitious paste. It is believed this ITZ is stronger between RCA and paste as
opposed to a natural aggregate and paste.
50
45
Figure 2.3 Compressive Strengths of RCA Concrete Mixtures Designed with the
FMV Method and ACI Method, assumed 28-Day age, Fathifazl ct al (2009).
Mirjana Malesev et al (2010) investigated the hardened concrete properties of
concrete made with RCA. The RCA replacement was investigated at 50 and 100% of
the total coarse aggregate content. Natural sand was used for the fine aggregate on
61


both these mixtures and for the 50% RCA mixture, the remaining 50% of coarse
aggregate was a natural aggregate. A control mixture was also made with the same
natural sand as the RCA mixtures and the same natural coarse aggregate as the 50%
RCA mixture. A Type-II PC was used for all mixtures and the w/c was started at
0.514 and varied depending on mixture. The compressive strength was tested at 2, 7
and 28 days of age. Figure 2.4 shows the average results of these tests.
-----Control-------50-RAC --------100-RAC
Figure 2.4 Average Compressive Strength of RCA Concrete and Control
Concrete, Mirjana Malesev et al (2010).
62


The results of this study show that early age strengths for both the 50 and
100% RAC mixtures was slightly below that of the control mixtures made from 100%
natural aggregates. After day 2 however the RAC mixtures gained strength more
rapidly and overtook the control mixture for both the 7-day and 28-day strengths.
This difference is small however and in comparison these mixtures were all quite
comparable in strength. The w/c for the control mixture, 50-RAC and 100-RAC were
0.514, 0.568 and 0.620 respectively. Although strengths are comparable between
these three mixtures, it was noted that the modulus of elasticity was lower for both of
the RAC mixtures.
Karthik Obla et al (2007) performed extensive testing to determine the effects
of CCA on concrete strength. They were interested in whether or not source concrete
strength had an impact on new CCA concrete strength. Multiple concrete mixtures
with different replacement levels of CCA ranging from 10 to 100% were developed
and tested. The types of CCA used also varied and were categorized based on original
design strength of the source concrete. Four types of CCA were used; 1000, 3000,
5000 psi (6.9, 20.7 and 34.5 MPa) and an uncontrolled combination. The 1000,
3000 and 5000 psi (6.9,20.7 and 34.5 MPa) CCA came from known sources and was
in essence controlled with regards to aggregate properties. The uncontrolled CCA
was labeled as pile 1 because it came from the general recycled concrete pile at the
batch plant and is a general mixture of various recycled concrete.
Two different methods of incorporating the CCA into the mixture were used;
one included the CCA in an as received state from the crusher and the other used
pre-designated Coarse CCA aggregate. The as received CCA had a percentage of
fine aggregate included within and therefore percentages of virgin fine aggregate
were subtracted from the mixture design. If the Coarse method was used, only a
percentage of the coarse aggregate was replaced and 100% virgin fine aggregates
63


were used. Table 2.12 summarizes the results from compressive strength tests
performed at 7, 28 and 90-days of age.
Table 2.12 Compressive Strengths of Concretes Made with Varying Amounts of
CCA, Karthik Obla et al (2007).
Mixture ID Source Concrete Strength (lb/in2) w/c 7-Day Strength (lb/in2) 28-Day Strength (lb/in2) 90-Day Strength (lb/in2)
1 Control-1 N/A 0.58 3080 4100 4740
2 Control-2 N/A 0.58 2980 3930 5350
3 10% T-CCA 1000 0.55 2910 3990 4670
4 20% T-CCA-1 1000 0.59 2410 3630 3790
5 20% T-CCA-2 3000 0.58 2800 3690 4450
6 20% T-CCA-3 3000 0.58 2610 3760 4570
7 20% T-CCA-4 3000 0.58 3900 4390
8 30% T-CCA 3000 0.59 2800 3890 4720
9 20% T-CCA-5 Pile 1 0.58 2590 3410 4530
10 50% C-CCA 1000 0.55 2640 3470 4330
11 100% C-CCA-1 1000 0.52 2460 3180 3630
12 100% C-CCA-2 3000 0.59 2730 3930 4270
13 100% C-CCA-3 3000 0.59 4220
14 100% C-CCA-4 5000 0.57 2740 3790 4810
15 100% C-CCA-5 Pile 1 0.68 2140 2690 3190
16 100% C-CCA-6 Pile 1 0.69 2840 3360
64


At 7 days of age all CCA concretes exhibited slightly reduced strengths when
compared with the two control mixtures. When comparing the coarse mixtures (10-
16) to the as received mixtures (3-9), the coarse mixtures tended to have slightly
lower strengths on average. When comparing source concrete strengths for the
coarse group there is a slight correlation between strength of new concrete versus
source concrete strength. However, this cannot be said for the as received group of
mixtures as the highest strength came from mixture 3. This may be due to the
decreased percentage of CCA as well. This same trend can be noticed in mixtures 10
and 11 which suggests a decrease in strength with increasing CCA content.
At 28 days of age most CCA concretes exhibited reduced strengths when
compared with the two control mixtures. Mixtures 3, 7 and 12 were comparable with
the controls having equal or slightly greater strengths. When comparing the coarse
mixtures (10-16) to the as received mixtures (3-9), there is not a strong correlation
between strengths. At 28 days of age most CCA concretes exhibited only slightly
lower strengths than the control mixtures. When comparing source concrete strengths
for the coarse group there is a more noticeable correlation between strength of new
concrete versus source concrete strength when comparing mixture 11 to mixtures 12
and 13. Yet this correlation diminishes when comparing 14 made with the 5000 psi
(34.5 MPa) source concrete to mixture 13. There is a correlation between percentage
of replacement and strength when comparing mixtures 3 to 4 and mixtures 10 to 11.
At 90 days of age most CCA concretes exhibited reduced strengths when
compared with the two control mixtures. Mixtures 3, 7 and 12 were comparable with
the controls having equal or slightly greater strengths. When comparing the coarse
mixtures (10-16) to the as received mixtures (3-9), the coarse mixtures tended to
have slightly lower strengths on average. When comparing source concrete strengths
for the coarse group there is a stronger correlation between strength of new
concrete versus source concrete strength. Strengths increased for all mixtures as
65


source concrete strength increased. There is a correlation between percentage of
replacement and strength when comparing mixture 3 to 4 and mixture 10 to 11.
However the opposite occurred for mixtures 7 and 8 as strength increased with
increasing CCA content.
The data from this research appears to be somewhat scattered. There are some
correlations between source concrete strength and CCA strength but this was not
always the case. There does also appear to be a correlation between percentage of
replacement and strength as most results do show a decrease in strength as CCA
content increases. Another notable observation is the substantial decrease in strengths
for almost all the CCA concrete mixtures made from the uncontrolled pile 1
aggregate (mixtures 9, 15 and 16). However, mixtures 15 and 16 also had the highest
w/c values and this would have definitely played a major role in the decreased
strengths.
Mark Reiner (2007) investigated the effects of recycled materials on concretes
strength. The mixtures incorporated 50% and 100% RCA as coarse aggregate and had
varying amounts of fly ash as well. Two mixtures were batched with 50% RCA
replacement and 30% fly ash replacement. One mixture utilized a class F fly ash and
the other a class C. Four mixtures were batched using 100% RCA replacement. Two
mixtures used class F fly ash at 30 and 40% replacement, and the other two mixtures
used class C fly ash at 30 and 40% replacement. All of these mixtures were compared
with control mixtures that had the same amount of cementitious materials and natural
virgin aggregates. The results of the 50% RCA mixtures showed that the strengths
were all slightly lower for all age groups up to 56-days when compared with the
control mixture. However, the strength trend showed more differences at early age as
opposed to later age strength.
66


The 100% RCA mixtures were based on a CDOT Class-B concrete mixture.
These mixtures were tested up to 90-days of age as opposed to the 50% RCA
mixtures which were only tested to 56-days. The compressive test results showed that
strengths of all mixtures made with RCA were lower than the control mixture.
I lowever, three out of the four mixtures did meet the CDOT Class-B required
strength of 3000 psi (21.4 MPa) at 28-days. Although tables were not provided
indicating results for the 100% RCA mixtures (only graphs), an easy comparison of
the 50% RCA and 100% RCA graphs shows that as RCA content increases, strengths
decrease.
2.4.4.2 Permeability
According to information provided in ACI 555R (2001), the permeability of concretes
incorporating RCA with w/c values in the range of 0.5 to 0.7 will be on the order of
two to five times higher than normal concretes. This may be offset by lowering the
w/c value by 0.05 to 0.10. Results reported by Mirjana Malesev et al (2010) support
this claim. They found that as RCA content increased the water absorption increased
when hardened specimens were tested at 28-days of age. The w/c for these mixtures
was within the range of 0.5 to 0.7 as well. The ASTM procedure, if followed, was not
reported in this study.
Results of rapid chloride ion penetrability studies performed by Karthik Obla
et al (2007) showed a general trend of increasing permeability with increasing CCA
content. They investigated the effects of crushed concrete aggregate (CCA) on
hardened concrete properties. Multiple concrete mixtures with different replacement
levels of CCA ranging from 10 to 100% were developed and tested. The types of
CCA used also varied and were categorized based on original design strength of the
source concrete. Four types of CCA were used; 1000, 3000, 5000 psi (6.9, 20.7 and
67


34.5 MPa) and an uncontrolled combination. The 1000, 3000 and 5000 psi (6.9,
20.7 and 34.5 MPa) CCA came from known sources and was in essence controlled
with regards to aggregate properties. T he uncontrolled CCA was labeled as pile 1
because it came from the general recycled concrete pile at the batch plant and is a
general mixture of various recycled concrete. Two different methods of incorporating
the CCA into the mixture were used; one included the CCA in an as received state
from the crusher and the other used pre-designated Coarse CCA aggregate. The as
received CCA had a percentage of fine aggregate included within and therefore
percentages of virgin fine aggregate were subtracted from the mixture design. If the
Coarse method was used, only a percentage of the coarse aggregate was replaced.
Table 2.13 summarizes the results from the RCP testing which was performed at 90-
days of age.
Specimens were prepared and tested for rapid chloride ion penetrability
according to ASTM C 1202. These results show that as CCA content increases, so
docs the permeability. The lowest results did come from the 10 and 20% total
replacement mixtures 3 and 4 which were made from 1000 psi (6.9 MPa) source
CCA. What is interesting is that as the source concrete strength increased, the
permeability increased as can be seen in mixtures 5, 6 and 7 when compared with
mixture 4. When comparing the 20 and 30% total replacements with 3000 psi (20.7
MPa) source concrete there is a significant increase in permeability with a 10%
increase in CCA content. Although mixture 9 was made with the general Pile-1 CCA,
this concrete exhibited lower permeability than both control mixtures at 20%
replacement. These trends of increasing permeability with increasing source strength
and CCA content was not repeated for mixtures designed with only coarse aggregate
replacements. When comparing mixtures 10 and 11 the permeability decreased as
CCA content increased. All mixtures designed with only coarse aggregate
replacements showed significantly higher permeabilitys than those designed with
68


total replacements. This would suggest a general trend of increasing permeability
with increasing RCA content. What can also be seen from these results is that it is
difficult to produce a dense concrete when w/c values as high as these are used:
Table 2.13 RCPT Test Results of Concretes Made with Varying Amounts of
CCA, Karthik Obla et al (2007).
Mixture ID Source Concrete Strength (lh/in2) w/c RCPT Test Results (Coulombs) Permeability Class
1 Control-1 N/A 0.58 3618 Moderate
2 Control-2 N/A 0.58 3424 Moderate
3 10% T-CCA 1000 0.55 2970 Moderate
4 20% T-CCA 1000 0.59 2984 Moderate
5 20% T-CCA 3000 0.58 3936 Moderate
6 20% T-CCA 3000 0.58 3316 Moderate
7 20% T-CCA 3000 0.58 3683 Moderate
8 30% T-CCA 3000 0.59 4276 High
9 20% T-CCA Pile 1 0.58 3232 Moderate
10 50% C-CCA 1000 0.55 5402 High
11 100% C-CCA 1000 0.52 5187 High
12 100% C-CCA 3000 0.59 6248 High
13 100% C-CCA 3000 0.59 5036 High
14 100% C-CCA 5000 0.57 4729 High
15 100% C-CCA Pile 1 0.68 6201 High
16 100% C-CCA Pile 1 0.69 6033 High
Gholamreza Fathifazl et al (2009) examined the effects of RCA on the
hardened properties of concrete. As previously stated, for this particular study a new
mixture design was developed to incorporate the existing mortar content within the
69


RCA. This new method accounts for this preexisting mortar in the design process and
subtracts this amount from the new cementitious quantities. This new method of
design is called the Equivalent Mortar Volume method (EMV). Several mixtures
were developed from two RCA sources and compared with a similar control mixture
composed of normal virgin aggregates and PC. To compare the significance of this
new design method two separate batches were designed and mixed, one using the
EMV method and the second using the traditional ACI method. Chloride penetration
tests were performed on each mixture using the acid bulk diffusion test according to
ASTM C 1556. They determined that the chloride diffusion coefficients for the RCA
mixtures, regardless of design method were higher than that of the control specimens
made from natural aggregates. When comparing the two different design methods, the
EMV designed specimens had higher coefficients (on the order of 8 %) than the ACI
designed specimens.
Mark Reiner (2007) investigated the effects of recycled materials on concretes
permeability. Four mixtures incorporating 100% RCA as coarse aggregate and
varying amounts of fly ash were tested. Two mixtures used class F fly ash at 30 and
40% replacement, and the other two mixtures used class C fly ash at 30 and 40%
replacement. All of these mixtures were compared with control mixtures that had the
same amount of cementitious materials and natural virgin aggregates. Testing was
conducted at 14, 28 and 90 days of age. The results showed that at 14 days, all four
mixtures with RCA had significantly higher results (charge passed). However, the
results at 28 days of age showed different results. The two mixtures made with class-
F fly ash had surpassed the control mixture and achieved lower ratings at 28-days.
The two mixtures with class-C fly ash still achieved higher results. At 90 days of age,
all but one RCA mixture achieved better results than the control mixture. The two
mixtures with class-F fly ash actually achieved 31 to 36% improvements when
compared with the control at 90-days.
70


2.4.4.3 Freeze-Thaw Durability
According to ACI 555R (2001), most studies performed on the freeze-thaw resistance
of concretes made with RCA show little if any difference when compared to normal
concretes made with natural aggregates. They did mention several Japanese studies
which indicated a decrease in freeze-thaw resistance for concretes made with RCA
but also noted that the studies performed in Japan used much lower quality of source
concrete as compared to studies from other sources.
Karthik Obla et al (2007) performed freeze-thaw durability tests on concretes
containing various replacement amounts of aggregate with crushed concrete
aggregate (CCA). Three concrete mixtures with different replacement levels of CCA
were developed and tested as well as a control mixture composed of natural
aggregates. Two mixtures contained 20% CCA and one mixture contained 100%
CCA for coarse aggregate. The types of CCA used also varied and were categorized
based on original design strength of the source concrete. Two types of CCA were
used; 1000 and 3000 psi (6.9 and 20.7 MPa). The 1000 and 3000 psi (6.9 and 20.7
MPa) CCA came from known sources and was in essence controlled with regards
to aggregate properties. Two different methods of incorporating the CCA into the
mixture were used; The first two mixtures included the CCA in an as received state
from the crusher and the third used pre-designated Coarse CCA aggregate. The as
received CCA had a percentage of fine aggregate included within and therefore
percentages of virgin fine aggregate were subtracted from the mixture design. If the
Coarse method was used, as was done for the third mixture, only a percentage of
the coarse aggregate was replaced. Two beam specimens were developed for each
mixture and tested at 56-days of age according to ASTM C 666 procedures. Table
2.14 summarizes the results from the freeze-thaw durability tests.
71


Mixture #4 had the highest air content for all mixtures but the air contents for
the other three mixtures were well within reasonable limits to resist the freezing and
thawing cycles. Both specimens from mixtures #2 and #3 that were mixed with the
CCA in as received condition failed during testing. Both specimens from mixture
#2 failed (i.e. the dynamic modulus fell below 60%) before completion of the test at
107 and 190 cycles respectively. It was also noted that visible cracks could be seen in
these specimens. Both specimens from Mixture #3 performed better than Mixture #2
specimens but also failed before completion of the test at 243 and 300 cycles
respectively. No visible deterioration was noted in these specimens however. Both
specimens from Mixture #4 performed well. These mixtures had 100% CCA as
coarse aggregate and the failure of the previous two mixtures was associated with the
CCA fines content that was not present in Mixture #4.
Table 2.14 Average Results of Rapid Freeze-Thaw Testing on Concretes Made
with CCA, Karthik Obla et al (2007).
Mixture ID Source Concrete Strength (Ih/itt2) w/c Air Content (%) Durability Factor Mass Loss (%)
1 Control N/A 0.45 6.4 92 0.52
2 20% T-CCA 1000 0.45 4.8 13 0.18
3 20% T-CCA 3000 0.45 5.6 9 0.73
4 100% C-CCA 3000 0.45 8.5 89 1.23
Gholamreza Fathifazl et al (2009) performed freeze-thaw durability tests on
concretes with 100% RCA as coarse aggregate. As previously stated, for this
particular study a new mixture design was developed to incorporate the existing
72


mortar content within the RCA. This new method accounts for this preexisting mortar
in the design process and subtracts this amount from the new cementitious quantities.
This new method of design is called the Equivalent Mortar Volume method (EMV).
Two mixtures were developed from one RCA source and compared with a similar
control mixture composed of normal virgin aggregates and PC. To compare the
significance of this new design method two separate batches were designed for the
RCA mixtures, one using the EMV method and the second using the traditional ACI
method. Freeze thaw durability testing was performed on each mixture using
procedure A of ASTM C 666. It is assumed these specimens were tested at 28-days of
age as this information was not provided. The durability factors (DF) for all mixtures,
including the control were above 90%. The specimens designed by the new EMV
method performed slightly better (3.3% higher DF) than the specimens designed by
the traditional ACI method. They attributed this difference to the lower total mortar
content of the EMV designed specimens.
Reiner (2007) investigated the effects of RCA content on the freeze-thaw
durability of concretes. Four mixtures incorporated 100% RCA as coarse aggregate
and had varying amounts of fly were tested. Two mixtures used class F fly ash at 30
and 40% replacement, and the other two mixtures used class C fly ash at 30 and 40%
replacement. All of these mixtures were compared with control mixtures that had the
same amount of cementitious materials and natural virgin aggregates. The mixtures
were tested according to ASTM C 666 procedures. Although not reported, it is
assumed these mixtures were tested at 28 days of age due to the inclusion of high
amounts of fly ash. After 300 cycles of testing the RCA mixtures performed well. All
four mixtures performed better than the control mixture and had durability factors
above 85.
73


2.4.4.4 Alkali-Silica Reactivity (ASR)
The potential for ASR when using RCA as aggregate should raise the same, if not
more concern to the designer when compared to natural aggregates. The potential for
ASR in RCA depends greatly on the source aggregate and this information is not
always known. If the source concrete information is available then a simple review of
the aggregate used will probably determine whether or not ASR could be a problem
in the new concrete. If the source concrete information is not available then caution
should be exercised and testing should be performed on the RCA to determine
potential for ASR. The source concrete aggregate may have tested positive for ASR
but low alkali cement or other techniques could have been used to reduce this
potential. The effects of any mitigation techniques used in the source concrete are no
longer applicable once the concrete is crushed and used as RCA.
2.4.5 Summary
This literature review showed that there can be neutral or negative effects from using
RCA as aggregate in a concrete. Neutral (or non-detrimental) effects are actually a
good quality when it comes to aggregates. Aggregates are by and large a filler
material in a concrete and do not contribute to strength until high strengths are
required. Therefore, it would actually be uncommon or untypical to find allot of
supporting information that would suggest RCA increases a concretes strength. Most
information found in regards to RCA has shown neutral effects but in some cases the
effects were detrimental or slightly beneficial.
It appeared that when researchers took care in choosing a reputable source
concrete for use as RCA, and performed testing as if the RCA were any other natural
aggregate, the results were positive and less scattered. The most critical aspect of
74


using RCA in a concrete is treating the RCA as an unknown aggregate source.
RCA fines also appear to be attributed to negative results in concrete. All properties
of concrete decreased as fines content increased from slump to compressive strength
and durability. The compressive strength in general may decrease or remain the same
but at best, will only increase slightly when compared with natural aggregate
concrete. This is reasonable because once a concrete is crushed and recycled one
cannot expect to achieve beneficial effects because the old concrete was a high
strength concrete. There also tends to be more scattering of results if concretes are
made from a general stock pile of recycled concrete as opposed to a known source
concrete.
Although there wasnt a strong correlation between source concrete strength
and new RCA concrete strength, there was minor evidence that suggested higher
strength source concretes would produce higher strength RCA mixtures, when
compared with other RCA mixtures made from lesser strength source concretes. This
was probably due to the mortar still attached to the aggregates and one can expect that
the higher strength concretes had a better bond between aggregate and paste and
would therefore provide a stronger ITZ between the old mortar and new. However,
the bond between the aggregate and old paste would still be the bond in question, not
the new bond between old and new paste. At best, the bond between old and new
paste can only be as strong as the bond between old paste and aggregate, since this
will remain the weakest link. The freeze-thaw resistance of concretes made with RCA
will most probably decrease. This decrease in freeze-thaw resistance does not appear
to be correlated with amount of RCA content however but more on the fines content.
The permeability of concretes made with RCA are expected to increase with
increasing RCA content, regardless of fines. Other durability factors such as ASR
may increase or decrease and this depends greatly on the source concrete and
aggregate.
75


2.5 Expectations
This research involves testing concretes made with varying amounts of combined
recycled materials. These replacements will be done together on all three ingredients
equally. The combined effects that GGBFS, RCA and waste glass will have on
concrete is unknown and can only be speculated based on the observed and reported
el'fects that each alone have had on concrete. Since aggregate volume is much higher
than paste volume it would be assumed that the aggregate qualities would have the
most pronounced effect. However, the cementitious paste is the heart of the mixture
and the GGBFS effects may govern. An attempt at theorizing the combined effects is
warranted and these effects will be limited to those that are being tested for this
research program.
The workability of a concrete was shown to increase with increasing GGBFS
and decrease with increasing RCA and waste glass. Therefore, it can be assumed that
since two of these materials decrease workability and only GGBFS increases
workability that the total workability will decrease as recycled materials content
increases. The AEA dosage of GGBFS mixtures tended to increase with increasing
content. The AEA requirements of RCA also tended to increase but not as
significantly and this was related to fines content. The air contents of waste glass
mixtures did not appear to be effected. Therefore, it can be assumed that the
combined effect will be a decrease in air contents for increasing recycled materials
when AEA dosage is held constant.
The strength of concretes made with GGBFS tend to increase up to a certain
percentage replacement and then decrease thereafter. The use of RCA tends to
decrease a concrete mixtures strength with increasing content. The use of waste glass
also tended to decrease a concretes strength with increasing content. Therefore, with
two ingredients that decrease strength and one ingredient that increases to a certain
76


degree, it can be assumed that as recycled materials increase a trend of decreasing
strength will be seen. Strength gains will probably be slower for increasing recycled
contents due to the GGBFS.
The permeability of concretes made with GGBFS significantly decreased with
increasing GGBFS content. The permeability of concretes made with RCA as coarse
aggregate tended to increase with increasing content. The use of waste glass was
shown to increase the permeability with increasing content. Whether or not the
GGBFS can offset the increased permeability caused by the recycled aggregates will
be interesting. However, with two ingredients that increase and only one that
decreases permeability it will be assumed that permeability will increase for
increasing recycled content. This seems reasonable since paste volume is significantly
less than aggregate volume. One would think that the effects that the recycled
aggregate have on the permeability will govern. However, the cementitious paste is
the area where water penetrates and the GGBFS content may prove to govern this
area. The freeze thaw durability of concretes made with GGBFS may decrease as
replacement levels increase. The freeze thaw durability of concrete was also shown to
decrease as RCA content increased as well as glass content when used as coarse and
fine aggregates respectively. Therefore, it is assumed that as recycled materials
content increases the freeze-thaw durability will decrease.
The alkali-silica reactivity of concretes made with glass aggregate has been
shown to increase as content increases. The size of the glass particles will also play a
major role. Since the glass aggregate used for this research is a mixed colored glass
and is a 3/8 inch minus aggregate, the potential for ASR is real. The ASR potential of
the RCA is unknown. The use of GGBFS has shown to decrease the ASR expansions
in concretes made with reactive aggregates and may help this from occurring. It is
expected that concrete mixtures with GGBFS will not be affected by ASR, but the
concrete mixture with only recycled aggregates and PC will undergo ASR expansion.
77


3.
Problem Statement
3.1 Statement
Todays society is changing. Gone are the days of excess and care-free living. Talk of
the environment is on everyones agenda, and global warming is something even
children know about. Recycling has become a common practice not only in
households across America, but in todays industry as well. Most products made
today consist of at least some form of recycled material, whether its the packaging it
comes in or the product ingredients themselves. Energy consumption is another area
of concern due to the increased awareness of global warming and the planets limited
supply of natural fuel resources. It is for these reasons that the potential benefits of
using recycled materials in concrete was chosen for the topic of this research.
Concrete is the most widely used construction material in the world. In
residential construction the foundations on which houses, apartment buildings, town
houses, etc. rest upon will most undoubtedly be made of concrete. This foundation
may be entirely formed and placed or it may be a composite structure composed of
cider blocks and placed concrete. Sky scrapers and other larger commercial or
residential buildings also have foundations of concrete, but unlike smaller residential
structures these larger building may be made primarily of concrete. Concrete has been
used to build many of the roadways and bridges throughout the world, not to mention
many other civil structures such as dams and waterways. The importance of concrete
in todays growing infrastructure is obvious. The use of concrete will only increase in
the future, and the materials and energy needed to fulfill this demand are becoming
more difficult to obtain.
The primary materials needed to produce concrete today consist of portland
cement, aggregates and water. Cement is the most costly component of concrete.
Unlike aggregates, cement cannot simply be mined from the earth. Cement
78


production is not only expensive but also produces large amounts of pollution. It is
estimated that for every ton (metric ton) of cement produced an equal amount of
carbon dioxide is emitted to the atmosphere (World Buisness Council for Sustainable
Development, 2002). The aggregates used in concrete must be quarried from the earth
and processed before being suitable for use. Aggregate quarries are not only
destructive from a natural beauty standpoint, but also require substantial amounts of
energy. An aggregate must also have suitable inert qualities before being used in
concrete. Not every rock and gravel source is acceptable. These quality aggregate
sources have diminished over the years and finding acceptable aggregates is
becoming difficult. Concrete used for a particular job is not a material that can be
shipped from great distances and concrete batch plants typically need to be located
somewhere nearby. Economically, the aggregate sources must also be located close to
the batch plant as well. Therefore, quality and proximity are key components to a
concrete batch plant. The focus of this research is to determine the potential benefits
of replacing portland cement and the aggregates of a concrete with recycled materials.
The purpose of this thesis is to design and batch concrete mixtures with
varying percentages of recycled materials. To fully understand the effects that
recycled materials have on the fresh and hardened concrete properties the percentages
ranged from a 25% replacement to a complete 100% recycled material concrete. The
compressive strength of a concrete is its most important property. Concrete used
today must be able to carry the loads it is subjected to and must be durable enough to
withstand harsh environments. The ultimate strength, permeability and freeze-thaw
durability was examined for all mixtures as well as the fresh properties. Alkali-silica
reactivity was also investigated for this research. The primary benefit of this research
is to examine the positive and negative effects of using recycled materials in a
concrete mixture, and determine at what point these materials no longer become
beneficial.
79


4. Experimental Plan
4.1 Design Summary
The purpose of this research program was to investigate the potential benefits of
using recycled materials in concrete mixtures. The three main ingredients that were
subject to replacement with recycled materials included cement, fine and coarse
aggregates. This research began by first obtaining the recycled materials and then
performing tests to determine their physical properties. Recycled concrete was
obtained to be used as a coarse aggregate replacement. Waste glass was obtained to
be used as a fine aggregate replacement and ground granulated blast furnace slag
(GGBFS) as a cement replacement.
In order to investigate the effects these recycled materials would have on the
fresh and hardened properties of the concrete, six mixtures were designed and
batched. These mixtures were designed based on a Class D, Colorado Department of
Transportation (CDOT) concrete mixture. A Class D CDOT concrete is designated as
a medium strength structural concrete that can be used in multiple applications
including bridge decks. Certain requirements for a Class-D CDOT concrete are
shown in Table 4.1. It should be noted that these mixtures were only loosely based
on a Class-D mixture because the use of GGBFS and replacement amounts used for
this research are prohibited. Recycled concrete and waste glass are also prohibited.
However, the cementitious content, percentages of coarse and fine aggregates, w/c
and air content guidelines were all followed for these mixtures.
The total cementitious contents and w/c were held constant for all mixtures.
All mixtures had similar ratios of coarse and fine aggregates as well. The research
began by first batching a mixture with 100% replacement of aggregates and cement
with recycled aggregates and GGBFS. A second mixture was batched at the same
time which had 100% of the aggregates replaced with recycled aggregates, but had
80


100% PC. These two mixtures were batched first to investigate any potential adverse
effects between the materials and any effects on fresh properties that may develop
from batching. The next stage of research involved designing and batching
intermediate replacement mixtures ranging from 25 to 75% replacements. A control
mixture composed of normal aggregates and 100% PC was also batched to be used
for comparison.
Table 4.1 Colorado Department of Transportation (CDOT) Class-D concrete
specifications and field requirements
28-Day Compressive Strength Total Cementitious Range Air Content Range Maximum Water to Cementitious Minimum Amount of Coarse Agg.
(lb/in2) (Ib/ycf) (percent) (ratio) (percent)
4500 615-660 5-8 0.44 55
4.2 Material Properties
4.2.1 Ground Granulated Blast Furnace Slag (GGBFS)
The GGBFS was obtained from Lehigh, Heidelberg Cement Group located in
Concord, California. Because this research involved developing a concrete mixture
that had a 100% replacement of cement with GGBFS, a Grade 120 slag was chosen
because of its higher reactivity compared to Grade 100. The GGBFS was tested in
accordance with ASTM C 989 and the results of this testing are shown in Table 4.2.
81


Table 4.2 Lehigh Grade 120 GGBFS Physical and Chemical Properties
ALLCEM GGBFS Grade 120
Chemical and Physical Properties Test Results ASTM C 989 Specifications
Sulfur Trioxide (SO3) (%) 2.192 4.0 Max
Sulfur Sulfide (S) (%) 0.560 2.5 Max
Blaine Fineness cm2/g 5240
Percent Retained on 325 Mesh (%) 0.2 20 Max
Air Content of Slag Mortar (%) 6.3 12 Max
Specific Gravity 2.93
Slag Activity Index
7-Day Individual (%) 134 90 Min
28-Day Individual (%) 138 110 Min
Compressive Strength
7-Day Reference Cement (lb/in2) 4890
28-Day Reference Cement (lb/in2) 6260 5000 Min
7-Day Slag + Ref. Cement (lb/in2) 6560
28-Day Slag + Ref. Cement (lb/in2) 8660
4.2.2 Portland Cement
1'hc cement used for this research was a Type-l-II portland cement. The cement was
supplied by Holcim Cement Company, located in Florence, Colorado. The cement
was tested in accordance to ASTM C 150 and the results are shown in Table 4.3.
82


Table 4.3 Holcim Type-I-II Cement, Physical and Chemical Properties
Holcim Type-I-II Portland Cement
Chemical and Physical Properties Test Results ASTMC 150 Specifications
SiO, (%) 19.6 ....
ai2o, (%) 4.7 6.0 max
Fe20? (%) 3.2 6.0 max
CaO (%) 63.4
MgO (%) 1.5 6.0 max
SO., (%) 3.4 3.0 max
n 0 t j 1 (%) 1.4
Limestone (%) 3.7 5.0 max
CaCOj in Limestone (%) 84.0 70 min
C,S (%) 59.0
C2S (%) 11.0
c2a (%) 7.0 8 max
c4af (%) 10.0
C,S + 4.75 C3A (%) 92.0 100 max
Loss of Ignition <%) 2.6 3.0 max
Blaine Fineness cm2/g 414 2600 4300
Air Content of PC Mortar (%) 6.3 12 max
Specific Gravity 3.15
83


4.2.3 (Virgin) Coarse and Fine Aggregates
Both the coarse and fine aggregates were obtained by University of Colorado Denver.
Both these aggregates came from Bestway Concrete and sources located in the
Colorado area. The material properties and gradations of these aggregates were
performed by WestTest Laboratories located in Denver, Colorado. The coarse
aggregate was tested in accordance to ASTM C 33 No. 57 and 67 Coarse Aggregate
and meets these requirements. The sand was tested in accordance to ASTM C33 Fine
Aggregate, and meets these requirements. These aggregates will be referred to as
UCD sand and UCD rock for the remainder of this thesis. The material properties
data and complete gradations for both aggregates are included in Appendix B.
4.2.4 (Recycled) Coarse and Fine Aggregates
The recycled concrete and waste glass used for replacement of aggregates was fully
tested prior to use in a concrete mixture. All aggregate samples used for testing
conformed to ASTM C 702 requirements. Table 4.4 summarizes the tests performed.
Table 4.4 Testing of Recycled Materials
Aggregate Test Type Performed ASTM Method
Recycled Concrete (Coarse Aggregate) Specific Gravity & Absorption Capacity ASTM C 127
Sieve Analysis (Gradation) ASTM C 136
Unit Weight & Moisture Contents ASTM C 29
Waste Glass Specific Gravity & Absorption Capacity ASTM C 128
(Fine Aggregate) Sieve Analysis (Gradation) ASTM C 136
84


4.2.4.1 Waste Glass as Fine Aggregate
The waste glass was obtained from Rocky Mountain Bottling Co., owned by Coors.
The waste glass was used as a replacement for the UCD sand. The waste glass was
produced mainly from beer bottles and consisted of various colors (clear, amber and
green). The process of recycling bottles involves multiple crushers which decrease the
size of the glass particles more after each crusher. The glass moves along a conveyer
belt from one crusher to the next while vacuums and magnets remove any labels or
other foreign objects. After all crushing has taken place the glass drops into a hopper
and is ready for the furnace where it will be melted down. It is from this hopper that
the glass for this research was taken. The glass was used as received from the
recycling plant and no washing took place. Only larger metallic objects (i.e. batteries)
were hand removed from the waste glass prior to use in the concrete mixtures.
Testing was performed on the waste glass to determine the physical and chemical
properties. Table 4.5 shows the properties of the waste glass as well as the UCD sand
fine aggregate for comparison.
Table 4.5 Fine Aggregate Properties of Waste Glass and UCD Sand
Aggregate Property Waste Glass UCD Sand
Absorption Capacity, (%) 0.18 0.70
Specific Gravity 2.47 2.63
Fineness Modulus 4.08 2.74
When comparing these two fine aggregates it can be seen there are similarities
and differences in physical properties. The specific gravity (SG) of both aggregates
85