TECHNICAL REPORT 0-6966-1

TxDOT PROJECT NUMBER 0-6966

Supplementary Cementitious Materials:

Assessment of Test Methods for New
and Blended Materials

Saif Al-Shmaisani
Ryan Kalina
Katelyn O’Quinn
Jae Kyeong Jang
Michael Rung
Raissa Ferron
Maria Juenger

December 2020; Published March 2021

http://library.ctr.utexas.edu/ctr-publications/0-6966-1.pdf
 Technical Report Documentation Page
1. Report No. 2. Government 3. Recipient’s Catalog No.
FHWA/TX-21/0-6966-1 Accession No.

4. Title and Subtitle 5. Report Date

Supplementary Cementitious Materials: Assessment of Test Submitted: December 2020; Published
Methods for New and Blended Materials March 2021
6. Performing Organization Code
7. Author(s) 8. Performing Organization Report No.
Saif Al-Shmaisani, Ryan Kalina, Katelyn O’Quinn, Jae Kyeong 0-6966-1
Jang, Michael Rung, Raissa Ferron, and Maria Juenger
9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
Center for Transportation Research 11. Contract or Grant No.
The University of Texas at Austin 0-6966
3925 W. Braker Lane, 4th Floor
Austin, TX 78759
12. Sponsoring Agency Name and Address 13. Type of Report and Period Covered
Texas Department of Transportation Technical Report
Research and Technology Implementation Division September 2017 – January 2021
P.O. Box 5080 14. Sponsoring Agency Code
Austin, TX 78763-5080
15. Supplementary Notes
Project performed in cooperation with the Texas Department of Transportation.
16. Abstract
In Texas, and most of the U.S., Class F fly ash is the most used supplementary cementitious material (SCM) due
to the many benefits it provides to concrete. In recent years, the availability of Class F fly ash has decreased as
many coal-fired power plants have shut down. Plants that have not shut down are required to install various
emission control systems that can significantly alter the type of fly ash produced. As the face of fly ash production
continues to change, the usability of non-traditional fly ashes and fly ash alternatives in concrete must be
evaluated. As the number of new SCM sources rises to meet demand, rapid tests are necessary to screen out
poor-performing materials and long-term performance testing is needed to qualify promising materials for use
in concrete mixtures. In this study, both rapid SCM screening tests and long-term performance of non-traditional,
blended fly ashes were examined. For rapid screening, both R3 testing (ASTM C1897) and a lime reactivity test
were successful at screening out inert materials. By pairing the R3 test with an extra step in bound water testing,
it is also possible to distinguish between pozzolanic and hydraulic reactivity of SCMs. With respect to non-
traditional, blended fly ashes, it was determined that fly ashes that do not meet the definition of a Class F fly ash
but do meet the chemical and physical property requirements performed comparably to a traditional Class F fly
ash in most cases. Most of the performance differences were negligible and can be remedied through the addition
of limestone, gypsum, or chemical admixtures, except for sulfate resistance. The poor sulfate resistance of some
non-traditional fly ashes can be directly linked to the presence of certain crystalline phases, making the ash
perform more like a Class C fly ash than a Class F fly ash.
17. Key Words 18. Distribution Statement
Concrete, Fly Ash, Supplementary Cementitious No restrictions. This document is available to the
Material, Pozzolan, Reactivity, Testing public through the National Technical Information
Service, Springfield, Virginia 22161; www.ntis.gov.
19. Security Classif. (of report) 20. Security Classif. (of this page) 21. No. of pages 22. Price
Unclassified Unclassified TBD
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
Supplementary Cementitious Materials: Assessment of
Test Methods for New and Blended Materials
Saif Al-Shmaisani
Ryan Kalina
Katelyn O’Quinn
Jae Kyeong Jang
Michael Rung
Raissa Ferron
Maria Juenger

CTR Technical Report: 0-6966-1

Report Date: Submitted: December 2020; Published March 2021
Project: 0-6966
Project Title: Assessment of Test Methods for New and Blended Materials
Sponsoring Agency: Texas Department of Transportation
Performing Agency: Center for Transportation Research at The University of Texas at Austin

Project performed in cooperation with the Texas Department of Transportation and the Federal
Highway Administration.
Disclaimers
Author’s Disclaimer: The contents of this report reflect the views of the authors,
who are responsible for the facts and the accuracy of the data presented herein. The
contents do not necessarily reflect the official view or policies of the Federal
Highway Administration or the Texas Department of Transportation (TxDOT).
This report does not constitute a standard, specification, or regulation.

Patent Disclaimer: There was no invention or discovery conceived or first actually

reduced to practice in the course of or under this contract, including any art, method,
process, machine manufacture, design or composition of matter, or any new useful
improvement thereof, or any variety of plant, which is or may be patentable under
the patent laws of the United States of America or any foreign country.

Engineering Disclaimer
NOT INTENDED FOR CONSTRUCTION, BIDDING, OR PERMIT
PURPOSES.

Research Supervisor: Maria C. Garci Juenger

iv
Acknowledgments
The authors express appreciation to the TxDOT Project Director, Andy Naranjo,
Research Project Manager, Joanne Steele, and members of the Project Monitoring
Committee: Michael Botzaritch, Rachel Cano, Clifton Coward, Jr., Phillip Hempel,
and Masoud Moradian. The authors would also like to thank the following for
helpful discussions: Thano Drimalas at the University of Texas at Austin, Prannoy
Suraneni and Sivakumar Ramanathan at the University of Miami, and Mahipal
Kasaniya and Michael Thomas at the University of New Brunswick. And finally,
special thanks go to Clifton Coward, Jr., Masoud Moradian, and the petrography
team at TxDOT and Lisa Hart, Harsh Mundra, and Tongren Zhu at the University
of Texas at Austin for collecting some of the experimental data presented herein.

v
Table of Contents
Chapter 1. Introduction ........................................................................................... 1
1.1. Motivation .................................................................................................... 1
1.2. Literature Review......................................................................................... 2
1.2.1. Supplementary Cementitious Materials ................................................ 2
1.2.2. SCM Testing ......................................................................................... 7
1.3. Materials Selected for Testing ................................................................... 12
Chapter 2. Material Characterization .................................................................... 15
2.1. Characterization Procedures and Results ................................................... 15
2.1.1. Oxide Composition ............................................................................. 15
2.1.2. Moisture Content and Loss on Ignition .............................................. 17
2.1.3. Density ................................................................................................ 17
2.1.4. Fineness............................................................................................... 18
2.1.5. Soundness ........................................................................................... 19
2.1.6. Strength-Activity Index and Water Requirement ............................... 19
2.1.7. Available Alkali .................................................................................. 20
2.1.8. Crystalline and Bulk Amorphous Composition .................................. 21
2.1.9. Particle Size Distribution .................................................................... 25
2.1.10. Scanning Electron Microscopy ......................................................... 26
Chapter 3. Reactivity Testing ............................................................................... 27
3.1. Reactivity Tests.......................................................................................... 27
3.1.1. Isothermal Calorimetry ....................................................................... 27
3.1.2. X-ray Diffraction ................................................................................ 33
3.1.3. Loss on Ignition .................................................................................. 34
3.1.4. Thermogravimetric Analysis .............................................................. 37
3.1.5. Compressive Strength ......................................................................... 40
Chapter 4. Admixture Interaction Testing ............................................................ 44
4.1. Interaction with Air-Entraining Agent ....................................................... 44
4.1.1. Foam Index Test ................................................................................. 44
4.1.2. Air Void Spacing Analysis ................................................................. 46
4.1.3. Air-Entrainment of Mortar .................................................................. 47
4.2. Interaction with High-Range Water Reducers ........................................... 48
4.2.1. Saturation Dosage ............................................................................... 49
4.2.2. Small Amplitude Oscillatory Shear Testing ....................................... 50

vi
Chapter 5. Concrete Property and Long-Term Durability Testing ....................... 53
5.1. Paste Testing .............................................................................................. 53
5.1.1. Isothermal Calorimetry ....................................................................... 53
5.1.2. Rheology ............................................................................................. 58
5.2. Mortar Testing ........................................................................................... 60
5.2.1. Drying Shrinkage ................................................................................ 60
5.2.2. Alkali-Silica Reaction ......................................................................... 61
5.2.3. Sulfate Resistance ............................................................................... 62
5.3. Concrete Testing ........................................................................................ 64
5.3.1. Fresh Concrete Properties ................................................................... 64
5.3.2. Compressive Strength ......................................................................... 67
5.3.3. Rapid Chloride Penetrability............................................................... 68
5.3.4. Bulk Electrical Resistivity .................................................................. 69
5.3.5. Alkali-Silica Reaction ......................................................................... 70
Chapter 6. Analysis and Recommendations ......................................................... 72
6.1. Reactivity Tests.......................................................................................... 72
6.2. Use of Non-Traditional Fly Ashes ............................................................. 73
References ............................................................................................................. 75
Appendix A. X-Ray Diffractograms ..................................................................... 84
Appendix B. Scanning Electron Microscopy Images ........................................... 93

vii
List of Tables
Table 1.1: Procured Class F Fly Ashes ................................................................. 13
Table 1.2: Procured Class C Fly Ash and Milled Bottom Ash ............................. 13
Table 1.3: In-House Blended Materials ................................................................ 13
Table 1.4: Materials Procured for Reactivity Testing........................................... 14
Table 2.1: Material Characterization Tests ........................................................... 15
Table 2.2: Oxide Compositions of Materials ........................................................ 16
Table 2.3: MC and LOI of Materials .................................................................... 17
Table 2.4: Physical Properties of Materials .......................................................... 18
Table 2.5: SAI and Water Requirement for Mortars ............................................ 20
Table 2.6: Available and Total Alkalis (Percent by Mass of SCM) ..................... 21
Table 2.7: Amorphous Contents and Main Crystalline Phases Present in
Coal Combustion Products (CCPs) ................................................................... 22
Table 2.8: Amorphous Contents and Main Crystalline Phases Present in
Blended or Remediated CCPs ........................................................................... 23
Table 2.9: Amorphous Contents and Main Crystalline Phases Present in
Natural Pozzolans and Slag ............................................................................... 23
Table 2.10: Bulk Amorphous Composition and NBO/T of Materials .................. 24
Table 2.11: Particle Size Distribution of Materials .............................................. 25
Table 3.1: Reactivity Tests ................................................................................... 27
Table 3.2: Main Crystalline Phases in Hydrated SCM-Alkaline Solution
Pastes ................................................................................................................. 33
Table 4.1: AEA Demand as Determined via the Foam Index Test ...................... 46
Table 4.2: F-G Concrete Mixture Design for Air Void Analysis ......................... 47
Table 4.3: Results from Air Void Spacing Analysis............................................. 47
Table 4.4: Air-Entrainment of Mortar Results...................................................... 48
Table 4.5: ViscoCrete® 2110 Saturation Dosages for Cement Pastes ................. 50
Table 4.6: Strain and Frequency Values for SAOS Testing ................................. 50
Table 5.1: Paste, Mortar, and Concrete Tests ....................................................... 53
Table 5.2: 3-Day Cumulative Heat Values of Cement Pastes .............................. 55
Table 5.3: Bingham Parameters ............................................................................ 60
Table 5.4: Drying Shrinkage and Weight Loss on Drying Relative to OPC
Control ............................................................................................................... 61

viii
Table 5.5: ASR Mortar Bar Expansion Percentage at 14 Days ............................ 62
Table 5.6: Sulfate Mortar Bar Expansion Percentages ......................................... 63
Table 5.7: ACI 201 Sulfate Exposure Classifications .......................................... 63
Table 5.8: OPC Concrete Mixture Design ............................................................ 64
Table 5.9: Fresh Concrete Properties .................................................................... 65
Table 5.10: Bulk Electrical Resistivity Results of Concrete Cylinders ................ 70
Table 5.11: ASTM C1293 Expansion Percentages............................................... 71

List of Figures
Figure 3.1: Rate of heat evolution of cement pastes containing control
materials and SCMs .......................................................................................... 28
Figure 3.2: Rate of heat evolution of cement pastes containing control
materials and natural pozzolans ........................................................................ 28
Figure 3.3: 7-day cumulative heat release of R3 mixtures at 40°C ....................... 30
Figure 3.4: Cumulative heat evolved of control and SCM-water pastes .............. 31
Figure 3.5: Cumulative heat evolved of control and natural pozzolan-water
pastes ................................................................................................................. 31
Figure 3.6: 3-day cumulative heat release of SCM-alkaline solution pastes
at 40°C ............................................................................................................... 32
Figure 3.7: 7-day bound water content of R3 mixtures at 40°C ............................ 34
Figure 3.8: R3P calcium hydroxide content of R3 mixtures at 7 days................... 35
Figure 3.9: Classifications of SCMs based on reactivity using the 7-day
heat release of R3 mixtures at 40°C and CH content of R3 mixtures using
the R3P method. Thresholds adapted from Suraneni et al. [53] and
Kalina et al. [35] ................................................................................................ 36
Figure 3.10: CH content of R3 mixtures at 7 days using DSC/TGA vs R3P......... 38
Figure 3.11: CH content of cement-SCM pastes at 25% replacement of the
cement by mass with gray dashed lines indicating Q paste calcium
hydroxide content with 5% error ....................................................................... 39
Figure 3.12: Mortar cube compressive strength results of cement-SCM
mortars ............................................................................................................... 41
Figure 3.13: Mortar cylinder compressive strength results of cement-SCM
mortars ............................................................................................................... 41
Figure 3.14: Compressive strength results of UNBPRT mortars at 7 days .......... 43

ix
Figure 4.1: Example of a stable foam ................................................................... 45
Figure 4.2: Storage modulus results for fly ash pastes at 23°C for 100
minutes .............................................................................................................. 51
Figure 4.3: Storage modulus results for BA-V pastes at 16, 23, and 30°C
for 100 minutes.................................................................................................. 51
Figure 5.1: Rate of heat evolution of cement pastes containing control
materials and blended fly ashes ......................................................................... 54
Figure 5.2: Rate of heat evolution of cement pastes containing control
materials and fly ashes ...................................................................................... 54
Figure 5.3: Rate of heat evolution of limestone cement pastes containing
control materials and blended fly ashes ............................................................ 56
Figure 5.4: Rate of heat evolution of limestone cement pastes containing
control materials and fly ashes .......................................................................... 57
Figure 5.5: Rate of heat evolution of PLC paste with gypsum additions ............. 57
Figure 5.6: Rate of heat evolution of BA-V paste with gypsum additions ........... 58
Figure 5.7: Flow curves of cement pastes containing control materials and
blended fly ashes ............................................................................................... 59
Figure 5.8: Flow curves of cement pastes containing control materials and
fly ashes ............................................................................................................. 59
Figure 5.9: Time of set of concrete containing control materials and
blended fly ashes ............................................................................................... 66
Figure 5.10: Time of set of concrete containing control materials and fly
ashes .................................................................................................................. 66
Figure 5.11: Compressive strength of concrete cylinders containing
control materials and blended fly ashes ............................................................ 67
Figure 5.12: Compressive strength of concrete cylinders containing
control materials and fly ashes .......................................................................... 68
Figure 5.13: RCPT results of 90-day concrete cylinders containing control
materials and SCMs with dashed lines indicating thresholds set by
ASTM C1202 .................................................................................................... 69

x
Chapter 1. Introduction

1.1. Motivation
Currently in Texas, fly ash, a coal combustion product, is the most commonly used
SCM and is typically used to replace 15–35% of portland cement by mass. ASTM
C618 [1] categorizes fly ash in two different classes: Class C fly ash, containing
greater than 18% calcium oxide, and Class F fly ash, consisting of a maximum of
18% of calcium oxide. The use of fly ash in concrete affects both the fresh state and
hardened state properties in positive ways by increasing the workability, reducing
bleeding, improving pumpability, reducing the heat of hydration, increasing
strength gain at later ages, and refining the pore structure to reduce permeability
[2]. Class F fly ash is used predominantly in Texas and in most of the U.S. due to
its ability to provide resistance to alkali-silica reaction (ASR) as well as sulfate
attack [3].

In 2019, the U.S. generated approximately 71.3 million metric tons of coal
combustion products, 37% (approximately 26.6 million metric tons) of which was
fly ash. Of the 26.6 million metric tons of fly ash, 16.1 million metric tons (61%)
were utilized in other industries [4]. Based on American Coal Ash Association
(ACAA) data [4,5], the percentage of fly ash utilized has increased by 5% between
2018 and 2019, while the amount produced has dropped 19%. This can be largely
attributed to competing fuel sources for electricity generation, such as natural gas,
as well as emission standards issued by the EPA in 2011 requiring coal-fired power
plants to install emission control systems, which, in turn, alters the composition of
the fly ash. Natural gas and renewable energy have become more price competitive,
so because of this, no new coal-fired power plants have been constructed in the U.S.
since 2013, with plant retirements forecasted through 2040 [6]. In the last five
years, there have been 324 coal-fired power generator closures in the U.S., and the
U.K. is set to have its remaining coal-fired power plants retired by 2025 [7]. This
presents a problem for the U.S. infrastructure, following a study conducted by the
American Road and Transportation Builders Association (ARTBA) [8], which
estimates that concrete production will increase more than 50% through 2033 [9].

In order to prepare for the upcoming shortage in fly ash supply, it is of interest to
explore potential alternatives for Class F fly ash. Many power plants have started
to burn blends of different coals, which results in fly ash with a different
composition than a traditional Class F fly ash. Suppliers have also started to blend
Class C and Class F fly ash to compensate for the lack of supply of Class F fly ash.
Since the final products of these two processes have different compositions than a

1
traditional Class F fly ash, they must be tested to qualify their use in concrete. With
the high demand for concrete and the low availability of Class F fly ash, rapid
screening methods must be developed to identify promising materials and reject
poor performers. This document presents an overview of research on fly ashes from
blended coal sources, blended fly ashes, and reactivity tests for supplementary
cementitious materials (SCMs).

Additionally, for these SCMs to be used in Texas Department of Transportation

(TxDOT) projects, they must meet TxDOT requirements. Permissible SCMs for
hydraulic cement concrete mixture designs are designated in Section 4.2.6 of Item
421 in the 2014 edition of the TxDOT Standard Specifications for Construction and
Maintenance of Highways, Streets, and Bridges [10]. Section 4.2.6 specifies the use
of Class C and F fly ashes, ground-granulated blast furnace slag (GGBFS), silica
fume, and metakaolin. Use of SCMs other than those outlined in Section 4.2.6
depends on the ability of the concrete mixture design utilizing that SCM to limit
expansion from alkali-silica reaction (ASR) to 0.08%, per testing procedures in
ASTM C1567 [11] when dealing with fine, intermediate, or coarse aggregate with
ASTM C1260 [12] values greater than 0.10%.

Besides durability improvements, TxDOT also recognizes other properties of

concrete mixtures that are changed with the use of SCMs, such as heat of hydration,
setting times, and water demand. Therefore, these properties will be discussed for
the materials of interest.

1.2. Literature Review

1.2.1. Supplementary Cementitious Materials

1.2.1.1. Blended Coal Fly Ashes

In 2011, the EPA issued emission standards requiring coal-fired power plants to
install emission control systems, which, in turn, reduces the quality of the fly ash.
Other methods to help meet these standards include changing the burning process
or changing the coal type that is burned. These methods alter the chemical
properties of the fly ash produced by the coal plants and change their effects on
cementitious mixtures. In Texas, power plants are increasingly burning Powder
River Basin (PRB) coal to comply with air quality regulations, either exclusively
or in blends with Texas Lignite Coal (TLC). PRB coal has a higher lime content
than TLC, so the fly ash produced has a higher calcium content and is classified as
Class C [13,14]. When the two coal sources are blended, the fly ash can be either a
Class F or a Class C, depending on the proportions used.

2
Class F fly ashes resulting from blended PRB and TL coals may have different
chemical properties than fly ashes made from burning 100% TLC. However, the
fly ashes can still meet the ASTM C618 criteria for Class F classification. Given
that Class C and Class F perform very differently with respect to control of heat of
hydration, sulfate attack, and alkali-silica reaction, it’s not clear if these blends will
perform like a Class F fly ash. To our knowledge, there is no published literature
comparing the performance of blended coal fly ashes to traditional Class F fly
ashes.

1.2.1.2. Blended Fly Ashes

ASTM C1697 [15] allows the blending of Class C and Class F fly ash. The standard
currently requires that the fly ashes to be blended meet ASTM C618 [1]
specifications. Similar to fly ashes resulting from blended coals, fly ashes resulting
from blended ashes may have similar composition to Class F fly ash, but it is not
clear that they will behave as a Class F fly ash in concrete.

In 1998, Naik et al. [16] made a case for blending fly ashes. The researchers
hypothesized that a blend of Class C and Class F fly ashes would produce a quicker
rate of hydration reaction in a cementitious mixture compared to Class F fly ash,
while maintaining a more favorable microstructure than a straight portland cement
mixture. Three fly ash blends were created: 75% Class C and 25% Class F fly ash,
50% Class C and 50% Class F fly ash, and 25% Class C and 75% Class F fly ash.
These blends were mixed at a 40% replacement of cement in concrete mixtures and
compared to two controls, a straight portland cement concrete and a concrete with
35% Class C fly ash. The concretes were tested for compressive strength, tensile
strength, flexural strength, drying shrinkage, abrasion resistance, salt scaling
resistance, and chloride permeability. Results showed that the blending of Class C
and Class F fly ash had a significant effect on the properties tested, demonstrating
that up to 50% of Class C fly ash can be replaced by Class F fly ash.

Antiohos et al. [17,18] also tested blends of Class C and Class F fly ashes from
Greece. Prior to blending, the fly ashes were ground using a ball mill to obtain ashes
of similar particle size. Similar to Naik et al. [16], the researchers created three fly
ash blends at the same proportions. Mortar compression tests were conducted on
these blended ashes at 20% and 30% cement replacement and compared to three
controls, a straight portland cement mortar as well as mortars containing both
original fly ashes at 20% and 30% cement replacement. Compression test results
showed that the 20% fly ash blend mortars outperformed the controls at later ages
with the highest strength mortar containing the blend with equal proportions of both
fly ashes. At 30% replacement, compressive strength had a direct correlation to
active silica content with the blend possessing the highest active silica content

3
having the highest compressive strength. Active silica content can be tested in
accordance to European Standard (EN) 197-1 [19] and correlates with the amount
of the total silica content that is soluble. Synergistic action was also claimed for the
fly ash blend mortars to account for the increment observed in compressive strength
testing. The authors believed that the instant hydration of free lime present in high-
lime ash temporarily increased the alkalinity of the mixture creating an “internal
activation” process. This increase in alkalinity assists in the dissociation of firm
glassy chains of the low-lime ash, releasing additional active centers, resulting in
this synergistic effect. In 2007, Antiohos et al. [20] tested an additional Class C fly
ash with lower active silica and higher CaO content than the original Class C fly
ash tested. This Class C fly ash was ground and blended with the same Class F fly
ash used in the previous study at 50% Class C-50% Class F fly ash and 25% Class
C-75% Class F fly ash. Mortar compression tests were conducted on the new
blended ashes and the previous blended ashes, blended at the same proportions, at
20% and 30% cement replacement and compared to four controls, a straight
portland cement mortar as well as mortars containing all three original fly ashes at
20% and 30% cement replacement. Results agreed with the previous study,
showing that blends with equal contributions from each fly ash were the most
effective at 20% replacement, while performance at 30% replacement was highly
dependent on the active silica content. In addition to compressive strength testing,
pozzolanicity and hydration product development tests were performed.

In a study by Tanikella [21], twenty different fly ashes were characterized for their
physical and chemical properties (thirteen Class C and seven Class F fly ashes) and
studied in both binary and ternary paste systems. Mixtures with a 20% cement
replacement by each individual fly ash were created and tested for the following
properties: the initial set time, the rate of strength gain (strength-activity index), the
heat of hydration, and the non-evaporable water and calcium hydroxide contents at
various ages. These mixtures were also compared to a control portland cement
mixture. Once testing was completed, a linear regression analysis was performed
for each individual test with the independent variables consisting of the physical
and chemical properties of the fly ash. For each test, a model was created used the
combination of independent variables that had the most significant effects. Once
models were created for both the Class C and Class F fly ashes, ternary pastes were
mixed using an orthogonal array that defined specific compositions for each
mixture and were analyzed and modeled in the same way as the binary systems. At
the end of testing, it was observed that the properties of the ternary binder systems
were not a weighted linear combination of the properties of binary pastes prepared
from individual fly ashes. Nonetheless, the most influencing variables of the ternary
blends were identified for each test.

4
Few studies have tested the performance of fly ash blends in sulfate attack and
alkali-silica reaction. Recently, Franklin and Rhodes [22] made a case for blending
fly ashes along the guidelines in ASTM C1697. They claimed that blended fly ashes
can be used where Class F fly ash is specified, as long as the sum of the oxides,
alkalis, and LOI are kept with the limits. They showed data demonstrating that the
blended ashes behave in between the Class F and Class C ashes with respect to
sulfate attack (C1012 [23]) and ASR (C1567 [11]).

In 1986, Mehta [24] conducted a study on the effect of fly ash composition on
sulfate resistance of cement. Sixteen fly ash samples were tested with varying
calcium contents and were blended with Type I portland cement at 25 and 40%
replacement of cement. Mehta concluded that the resistance to sulfate attack
depended solely on the type of aluminate phase present in the hydrated system at
the time of exposure. In cases when hydrated cement pastes contained monosulfate
hydrate or calcium aluminate hydrates prior to sulfate immersion, expansion would
take place due to ettringite formation. When hydrated cement pastes contained
ettringite prior to sulfate immersion, they would perform satisfactorily.

In 2011, Dhole et al. [25] evaluated the sulfate resistance of mortars containing
high-calcium and low-calcium fly ash, along with ternary blends containing both
high and low-calcium fly ash. The results showed that binary mixtures containing
Class F fly ash showed significantly improved sulfate resistance compared to the
control portland cement mixture. As the level of Class F fly ash increased, the
expansion values decreased. Binary mixtures containing Class C fly ash revealed
the opposite behavior. These mortars contained significant quantities of reactive
calcium aluminate and monosulfate, which produced abundant ettringite when
immersed in sulfate solution, leading to expansion in the bars. Formation of
monosulfate and calcium aluminate hydrates at early ages can be linked to
increased contents of phases such as C3A, C4A3Š, alkali sulfates, and reactive
calcium aluminate glass in fly ashes. Analysis of the fly ashes used in the study
found that fly ashes containing a calcium content of up to 15% CaO had
insignificant amounts of C3A. However, the C3A content increased significantly as
the calcium content increased past 15%. In ternary mixtures made with Class C fly
and Class F fly ashes, with lower amounts of reactive alumina phases, sulfate
resistance was improved compared to the control mixture. Ternary mixtures
containing high percentages of Class C fly ash and low percentages Class F fly ash
were able to control expansion better than the control cement mortar, but not
enough to provide adequate protection. Mixtures containing high percentages of
Class F fly ash and low percentages of Class C fly ash had good performance and
were crack-free even after a year. Additionally, it was found that a combination of

5
50:50 Class C and Class F fly ash in a mixture with 25 to 30% cement replacement
would perform satisfactorily in sulfate exposure.

In a study by Shehata and Thomas [26], eighteen different fly ashes were tested for
their ability to prevent expansion due to ASR. Data showed that, at a 25%
replacement level, low calcium, low alkali fly ashes (<20% CaO and <4% Na2Oeq)
controlled ASR expansion very well. The researchers concluded that a fly ash’s
ability to mitigate expansion is directly correlated with calcium, alkali, and silica
content, where the minimum level of replacement required to control expansion
generally increases as calcium or alkali content of the fly ash increases, or as the
silica content decreases. Accordingly, blending coal or blending fly ashes should
result in predictable performance in concrete since oxide composition plays a major
role. In another study by Shehata and Thomas [27], the alkali release characteristics
were studied for blended cements. Results concluded that the total alkalis
contributed from a fly ash with a CaO content greater than 20% were greater than
those contributed from fly ashes with a CaO content less than 20%. Additionally,
the total available alkalis for samples containing high-alkali fly ash were much
higher than those contributed by lower alkali fly ash. This should be considered
when blending fly ashes of high alkalinity with lower alkalinity fly ashes.

The potential problem with blending stems from the expectation that blends of coal
or blends of ashes will perform according to the weighted average of their oxides.
Fly ashes are generally characterized using x-ray fluorescence to obtain an oxide
analysis. However, these oxides are bound in different phases in the fly ash
depending on the coal burned and the burning conditions. While the majority of the
oxides are bound in glassy phases, there are several types of glassy phases present
and the composition of these phases impacts reactivity [28–32]. The crystalline
phases present in Class C fly ashes are quite different than in Class F fly ashes.
Durdzinski et al. [32] characterized four Class C fly ashes and found three to
contain both C3A and C4AF in quantities from 1-9%. Aughenbaugh [29] likewise
found C3A in a Class C fly ash, but none of the eight Class F fly ashes tested
contained cementitious phases. McCarthy et al. [33] claim that nearly all western
U.S. (PRB) fly ashes contain C3A or similar phases. The presence of crystalline
aluminate phases can affect the gypsum balance of the cementitious system, which
can affect setting [34] as well as affect the response of the system to sulfate attack,
as was shown by Mehta [24] and Dhole et al. [25]. Additionally, interactions with
chemical admixtures, particularly ones that adsorb on aluminate phases, are also
likely to be impacted due to these phases present in blended ash systems. Blending
these ashes with a Class F fly ash or other SCM only reduces the amount of the
phases but does not eliminate them. Therefore, it is possible that some blends will

6
not behave as their weighted averages of oxide components but will be dominated
by the phases present. This issue merits further investigation.

1.2.2. SCM Testing

ASTM C618 provides the specifications for fly ash and natural pozzolans for use
in concrete. Some parts of the specification are prescriptive, and others are
performance-based. The standard is problematic for natural pozzolans because it
fails to screen out inert materials like finely ground quartz [35]. In order to qualify
new sources of SCMs for use in Texas to accommodate the shortage in Class F fly
ash, including blended fly ashes, the materials need appropriate screening, beyond
the tests included in ASTM C618.

TxDOT Project 0-6717 [36] demonstrated that it is necessary to test SCMs more
extensively than ASTM C618 demands. Testing in that project included material
characterization such as x-ray diffraction and laser particle size analysis,
pozzolanicity testing through measurement of calcium hydroxide consumption, and
performance testing for ASR using ASTM C1293 (among many other tests
performed). The tests provided assurance that the SCMs that performed well were
appropriate for use in concrete. However, while such extensive testing is necessary
for new SCMs (and is even recommended by ASTM C1709 [37] for new,
alternative SCMs), it is impractical for testing every fly ash variant or blend that
occurs in Texas when power plants change operations. Therefore, there is a need
for simple, rapid screening tests.

1.2.2.1. Reactivity Tests

The prescriptive compositional (sum of SiO2 + Al2O3 + Fe2O3) requirements in
ASTM C618 are intended to ensure that the fly ash or natural pozzolan will react
with calcium hydroxide in a pozzolanic reaction [38]. However, given that at least
some part of these oxides may be bound in unreactive, crystalline phases (e.g.
quartz or mullite), it is not a particularly effective requirement. Measuring the
crystalline phases can be done using x-ray diffraction, but the glassy phases of fly
ash and most SCMs are the reactive ones, so any prescriptive compositional
specification would have to assess the composition of the glassy phases. The easiest
way to evaluate glassy phase composition is by subtracting crystalline phases from
the oxide composition to determine a bulk oxide composition of a glass [39]. This
value can then be used to calculate characteristics of a glass, such as the glassy
SiO2/Al2O3 ratio or non-bridging oxygen/tetrahedral (NBO/T) that give indications
of glass reactivity in alkaline solutions, such as in hydrating portland cement [39].

7
The best way to ensure pozzolanicity, however, is to test it. Since pozzolanicity
reflects the ability of the SCM to consume calcium hydroxide (CH) to form calcium
silicate hydrate (C-S-H), typically pozzolanicity tests involve measuring CH
consumption, either directly or indirectly. Traditional accelerated methods include
the modified Chapelle test [40], Frattini test [41], and lime reactivity test [42,43].
This section aims to provide background information on current reactivity tests and
examine their reliability.

1.2.2.1.1. Active Silica

The amount of reactive silica in an SCM is commonly associated with the reactivity
of the material. European Standard (EN) 197-1 defines the reactive silica content
as “the fraction of the silicon dioxide which is soluble after treatment with
hydrochloric acid (HCl) and with boiling potassium hydroxide (KOH) solution”
[19]. The quantity of reactive silica is determined as the difference between the total
silica content and the dry insoluble residues. In a study by Snellings and Scrivener
[44], it was determined that the test for active silica cannot be simply extrapolated
to simulate portland cement conditions and that this test is not suitable as an
estimator for SCM reactivity.

1.2.2.1.2. Modified Chapelle Test

The modified Chapelle test aims to measure the amount of CH consumed by a
pozzolan. As per instructions given in NFP 18-513 [40], 1 g of pozzolan and 2 g of
CaO are combined in a volumetric flask with 250 mL of deionized water and then
stirred at 85 ± 5 °C for 16 hours along with another flask that contains no pozzolan.
The flasks are then cooled to ambient temperature using running water and a
solution of 250 mL of deionized water and 60 g of sucrose are added to each flask.
Both flasks are then stirred for 15 minutes. Following stirring, 200 mL of each
solution is filtered, and 25 mL of the filtered solutions are removed to be titrated
with 0.1 N hydrochloric acid. The volume to titrate both solutions is used to
calculate the CH consumed by the pozzolan. In a study by Seraj and Juenger [45],
it was observed that when compared to compressive strength results, the Chapelle
test was efficient at identifying pozzolans with high early reactivity (7 and 28-day
strengths) but was poor at identifying pozzolans with later reactivity (90-day
strength). The Chapelle test also had high sensitivity to the particle size distribution
of the SCMs, as well as the cooling rate and filter paper used. These inconsistencies
make it more difficult to standardize the test. Additionally, quartz was tested as an
inert filler and resulted in a non-zero value for CH consumption and must be tested
more in order to determine if this is an issue with the test or the material.

8
1.2.2.1.3. Differential Scanning Calorimetry/Thermal Gravimetric Analysis
(DSC/TGA) for Pozzolanicity
The objective of DSC/TGA testing is to determine the mass loss of water during
the degradation of CH, which typically occurs between approximately 400 and 500
ºC [46]. TxDOT Projects 0-6717 [36] and 5-6717 [47] both implemented this
technique to quantify the CH content in control pastes and pastes containing SCMs.
The procedure in the study consisted of preparing sample mixtures with a
predetermined cement replacement percentage and curing them until their target
age was reached. Upon reaching the target age, the edges of the sample were
removed and discarded in order to remove calcium hydroxide crystals that orient
along the edges of the sample and walls of the container. Next, the sample was
crushed to pass the No. 100 sieve (150 μm) and then placed in a vacuum desiccator
for a minimum of two weeks to cease hydration. The samples were then ground to
pass the No. 325 sieve (45 μm) and placed back under vacuum prior to DSC/TGA
testing to prevent carbonation of the materials [47]. The DSC/TGA testing was
performed utilizing a Mettler Thermogravimetric Analyzer, Model TGA/DSC 1.
The heat flow and mass loss were recorded as the materials were heated from 40‒
1000 ºC at a rate of 20 ºC/min within a chamber with N2 gas to prevent carbonation
of the material while the test was underway. The DSC curve was used to pinpoint
the exact start and end temperatures when the mass loss occurred [48]. The mass
loss within this region is then used to calculate the amount of CH present in the
tested sample, which was converted to the amount of CH per gram of cement by
normalizing by ignited weight and accounting for the mass percent of SCM.

A significant issue with this test is the abundant opportunities for the sample to
carbonate before being measured in the thermogravimetric analyzer. Other issues
occur with the interpretation of the TGA curve, which typically significantly
overestimates the actual content of CH in the samples [46,48]. Another drawback
comes from the reaction between the SCMs and the calcium from existing C-S-H
in the system [49]. This means that the consumption of CH would be much higher
if no calcium was contributed from the existing C-S-H.

1.2.2.1.4. Frattini Test

The Frattini test aims to assess the pozzolanicity of a material by comparing the
concentration of calcium present in a solution in contact with hydrated cement with
the quantity of calcium capable of saturating a solution of the same alkalinity [41].
The procedure consists of mixing a blend of portland cement with a 30%
replacement of the cement by an SCM at a water to solid ratio of five and is kept at
40°C for seven days. The suspension is then filtered and titrated to determine the
saturation of the filtrate with respect to portlandite. Snellings and Scrivener [44],

9
performed the Frattini test on multiple different SCMs and concluded that it is
unclear how the results can be translated to a quantitative indication of compressive
strength since only the solution saturation with respect to portlandite is assessed.

1.2.2.1.5. Lime Reactivity Test

The Canadian standard for alternative SCMs [43] uses a lime reactivity test, and an
Indian standard uses a similar one [42]. The test involves mixing the SCM with
calcium hydroxide, sand, and an alkali source, then testing for compressive strength
after a specified period. The assumption is that strength gain is due to formation of
C-S-H, so this is an indirect method of evaluating pozzolanicity. The method is
easy to perform and repeatable. Kasaniya et al. [50] have modified the lime
reactivity test in what they call the University of New Brunswick Pozzolanic
Reactivity Test (UNBPRT). The UNBPRT was optimized for calcium hydroxide
to pozzolan ratio, water to binder ratio, curing time, solution composition, and
temperature.

1.2.2.1.6. Rapid, Relevant, and Reliable (R3) Test

A new reactivity test called the rapid, relevant, and reliable (R3) test method was
recently standardized in ASMT C1897 [51]. The goal of the R3 test is to create an
environment that isolates the reactivity of the SCMs separate from the reaction of
portland cement clinker. The procedure consists of mixing portlandite, deionized
water, potassium hydroxide (KOH), potassium sulfate (K2SO4), and calcite along
with the SCM. These pastes are then cured at 40°C and tested in two methods:
cumulative heat using isothermal calorimetry or bound water content using loss-
on-ignition. The R3 method has shown to be effective at distinguishing between
inert and reactive materials [52]. Measuring calcium hydroxide consumption in
theses pastes has also been able to separate hydraulic and pozzolanic reactivity
[53,54]. Bound water tests are preferred in lieu of isothermal calorimetry to provide
a cost-effective method to determine reactivity since the only equipment necessary
is an oven that can reach 400°C.

1.2.2.1.7. Compressive Strength Development

Another means of evaluating pozzolanicity can be through compressive strength
development of cement-SCM pastes or mortars. ASTM C618 tries to get at this
through the strength activity index (SAI), but this test is ineffective at assessing
pozzolanicity since it confounds reactivity with water demand by not having a
constant water content and is too generous in its strength criteria (75% strength with
a 20% replacement of cement with SCM)[35,38,55]. The SAI is so generous, in
fact, that an inert quartz filler passes [35]. The SAI test is in the spirit that the SCM

10
should “do no harm” to compressive strength or workability. The test requires that
mortars have a constant flow, not water-to-cementitious materials ratio. So, SCMs
that cause poor workability will require high water contents, which reduces strength
so much that the SCM does not pass the SAI. Because the test links workability and
strength, it is not a good measure of the impact of the SCM on strength. Because of
this, Bentz et al. [55] proposed a constant volume method for strength testing. The
ASTM task group working on developing a new specification for natural pozzolans
is considering a constant water-to-cementitious materials ratio by mass with a water
reducer specified, similar to the ASTM C1240 standard for silica fume. Workability
is still evaluated by mortar flow, but poor workability is reported and then
compensated through the use of a water reducer. Workability in ASTM C618 is
assessed through a mortar flow test.

There are other, more rigorous, means of using strength to evaluate pozzolanicity,
including a method used by some Europeans called the k-value efficiency factor
[56]. The k-value is defined as the portion of SCM in a cementitious mixture, which
can be considered as equivalent to portland cement, having the same properties as
a cementitious mixture without SCM. In this case, a portland cement mixture would
have a k-value of 1. Papadakis and Tsimas [56] estimated these k-values using an
empirical equation for compressive strength related to the water content, cement
content, SCM content, cement type, and time and curing. This shows that the SCMs
can easily substitute, equivalently, for portland cement up to a certain level. Also,
it may be possible to insist that strength of an SCM-containing mixture increase at
a faster rate than a cement-only mixture, a condition that is being considered in a
new ASTM specification under development for ground glass pozzolans.

1.2.2.1.8. Issues with Reactivity Tests

One thing to consider when evaluating reactivity is that it can be difficult to
differentiate in common test methods between the pozzolanic reaction and
cementitious reactions (e.g. from Class C fly ash or slag) and filler effects (e.g. from
ground limestone or quartz) [57]. Methods such selective dissolution and image
analysis can be used to determine the degree of reaction of an SCM, but the former
is prone to error and the latter is expensive and time consuming [58]. De-
convoluting these reactions is an area of research that needs attention.

1.2.2.2. Other Tests

Interactions between SCMs and air entraining agents are addressed by standards
for SCMs in different ways. ASTM C1709 [37] suggests using a foam index test,
the procedures of which are outlined in the specification. ASTM C1240 [59] for
silica fume includes a procedure for making a mortar with the SCM and an air

11
entraining agent and measuring the air content. These tests are important for fly ash
characterization because the loss-on-ignition value does not correlate directly with
air entrainer absorption. This is because activated carbon injection results in the
presence of activated carbon, which can have low loss on ignition, yet a large
impact on air entrainment [9]. Therefore, it is important to test air entrainment in
addition to loss on ignition.

None of the tests specified in ASTM C618 evaluate interactions with water
reducing admixtures. This is problematic because it is typical for concrete mixtures
to contain at least one chemical admixture and it is well known that admixture
incompatibility issues (oftentimes manifested in rheology, setting, strength gain, or
cracking problems) can occur in the presence of SCMs.

Finally, a comment should be made about the performance of SCMs with cements
that contain limestone fillers. There is evidence that limestone-SCM blends
improve compressive strength in a synergistic way because of the formation of
carboaluminate phases, particularly with high-alumina content SCMs [60].
Limestone can also improve setting time delays in high volume fly ash mixtures
[61]. However, limestone can disrupt the gypsum balance of an SCM-containing
mixture [60], so the interaction of SCMs and limestone-containing cements merits
attention.

1.3. Materials Selected for Testing

To find a potential substitute for traditional Class F fly ash, several ashes were
procured that meet ASTM C618 [1] Class F physical and compositional
requirements. The materials were classified into the following categories:

• Production Class F fly ash produced from 100% Texas lignite coal

• Class F fly ashes produced from blends of Powder River Basin (PRB) and
Texas lignite coal

• Fly ash blends of ASTM C618-conforming Class C and Class F fly ash,
blended to meet the compositional requirements of an ASTM C618 Class F
fly ash

• Fly ashes that do not meet ASTM C618 specifications and have been
remediated to meet the compositional and physical requirements

• A Class F fly ash procured from an international source

12
The material designations, classifications, and sources are presented in Table 1.1
for these materials.

Table 1.1: Procured Class F Fly Ashes

Designation Material Classification Source
F-G Texas
Production Class F Fly Ash
F-Z Ohio
BA-P Alabama
BA-S Blended Class C and F Fly Ash Arizona
BA-V Texas
BC-B Texas
Blended Coal Fly Ash
BC-M Texas
RM-S3 Arizona
Remediated Class F Fly Ash
RM-S9 Arizona
I-S International Class F Fly Ash Spain

In addition to these Class F fly ashes procured directly from suppliers, a Class C
fly ash and a milled bottom ash, produced at the same parent plant as F-G, were
procured to be blended in-house at the University of Texas at Austin. Table 1.2
shows these materials, and Table 1.3 shows the blends made using them.

Table 1.2: Procured Class C Fly Ash and Milled Bottom Ash
Designation Material Classification Source
C-H Class C Fly Ash Oklahoma
MBA Milled Bottom Ash Texas

Table 1.3: In-House Blended Materials

Designation Material Classification Blend Proportions
Blended Milled Bottom Ash and
BA-B 40% MBA/60% F-G
Class F Fly Ash
BA-H Blended Class C and F Fly Ash 54% C-H/46% F-G

To determine the effectiveness of reactivity testing, a variety of materials were

procured, including materials that possess pozzolanic or hydraulic reactivity, or a
combination of the two, or are inert. Table 1.4 shows the material designations,
classifications, and sources.

13
 Table 1.4: Materials Procured for Reactivity Testing
Designation Material Classification Source
Q Inert Quartz Powder West Virginia
C-H Class C Fly Ash Oklahoma
F-G Production Class F Fly Ash Texas
MBA Milled Bottom Ash Texas
M-D Metakaolin Missouri
P-O Overburden Pumicite New Mexico
P-P Pure Pumicite New Mexico
RM-S9 Remediated Class F Fly Ash Arizona
S Slag Grade 100 Illinois
T-P Pumiceous Tuff Arizona

To perform paste, mortar, and concrete testing, cements and aggregates were
procured. Two cements, one with a low limestone content (<5%) and one with a
higher limestone content (>5%), were procured. Both cements were sourced in
Texas and are designated OPC and PLC, respectively. The low limestone cement
was used for the majority of the testing, while the higher limestone cement was
used on select samples. Additionally, a fine aggregate from Texas that was
previously confirmed to be reactive using ASTM C1260 [12] was procured for
alkali-silica reactivity testing. For ASTM C1293 [62] testing, a limestone coarse
aggregate from Texas was procured. For all other concrete testing, a Colorado River
sand and gravel from Texas were used as the fine and coarse aggregate,
respectively.

14
Chapter 2. Material Characterization

Prior to conducting performance and reactivity testing, compositional and physical

analysis was performed on the supplementary cementitious materials (SCMs).
Characterization was completed to compare material properties with performance.
Additionally, testing in accordance with ASTM C311 [63] was performed to ensure
fly ash alternatives met the chemical and physical requirements specified in ASTM
C618 [1] for Class F fly ash. Table 2.1 outlines the characterization tests performed.
Testing was performed at the University of Texas at Austin, unless otherwise noted.
Material characterization was not performed on materials blended in-house. For
these materials, compositional and physical properties were determined as the
weighted sum of their individual components.

Table 2.1: Material Characterization Tests

Test Method Property Measured
X-ray Fluorescence Oxide Composition
Oven Drying Moisture Content
Oven Drying Loss on Ignition
Pycnometer Density
No. 325 Wet Sieve Fineness
Autoclave Soundness
Mortar Compression Testing Strength Activity Index
Mortar Flow Water Requirement
Flame Photometer Available Alkali
X-ray Diffraction Crystalline and Bulk Amorphous Composition
Laser Diffraction Particle Size Distribution
Scanning Electron Microscopy Particle Shape

2.1. Characterization Procedures and Results

2.1.1. Oxide Composition

X-ray fluorescence (XRF) testing, performed at the Texas Department of
Transportation (TxDOT), was used to determine the oxide compositions of the
materials. Non-traditional Class F fly ashes must meet the oxide composition
requirements set by ASTM C618 [1]. This includes a 50% minimum sum of the
silicon dioxide (SiO2), aluminum oxide (Al2O3), and ferric oxide (Fe2O3) and a
maximum calcium oxide (CaO) content of 18%. A maximum sulfur trioxide (SO3)

15
content of 5% is also included in the standard. Table 2.2 shows the results from
XRF testing.

Table 2.2: Oxide Compositions of Materials

SiO2+
SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O Al2O3+
Material
(%) (%) (%) (%) (%) (%) (%) (%) Fe2O3
(%)
F-G 54.81 21.19 4.72 11.24 2.52 0.45 0.15 0.99 80.71
F-Z 43.89 19.20 18.05 7.53 0.80 2.69 0.56 2.09 81.14
BA-B 54.65 21.34 5.01 11.17 2.47 0.40 0.13 0.95 80.99
BA-H 46.64 20.33 5.05 16.97 4.23 0.94 0.89 0.76 72.02
BA-P 45.30 20.26 9.93 12.69 2.88 1.06 1.19 1.58 75.50
BA-S 50.86 22.80 5.15 10.82 2.42 0.88 1.83 0.93 78.81
BA-V 45.40 17.76 7.33 18.58 4.44 1.18 1.28 0.87 70.49
BC-B 55.22 18.24 6.12 11.64 2.82 0.50 0.46 1.25 79.58
BC-M 51.10 19.26 5.05 14.10 3.76 0.72 0.68 1.42 75.41
RM-S3 52.36 22.91 5.71 8.25 0.90 3.92 1.29 0.93 80.97
RM-S9 62.60 18.83 4.10 2.87 0.64 0.23 2.39 2.46 85.52
I-S 57.41 20.08 9.32 2.87 1.65 0.57 2.24 1.82 86.81
Q 99.20 0.21 0.13 0.18 0.03 0.09 0.02 0.03 99.54
C-H 39.69 19.61 5.32 21.85 5.68 1.36 1.52 0.56 64.62
MBA 54.40 21.58 5.43 11.06 2.39 0.33 0.10 0.89 81.41
M-D 54.57 36.02 2.90 0.41 0.35 0.05 0.00 1.37 93.49
P-O 64.89 11.75 2.55 1.09 0.31 0.03 3.78 3.97 79.18
P-P 73.99 13.08 2.08 0.33 0.00 0.04 4.40 4.27 89.14
S 35.24 10.61 1.50 39.01 10.74 2.63 0.30 0.54 47.35
T-P 72.47 14.14 1.80 1.96 1.18 0.05 2.82 5.05 88.41

Most of the fly ash alternatives meet the requirements for a Class F fly ash as set
by ASTM C618, except for BA-V due to its CaO content greater than the 18%
maximum. The 18% limit set on CaO content for Class F fly ash was recently
added, after performance testing had started on the procured materials. Prior to this
addition, BA-V met the previous requirements for a Class F fly ash. Therefore, BA-

16
V is presented throughout the report as an alternative to traditional Class F fly ash,
although it no longer meets the criteria for a Class F fly ash in ASTM C618.

2.1.2. Moisture Content and Loss on Ignition

Moisture content (MC) and loss on ignition (LOI) of the materials were measured
in accordance with ASTM C311 [63] to ensure the fly ash alternatives met the
requirements set by ASTM C618 [1] for Class F fly ash. ASTM C618 limits MC
and LOI of Class F fly ash to 3% and 6%, respectively. Results from MC and LOI
testing on the fly ash alternatives are shown in Table 2.3.

Table 2.3: MC and LOI of Materials

Material MC (%) LOI (%)

F-G 0.46 ± 0.09 0.17 ± 0.01

F-Z 0.74 ± 0.04 2.15 ± 0.01
BA-B 0.57 ± 0.13 0.22 ± 0.01
BA-H 0.49 ± 0.05 0.34 ± 0.10
BA-P 0.65 ± 0.02 1.06 ± 0.02
BA-S 3.30 ± 0.09 1.40 ± 0.10
BA-V 0.70 ± -- 0.70 ± --
BC-B 0.62 ± 0.01 0.25 ± 0.05
BC-M 0.62 ± 0.09 0.16 ± 0.05
RM-S3 0.93 ± 0.03 1.45 ± 0.01
RM-S9 0.54 ± 0.06 0.38 ± 0.04
I-S 0.66 ± 0.08 1.00 ± 0.15
Red text indicates material that does not meet ASTM C618 requirements for a Class F fly ash

As-received, BA-S did not meet the moisture content requirements set by ASTM
C618. This can be remedied by the supplier by drying the fly ash prior to
distribution or adjusting blend proportions.

2.1.3. Density
Density measurements are performed on fly ashes to ensure uniformity compliance
between separate batches of materials. Since the materials procured for this study
came from a single batch, density was only measured for concrete mixture

17
proportioning. Measurements were completed at TxDOT using a gas-comparison
pycnometer and results are presented in Table 2.4.

Table 2.4: Physical Properties of Materials

Material Density (g/cm3) Fineness (%) Soundness (%)
F-G 2.34 24.43 -0.02
F-Z 2.57 12.34 -0.04
BA-B 2.47 22.01 -0.01
BA-H 2.50 19.33 0.02
BA-P 2.51 14.21 0.02
BA-S 2.33 11.19 0.02
BA-V 2.71 6.30 -0.06
BC-B 2.49 15.34 0.02
BC-M 2.50 16.54 -0.03
RM-S3 2.23 21.44 -0.04
RM-S9 2.25 17.21 -0.04
I-S 2.43 15.04 -0.02
Q 2.64 -- --
C-H 2.63 14.98 0.06
MBA 2.66 18.39 0.01
M-D 2.72 -- --
P-O 2.57 -- --
P-P 2.40 -- --
S 2.93 -- --
T-P 2.47 -- --

2.1.4. Fineness
Fineness of the materials was measured in accordance with ASTM C311 [63] to
ensure the fly ash alternatives met the requirements set by ASTM C618 [1] for
Class F fly ash. Testing is conducted by wet sieving a material through a 45µm (No.
325) sieve and. The amount of material retained on the sieve is then calculated and
cannot exceed 34% to meet Class F fly ash criteria. Results from fineness testing

18
are shown only for the fly ash alternatives in Table 2.4. All the fly ash alternatives
met the fineness criteria for a Class F fly ash.

2.1.5. Soundness
Soundness of the materials was measured in accordance with ASTM C311 [63] to
ensure the fly ash alternatives met the requirements set by ASTM C618 [1] for
Class F fly ash. Testing is conducted by measuring the amount of autoclave
expansion of paste bars. The expansion of the paste bars must remain below 0.8%
to meet Class F fly ash criteria. Results from soundness testing were provided by
the material suppliers and are shown only for the fly ash alternatives in Table 2.4.

2.1.6. Strength-Activity Index and Water Requirement

The strength-activity index (SAI) and water requirement for the materials were
measured in accordance with ASTM C311 [63] to ensure the fly ash alternatives
met the requirements set by ASTM C618 [1] for Class F fly ash. Mortar mixtures
were made with a varying water to cementitious materials ratio (w/cm) to achieve
a flow within ± 5 of the control portland cement mixture. Once the flow
requirements were met, mortar was cast into 2 in. cubes and tested in compression
at 7 and 28 days in accordance with ASTM C109 [64]. Compressive strength of the
mortar cubes must be at a minimum of 75% of the portland cement control at 7 or
28 days to meet Class F fly ash criteria. Results from SAI and water requirement
testing were provided by the material suppliers and are shown only for the fly ash
alternatives in Table 2.5. Testing was not performed on in-house blends but results
for their individual components are presented.

19
 Table 2.5: SAI and Water Requirement for Mortars
SAI (%)
Material Water Requirement (%)
7 Days 28 Days
F-G 82 92 95
F-Z 76 80 101
BA-P 93 92 95
BA-S 79 88 97
BA-V 93 112 91
BC-B 86 91 94
BC-M 89 103 94
RM-S3 81 85 99
RM-S9 79 90 101
I-S 89 89 97
C-H 104 102 96
MBA 87 88 98

2.1.7. Available Alkali

ASTM C311 [63] procedures were followed to determine available alkali of the fly
ashes presented as the equivalent sodium oxide content (Na2Oe). The samples were
prepared by dry mixing 5 g of SCM with 2 g of hydrated lime, adding 10 mL of
water, and mixing until the sample was uniform. The mixtures were then sealed in
a plastic vial and stored at 38°C for 28 days.

When the samples were ready for testing, the contents of the vial were removed,
broken up, and ground with water to create a uniform slurry containing no lumps.
The slurry was transferred into a beaker where water was added to bring the total
volume of the sample to 200 mL. The sample was left for 1 hour on a magnetic
stirrer, after which the sample was then filtered through a medium-textured filter
paper into a 500 mL volumetric flask and washed with hot water 8 to 10 times. The
filtrate was neutralized with dilute HCl (1+3) using 1 to 2 drops of phenolphthalein
solution as the indicator. The solution was cooled to room temperature and then
deionized water was added to fill the remaining portion of the volumetric flask.
Flame photometry was utilized to determine the amount of sodium and potassium
oxides in the solution, measured in ppm. The values obtained were then converted
to percent oxides for sodium and potassium and then converted to the equivalent

20
sodium oxide (Na2Oe = Na2O + 0.658 K2O). The soluble alkali contents of the fly
ashes are presented in Table 2.6.

Table 2.6: Available and Total Alkalis (Percent by Mass of SCM)

Available Total
Material Na2O (%) K2O (%) Na2Oe (%) Na2Oe (%)
F-G 0.19 0.46 0.49 0.79
F-Z 0.36 1.00 1.02 1.93
BA-B 0.16 0.42 0.43 0.75
BA-H 0.81 0.47 1.12 1.39
BA-P 0.90 0.93 1.51 2.23
BA-S 1.15 0.37 1.39 4.01
BA-V 0.70 0.36 0.94 1.85
BC-B 0.46 0.55 0.82 1.02
BC-M 0.47 0.77 0.98 1.61
RM-S3 0.73 0.41 1.00 1.90
RM-S9 1.01 0.47 1.32 2.44
I-S 1.09 0.86 1.66 3.45

2.1.8. Crystalline and Bulk Amorphous Composition

Quantitative X-ray diffraction (XRD) was performed to determine the crystalline
composition and bulk composition of the amorphous phases in the materials.
Samples were prepared by first grinding 2.5 g of SCM using a ceramic mortar and
pestle until all the material passed through a No. 325 sieve to ensure adequate
packing in the XRD sample holder. The material was then mixed with a reference
material, zincite (ZnO), at 10% by mass in an agate mortar and pestle. A small
amount of isopropanol was added to the powder mixture to ensure a homogenous
dispersion of the powders. Samples were then dried in an oven at 60°C for a short
period of time (less than 10 minutes) and placed in a vacuum desiccator until
testing.

Once the samples were ready for testing, the powders were packed into discs for
analysis in a Rigaku MiniFlex II. To minimize preferential orientation of the
crystals, a razor blade was used to form perpendicular cuts along the surface of the
powder before a glass slide was used to press down the powder without twisting.

21
Samples were then run using a continuous scan at a measurement range of 5° to 75°
2θ, a step size of 0.02° 2θ at a rate of 0.5° 2θ per second, a tube voltage of 40 kV,
and a tube current of 15 mA. The XRD pattern provides a plot (diffractogram) of
x-ray intensities throughout the measurement range, which can be found for all
samples in Appendix A. Crystalline phases were first identified using Jade before
completing the quantitative analysis using PDXL2. The composition of the
amorphous content was then determined as the difference of the oxides present in
the crystalline phases from the XRF data. Crystalline compositions of the samples
are presented in Table 2.7-Table 2.9. The bulk amorphous compositions for all the
materials are shown in Table 2.10.

Table 2.7: Amorphous Contents and Main Crystalline Phases Present in Coal
Combustion Products (CCPs)
C-H F-G F-Z I-S BC-B BC-M MBA
Phase Name
(%) (%) (%) (%) (%) (%) (%)
Amorphous 75.39 82.84 77.07 80.46 82.24 75.10 65.85
Albite -- -- -- -- -- -- --
Anhydrite 0.93 -- 2.39 -- -- -- 0.50
Anorthite -- -- -- -- -- -- 18.47
Augite -- -- -- -- -- -- 2.63
Clinoptilolite-Na -- -- -- -- -- -- --
Cristobalite -- -- -- -- -- -- --
Gehlenite 3.43 -- -- -- -- -- --
Hematite -- 0.47 2.49 0.77 0.32 -- --
Kalsilite -- -- -- -- -- -- --
Lime 0.47 0.40 0.75 -- -- -- --
Magnetite -- -- 4.43 6.83 0.42 -- --
Mullite 2.25 6.73 6.78 3.46 3.06 4.14 5.12
Periclase 2.32 -- -- -- -- 1.98 0.37
Portlandite -- -- -- -- -- -- --
Tricalcium
7.53 -- -- -- -- 3.89 --
Aluminate
Quartz 7.68 9.56 6.10 8.48 13.96 14.90 7.56

22
Table 2.8: Amorphous Contents and Main Crystalline Phases Present in Blended or
Remediated CCPs
BA-B BA-H BA-P BA-S BA-V RM-S3 RM-S9
Phase Name
(%) (%) (%) (%) (%) (%) (%)
Amorphous 76.04 78.82 84.89 66.34 80.22 71.83 83.04
Albite -- -- -- 4.16 -- -- --
Anhydrite -- 0.50 0.55 -- 0.62 -- --
Anorthite 7.39 -- -- -- -- -- --
Augite 1.05 -- -- -- -- -- --
Clinoptilolite-Na -- -- -- 13.52 -- -- --
Cristobalite -- -- -- 3.58 -- -- --
Gehlenite -- 1.85 -- -- -- -- --
Hematite 0.28 0.22 1.39 -- -- -- --
Kalsilite -- -- -- -- -- 2.36 --
Lime 0.24 0.44 0.20 -- 0.34 -- 0.41
Mullite 6.09 4.31 3.20 5.51 3.52 15.37 9.58
Periclase 0.15 1.25 1.34 -- 1.67 -- 0.84
Portlandite -- -- -- -- -- 2.46 --
Tricalcium
-- 4.07 2.02 -- 3.80 -- 0.12
Aluminate
Quartz 8.76 8.54 6.42 6.89 9.85 7.98 6.01

Table 2.9: Amorphous Contents and Main Crystalline Phases Present in Natural
Pozzolans and Slag
M-D P-O P-P S T-P
Phase Name
(%) (%) (%) (%) (%)
Amorphous 77.01 84.89 90.47 95.46 66.84
Albite -- 4.27 -- -- 7.84
Anatase 1.19 -- -- -- --
Anorthoclase -- 15.51 -- -- 14.70
Hematite 0.39 -- -- -- --
Lime 0.20 -- -- -- --
Merwinite -- -- -- 4.54 --
Mullite 1.94 -- -- -- --
Muscovite 7.29 -- -- -- --
Sanidine -- 4.14 8.38 -- --
Quartz 12.17 7.23 1.15 -- 10.62

23
 Table 2.10: Bulk Amorphous Composition and NBO/T of Materials
SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O SiO2/
Material NBO/T
(%) (%) (%) (%) (%) (%) (%) (%) Al2O3
F-G 43.36 16.35 4.25 10.83 2.52 0.45 0.15 0.99 4.39 0.12
F-Z 35.88 14.33 11.13 5.80 0.80 1.29 0.56 2.09 3.88 -0.01
BA-B 40.41 14.27 4.65 9.17 2.17 0.40 0.13 0.95 4.35 0.11
BA-H 36.48 15.02 4.83 13.03 2.97 0.65 0.89 0.76 3.87 0.50
BA-P 37.98 17.20 8.55 11.01 1.54 0.73 1.19 1.58 3.79 0.27
BA-S 27.20 16.35 5.09 10.64 2.42 0.88 0.75 0.63 3.79 0.12
BA-V 34.56 13.80 7.33 15.63 2.77 0.82 1.28 0.87 4.34 0.61
BC-B 40.40 16.05 5.37 11.64 2.82 0.50 0.46 1.25 5.14 0.21
BC-M 35.03 14.83 5.05 11.68 1.78 0.72 0.68 1.42 4.50 0.34
RM-S3 39.15 11.12 5.71 6.39 0.90 3.92 1.29 0.23 3.88 -0.04
RM-S9 53.88 11.91 4.10 2.39 0.00 0.23 2.39 2.46 5.64 -0.08
I-S 47.95 17.60 1.73 2.87 1.65 0.57 2.24 1.82 4.85 -0.08
C-H 30.63 13.88 5.32 14.90 3.36 0.81 1.52 0.56 3.43 0.83
MBA 36.00 11.13 5.24 6.67 1.64 0.33 0.10 0.89 4.28 0.10
M-D 37.90 32.30 2.51 0.41 0.35 0.05 0.00 0.69 2.57 -0.50
P-O 41.49 7.18 2.55 1.09 0.31 0.03 1.80 2.75 9.37 0.02
P-P 67.34 11.52 2.08 0.33 0.00 0.04 4.16 3.19 9.60 -0.01
S 33.58 10.61 1.50 36.69 10.19 2.63 0.30 0.54 5.64 2.48
T-P 46.51 9.80 1.80 1.96 1.18 0.05 0.61 4.40 8.70 0.04

Once the amorphous composition has been calculated, the network modifier
content can be used to express the level of reactivity of the SCMs [39,65]. In a pure
silicate glass, silicate tetrahedra are linked at each corner to exactly one other
silicate tetrahedron (T). This structure can be affected by the introduction of alkali
and alkaline earth metals (M2O and M’O), which are considered network modifiers
[65,66]. The M+ cations, such as potassium and sodium, destabilize the glass
network by bonding to oxygen atoms for charge balance, preventing them to act as
a bridge between silicon atoms, thus creating non-bridging oxygens (NBO) [65,66].
The glass present in fly ash and pozzolans typically consists of alkali and alkaline
earth aluminosilicate glass. Aluminum is known to act as a network former due to
its ability to be tetrahedrally coordinated in place of silicon, resulting in a net charge
of -1 on the tetrahedron, which must be balanced by a cation in the interstices [66].
By using the NBO/T parameter, the reactivity of the fly ash and pozzolans can be
assessed. NBO/T (shown in Table 2.10) was calculated using the amorphous oxide
atomic percentages, which were converted from the mass percentages, using
Equation 1.1 from Diaz-Loya et al. [65].

24
 𝑁𝑁𝑁𝑁𝑁𝑁 2(𝑁𝑁𝑁𝑁2 𝑂𝑂+𝐾𝐾2 𝑂𝑂+𝐶𝐶𝐶𝐶𝐶𝐶+𝑀𝑀𝑀𝑀𝑀𝑀−𝐴𝐴𝐴𝐴2 𝑂𝑂3 )
= (1.1)
𝑇𝑇 (𝑆𝑆𝑆𝑆𝑆𝑆2 +𝐴𝐴𝐴𝐴2 𝑂𝑂3 +𝐹𝐹𝐹𝐹2 𝑂𝑂3 )

NBO/T values typically range from 0 to 4, where a value of 0 represents a fully

polymerized network of silicon tetrahedra with little reactivity, and a value of 4
implies that there are no oxygen-bridged tetrahedral, so the glass has a high
reactivity [39,65]. Negative NBO/T values are due to insufficient amounts of
network modifiers to charge balance all the Al3+, resulting in no network modifiers
available to prevent oxygen atoms from bridging neighboring tetrahedra [39]. This
causes the aluminum to take on 6-coordination, which has a lower reactivity than
the 4-coordination aluminum that occurs when its charge is sufficiently balanced
[39,67].

The vitreous SiO2/Al2O3 ratio can also be used to evaluate the reactivity of fly ash
and pozzolans in high alkaline environments, such as in portland cement mixtures
and geopolymers. Optimum ratios have been identified for geopolymers, with
metakaolin having an optimum range of 2.0-4.3, Class C fly ash performing best
when the ratio is below 4.3, and Class F fly ash showing improved reactivity at low
SiO2/Al2O3 ratios [39]. The amorphous oxide atomic percentages were used to
calculate the SiO2/Al2O3 ratios, which are shown in Table 2.10.

2.1.9. Particle Size Distribution

Particle size distribution of the materials was determined through laser diffraction.
Samples were dispersed in isopropanol and particle size was measured using a
Malvern Mastersizer 2000 Laser Diffraction Particle Size Distribution Analyzer.
The results are presented as the particle diameters at 10%, 50%, and 90% of the
cumulative particle size distribution in Table 2.11.

Table 2.11: Particle Size Distribution of Materials

Material d10 (µm) d50 (µm) d90 (µm)
F-G 2.1 16.7 96.3
F-Z 2.2 17.6 80.0
BA-B 3.2 23.2 88.3
BA-H 2.1 14.4 80.9
BA-P 1.5 11.8 73.2
BA-S 2.1 12.4 56.8
BA-V 1.5 11.9 50.5
BC-B 1.5 11.4 67.2
BC-M 1.4 11.6 66.1

25
 Material d10 (µm) d50 (µm) d90 (µm)
RM-S3 2.3 20.2 94.3
RM-S9 1.9 18.1 84.5
I-S 1.5 11.5 72.3
Q 4.4 16.2 40.3
C-H 2.0 12.4 67.8
MBA 4.8 32.8 76.4
M-D 2.4 19.5 55.8
P-O 1.4 5.9 28.6
P-P 1.6 5.2 14.0
S 1.5 8.6 25.2
T-P 1.6 7.3 25.2

2.1.10. Scanning Electron Microscopy

Scanning electron microscopy (SEM) was performed on the fly ashes to determine
the particle shape of the materials. It has been shown that particle shape and size of
fly ash can affect the performance of cement-based materials, which can be altered
during the blending or remediation process [68]. SEM samples were mounted using
carbon conductive tape on a SEM stub. To reduce charging, a sputter coater
(Electron Microcopy Sciences) coated specimens for a 30-second gold/palladium
(60:40) at a current of 40 mA. Samples were loaded in a FEI Quanta 650 SEM and
images were taken under high vacuum and an accelerating voltage of 20 kV. SEM
images can be found in Appendix B.

26
Chapter 3. Reactivity Testing

With the decrease in availability of Class F fly ash in Texas, new sources and
materials are being tested as potential alternatives. Due to the high demand for these
materials, rapid screening methods are necessary to expedite their use in concrete.
Table 3.1 shows the tests selected to assess the filler effects, cementitious reactions,
and pozzolanic reactions of the supplementary cementitious materials (SCMs)
selected for this testing (Table 1.4).

Table 3.1: Reactivity Tests

Test Method Sample Type Reactions Measured
Cement-SCM Pastes Filler
Isothermal Calorimetry R3 Mixtures Pozzolanic
SCM-Water/Alkaline Pozzolanic
X-Ray Diffraction Hydrated Pastes Cementitious
Loss on Ignition R3 Mixtures Pozzolanic
Thermogravimetric R3 Mixtures Pozzolanic
Analysis Cement-SCM Pastes Pozzolanic
Cement-SCM Mortars Strength Development
Compressive Strength
Lime-SCM Mortars Pozzolanic

3.1. Reactivity Tests

3.1.1. Isothermal Calorimetry

3.1.1.1. Cement-SCM Pastes

Isothermal calorimetry was performed on cement-SCM pastes to examine the
effects the procured SCMs have on the hydration kinetics of cement, including filler
effects. Filler effects can be attributed to two main mechanisms: increased space
and enhanced nucleation [69]. Since fillers do not produce any hydrates, more space
is available for cement hydrates to form. Enhanced nucleation occurs when the filler
is finer than the cement, which provides additional nucleation sites on the surfaces
of the inert material for cement phases to form. It should be noted that isothermal
calorimetry results for SCM-cement pastes cannot identify whether or not an SCM
is pozzolanic. These tests only determine if the SCM alters the hydration kinetics
of cement through filler effects or through the SCM’s own hydraulic reactivity.

27
Pastes consisted of 25 g cementitious material and a w/cm of 0.45. For pastes
containing SCMs, 25% of the cement was replaced by mass. Pastes were mixed for
2 min using an overhead laboratory mixer at 1600 rpm and then placed into a glass
ampoule, sealed, and inserted into a TAM Air (Thermometric or TA Instruments)
isothermal calorimeter at 23°C. Heat release of the pastes was then measured for
72 h. Figure 3.1 and Figure 3.2 show the rates of heat evolution of the pastes.

24
OPC
Rate of Heat Evolved per gram of Cement

22
F-G
20
Q
18
C-H
16
MBA
14
J/(h*g)

RM-S9
12
S
10
8
6
4
2
0
0 10 20 30 40 50 60 70
Time (h)

Figure 3.1: Rate of heat evolution of cement pastes containing control materials and
SCMs

24
OPC
Rate of Heat Evolved per gram of Cement

22
F-G
20
Q
18
M-D
16
P-P
14
J/(h*g)

P-O
12
T-P
10
8
6
4
2
0
0 10 20 30 40 50 60 70
Time (h)

Figure 3.2: Rate of heat evolution of cement pastes containing control materials and
natural pozzolans

28
In Figure 3.1, Q increases the rate of heat evolution in paste per gram of cement,
despite being an inert material due to filler effects. At early ages, slow reacting and
low reactivity materials act as a filler, including low-calcium fly ashes and milled
bottom ash [69]. Therefore, it is expected that F-G and MBA pastes would behave
similarly to Q paste due to filler effects. Differences in Figure 3.1 between F-G,
MBA, and Q are due to the particle sizes of these materials, as shown in Table 2.11.
Since F-G and MBA have larger particle sizes than Q, they do not provide increased
nucleation sites, although they still provide increased space for cement hydrates to
form.

C-H is a high calcium fly ash that contains tricalcium aluminate and free lime,
which are reactive crystalline phases that can accelerate and alter cement hydration
reactions [69]. This results in a heat evolution curve for C-H paste that has an
amplified second peak that starts to overlap the main hydration peak in Figure 3.1.
The increase in rate of heat evolution per gram of cement of S paste compared to
OPC paste can be attributed to the hydraulic properties of slag.

In Figure 3.2, the main hydration peak of natural pozzolan pastes occurred sooner
than the control OPC paste due to the smaller particle sizes of the materials as
shown in Table 2.11. Additionally, an amplification of the second peak is observed
for the natural pozzolan pastes. This is likely due to a combination of filler effects
and reactive alumina in the natural pozzolans while attributing no additional
sulfates. The increase in aluminates in the system results in a quicker depletion of
the sulfates in the system, which can result in an amplification of the second peak
[69].

The observed filler and hydraulic behaviors of the SCMs in Figure 3.1 and Figure
3.2 provide information on how the materials will affect early-age concrete
performance, but do not serve as screening tools for SCM reactivity. In order to
measure SCM reactivity, both pozzolanic and hydraulic, other types of tests are
needed.

3.1.1.2. Rapid, Relevant, and Reliable (R3) Mixtures

An isothermal calorimetry test protocol for SCMs called the rapid, relevant, and
reliable (R3) test method recently standardized in ASTM C1897 [51] is being used
to evaluate the reactivity of SCMs. The goal of the R3 test is to create an
environment that isolates the reaction of the SCMs without the use of portland
cement. Mixtures are composed of five different solid components: (1) one of the
SCMs or quartz, (2) portlandite (Ca(OH)2), or CH in cement chemistry notation),
(3) calcite, (4) potassium hydroxide (KOH), and (5) potassium sulfate (K2SO4). CH
is added at a 3:1 mass ratio to the SCM to provide an excess amount of CH for

29
consumption by the SCM. Calcite is added as an excess counter anion for the AFm
phase. KOH and K2SO4 are added to deionized water to produce an activating
solution, while the solution-to-binder ratio of the system was 1.2. The cumulative
heat is measured using an isothermal calorimeter at 40ºC for 7 days, and higher
values suggest greater SCM reactivity. Similar to cement-SCM paste calorimetry,
R3 mixtures are mixed for 2 minutes using an overhead laboratory mixer at 1600
rpm. The cumulative heats for the R3 mixtures at 7 days are shown in Figure 3.3.

700
7-Day Cumulative Heat (J/g SCM)

600

500

400

300

200

100

0
OPC F-G Q C-H MBA M-D P-O P-P RM-S9 S T-P

Figure 3.3: 7-day cumulative heat release of R3 mixtures at 40°C

In Figure 3.3, the inert quart mixture, Q, had a substantially lower cumulative heat
value at 7 days than the rest of the mixtures, indicating that the test is effective at
distinguishing between inert and reactive materials. OPC and S mixtures had higher
cumulative heat values at 7 days than most of the mixtures, except for M-D, due to
their hydraulic properties. The hydraulic properties of the Class C fly ash mixture,
C-H, also attributed to the higher cumulative heat value than the other ashes and
pumices. M-D mixture had the highest cumulative heat value at 7 days, which is
due to metakaolin being a silica and alumina-rich material. CH, in the R3 mixture,
reacts with the silica and alumina present in M-D to form both calcium aluminate
and calcium silicate hydrates. The formation of calcium aluminate hydrates
generates more heat than calcium silicate hydrates, leading to higher heat release
values for M-D since it has a higher alumina content than the other materials, as
shown in Table 2.2 [70,71]. It is clear from the results in Figure 3.3 that the R3
isothermal calorimetry test can correctly identify inert and reactive SCMs, but does
not distinguish between hydraulic and pozzolanic reactivity.

30
3.1.1.3. SCM-Water/Alkaline Solution Pastes
Isothermal calorimetry was also performed on pastes containing only the SCM
mixed with water to identify the self-cementitious reactions of the SCMs. The idea
was to create an environment that only measures heat from hydraulic reactions in
order to separate hydraulic from pozzolanic SCMs. The same mixture proportions
and procedure were used as for cement-SCM pastes, but with a 100% cement
replacement of the cement with the SCM by mass. The cumulative heats for the
SCM-water pastes are shown in Figure 3.4 and Figure 3.5.

350
OPC
F-G
300
Q
Rate of Heat Evolved per gram of

C-H
250
MBA
Cementitious (J/g)

RM-S9
200
S
150

100

50

0
0 10 20 30 40 50 60 70
Time (h)

Figure 3.4: Cumulative heat evolved of control and SCM-water pastes

350
OPC
F-G
300
Q
Rate of Heat Evolved per gram of

M-D
250
P-P
Cementitious (J/g)

200 P-O
T-P
150

100

50

0
0 10 20 30 40 50 60 70
Time (h)

Figure 3.5: Cumulative heat evolved of control and natural pozzolan-water pastes

31
It is apparent from both Figure 3.4 and Figure 3.5 that only OPC, C-H, and F-G
pastes released any measurable heat by the end of 3 days. Slag paste was expected
to have a higher cumulative heat value than both C-H and F-G due to its latent
hydraulic properties, but it did not generate any heat at the end of 3 days. It was
determined that in order for the latent hydraulic properties of slag to be apparent,
an alkaline solution would be necessary. Therefore, it was concluded that
isothermal calorimetry of SCM-water paste mixtures is ineffective at distinguishing
between hydraulic and pozzolanic SCMs.

In the next test, a 0.3 M potassium hydroxide solution was used for SCM-alkaline
solution pastes, to match the potassium concentration of R3 mixtures, at a 0.9
solution-to-SCM ratio. The samples were prepared using the same procedure as for
cement-SCM pastes, and heat release was measured for 3 days at 40°C. The
temperature was increased from 23 to 40°C to help facilitate the reaction of the
SCMs. The cumulative heat values at 3 days of the SCM-alkaline solution pastes
are presented in Figure 3.6.

400

350
3-Day Cumulative Heat (J/g SCM)

300

250

200

150

100

50

0
OPC F-G Q C-H MBA M-D P-O P-P RM-S9 S T-P

Figure 3.6: 3-day cumulative heat release of SCM-alkaline solution pastes at 40°C

Both the OPC and S pastes in Figure 3.6 have the highest cumulative heat values at
3 days showing that the test clearly identifies hydraulic materials. However, pastes
containing F-G and RM-S9, both low-calcium fly ashes (Table 2.2), show a
measurable heat release due to some form of geopolymerization. This can lead to
false positives when identifying hydraulic materials. Additionally, the Class C fly
ash paste, C-H, had a lower cumulative heat value than F-G despite having a higher
calcium content. C-H paste began to stiffen prior to being inserted in the
calorimeter. The rapid reaction of C-H prevented a complete measurement of the
heat release. This can lead to false negatives in the test as well, with quick reacting

32
materials appearing non-hydraulic. Therefore, it was concluded that calorimetry
testing on SCM-alkaline solutions is ineffective at distinguishing between
hydraulic and pozzolanic SCMs.

3.1.2. X-ray Diffraction

3.1.2.1. Hydrated Pastes

X-ray diffraction (XRD) analysis was on conducted on select pastes from SCM-
alkaline solution calorimetry testing to determine through the presence of
cementitious reaction products if hydraulic reactions were indeed occurring under
these testing conditions. After 3 days, samples were removed from the calorimeter
and paste was extracted from the glass ampoule. Hydrated pastes were then placed
under vacuum at 30 mm-Hg for a minimum of 2 weeks to stop hydration [48].
Samples were then ground using a ceramic mortar and pestle until all the material
passed through a No. 325 sieve to ensure adequate packing in the XRD sample
holder. The powders were packed into sample holders for analysis in a Bruker D8.
To minimize preferential orientation of the crystals, a razor blade was used to form
perpendicular cuts along the surface of the powder before a glass slide was used to
press down the powder without twisting. Samples were then run using a continuous
scan at a measurement range of 5° to 75° 2θ, a step size of 0.02° 2θ at a rate of 2°
2θ per second, a tube voltage of 40 kV, and a tube current of 30 mA. The main
crystalline phases present in the hydrated pastes are presented in Table 3.2.

Table 3.2: Main Crystalline Phases in Hydrated SCM-Alkaline Solution Pastes

Designation Main Crystalline Phases
OPC Portlandite, C3S, C2S, C-S-H, Calcite, Ettringite, Gypsum
C-H Quartz, Mullite, Periclase, Calcite
F-G Quartz, Mullite, Hematite
S C-S-H, Merwinite

Only OPC and S alkaline solution pastes contained cementitious reaction products
when analyzed in XRD. This agrees with the results from calorimetry testing, where
OPC and S pastes had the highest measurable heat values. Although F-G and C-H
pastes produced heat in calorimetry testing, they contained no hydration products.
Since the sodium aluminosilicate hydrate (N-A-S-H) gel formed during
geopolymerization is amorphous and, therefore, not easily measured by XRD, it is
possible that the heat released from F-G and C-H is from geopolymerization. This
confirms that calorimetry on SCM pastes with alkaline solutions is not appropriate
for separating pozzolanic from hydraulic reactions.

33
3.1.3. Loss on Ignition

3.1.3.1. R3 Mixtures
ASTM C1897 [51] includes an alternative method to isothermal calorimetry to
assess SCM reactivity using the same R3 mixtures and curing time. The procedure
involves measuring the loss on ignition in a furnace between 40°C and 350°C. This
mass loss is related to the water chemically bound to hydration products, such as
calcium silicate hydrates, calcium aluminate hydrates, and other minor hydrates
[46,72]. A higher mass loss indicates a higher amount of chemically bound water,
meaning that the SCM has greater reactivity.

Excess R3 mixtures from isothermal calorimetry testing were cast into sealed plastic
vials and cured in an oven at 40°C for 7 days. After the curing period, the samples
were removed from the vials and crushed, resulting in a paste. Following crushing,
10 g of the paste was spread evenly on a petri dish and placed back in the 40°C
oven to dry for 24 ± 1 hours. The samples were weighed then calcined at 350°C for
2 hours to determine the mass loss between the two temperatures. The results from
this bound water testing are shown in Figure 3.7.

14
Bound Water (g/100g of the total dried mixture)

12

10

0
OPC F-G Q C-H MBA M-D P-O P-P RM-S9 S T-P

Figure 3.7: 7-day bound water content of R3 mixtures at 40°C

Similar to results from R3 isothermal calorimetry testing, the inert quartz mixture,
Q, had the lowest bound water value of all the materials, showing that the test is
effective at separating inert and reactive materials. The bound water values of OPC,
S, and M-D mixtures were lower relative to other materials than their cumulative
heat values, while the pumices were higher. This is likely due to some hydration
reactions generating more heat, while binding similar amounts of water to hydration

34
products with a lower heat release. M-D mixture had the highest bound water value
at 7 days, agreeing with its high cumulative heat value in Figure 3.3. The formation
of calcium aluminate hydrates bind more water than calcium silicate hydrates,
leading to higher bound water values [70,71]. Therefore, the bound water method
in ASTM C1897 [51] is equally effective as the calorimetry method at screening
inert materials, though there are some differences in the methods if one wants to
make an interpretation regarding the “degree” of reactivity.

While the R3 tests in ASTM C1897 [51], both through calorimetry and bound water
measurement, successfully screen inert SCMs from reactive ones, neither can
distinguish between pozzolanic and hydraulic reactivity. Therefore, an additional
step was added to the standardized bound water test to measure the water bound to
the remaining CH in the mixture [73]. After weighing the samples at the end of the
bound water test, the samples were returned to the oven and heated to 500°C for an
additional 2 hours. The mass loss between 350 and 500°C can be used as an
estimate for unreacted CH in the mixture. The results from this test, herein denoted
as the R3P test, are presented in Figure 3.8.

60.0
CH Content (g/100g of the total dried mixture)

50.0

40.0

30.0

20.0

10.0

0.0
OPC F-G Q C-H MBA M-D P-O P-P RM-S9 S T-P

Figure 3.8: R3P calcium hydroxide content of R3 mixtures at 7 days

It is apparent in Figure 3.8 that OPC mixture has a higher CH content than the inert
material. That is because OPC is producing additional CH as it reacts and does not
react with any of the CH present in the mixture, so there is more CH in the final
mixture. The remaining mixtures all had CH content lower than Q, showing that
they are reacting pozzolanically, consuming CH through the pozzolanic reaction.
The mixtures with slag (S) and Class C fly ash (C-H) consume the least CH of the
SCMs, as expected since their reactivity is more from hydraulic reactions than
pozzolanic ones. It can be concluded that running the R3P test can help distinguish

35
between pozzolanic and hydraulic materials, but the analysis of results needs
careful attention.

The results from R3 isothermal calorimetry and R3P testing can be paired to
distinguish between inert and reactive materials, and hydraulic and pozzolanic
reactivity. Thresholds, adapted from Suraneni et al. [53] and Kalina et al. [35], were
used to separate reactivity types into categories, shown in Figure 3.9. A material is
classified as inert if it had a CH content in R3P testing greater than 40 g/100 g of
the total dried mixture and a 7-day cumulative heat in R3 testing below 100 J/g
SCM. A material is classified as pozzolanic, less reactive if it had a CH content
between 25 and 45 g/100 g of the total dried mixture and a 7-day cumulative heat
between 100 and 350 J/g SCM. A material is classified as pozzolanic, more reactive
if it had a CH content below 35 g/100 g of the total dried mixture and a 7-day
cumulative heat greater than 350 J/g SCM. A material is classified as hydraulic,
less reactive if it had a CH content greater than 45 g/100 g of total dried mixture
and a 7-day cumulative heat between 100 and 350 J/g SCM, and hydraulic, more
reactive if it had a CH content greater than 35 g/100 g of total dried mixture and a
7-day cumulative heat greater than 350 J/g SCM. While more work is needed to
define and refine the thresholds in Figure 3.9, it is apparent that this is a promising
method to rapidly categorize SCM reactivity. Furthermore, the same type of plot in
Figure 3.9 can be developed for R3 bound water and R3P test results, with similar
effectiveness.

700
Pozzolanic, Hydraulic,
More Reactive
600
7-Day Heat Release (J/g SCM)

500

400 OPC
F-G
300 Q Hydraulic,
C-H Less Reactive
MBA
200 M-D
P-O Pozzolanic,
P-P Less Reactive
100 RM-S9
S Inert
T-P
0
0 10 20 30 40 50 60
CH Content (g/100g of the total dried mixture)

Figure 3.9: Classifications of SCMs based on reactivity using the 7-day heat release of R3
mixtures at 40°C and CH content of R3 mixtures using the R3P method. Thresholds
adapted from Suraneni et al. [53] and Kalina et al. [35]

36
3.1.4. Thermogravimetric Analysis

3.1.4.1. R3 Mixtures
The R3P test assumes that the water lost between 350°C and 500°C is due to the
decomposition of CH. Thermogrametric analysis (TGA) testing was done on the
same samples to verify this assumption.

After crushing samples down for R3 bound water testing, excess sample was placed
under vacuum at 30 mm-Hg for a minimum of 2 weeks to stop hydration [48].
Samples were then ground using a ceramic mortar and pestle until all the material
passed through a No. 325 prior to running differential scanning calorimetry (DSC)
and thermogravimetric analysis (TGA). 10 mg of sample was measured for mass
loss between 40 and 550°C using a Mettler Thermogravimetric Analyzer, Model
TGA/DSC 1 at a rate of 20°C/min within a N2 gas filled chamber flowing at a rate
of 50 mL/s to prevent carbonation. The DSC curve was used to determine the start
and end of CH decomposition along the TGA curve, typically between 400 and
500°C [46]. This mass loss was used to determine the CH content of R3 mixtures
and was then compared to the results in R3P testing (referred to as single point
measurement) in Figure 3.10.

37
 60.00

CH Content Using DSC/TGA (g/100g of the total dried mixture)

R² = 0.89

50.00

40.00

30.00

20.00

10.00

0.00
0.00 10.00 20.00 30.00 40.00 50.00 60.00
CH Content Using Single Point Measurement (g/100g of the total dried
mixture)

Figure 3.10: CH content of R3 mixtures at 7 days using DSC/TGA vs R3P

The results in Figure 3.10 show that the CH contents of R3 mixtures were higher
when measured using the DSC/TGA method than through loss on ignition in a
furnace (single point measurement); this is apparent by the data points being
skewed higher than the black dotted line of equality in Figure 3.10. This difference
in measured CH contents could be due to carbonation during the 24-hour drying
period prior to calcining in the R3P method. To determine if the R3P method is a
suitable replacement for using DSC/TGA, a least squares regression line was fit to
the data. A strong correlation was found between the results, suggesting that while
the values measured are different, the trends are the same, validating the use of the
R3P method as a quicker, cheaper, and more user-friendly method to DSC/TGA.

3.1.4.2. Cement-SCM Pastes

A traditional way to assess pozzolanic reactivity is by measuring the CH content of
cement-SCM pastes over time through DSC/TGA, as was performed in TxDOT
Projects 0-6717 [36] and 5-6717 [47]. When pastes have lower CH contents than a
control cement-only paste or a paste containing cement and quartz, this suggests
that the pozzolanic reaction has occurred, reducing CH in the mixture.

38
Pastes consisted of a 25% replacement of the cement with SCM by mass and a
w/cm of 0.45 and were mixed using the same procedure as was used for isothermal
calorimetry for cement-SCM pastes. After mixing, pastes were poured into four
separate plastic containers and placed in a curing room at 23°C and 100% relative
humidity to be removed at 1, 7, 28, or 90 days. Upon reaching a testing age, one of
the containers was removed from the curing room and the outer edges of the sample
were cut and discarded to prevent bias in the results due to CH crystals that orient
along the surface of the container [74]. The remaining sample was then crushed to
pass a No. 100 sieve (150 µm) and placed under vacuum at 30 mm-Hg for a
minimum of 2 weeks to stop hydration [48]. Samples were then prepared and
measured in the same manner as TGA testing on R3 mixtures, except the testing
temperature range was increased from 40 to 1000°C. CH content of cement-SCM
pastes normalized per gram of anhydrous cement are presented in Figure 3.11.

25
OPC
CH Content per gram Anhydrous Cement

F-G
20
C-H

MBA
15
M-D
(%)

P-O
10
P-P

RM-S9
5
S

T-P
0
1 7 28 90
Days

Figure 3.11: CH content of cement-SCM pastes at 25% replacement of the cement by

mass with gray dashed lines indicating Q paste calcium hydroxide content with 5% error

Due to limited replicate testing, a representative error of 5% was applied to the

results, adapted from Kim and Olek [48]. The CH content of Q paste is represented
by the gray dashed lines at each age with an upper and lower line indicating ± 5%
of the measured value. At 1 day, none of the SCMs are reacting to form or consume
CH, resulting in all pastes having similar CH contents to each other and the OPC
and Q pastes. By 7 days, both M-D and S pastes had lower CH contents than OPC
and Q pastes, while the remaining pastes did not. Metakaolin is known to rapidly
consume CH when used in cement-based systems due to its high pozzolanic
reactivity [75]. Although S paste had a comparable CH content to M-D at 7 days,
it has been shown that this is not due to pozzolanic reaction, but rather a decrease

39
in the initial formation of CH in slag-cement systems [75]. At later ages, the CH
content of S paste does not change and remains below that of OPC and Q pastes.
This can make it difficult to determine whether low CH content of a paste is related
to decreased initial formation or consumption of CH.

By 90 days, all the cement-SCM pastes had lower CH contents than Q paste,
indicating that they are all pozzolanic. Both P-O and P-P pastes had considerable
decreases in CH content by 90 days, resulting in similar CH contents to M-D paste.
M-D paste had similar CH contents between 28 and 90 days, which is likely due to
limited reactants and space for hydrates to precipitate in [76]. Additionally, most of
the pastes had lower CH contents than OPC paste at 90 days as well, except for
MBA paste, indicating that it is the least pozzolanic material of the SCMs tested.

While measuring CH content of cement-SCM pastes at 90 days gives a very good

indication of whether or not an SCM is pozzolanic, there are some problems with
universally suggesting this method for pozzolanicity assessment: (1) the test is not
standardized and would be difficult to standardize, (2) the test takes 90 days, (3)
the test uses expensive DSC/TGA equipment, and (4) the test can be misleading for
materials such as slag. Therefore, this test is recommended for researchers
interested in understanding SCM behavior, but not for routine screening of
materials.

3.1.5. Compressive Strength

3.1.5.1. Cement-SCM Mortars

ASTM C618 [1] uses compressive strength development to assess SCM reactivity
through the strength activity index (SAI). This test is problematic, as discussed
earlier, because of the variable w/cm used, making it difficult to isolate the role of
the SCM on strength from the role of the water content. Therefore, for this study,
w/cm was fixed when evaluating the ability of compressive strength to assess SCM
reactivity. It should be noted that the impact of SCMs on strength development is
related not just to SCM reactivity, but also to their ability to impact cement
hydration kinetics among other factors.

Compressive strength testing of mortars was conducted in accordance with ASTM

C109 [64]. Mortars consisted of a 25% replacement of the cement with SCM by
mass and a fixed w/cm of 0.485. Standard graded sand, as specified by ASTM C109
[64], was used at a ratio of 2.75 to the cementitious material. Mortars were cast into
2 in. cubes to be tested at 1, 3, 7, 28, 56, and 90 days. Figure 3.12 shows the results
from mortar compressive strength testing.

40
 9000

8000 OPC
F-G
Compressive Strength (psi) 7000
Q
6000
C-H
5000 MBA
4000 M-D
P-O
3000
P-P
2000
RM-S9
1000 S

0 T-P
1 3 7 28 56 90
Days

Figure 3.12: Mortar cube compressive strength results of cement-SCM mortars

Results from mortar cube compressive strength testing showed unreliable results
with some mortars decreasing in compressive strength at later ages. To verify the
results, mortar compressive strength testing was repeated using 2 in. by 4 in.
cylinders since they have been found to have better precision [77]. Mortar cylinders
were tested at 3, 7, 28, and 90 days and the results are presented in Figure 3.13.

9000
OPC
8000 F-G
Compressive Strength (psi)

7000 Q

6000 C-H
MBA
5000
M-D
4000
P-O
3000 P-P
2000 RM-S9
S
1000
T-P
0
3 7 28 90
Days

Figure 3.13: Mortar cylinder compressive strength results of cement-SCM mortars

At 3 days, all the SCM mortars had comparable or lower compressive strength to
Q mortar (Figure 3.13). This is expected since most SCMs act as fillers at early
ages. By 7 days, only C-H, M-D, and S mortars had higher compressive strengths
than Q mortar. The early age strength gain for C-H and S mortar can be attributed

41
to the hydraulic properties of Class C fly ash and slag. Metakaolin is a fast-reacting
pozzolan, which allows M-D mortar to increase strength at early ages. Since M-D
mortar had similar compressive strength values at 7 days as C-H and S, it can be
difficult to determine if early strength gain is related to hydraulic reactivity or fast
pozzolanic reactivity. At 28 days, most of the mortars had higher compressive
strengths than Q mortar, except for MBA mortar. However, MBA mortar saw a
larger strength gain between 7 and 28 days. The lower early age compressive
strength of MBA mortar compared to Q mortar can be related to the larger particle
size of MBA than Q, as shown in Table 2.11. At 90 days, all the mortars had higher
compressive strengths than Q mortar, with most also having higher compressive
strength than OPC mortar, except for MBA mortar. This indicates that MBA is the
least reactive material tested, as was seen in other reactivity tests. Additionally, by
90 days many other mortars had comparable strengths to M-D mortar showing that
slower reacting materials can reach the same degree of reaction as fast-reacting
materials at later ages.

It can be concluded from mortar compressive strength tests that the 90 day strength
of mortars is a good indicator of SCM reactivity. Using earlier ages to assess SCM
reactivity can be misleading since the pozzolanic reaction is slow for many SCMs.

3.1.5.2. Lime-SCM Mortars

Recently, a modified lime reactivity test method was developed, known as the
University of New Brunswick Pozzolanic Reactivity Test (UNBPRT) [50].
Reactivity of materials is determined through compressive strength of lime mortars.
Similar to R3 testing, UNBPRT binder contains SCM, CH, and calcium carbonate.
CH is added to the binder at a 1:2 ratio to the SCM and calcium carbonate is added
at a 1:15 ratio to the binder. The same potassium sulfate solution used in R3 testing
is used for UNBPRT mortar at a water-to-binder mass ratio of 0.65 and standard
graded sand is added at a mass ratio of 2.5 to the binder. Although the test was
designed using 2 in. cubes, 2 x 4 in. cylinders were used instead for their improved
precision [77], as was seen in cement-SCM mortar testing. After casting, cylinders
were stored in a sealed container over water at 23°C for 24 hours. After the 24-hour
period, the container was moved to an oven at 40°C for 2 days and then demolded.
Once demolded, the cylinders were submerged in distilled water in sealed
containers at 40°C for an additional 4 days. Some mortars were too soft to be
removed from their molds at 3 days and were left in the mold over water at 40°C
until they were stiff enough to be removed, or at the end of the 7-day curing process.
At 7 days, the containers were removed from the oven and allowed to reach room
temperature before being tested in compression. Figure 3.14 shows the results from
UNBPRT. Thresholds were proposed by Kasaniya [78], where an inert material

42
would have a compressive strength below 2 MPa at 7 days, low reactivity would
have a strength of 2-5 MPa, moderate reactivity 5-10 MPa, high reactivity 10-20
MPa, and very high reactivity would have a compressive strength greater than 20
MPa.

18

16
Compressive Strength (MPa)

14
High Reactivity
12

10

8
Moderate Reactivity
6

4
Low Reactivity
2
Inert
0
OPC F-G Q C-H MBA M-D P-O P-P RM-S9 S T-P

Figure 3.14: Compressive strength results of UNBPRT mortars at 7 days

At 3 days, F-G, Q, and MBA UNBPRT mortars were too soft to be removed from
their molds at 3 days and were not demolded until the 6 days, for F-G and MBA
mortars, and 7 days for Q mortar. At 7 days, Q UNBPRT mortar was not set enough
to be tested in compression and was assigned a compressive strength value of zero.
All other UNBPRT mortars were able to be tested in compression, indicating that
the materials are reactive. As was seen in other reactivity tests, MBA was
determined to have the lowest reactivity in the UNBPRT while M-D had the
highest. The UNBPRT does not separate between pozzolanic and hydraulic
reactivity, despite being called a pozzolanicity test. However, it uses the same
equipment as the SAI test, in addition to a 40°C oven, and is more effective at
screening out inert materials. Therefore, the UNBPRT is a promising method of
screening inert from reactive materials in a 7-day period with limited need for new
or advanced testing equipment.

43
Chapter 4. Admixture Interaction Testing

The objective of the testing presented in Chapter 4 was to investigate interactions

between the SCMs and chemical admixtures since some SCMs are known to cause
admixture incompatibilities as discussed in Section 1.2.2.2. The first phase was to
assess the interaction effects of the fly ashes with an air-entraining agent (AEA) in
accordance with the foam index test, air-entrainment of mortar test, and air void
spacing analysis. The second phase was to evaluate interaction effects between the
SCMs and high-range water reducers via shear rheology tests.

4.1. Interaction with Air-Entraining Agent

4.1.1. Foam Index Test

Initial testing of the interaction effects between the SCMs and an AEA was
performed using the foam index test. The foam index test is a test that is used to
rapidly determine the relative levels of AEA needed for materials that affect air-
entrainment in concrete. As there are different variations of the test, the procedure
used is summarized here:

1) Determine the initial solution concentration for the AEA to use for the test.

a) A dilution ratio of 1:20 was selected. This results in a 5 vol.% AEA solution.
For example, this can be achieved by adding 10 mL of AEA to 200 mL of
water.

2) In a small, capped bottle, add 25 g of cementitious material and mix it well via
shaking. (Note, when an SCM is used, add 16.75 g of cement and 8.25 g of
SCM to the bottle).

3) In a different small, capped bottle (e.g., a 250 ml wide-mouth Nalgene®-type

container with a tight-fitting cap) add 50 g water.

4) Add the cementitious material from Step 2 to the water in Step 3. Using the cap,
close the bottle and shake the bottle for 10 seconds. This is the initial agitation
phase.

5) Stop shaking the bottle. Open the lid and add 20 µL of 5% AEA solution.

6) Close the lid of the bottle and shake it for 20 seconds.

44
7) Open the lid and allow the sample to sit for 20 seconds. If a metastable foam is
noticed after 20 seconds (see Figure 4.1), then the test is done. If not, proceed
to Step 8.

8) Open the lid of the bottle and add an additional 20 µL of 5% AEA solution to
the sample.

9) Repeat steps 7 and 8 until a stable foam is observed.

10) Record the total number of drops of air entraining admixture solution added to
achieve a stable foam (n) and the solution concentration of the air entraining
admixture solution used (C), and the drop volume (d).

Figure 4.1: Example of a stable foam

SCMs were evaluated using Sika® AIR. The manufacturer’s recommended dosage
is 16-195 mL per 100 kg of cementitious material (cm). Table 4.1 presents the
results from the foam index test. The total volume of AEA per 100 kg of sample
was calculated using Equation 4.1:

𝐴𝐴𝐴𝐴𝐴𝐴 = 𝑛𝑛 𝑥𝑥 𝑑𝑑 𝑥𝑥 𝐶𝐶 (4.1)

where n = total number of drops to achieve a stable foam, d=20 µL, and C = 0.05.

45
 Table 4.1: AEA Demand as Determined via the Foam Index Test
Total AEA
Designation
(mL/100 kg cm)
OPC 0
PLC 4
Q 4
F-G 12
F-Z 32
BA-B 12
BA-H 12
BA-P 44
BA-S 8
BA-V 8
BC-B 8
BC-M 8
RM-S3 80
RM-S9 12
I-S 32
Note: One 20 µL drop of AEA diluted to 5% is equivalent to 4 mL of AEA per 100 kg of cm

Most of the samples had AEA demand of 12 mL or less, but there is a substantial
difference between the samples that had the lowest and highest AEA demand. The
RM-S3 paste had an AEA demand significantly larger than the other samples. This
may be due to the sulfate treatment process used for this fly ash. Air void analysis
was then conducted on concrete samples containing RM-S3 and BA-P due to their
high foam index values.

4.1.2. Air Void Spacing Analysis

Air void analysis was completed on concrete specimens containing RM-S3 and BA-
P due to their high foam index values, as well as F-G as a control. Concrete samples
were dosed with 106 mL/100 kg cm of Sika® AIR. This dosage was selected
because it is the median value of the product recommended dosage range. Concrete
mixture design consisted of a 25% cement replacement with the SCM by mass and
a 0.485 w/cm. The cementitious content was designed to contain 564 lb/yd3.
Colorado River gravel was used as the coarse aggregate at a fixed content of 1800
lb/yd3. An assumed value of 2 vol.% was used for entrapped air. The Colorado
River sand was used as the fine aggregate, and its content was adjusted to complete
the remaining cubic yard volume in the concrete mixture. Table 4.2 shows an
example of a concrete mixture design for air void analysis.

46
 Table 4.2: F-G Concrete Mixture Design for Air Void Analysis
Component Amount (lb/yd3)
OPC 423
F-G 141
Water 296
Coarse Aggregate 1804
Fine Aggregate 1258
Air 2 vol.%
AEA 270 mL

After mixing, concrete was cast into a 4 x 8 in. cylinder and cured for 56 days prior
to being sent to TxDOT for testing. The entrained air content determined through
air void spacing analysis is shown in Table 4.3. Desirable spacing factors are less
than 0.02 mm (0.008 in.), since values less than 0.02 mm indicate that the concrete
should be adequately protected against freeze thaw damage.

Table 4.3: Results from Air Void Spacing Analysis

Designation Entrained Air Content (%) Spacing Factor (in.)
F-G 7.73% 0.00334
BA-P 6.04% 0.00437
RM-S3 5.32% 0.00709

The results shown in Table 4.3 reflect the foam index results in Table 4.1. Since
BA-P and RM-S3 required more AEA to achieve a stable foam in the foam index
test, less air was entrained in concrete containing these fly ashes when compared to
F-G concrete at the same AEA dosage. As a result, it should be noted that more
AEA is required to achieve a target entrained air content when using these fly ashes.

4.1.3. Air-Entrainment of Mortar

Admixture interactions were further explored with the air-entrainment of mortar
test in accordance with ASTM C311 [63]. This was conducted using a commercial
neutralized Vinsol resin solution prepared in accordance with ASTM C226 [79].
Mixtures consisted of 300 g of cement, 75 g of SCM, and 1125 g of 20-30 mesh
size standard sand. First, the water content was adjusted for each mixture to give a
flow of 80 to 95, and then the amount of neutralized Vinsol resin solution was
varied to produce an air content of 18 ± 3%. In this test, two test mixtures were
performed with target air contents of 15-18% and 18-21%. Then, the amount of
neutralized Vinsol resin to produce 18% air content was determined by linear
interpolation between the two test mixtures. The air content of the test mixtures was

47
calculated using Equations 4.2 and 4.3, where P is the percent of mixing water plus
Vinsol resin solution based on mass of cement, D is the density of the SCM in
mg/m3 and Wa is the mass per unit volume (g/mL) of mortar determined by ASTM
C185 [3].
𝑊𝑊𝑎𝑎
𝐴𝐴𝐴𝐴𝐴𝐴 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐, 𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣 % = 100(1 − ) (4.2)
𝑊𝑊𝑐𝑐

300+1125+75+(300×𝑃𝑃×0.01)
𝑊𝑊𝑐𝑐 = 300 1125 75 300×𝑃𝑃×0.01 (4.3)
+ + +
3.15 2.65 𝐷𝐷 1

The air-entrainment of mortar results are presented in Table 4.4. This test procedure
estimates a linear trend for AEA needed for an air content of 18 ± 3%. When three
points were determined within the range, it is questionable whether a linear
approximation is valid, especially at air contents near 15%. Additionally, the water
content is iterated to produce a given flow from 80 to 95, which can result in two
test mixtures with different P values. The mass per volume of the mortar has a large
effect on the air content calculation, causing concern as to whether one can
interpolate between two mixtures with different P values accurately due to the
difference in water content. In initial tests, Sika® AIR was used because it is more
available; however, concerns about the test results occurred due to the high dosages
of AEA required and additional water needed. Due to the availability of the
materials and concerns about the accuracy of the test, only four samples were tested.
When comparing the results from these tests with the foam index testing results,
there was not good agreement. For example, in the foam index test, the F-Z paste
required four times the amount of AEA to reach a stable foam than the BA-S paste.
However, in the ASTM C311 air entrainment test, the F-Z mortar required 1.55 mL
less of AEA than the BA-S mortar to achieve an air content of 18%. This
discrepancy was another reason for discontinuing the test.

Table 4.4: Air-Entrainment of Mortar Results

AEA to Achieve 18% Air Content
Designation
(mL)
OPC 1.48
F-Z 3.43
BA-S 4.98
RM-S9 4.89

4.2. Interaction with High-Range Water Reducers

Paste rheology was conducted on select materials that could have potential issues
with high-range water reducers. Testing was conducted using an Anton Paar MCR

48
301 rheometer. For all paste rheology testing, mixtures consisted of 500 g of
cementitious materials with a 25% replacement of cement by mass for each SCM
and a water-to-cementitious materials (w/cm) ratio of 0.45. Mixing procedures
followed ASTM C1738 [80] using a high-shear mixer. Water chilled to 5°C was
used for mixing in order to maintain a paste temperature of 23 ± 3°C due to the heat
generated from the high-shear mixer.

4.2.1. Saturation Dosage

Saturation dosages for a high-range water reducer (HRWR) were determined for
each paste using Sika® ViscoCrete® 2110. The HRWR was added to the mixing
water prior to the addition of the dry material. Paste mixtures were created starting
at admixture dosages of 0.1 or 0.2% by weight of the cementitious materials (% wt.
cm). After mixing, 19 mL of paste was transferred to the rheometer cup measuring
system and subjected to a shear rate of 50 s-1 using a helical bob geometry at a
controlled temperature of 23°C for 90 s. Once the 90 s were complete, a 3-minute
rest period followed. After the rest period, the shear rate was increased from 1 s-1
to 50 s-1 (i.e., the up curve) and then decreased from 50 s-1 to rest (i.e., the down
curve). Each shear rate was held for 45 s prior to changing to allow for the
equilibrium stress to be achieved. The downward paste flow curves were then
analyzed using the Bingham model (using the equilibrium stress values at each
shear rate) to determine the paste’s viscosity and yield stress values. The admixture
dosage was increased in incremental dosages of 0.1%, and the rheology test was
repeated (note: a new paste was prepared for each admixture dosage). The HRWR
saturation dosage was determined as the minimum dosage to reach the stable
minimum yield stress of the paste. The saturation dosage provides the maximum
amount of HRWR to use in a mixture before it becomes unstable; however, dosages
less than the saturation dosage are used in actual concrete mixtures to limit the risk
of bleeding and segregation of the concrete mixture. Table 4.5 shows the saturation
dosages for the ViscoCrete® 2110 for most of the SCMs selected for this testing.

49
 Table 4.5: ViscoCrete® 2110 Saturation Dosages for Cement Pastes
Admixture Dosage
Designation
(% wt. cm)
OPC 0.4
PLC 0.3
F-G 0.2
F-Z 0.3
BA-B 0.2
BA-V 0.2
RM-S3 0.2
RM-S9 0.4

The results shown in Table 4.5 indicate that the materials selected had no impact
on the ViscoCrete® 2110 saturation dosage amounts. Most of the fly ashes had
lower saturation dosages than the OPC, except for RM-S9, which required the same
amount of HRWR.

4.2.2. Small Amplitude Oscillatory Shear Testing

Small amplitude oscillatory shear (SAOS) testing examines potential issues in the
stiffening and setting behavior of the mixtures in the presence of an HRWR. An
HRWR dosage of 0.1% was chosen for the pastes to ensure the dosage is not beyond
the saturation point of the pastes. Three materials were selected for SAOS testing:
F-G as a control, BA-V because of the fluidity of BA-V paste and its high calcium
content, and RM-S3 for its high sulfate content. To ensure that the testing does not
significantly disrupt the microstructure of the paste, the test is conducted in the
linear viscoelastic region (LVER) so particles can recover elastically [81]. Testing
was conducted once for each mixture at 16, 23, and 30°C to determine the impact
of temperature on cement-HRWR interactions. Since BA-V paste was the most
fluid of the three pastes, and therefore had the smallest LVER, a strain sweep was
conducted to determine the critical strain where particles are no longer able to
recover elastically. Once the critical strain was determined, a frequency sweep was
performed to find the critical frequency. Table 4.6 shows the strain and frequency
values chosen for SAOS testing at each temperature.

Table 4.6: Strain and Frequency Values for SAOS Testing

Temperature (°C) Strain (%) Frequency (Hz)
16 0.5 1
-6
23 5x10 1
30 0.0001 1

50
Once the strain and frequency values within the LVER were determined, SAOS
tests were performed on the pastes for a test duration of 100 minutes and the storage
modulus was measured. Figure 4.2 and Figure 4.3 show the evolution of the storage
modulus over the duration of the testing period at each temperature for the tested
pastes.

10000

1000
Storage Modulus (kPa)

100 F-G
BA-V
10 RM-S3

0.1
0.1 1 10 100
Time (min)

Figure 4.2: Storage modulus results for fly ash pastes at 23°C for 100 minutes

10000

1000
Storage Modulus (kPa)

100
BA-V 16C
10 BA-V 23C
BA-V 30C
1

0.1

0.01
0.1 1 10 100
Time (min)

Figure 4.3: Storage modulus results for BA-V pastes at 16, 23, and 30°C for 100 minutes

SAOS testing at 23°C (Figure 4.2) showed that the fly ash pastes all behaved
similarly, with the storage modulus increasing over time. The increase in storage
modulus can be attributed to both soft colloidal interactions (i.e. electrostatic
interactions) and formation early hydration products (e.g. calcium silicate hydrate).

51
Figure 4.3 shows the effects of temperature on storage modulus of BA-V cement
paste. At lower temperatures, there is a slower gain in storage modulus in BA-V
paste at 16°C than at higher temperatures. Thus, this indicates that the setting time
of the mixture has increased. The structural rigidity of the system is higher than at
lower temperatures, which is reflected by the rapid increase in storage modulus of
BA-V paste at 30°C. While only the BA-V paste is shown here, all of the fly ashes,
even the control, behaved similarly in SAOS testing. This indicates that no
incompatibilities occurred.

52
Chapter 5. Concrete Property and Long-Term
Durability Testing

Cement paste, mortar, and concrete testing was conducted on materials to examine
property development and long-term durability. This testing is important for
qualifying materials for use in concrete mixtures. Tests included isothermal
calorimetry, rheology, and standardized ASTM tests as outlined in Table 5.1.

Table 5.1: Paste, Mortar, and Concrete Tests

Sample Type Test Method Property Measured
Heat of Hydration
Isothermal Calorimetry
Pastes Sulfate Optimization
Rheology Water Demand/Workability
Drying Shrinkage (ASTM C311) Drying Shrinkage
Mortars ASTM C1567 ASR Control
ASTM C1012 Sulfate Attack Control
ASTM C143 Slump
ASTM C231 Air Content
ASTM C403 Setting Time
Concrete ASTM C39 Compressive Strength
ASTM C1202 Chloride Penetrability
ASTM C1876 Bulk Electrical Resistivity
ASTM C1293 ASR Control

5.1. Paste Testing

5.1.1. Isothermal Calorimetry

5.1.1.1. Heat of Hydration

Isothermal calorimetry was performed on cement-SCM pastes to determine the
ability of the SCMs to reduce heat of hydration for use in thermal control plans.
Mixture design and mixing procedure of calorimetry pastes are outlined in Section
3.1.1.1. Figure 5.1 and Figure 5.2 show the rate of heat evolution of cement pastes
normalized per mass of paste, and Table 5.2 shows the cumulative heat after 3 days.

53
 12
OPC

Rate of Heat Evolved per gram of Paste

F-G
10
Q
BA-B
8
BA-H
J/(h*g)

BA-P
6
BA-S
BA-V
4

0
0 10 20 30 40 50 60 70
Time (h)

Figure 5.1: Rate of heat evolution of cement pastes containing control materials and
blended fly ashes

12
OPC
Rate of Heat Evolved per gram of Paste

F-G
10 Q
F-Z
8 BC-B
BC-M
J/(h*g)

RM-S3
6
RM-S9
I-S
4

0
0 10 20 30 40 50 60 70
Time (h)

Figure 5.2: Rate of heat evolution of cement pastes containing control materials and fly
ashes

54
 Table 5.2: 3-Day Cumulative Heat Values of Cement Pastes
Designation 3-Day Cumulative Heat per gram of Paste (J/g)
OPC 208
Q 173
F-G 173
F-Z 181
BA-B 172
BA-H 184
BA-P 168
BA-S 173
BA-V 190
BC-B 176
BC-M 178
RM-S3 172
RM-S9 183
I-S 175

From Table 5.2, it is clear that all of the materials, including Q, reduce the
cumulative heat released by the pastes in the first three days, suggesting that they
are all appropriate for use in concrete thermal control plans.

In Figure 5.1, most of the materials reduced the maximum rate of heat released from
the paste, except for BA-V. BA-V has the highest calcium content of the tested
materials (Table 2.2) and has abundant tricalcium aluminate and free lime (Table
2.8), which are reactive crystalline phases that can accelerate and alter cement
hydration reactions [69]. This results in a heat evolution curve for BA-V paste that
has an amplified second peak that starts to overlap the main hydration peak. The
hydration curve of BA-S paste also differs from the control Class F fly ash paste,
F-G. The main hydration peak for BA-S paste occurs earlier than the other cement-
SCM pastes. This is likely due to its smaller particle size in comparison to the other
fly ashes as shown in Table 2.11. Additionally, BA-S paste has an amplified second
peak, which can be attributed to the reactive alumina in the clinoptilolite phase
(Table 2.8) [82]. The remaining pastes in Figure 5.1 had similar hydration curves
to F-G paste.

In Figure 5.2, most of the cement-SCM pastes had similar hydration curves to F-G
paste, except for RM-S3. RM-S3 is a sulfate-treated fly ash that comes from a coal-
fired power plant that injects calcium hydroxide into the flux to react with SO2 to
reduce sulfur emissions. This reaction forms calcium sulfite (CaSO3), which has a
low solubility rate [83]. The presence of calcium sulfite in RM-S3 contributed to
the delayed hydration peak and prolonged induction period of RM-S3 paste.

55
Amplification of the second peak and prolonged induction periods of cement paste
are related to sulfate imbalances in the system. Sulfate imbalances can cause issues
with setting time, sulfate resistance, and chemical admixture compatibility. This
can be remedied through the addition of gypsum or limestone to increase or
decrease the sulfates in the system, respectively [83–85].

5.1.1.2. Sulfate Optimization

To identify problems with sulfate optimization when SCMs are used with high
limestone cements, isothermal calorimetry tests were repeated with a high
limestone content cement. In this case, since the interest was on examining the
impact of the SCMs on cement hydration, the rate of heat released was normalized
against the mass of cement rather than the mass of paste. The rates of heat evolution
are shown in Figure 5.3 and Figure 5.4.

24
PLC
Rate of Heat Evolved per gram of Cement

22
PLC - F-G
20
PLC - Q
18
PLC - BA-B
16
PLC - BA-H
14
J/(h*g)

PLC - BA-P
12
PLC - BA-S
10
PLC - BA-V
8
6
4
2
0
0 10 20 30 40 50 60 70
Time (h)

Figure 5.3: Rate of heat evolution of limestone cement pastes containing control
materials and blended fly ashes

56
 24
PLC

Rate of Heat Evolved per gram of Cement

22
PLC - F-G
20
PLC - Q
18
PLC - F-Z
16
PLC - BC-B
14
J/(h*g)

PLC - BC-M
12
PLC - RM-S3
10
PLC - RM-S9
8
PLC - I-S
6
4
2
0
0 10 20 30 40 50 60 70
Time (h)

Figure 5.4: Rate of heat evolution of limestone cement pastes containing control
materials and fly ashes

As mentioned previously, the addition of limestone to a cement system can cause a

sulfate imbalance. This is evident by the amplified second peak in most mixtures
in Figure 5.3 and Figure 5.4. To improve the sulfate balance of the system, gypsum
was added to the control PLC paste and the paste with the highest second peak, BA-
V. Lab grade gypsum was added as a percentage of the cement weight at 1%
increments from 1-3%. Heat evolution curves are shown in Figure 5.5 and Figure
5.6 for gypsum additions for PLC and BA-V pastes, respectively.

24
PLC-0%
Rate of Heat Evolved per gram of Cement

22
PLC-1%
20
PLC-2%
18
PLC-3%
16
14
J/(h*g)

12
10
8
6
4
2
0
0 10 20 30 40 50 60 70
Time (h)

Figure 5.5: Rate of heat evolution of PLC paste with gypsum additions

57
 24
BA-V-0%

Rate of Heat Evolved per gram of Cement

22
BA-V-1%
20
BA-V-2%
18
BA-V-3%
16
14
J/(h*g)

12
10
8
6
4
2
0
0 10 20 30 40 50 60 70
Time (h)

Figure 5.6: Rate of heat evolution of BA-V paste with gypsum additions

Results in Figure 5.5 and Figure 5.6 show that gypsum additions were effective at
reducing the height of the secondary peak. However, this resulted in a delay in the
secondary peak. It may not be possible to reduce the secondary peak without
delaying its peak height. Eliminating the secondary peak with higher gypsum
additions may be able to improve sulfate balance of the system [84].

5.1.2. Rheology
To characterize the water demand of the mixtures, rheological properties were
measured. Paste mixture design and mixing procedure are outlined in Section 4.2.
All mixtures containing SCMs were completed with a 25% replacement of the
cement by mass. After mixing, paste samples were inserted into the rheometer and
testing followed the procedure outlined in Section 4.2.1. Once testing was
completed, the Bingham model was applied to the paste flow curves to determine
the yield stress and viscosity of the pastes [86]. The paste flow curves are presented
in Figure 5.7 and Figure 5.8. The yield stress was determined as the y-intercept of
the trend line and the viscosity the slope. Table 5.3 shows the yield stress and
viscosity for each paste.

58
 140
OPC
120
PLC
Average Shear Stress, τ (Pa)
100 F-G

Q
80
BA-B
60
BA-H
40 BA-P

20 BA-S

BA-V
0
0 10 20 30 40 50
Shear Rate (s-1)

Figure 5.7: Flow curves of cement pastes containing control materials and blended fly
ashes

140
OPC
120 PLC
Average Shear Stress, τ (Pa)

F-G
100
Q
80
F-Z

60 I-S

BC-B
40
BC-M
20 RM-S3

RM-S9
0
0 10 20 30 40 50
Shear Rate (s-1)

Figure 5.8: Flow curves of cement pastes containing control materials and fly ashes

59
 Table 5.3: Bingham Parameters
Designation Yield Stress (Pa) Viscosity (Pa·s)
OPC 73.55 0.16
PLC 90.45 0.19
Q 79.76 0.14
F-G 28.95 0.15
F-Z 45.52 0.35
BA-B 29.58 0.17
BA-H 36.77 0.12
BA-P 43.14 0.14
BA-S 129.28 0.11
BA-V 37.98 0.08
BC-B 40.16 0.14
BC-M 37.54 0.11
RM-S3 40.88 0.17
RM-S9 36.56 0.16
I-S 42.52 0.11

All the SCM-containing pastes had similar or lower viscosities than the control fly
ash, F-G, except for F-Z paste. This may cause issues with flowability and
pumpability of concrete. Additionally, BA-S had the highest yield stress of all
pastes, despite being a blend of Class C and F fly ash. This could be due to the
smaller particle size of BA-S, as shown in Table 2.11. Additionally, BA-S contains
angular particles, as shown in SEM images in Appendix B, which can increase the
yield stress of the mixture due to interlocking of angular particles [68]. If a similar
flow to a traditional Class F fly ash is desired when using BA-S, an HRWR can be
used.

5.2. Mortar Testing

5.2.1. Drying Shrinkage

Drying shrinkage testing was performed in accordance with ASTM C311 [63] on
mortar bars. The control cement mortar consisted of 500 g of cement at a constant
w/cm of 0.485 and 1375 g of graded standard sand. Mortars containing SCMs had
125 g of the sand replaced with the SCM and w/cm was held constant at 0.485.
Although the test proportions water based on flow, a constant w/cm was utilized to
limit the number of variables in testing. Mortar bars were cured for 7-days in a lime
water bath in accordance with ASTM C157 [87], and then the length and mass of
the mortar bars were measured. After measuring, the bars were stored in air at 23°C
for an additional 28 days. The relative increase in drying shrinkage and weight loss

60
compared to the control mixture were then calculated and are presented in Table
5.4.

Table 5.4: Drying Shrinkage and Weight Loss on Drying Relative to OPC Control
Relative Drying Shrinkage Relative Weight Loss
Designation
(%) (%)
Q 0.02 5.00
F-G 0.02 4.26
F-Z 0.00 5.29
BA-B 0.01 5.21
BA-H 0.03 3.91
BA-P 0.03 4.59
BA-S 0.03 4.90
BA-V 0.03 4.22
BC-B 0.01 3.34
BC-M 0.02 3.47
RM-S3 0.00 4.85
RM-S9 0.02 4.21
I-S 0.02 4.27

Since the w/cm was held constant at 0.485, replacing the sand with SCM resulted
in a higher overall water content in the mortar compared to the control cement. This
resulted in increased shrinkage and weight loss in most of the mortars. F-Z and RM-
S3. F-Z and RM-S3 mortars did not have an increase in drying shrinkage compared
to the control cement mortar. This is likely due to the high sulfate content of these
fly ashes (Table 2.2), which leads to the formation of more ettringite, resulting in a
greater volume of restraining, non-shrinking solid phases [88]. This helps reduce
the amount of shrinkage occurring in these mortars. Although the shrinkage did not
increase, the weight loss for both F-Z and RM-S3 mortars was comparable to the
other fly ash mortars due to the increased water content compared to the control
cement mortar. It can be concluded that the impact of any of these SCMs on drying
shrinkage in concrete mixtures would be negligible.

5.2.2. Alkali-Silica Reaction

ASTM C1567 [11] was followed to evaluate the ability of the materials to control
deleterious expansion due to alkali-silica reaction (ASR). Mixtures consisted of a
graded fine aggregate, which was earlier found to be reactive through ASTM C1260
[12], at a ratio of 2.25 parts to the cementitious material by weight with a w/cm of
0.47. A cement replacement of 25% by mass was used for all SCMs. Expansion of
the mortar bars was measured at 3, 7, 10, and 14 days. ASTM C1567 sets an

61
expansion threshold of 0.1% at 14 days. Mortar bars that expand beyond the
threshold are deemed ineffective at mitigating ASR. Table 5.5 shows the results
from ASTM C1567 testing. Red text indicates that the mortar bars have surpassed
an expansion limit of 0.1%.

Table 5.5: ASR Mortar Bar Expansion Percentage at 14 Days

Designation Percent Expansion
OPC 0.41 ± 0.01
Q 0.14 ± 0.01
F-G 0.02 ± 0.00
F-Z 0.10 ± 0.00
BA-B 0.03 ± 0.00
BA-H 0.08 ± 0.01
BA-P 0.01 ± 0.01
BA-S 0.04 ± 0.01
BA-V 0.09 ± 0.01
BC-B 0.03 ± 0.00
BC-M 0.05 ± 0.01
RM-S3 0.03 ± 0.00
RM-S9 0.02 ± 0.00
I-S 0.03 ± 0.01

All the fly ashes were able to mitigate expansion due ASR in mortar. F-Z mortar
expansion at 14 days was at the expansion limit, but it did not exceed the threshold.

5.2.3. Sulfate Resistance

Mortars were tested using ASTM C1012 [23] to assess the ability of the materials
to reduce expansion due to sulfate attack. Mixtures consisted of standard graded
sand at a ratio of 2.75 to the cementitious materials by mass and a w/cm of 0.485.
A 25% replacement level by mass was used for all SCMs. Mortar bar expansion
was evaluated for 18 months in accordance with ASTM C1012. Results from
ASTM C1012 testing are presented in Table 5.6. Red text indicates that the mortar
bars have surpassed an expansion limit of 0.1%. Yellow text for the F-G control fly
ash means that the result is at the expansion limit of 0.1%.

62
 Table 5.6: Sulfate Mortar Bar Expansion Percentages
Weeks Months
ID
1 2 3 4 8 13 15 4 6 9 12 18
Q 0.01 0.01 0.01 0.02 0.23 1.36 -- -- -- -- -- --
F-G 0.01 0.01 0.01 0.01 0.02 0.02 0.03 0.03 0.04 0.06 0.10 0.16
F-Z 0.01 0.01 0.02 0.02 0.03 0.03 0.03 0.03 0.04 0.06 0.09 0.15
BA-B 0.01 0.01 0.02 0.01 0.02 0.03 0.05 0.08 0.22 0.51 0.74 1.50
BA-H 0.01 0.01 0.02 0.02 0.07 0.29 0.54 0.89 -- -- -- --
BA-S 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.02 0.02 0.04 0.03 0.03
BA-P 0.01 0.02 0.02 0.02 0.04 0.14 0.25 0.49 -- -- -- --
BA-V 0.01 0.02 0.02 0.02 0.06 0.24 0.46 -- -- -- -- --
BC-B -0.01 0.00 0.01 0.01 0.02 0.02 0.03 0.04 0.08 0.19 0.32 0.45
BC-M -0.02 -0.01 0.00 0.01 0.02 0.04 0.08 0.16 0.40 -- -- --
RM-S3 0.01 0.02 0.02 0.02 0.03 0.03 0.03 0.04 0.05 0.07 0.08 0.13
RM-S9 -0.02 -0.01 -0.01 -0.01 0.00 0.01 0.02 0.03 0.14 0.42 0.55 0.89
I-S 0.00 0.01 0.00 0.01 0.01 0.02 0.02 0.02 0.02 0.02 0.03 0.04

ACI 201 Guide to Durable Concrete [89] establishes expansion requirements for
certain exposure classes, as shown in Table 5.7. Only the F-G, F-Z, BA-S, BC-B,
RM-S3, and I-S fly ashes were able to meet the criteria for use in a sulfate
environment. BC-B meets the requirements to be used in a Class 1 exposure
condition; F-G, F-Z, and RM-S3 meet the requirements to be used in a Class 2
exposure condition; and BA-S and I-S were able to meet the requirements to be
used in a Class 3 exposure condition.

Table 5.7: ACI 201 Sulfate Exposure Classifications

Exposure Class Expansion Requirement
Class 1 Below 0.10% at 6 months
Class 2 Below 0.05% at 6 months or 0.10% at 12 months
Class 3 Below 0.10% at 18 months

Calcium aluminosilicate glass and reactive crystalline phases, such as tricalcium

aluminate, gehlenite, and anhydrite have been shown to form sulfate attack-
vulnerable hydration products. These reactive crystalline phases were found to be
present in most of the fly ashes (Table 2.7 and Table 2.8) that failed to meet an
exposure condition in sulfate resistance testing, except for BA-B. BA-B has a lower

63
amorphous content than most of the other fly ashes and consists of 40% MBA,
which was found to have low reactivity in Chapter 3.

5.3. Concrete Testing

Concrete mixture design consisted of a 25% cement replacement with the SCM by
mass and a 0.485 w/cm. The cementitious materials content was designed to be 564
lb/yd3. Colorado River gravel was used as the coarse aggregate at a fixed content
of 1800 lb/yd3. An assumed value of 2 vol.% was used for entrapped air. The
Colorado River sand was used as the fine aggregate, and its content was adjusted
to complete the remaining cubic yard volume in the concrete mixture. Table 5.8
shows an example of a concrete mixture design for a straight cement mixture. This
concrete mixture design was used for most concrete testing, apart from ASR testing.

Table 5.8: OPC Concrete Mixture Design

Component Amount (lb/yd3)
OPC 564
Water 297
Coarse Aggregate 1804
Fine Aggregate 1298
Air 2 vol.%

5.3.1. Fresh Concrete Properties

Slump and air content testing were performed in accordance with ASTM C143 [90]
and C231 [91], respectively, on fresh concrete. Slump was measured once for each
concrete mixture. Concrete used to measure slump was returned to the rotary drum
mixer and mixed for an additional 30 seconds to reintegrate the sample prior to
performing additional testing. When performing air content testing, unit weight of
the concrete mixture was also measured. Similar to slump testing, air content testing
was performed only once for each mixture. Table 5.9 shows the fresh concrete
properties for each mixture.

64
 Table 5.9: Fresh Concrete Properties
ID Slump (in.) Air Content (%) Unit Weight (lb/ft3)
OPC 7.4 3.9 145
Q 7.5 3.0 145
F-G 8.5 1.5 147
F-Z 8.5 1.8 148
BA-B 8.5 2.1 146
BA-H 8.5 1.7 146
BA-S 6.5 2.9 144
BA-P 9.5 1.6 147
BA-V 9.3 1.8 147
BC-B 9.5 1.2 148
BC-M 9.0 1.1 148
RM-S3 8.8 1.4 147
RM-S9 9.3 1.5 147
I-S 8.8 1.4 147

Most of the fly ash concretes performed similarly to the control Class F fly ash
concrete, F-G, except for BA-S, which had the lowest slump of all the concrete
mixtures. This confirms the results seen in paste rheology testing, which showed
that BA-S paste had the highest yield stress (Table 5.3). This is likely due to the
angular particles present in BA-S (Appendix B) and its relatively small particle size
(Table 2.11). An HRWR can be used if a slump similar to F-G concrete is desired.

Time of set was also determined for all mixtures following ASTM C403 [92].
Mortar was separated from the coarse aggregate by sieving fresh concrete using a
vibrating plate. The mortar was then stored in a cylindrical container and
consolidated. ASTM C403 instructs to take the initial measurement 3 to 4 hours
after concrete mixing and continue measurements at 30 min to 1-hour intervals until
the concrete reaches final set. The results are shown in Figure 5.9 and Figure 5.10.
Initial set occurs once the penetration resistance reaches 500 psi, whereas final set
occurs once the penetration resistance reaches 4000 psi. Setting time was measured
once for each of the concrete mixtures.

65
 8000
OPC F-G
Q BA-B
7000
BA-H BA-P
BA-S BA-V
6000
Strength (psi)

5000
Final Set
4000

3000

2000

1000
Initial Set
0
180 210 240 270 300 330 360 390 420 450 480
Time (min)

Figure 5.9: Time of set of concrete containing control materials and blended fly ashes

8000
OPC F-G
Q F-Z
7000
BC-B BC-M
6000 RM-S3 RM-S9
I-S
Strength (psi)

5000
Final Set
4000

3000

2000

1000
Initial Set
0
180 210 240 270 300 330 360 390 420 450 480
Time (min)

Figure 5.10: Time of set of concrete containing control materials and fly ashes

The initial measurement for OPC concrete occurred within the recommended time
period of 3 to 4 hours after initial mixing but resulted in an initial penetration
measurement above the initial set threshold. An initial set time of 224 minutes was
estimated for OPC concrete by fitting a trend line to the curve. All SCMs delayed
setting time compared to the control OPC mixture, with the RM-S3 fly ash causing
significant delays. This agrees with results seen in heat of hydration testing in
Figure 5.2, where the induction period of RM-S3 was prolonged. This is related to
the high sulfate content of RM-S3 (Table 2.2). A delayed setting time can have
negative effects on construction, causing plastic shrinkage and longer finishing

66
times [93]. It is possible that limestone or an accelerating admixture can reduce the
final set time for RM-S3 concrete; however, further testing is warranted to verify
that.

5.3.2. Compressive Strength

Once concrete mixing was completed, concrete was cast into 4 x 8 in. cylinder
molds. Cylinders were stored at a 23°C for 24 hours before demolding and then
moved to a curing room at 23°C and 100% relative humidity to be tested at 7, 28,
56, and 90 days. Compressive strength testing of concrete cylinders was performed
in accordance with ASTM C39 [94]. Results are shown in Figure 5.11 and Figure
5.12.

8000

7000
OPC
Compressive Strength (psi)

6000
FA
5000 Q
BA-B
4000
BA-H
3000 BA-P
BA-S
2000
BA-V
1000

0
7 28 56 90
Days

Figure 5.11: Compressive strength of concrete cylinders containing control materials and
blended fly ashes

67
 8000

7000
Compressive Strength (psi) OPC
6000 FA
Q
5000
F-Z
4000 BC-B
BC-M
3000
RM-S3
2000 RM-S9
I-S
1000

0
7 28 56 90
Days

Figure 5.12: Compressive strength of concrete cylinders containing control materials and
fly ashes

At 7 days, only BA-B concrete had a lower compressive strength than Q concrete.
This is likely due to the larger particle size of BA-B compared to Q (Table 2.11) as
well as the lower reactivity of MBA, which makes up 40% of BA-B. At 28 days
and beyond, all concrete mixture, including BA-B, had higher compressive
strengths than Q concrete. By 56 days, all the fly ash containing concretes had
comparable or higher compressive strengths than OPC concrete. The results from
concrete compressive strength testing indicate that all the fly ashes tested are
pozzolanic and can increase long-term strength.

5.3.3. Rapid Chloride Penetrability

Rapid chloride penetrability testing (RCPT) was performed as outlined in ASTM
C1202 [95]. RCPT was performed on 90 days concrete cylinders. The concrete
cylinders were cut into 2-in. samples using an oil-lubricated concrete saw then were
washed with soap and water to remove oil left from the saw and left to soak in
soapy water overnight. Concrete samples were then rinsed and prepared for testing
following ASTM C1202. The results are shown in Figure 5.13. Measurements were
only completed once.

68
 3000

2500
Charge Passed (coulombs) Moderate
2000

Low
1500

1000
Very Low

500

0
OPC F-G Q BA-B BA-H BA-P BA-S BA-V F-Z BC-B BC-M RM-S3 RM-S9 I-S

Figure 5.13: RCPT results of 90-day concrete cylinders containing control materials and
SCMs with dashed lines indicating thresholds set by ASTM C1202

Despite the name of the test, RCPT measures the conductivity of a concrete sample
rather than chloride penetrability. The conductivity can be correlated with
permeability of the concrete, which gives an indication for how likely chloride ions
will penetrate. Pozzolanic materials reduce the permeability of concrete by
consuming CH and precipitating reaction products in available pore space, resulting
in decreased pore connectivity [69,96]. Results from RCPT indicate that all the fly
ashes are pozzolanic and were able to densify the pore structure of the concrete
resulting in very low chloride ion penetrability based on thresholds set by ASTM
C1202.

5.3.4. Bulk Electrical Resistivity

Bulk electrical resistivity was also measured for the concrete. Testing was
conducted in accordance with ASTM C1876 [97]. For some of the cylinders,
resistivity was measured directly on removal from the fog room as the cylinders
reached 90 days (Table 5.10). For all mixtures, formation factor was also
determined (Table 5.10). Once concrete cylinders reached 90 days of curing, they
were submerged in a simulated pore solution with a resistivity of 0.127 ohm·m for
6 days. At the end of the 6-day period, samples were removed from the solution
and blotted to remove excess liquid before being placed into the concrete electrical
resistivity meter, and the resistivity was recorded. The formation factor was then
calculated for the mixtures as a ratio of the bulk resistivity of the concrete to the
resistivity of the simulated pore solution.

69
 Table 5.10: Bulk Electrical Resistivity Results of Concrete Cylinders
ID Resistivity (kΩ) Formation Factor
OPC -- 540.4
Q 1.58 424.1
F-G 7.02 1325.8
F-Z 5.05 1297.5
BA-B -- 1178.1
BA-H -- 1046.2
BA-P -- 1109.0
BA-S -- 1460.8
BA-V -- 1008.5
BC-B 6.52 1266.1
BC-M 6.00 1303.8
RM-S3 6.08 1322.6
RM-S9 5.14 1228.4
I-S 8.60 1498.5

Resistivity of concrete is commonly associated with the permeability of the

concrete, where a permeable concrete contains an interconnected pore structure that
allows liquids and gases to pass through. The less permeable the concrete is, the
less connectivity between pores, resulting in less liquid or gas passing through the
matrix. This results in a more resistive and less permeable concrete. Measured
resistivity values in concrete are also dependent on the resistivity of the pore
solution. Therefore, measuring resistivity on concrete directly after curing
combines information on both permeability and pore solution conductivity.
Submerging samples in simulated pore solution results in all concrete samples
having the same pore solution conductivity, thereby removing this variable. The
formation factor, therefore, is only dependent on pore connectivity and is, thus, a
good estimate of permeability. From Table 5.10, it is apparent that all of the SCM-
containing mixtures had much higher formation factors than the control OPC and
Q concretes, indicating that the SCMs reduce permeability.

5.3.5. Alkali-Silica Reaction

Alkali-silica reaction concrete mixtures varied slightly from other concrete mixture
designs. A high alkali cement was used, and sodium hydroxide (NaOH) was added
to the mixing water to increase the sodium equivalent (Na2Oe) to 1.25% by mass of
cement. The fine aggregate consisted of the same reactive sand used in ASR mortar

70
bar testing, while the coarse aggregate was non-reactive limestone that was graded
in accordance with ASTM C1293 [62]. A w/cm of 0.45 was used for all mixtures,
and a cement replacement of 25% by mass was used for all SCMs. Measurements
were taken over the span of 2 years. Table 5.11 shows the expansion percentages
of the mixtures. Red text indicates that the mortar bars surpassed the expansion
limit of 0.04%.

Table 5.11: ASTM C1293 Expansion Percentages

Day Month
ID
7 28 56 3 6 9 12 18 24
OPC 0.00 0.01 0.01 0.02 0.10 0.17 0.21 0.22 0.22
Q 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.01 0.01
F-G 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01
F-Z 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01
BA-B 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
BA-H 0.00 0.00 0.00 0.00 0.00 0.01 0.01 0.01 0.01
BA-S -0.01 0.00 0.00 0.00 0.00 0.00 0.01 0.01 0.00
BA-P 0.00 0.00 0.00 0.00 0.00 0.01 0.01 0.01 0.01
BA-V -0.01 -0.01 -0.01 -0.01 0.00 0.00 0.00 0.00 0.00
BC-B -0.01 -0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00
BC-M -0.01 -0.01 0.00 0.00 0.00 0.01 0.01 0.01 0.01
RM-S3 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.01 0.01
RM-S9 0.00 0.00 0.00 0.00 0.00 0.01 0.01 0.01 0.01
I-S 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.02 0.02

Only the control cement concrete, OPC, expanded beyond the 0.04% expansion
limit set by ASTM C1293 after 2 years. Q, which failed ASR testing in mortar, was
able to pass ASR testing in concrete. This may be due to leaching of alkalis during
the testing duration in Q concrete, since Q concrete has a higher permeability than
the other concrete mixtures, as was seen in RCPT (Figure 5.13) and bulk electrical
resistivity testing (Table 5.10). The concrete prism test (ASTM C1293) has been
considered to be more reliable than the mortar bar test (ASTM C1567) due to the
harsh, unrealistic conditions that the mortar bars are subjected to in ASTM C1567
[98,99]. However, it has been shown that ASTM C1293 also has its shortcomings,
such as leaching of alkalis and lower alkali loading for SCM-containing concrete
mixtures [100]. Recent research [100] has shown that the newly standardized
miniature concrete prism test (AASHTO T380) [101] may be a better test method
to determine a materials ability to suppress deleterious expansion due to ASR.

71
Chapter 6. Analysis and Recommendations

6.1. Reactivity Tests

This study evaluated several available SCM reactivity tests to determine their
ability to screen out inert materials and differentiate between pozzolanic and latent
hydraulic materials. Successful, rapid reactivity testing will be helpful for rapidly
screening poor performing materials and pre-qualifying materials for use. It was
found that R3 testing, newly standardized in ASTM C1897 [51], and the University
of New Brunswick Pozzolanic Reactivity Test [50] were the most effective at
screening out inert materials. Additionally, by measuring CH content of R3 pastes
using thermogravimetric analysis or single-point mass loss between 350 and 500°C,
as done in the R3P test, the values can be paired with heat release in R3 isothermal
calorimetry or R3 bound water testing to distinguish between hydraulic and
pozzolanic reactivity.

The following recommendations are made for screening out inert materials. The
method used is dependent on the equipment available in a testing laboratory.

• If an isothermal calorimeter is available, the R3 calorimetry method would

be recommended for its reduced labor time and good reproducibility.
Combining the results with the R3P test would be recommended to further
separate between hydraulic and pozzolanic reactivity.

• If there is not access to a calorimeter, either the UNBPRT or the R3 bound

water test are recommended. The UNBPRT requires the same equipment as
the SAI test, with the addition of an oven at 40°C. This may be favorable to
the R3 bound water test, which additionally requires an oven that can hold
stable temperatures at 350°C and a desiccant-filled chamber. It may be
possible to combine the UNBPRT test with the R3P test to separate between
pozzolanic and hydraulic reactivity of materials; however, the UNBPRT is
unable to differentiate between pozzolanic and hydraulic reactivity on its
own.

The R3 matrix and UNBPRT lime reactivity tests were the most effective at
screening out inert materials in a short period of time. Since both tests are conducted
in simulated environments to accelerate the material, additional testing should be
conducted to confirm material performance in cement-based materials. This allows
for a better understanding of the limitations of material reactivity in concrete than
is provided in the R3 and UNBPR tests. Mortar compressive strength testing and
CH content of cement pastes can be used as supplements to reactivity testing if
further confirmation of performance is desired.

72
6.2. Use of Non-Traditional Fly Ashes
Most of the non-traditional fly ashes performed comparably to traditional Class F
fly ash in most tests, with a few notable exceptions. The following
recommendations apply to the use of non-traditional fly ashes in concrete.

• Sulfate balance: If using Class F-Class C fly ash blends in concrete,

additional gypsum may be needed to modify the sulfate balance. Cement
pastes containing these fly ashes exhibited an amplification in the second
hydration peak as measured using isothermal calorimetry. This is likely due
to the increased number of aluminates in the system coming from crystalline
phases in the Class C fly ash in the blend. Since this slight sulfate imbalance
is unlikely to cause problems, it is recommended that no action be taken
with these materials.

• Sulfate balance: On the other hand, it should be cautioned that some non-
traditional fly ashes can cause the concrete mixture to be oversulfated. RM-
S3 fly ash, which was treated for high sulfate content, led to a delayed
setting time, which can cause construction problems in the field. A possible
solution when using fly ashes with high sulfate content like RM-S3 could
be to add limestone powder to reduce the setting time [83]. This was not
done as part of this study but could be the subject of future work to validate
this concept. It is recommended that fly ashes treated for high sulfate
content be avoided until a solution for the setting time delay is found.

• Workability: Most fly ashes had the same impact on workability as

traditional Class F fly ash, except for BA-S, a blended Class F-Class C fly
ash. This is due to the presence of angular particles in the fly ash, which
could have been caused by the blending process. A water reducer can be
added to concrete mixtures when using BA-S if a similar workability as a
traditional Class F fly ash is desired. Since most blended Class F-Class C
fly ashes did not cause workability differences, it is recommended that no
action is necessary with regard to workability, except on a case-by-case
basis identified through trial mixtures before implementation. In these
cases, a water reducing admixture can be utilized.

• Sulfate resistance: Fly ashes that contain reactive crystalline phases (C3A,
gehlenite, and anhydrite) were not able to meet the requirements for an ACI
201 [89] sulfate exposure class. The Class F fly ash-milled bottom ash
blend, BA-B, also did not meet an ACI 201 sulfate exposure class. This is
likely due to the lower reactivity of BA-B compared to other fly ash blends,
as expected due to its low amorphous content and observed by its slow

73
 strength gain in mortar and concrete mixtures. Fly ash blends containing
reactive crystalline phases or having low reactivity should not be used in
concrete that requires sulfate resistance.

Class F fly ash produced through firing of blended coal sources, blending different
CCPs, remediation of non-compliant fly ash, or from other countries all show
promise as potential sources to extend the supply of fly ash as the availability of
traditional Class F fly ash declines. It is recommended that the sulfate content from
oxide analysis not exceed the limits in the ASTM C618 [1] specification, and it is
suggested that qualitative X-ray diffraction (XRD) testing be conducted on fly
ashes that are intended for use in sulfate environments.

74
References

[1] ASTM C618, Standard Specification for Coal Fly Ash and Raw or
Calcined Natural Pozzolan for Use, ASTM International, West
Conshohocken, PA, 2019. https://doi.org/10.1520/C0618-19.2.
[2] ACI Committee 232, Use of Fly Ash in Concrete (ACI 232.2R-18),
Farmington Hills, MI, 2018. https://doi.org/10.1021/cen-v041n021.p102.
[3] M. Thomas, R. Jewell, R. Jones, Coal fly ash as a pozzolan, in: T. Robl,
A. Oberlink, R. Jones (Eds.), Coal Combust. Prod. Charact. Util. Benef.,
Elsevier Ltd., 2017: pp. 121–154. https://doi.org/10.1016/B978-0-08-
100945-1.00005-8.
[4] ACAA, 2019 Coal Combustion Product (CCP) Production & Use Survey
Report, Farmington Hills, MI, 2020. https://www.acaa-
usa.org/Portals/9/Files/PDFs/2019-Survey-Results.pdf.
[5] ACAA, 2018 Coal Combustion Product (CCP) Production & Use Survey
Report, Farmington Hills, MI, 2019. https://www.acaa-
usa.org/Portals/9/Files/PDFs/2018-Survey-Results.pdf.
[6] U.S. Energy Information Administration, Annual Energy Outlook 2020:
with projections to 2050, Washington, DC, 2020.
https://www.eia.gov/outlooks/aeo/pdf/AEO2020 Full Report.pdf.
[7] M.J. McCarthy, T. Robl, L.J. Csetenyi, Recovery, processing, and usage of
wet-stored fly ash, Elsevier Ltd., 2017. https://doi.org/10.1016/B978-0-08-
100945-1.00014-9.
[8] American Road & Transportation Builders Association, Production and
Use of Coal Combustion Products in the U.S. - Market Forecast Through
2033, 2015. https://www.acaa-
usa.org/Portals/9/Files/PDFs/ReferenceLibrary/ARTBA-final-
forecast.compressed.pdf.
[9] I. Diaz-Loya, M. Juenger, S. Seraj, R. Minkara, Extending supplementary
cementitious material resources: Reclaimed and remediated fly ash and
natural pozzolans, Cem. Concr. Compos. 101 (2019) 44–51.
https://doi.org/10.1016/j.cemconcomp.2017.06.011.
[10] Texas Department of Transportation, Standard Specifications for
Construction and Maintenance of Highways, Streets, and Bridges, 2014.
ftp://ftp.dot.state.tx.us/pub/txdot-info/des/spec-book-1114.pdf.
[11] ASTM C1567, Standard Test Method for Determining the Potential Alkali-
Silica Reactivity of Combinations of Cementitious Materials and
Aggregate (Accelerated Mortar-Bar Method), ASTM International, West
Conshohocken, PA, 2013. https://doi.org/10.1520/C1567-13.2.
[12] ASTM C1260, Standard Test Method for Potential Alkali Reactivity of
Aggregates (Mortar-Bar Method), ASTM International, West

75
 Conshohocken, PA, 2014. https://doi.org/10.1520/C1260-14.2.
[13] A. Naranjo, E. Lukefahr, Fly Ash Supply - Full Version, Texas Dep.
Transp. Tech. Advis. (2010) 1–4. https://ftp.dot.state.tx.us/pub/txdot-
info/cst/tips/fly_ash.pdf.
[14] G.J. McCarthy, J.K. Solem, O.E. Manz, D.J. Hassett, Use of a Database of
Chemical, Mineralogical and Physical Properties of North American Fly
Ash to Study the Nature of Fly Ash and its Utiliziation as a Mineral
Admixture in Concrete, in: Mater. Res. Soc. Proc., 1990: pp. 3–33.
[15] ASTM C1697, Standard Specification for Blended Supplementary
Cementitious Materials, ASTM International, West Conshohocken, PA,
2018. https://doi.org/10.1520/C1697-18.2.
[16] T.R. Naik, S. Singh, B. Ramme, Mechanical Properties and Durability of
Concrete Made with Blended Fly Ash, Cem. Concr. Res. 95 (1998) 454–
460.
[17] S. Antiohos, V.. Papadakis, K. Maganari, S. Tsimas, the Development of
Blended Supplementary Cementing Materials Consisting of High and Low
Calcium Fly Ashes, 11th Int. Congr. Chem. Cem. (2003) 747–757.
[18] S. Antiohos, K. Maganari, S. Tsimas, Evaluation of blends of high and low
calcium fly ashes for use as supplementary cementing materials, Cem.
Concr. Compos. 27 (2005) 349–356.
https://doi.org/10.1016/j.cemconcomp.2004.05.001.
[19] EN 197-1, Cement - Part 1: Composition, specifications and conformity
criteria for common cements, European Committee for Standardization,
Brussels, Belgium, 2011.
[20] S.K. Antiohos, V.G. Papadakis, E. Chaniotakis, S. Tsimas, Improving the
performance of ternary blended cements by mixing different types of fly
ashes, Cem. Concr. Res. 37 (2007) 877–885.
https://doi.org/10.1016/j.cemconres.2007.02.017.
[21] P. Tanikella, Incorporating Physical and Chemical Characteristics of Fly
Ash in Statistical Modeling of Binder Properties, Purdue University, 2009.
[22] B. Franklin, D. Rhodes, Influences of Class F and C Fly Ash Blending,
2015 World Coal Ash Conf. Nasvhille. (2015).
http://www.flyash.info/2015/087-franklin-2015.pdf.
[23] ASTM C1012/C1012M, Standard Test Method for Length Change of
Hydraulic-Cement Mortars Exposed to a Sulfate Solution, ASTM
International, West Conshohocken, PA, 2018.
https://doi.org/10.1520/C1012.
[24] P.K. Mehta, Effect of Fly Ash Composition on Sulfate Resistance of
Cement, ACI J. 83 (1986) 994–1000. https://doi.org/10.14359/1892.
[25] R. Dhole, M.D.A. Thomas, K.J. Folliard, T. Drimalas, Sulfate Resistance
of Mortar Mixtures of High-Calcium Fly Ashes and Other Pozzolans, ACI

76
 Mater. J. 108 (2011) 645–654. https://doi.org/10.14359/51683468.
[26] M.H. Shehata, M.D.A. Thomas, Effect of fly ash composition on the
expansion of concrete due to alkali-silica reaction, Cem. Concr. Res. 30
(2000) 1063–1072. https://doi.org/10.1016/S0008-8846(00)00283-0.
[27] M.H. Shehata, M.D.A. Thomas, Alkali release characteristics of blended
cements, Cem. Concr. Res. 36 (2006) 1166–1175.
https://doi.org/10.1016/j.cemconres.2006.02.015.
[28] K.L. Aughenbaugh, R.T. Chancey, P. Stutzman, M.C. Juenger, D.W.
Fowler, An examination of the reactivity of fly ash in cementitious pore
solutions, Mater. Struct. 46 (2013) 869–880.
https://doi.org/10.1617/s11527-012-9939-6.
[29] K.L. Aughenbaugh, P. Stutzman, M.C.G. Juenger, Identifying Glass
Compositions in Fly Ash, Front. Mater. 3 (2016).
https://doi.org/10.3389/fmats.2016.00001.
[30] R.T. Chancey, P. Stutzman, M.C.G. Juenger, D.W. Fowler, Comprehensive
phase characterization of crystalline and amorphous phases of a Class F fly
ash, Cem. Concr. Res. 40 (2010) 146–156.
https://doi.org/10.1016/j.cemconres.2009.08.029.
[31] S. Diamond, On the Glass Present in Low-Calcium and in High-Calcium
Flyashes, Cem. Concr. Res. 13 (1983) 459–464.
[32] P.T. Durdziński, C.F. Dunant, M. Ben Haha, K.L. Scrivener, A new
quantification method based on SEM-EDS to assess fly ash composition
and study the reaction of its individual components in hydrating cement
paste, Cem. Concr. Res. 73 (2015) 111–122.
https://doi.org/10.1016/j.cemconres.2015.02.008.
[33] G.J. McCarthy, K.D. Swanson, L.P. Keller, W.C. Blatter, Mineralogy of
Western Fly Ash, Cem. Concr. Res. 14 (1984) 471–478.
https://doi.org/10.1016/0008-8846(84)90121-2.
[34] P.J. Sandberg, L.R. Roberts, Cement-Admixture Interactions Related to
Aluminate Control, J. ASTM Int. 2 (2005) 219–232.
https://doi.org/10.1520/jai12296.
[35] R.D. Kalina, S. Al-Shmaisani, R.D. Ferron, M.C.G. Juenger, False
Positives in ASTM C618 Specifications for Natural Pozzolans, ACI Mater.
J. 116 (2019) 1–8. https://doi.org/10.14359/51712243.
[36] S. Seraj, R. Cano, S. Liu, D. Whitney, D. Fowler, R. Ferron, J. Zhu, M.
Juenger, Evaluating the Performance of Alternative Supplementary
Cementing Material in Concrete, Austin, TX, 2014.
[37] ASTM C1709, Standard Guide for Evaluation of Alternative
Supplementary Cementitious Materials (ASCM) for Use in Concrete,
ASTM International, West Conshohocken, PA, 2018.
https://doi.org/10.1520/C1709-18.2.

77
[38] A.R. Pourkhorshidi, M. Najimi, T. Parhizkar, F. Jafarpour, B. Hillemeier,
Applicability of the standard specifications of ASTM C618 for evaluation
of natural pozzolans, Cem. Concr. Compos. 32 (2010) 794–800.
https://doi.org/10.1016/j.cemconcomp.2010.08.007.
[39] K.L. Aughenbaugh, T. Williamson, M.C.G. Juenger, Critical evaluation of
strength prediction methods for alkali-activated fly ash, Mater. Struct. 48
(2015) 607–620. https://doi.org/10.1617/s11527-014-0496-z.
[40] NF P 18-513 Annexe A, Détermination de la quantité d’hydroxyde de
calcium fixé (essai Chapelle modifié), 2010.
[41] EN 196-5, Method of testing cement - Part 5: Pozzolanicity test for
pozzolanic cement, European Committee for Standardization, Brussels,
Belgium, 2011.
[42] IS 1727-1967, Methods of Test for Pozzolanic Materials, Bureau of Indian
Standards, New Delhi, India, 2004.
[43] CSA A3004-E1, Standard Practice for the Evaluation of Alternative
Supplementary Cementing Materials (ASCMs) for use in Concrete,
Canadian Standards Association, Mississauga, ON, Canada, 2013.
[44] R. Snellings, K.L. Scrivener, Rapid screening tests for supplementary
cementitious materials: past and future, Mater. Struct. 49 (2016) 3265–
3279. https://doi.org/10.1617/s11527-015-0718-z.
[45] S. Seraj, M.C.G. Juenger, Evaluation of an accelerated characterization
method for pozzolanic reactivity, in: Am. Concr. Institute, ACI Spec. Publ.,
2016.
[46] B. Lothenbach, P. Durdziński, K. De Weerdt, Thermogravimetric analysis,
in: K. Scrivener, R. Snellings, B. Lothenbach (Eds.), A Pract. Guid. to
Microstruct. Anal. Cem. Mater., Taylor & Francis Group, 2016: pp. 177–
211. https://doi.org/10.1201/b19074.
[47] S. Al-Shmaisani, R. Kalina, M. Rung, R. Ferron, M. Juenger,
Implementation of a Testing Protocol for Approving Alternative
Supplementary Cementitious Materials (SCMs): Natural Minerals and
Reclaimed and Remediated Fly Ashes, Austin, TX, 2018.
http://library.ctr.utexas.edu/ctr-publications/5-6717-01-1.pdf.
[48] T. Kim, J. Olek, Effects of Sample Preparation and Interpretation of
Thermogravimetric Curves on Calcium Hydroxide in Hydrated Pastes and
Mortars, Transp. Res. Rec. (2012) 10–18. https://doi.org/10.3141/2290-02.
[49] P.T. Durdziński, M. Ben Haha, S.A. Bernal, N. De Belie, E. Gruyaert, B.
Lothenbach, E. Menéndez Méndez, J.L. Provis, A. Schöler, C. Stabler, Z.
Tan, Y. Villagrán Zaccardi, A. Vollpracht, F. Winnefeld, M. Zając, K.L.
Scrivener, Outcomes of the RILEM round robin on degree of reaction of
slag and fly ash in blended cements, Mater. Struct. 50 (2017).
https://doi.org/10.1617/s11527-017-1002-1.

78
[50] M. Kasaniya, M.D.A. Thomas, E.G. Moffatt, Development of Rapid and
Reliable Pozzolanic Reactivity Test Method, ACI Mater. J. 116 (2019)
145–154. https://doi.org/10.14359/51716718.
[51] ASTM C1897, Standard Test Methods for Measuring the Reactivity of
Supplementary Cementitious Materials by Isothermal Calorimetry and
Bound Water Measurements, ASTM International, West Conshohocken,
PA, 2020. https://doi.org/10.1520/C1897-20.2.
[52] X. Li, R. Snellings, M. Antoni, N.M. Alderete, M. Ben Haha, S. Bishnoi,
Ö. Cizer, M. Cyr, K. De Weerdt, Y. Dhandapani, J. Duchesne, J. Haufe, D.
Hooton, M. Juenger, S. Kamali-Bernard, S. Kramar, M. Marroccoli, A.M.
Joseph, A. Parashar, C. Patapy, J.L. Provis, S. Sabio, M. Santhanam, L.
Steger, T. Sui, A. Telesca, A. Vollpracht, F. Vargas, B. Walkley, F.
Winnefeld, G. Ye, M. Zajac, S. Zhang, K.L. Scrivener, Reactivity tests for
supplementary cementitious materials: RILEM TC 267-TRM phase 1,
Mater. Struct. Constr. 51 (2018). https://doi.org/10.1617/s11527-018-1269-
x.
[53] P. Suraneni, A. Hajibabaee, S. Ramanathan, Y. Wang, J. Weiss, New
insights from reactivity testing of supplementary cementitious materials,
Cem. Concr. Compos. 103 (2019) 331–338.
https://doi.org/10.1016/j.cemconcomp.2019.05.017.
[54] P. Suraneni, J. Weiss, Examining the pozzolanicity of supplementary
cementitious materials using isothermal calorimetry and thermogravimetric
analysis, Cem. Concr. Compos. 83 (2017) 273–278.
https://doi.org/10.1016/j.cemconcomp.2017.07.009.
[55] D.P. Bentz, A. Durán-Herrera, D. Galvez-Moreno, Comparison of ASTM
C311 strength activity index testing versus testing based on constant
volumetric proportions, J. ASTM Int. 9 (2011).
https://doi.org/10.1520/JAI104138.
[56] V.G. Papadakis, S. Tsimas, Supplementary cementing materials in concrete
Part I: efficiency and design, Cem. Concr. Res. 32 (2002) 1525–1532.
[57] M.C.G. Juenger, R. Siddique, Recent advances in understanding the role of
supplementary cementitious materials in concrete, Cem. Concr. Res. 78
(2015) 71–80. https://doi.org/10.1016/j.cemconres.2015.03.018.
[58] K.L. Scrivener, B. Lothenbach, N. De Belie, E. Gruyaert, J. Skibsted, R.
Snellings, A. Vollpracht, TC 238-SCM: hydration and microstructure of
concrete with SCMs: State of the art on methods to determine degree of
reaction of SCMs, Mater. Struct. 48 (2015) 835–862.
https://doi.org/10.1617/s11527-015-0527-4.
[59] ASTM C1240, Standard Specification for Silica Fume Used in
Cementitious Mixtures, ASTM International, West Conshohocken, PA,
2020. https://doi.org/10.1520/C1240-20.2.
[60] M. Antoni, J. Rossen, F. Martirena, K. Scrivener, Cement substitution by a

79
 combination of metakaolin and limestone, Cem. Concr. Res. 42 (2012)
1579–1589. https://doi.org/10.1016/j.cemconres.2012.09.006.
[61] D.P. Bentz, T. Sato, I. De La Varga, W.J. Weiss, Fine limestone additions
to regulate setting in high volume fly ash mixtures, Cem. Concr. Compos.
34 (2012) 11–17. https://doi.org/10.1016/j.cemconcomp.2011.09.004.
[62] ASTM C1293, Standard Test Method for Determination of Length Change
of Concrete Due to Alkali-Silica Reaction, ASTM International, West
Conshohocken, PA, 2020. https://doi.org/10.1520/C1293-20A.2.
[63] ASTM C311/C311M, Standard Test Methods for Sampling and Testing Fly
Ash or Natural Pozzolans for Use in Portland-Cement Concrete, ASTM
International, West Conshohocken, PA, 2018.
https://doi.org/10.1520/C0311.
[64] ASTM C109/C109M, Standard Test Method for Compressive Strength of
Hydraulic Cement Mortars (Using 2-in. or [50 mm] Cube Specimens),
ASTM International, West Conshohocken, PA, 2020.
https://doi.org/10.1520/C0109.
[65] E.I. Diaz-Loya, F. Kinney, C.A.O. Rios, Reactivity indicators for activated
high-calcium fly ash-based binders, ACI Spec. Publ. 294 (2013) 1–22.
[66] T. Oey, C. Huang, R. Worley, S. Ho, J. Timmons, K.L. Cheung, A. Kumar,
M. Bauchy, G. Sant, Linking Fly Ash Composition to Performance in
Cementitious Systems, 2015 World Coal Ash Conf. Nasvhille. (2015).
http://www.flyash.info/.
[67] P. Duxson, J.L. Provis, Designing Precursors for Geopolymer Cements, J.
Am. Ceram. Soc. 91 (2008) 3864–3869.
[68] S. Al-Shmaisani, R.D. Kalina, R.D. Ferron, M.C.G. Juenger, Evaluation of
Beneficiated and Reclaimed Fly Ashes in Concrete, ACI Mater. J. 116
(2019) 79–87. https://doi.org/10.14359/51716713.
[69] B. Lothenbach, K. Scrivener, R.D. Hooton, Supplementary cementitious
materials, Cem. Concr. Res. 41 (2011) 1244–1256.
https://doi.org/10.1016/j.cemconres.2010.12.001.
[70] R. Snellings, X. Li, F. Avet, K. Scrivener, Rapid, Robust, and Relevant
(R3) Reactivity Test for Supplementary Cementitious Materials, ACI
Mater. J. 116 (2019) 155–162. https://doi.org/10.14359/51716719.
[71] B. Lothenbach, D.A. Kulik, T. Matschei, M. Balonis, L. Baquerizo, B.
Dilnesa, G.D. Miron, R.J. Myers, Cemdata18: A chemical thermodynamic
database for hydrated Portland cements and alkali-activated materials,
Cem. Concr. Res. 115 (2019) 472–506.
https://doi.org/10.1016/j.cemconres.2018.04.018.
[72] B.K. Marsh, R.L. Day, Pozzolanic and cementitious reactions of fly ash in
blended cement pastes, Cem. Concr. Res. 18 (1988) 301–310.
[73] S. Ramanathan, P. Suraneni, Cheap Reactivity Testing, (2020).

80
[74] J.A. Larbi, J.M.J.M. Bijen, Orientation of calcium hydroxide at the
portland cement paste-aggregate interface in mortars in the presence of
silica fume: A contribution, Cem. Concr. Res. 20 (1990) 461–470.
https://doi.org/10.1016/0008-8846(90)90037-X.
[75] B. Lothenbach, M. Zajac, Application of thermodynamic modelling to
hydrated cements, Cem. Concr. Res. 123 (2019).
https://doi.org/10.1016/j.cemconres.2019.105779.
[76] J. Skibsted, R. Snellings, Reactivity of supplementary cementitious
materials (SCMs) in cement blends, Cem. Concr. Res. 124 (2019).
https://doi.org/10.1016/j.cemconres.2019.105799.
[77] J.K. Jang, A New Approach to Measuring Pozzolanicity of Supplementary
Cementitious Materials Using Existing ASTM Standards, The University
of Texas at Austin, 2020.
[78] M. Kasaniya, UNBPRT Thresholds, (2020).
[79] ASTM C226, Standard Specification for Air-Entraining Additions for Use
in the Manufacture of Air- Entraining Hydraulic Cement, ASTM
International, West Conshohocken, PA, 2019.
https://doi.org/10.1520/C0226-19.2.
[80] ASTM C1738/C1738M, Standard Practice for High-Shear Mixing of
Hydraulic Cement Pastes, ASTM International, West Conshohocken, PA,
2019. https://doi.org/10.1520/C1738.
[81] M.A. Schultz, L.J. Struble, Use of oscillatory shear to study flow behavior
of fresh cement paste, Cem. Concr. Res. 23 (1993) 273–282.
https://doi.org/10.1016/0008-8846(93)90092-N.
[82] V.F. Rahhal, Z. Pavlík, A. Tironi, C.C. Castellano, M.A. Trezza, R. Černý,
E.F. Irassar, Effect of cement composition on the early hydration of
blended cements with natural zeolite, J. Therm. Anal. Calorim. 128 (2017)
721–733. https://doi.org/10.1007/s10973-016-6007-4.
[83] F. Zunino, D.P. Bentz, J. Castro, Reducing setting time of blended cement
paste containing high-SO3 fly ash (HSFA) using chemical/physical
accelerators and by fly ash pre-washing, Cem. Concr. Compos. 90 (2018)
14–26. https://doi.org/10.1016/j.cemconcomp.2018.03.018.
[84] J.O. Wheeless, Improving the Sulfate Resistance of Class C Fly Ash : A
Scientific Approach to Making Bad Ash Concrete, The University of Texas
at Austin, 2018.
[85] L. Wadsö, F. Winnefeld, K. Riding, P. Sandberg, Calorimetry, in: K.
Scrivener, R. Snellings, B. Lothenbach (Eds.), A Pract. Guid. to
Microstruct. Anal. Cem. Mater., Taylor & Francis Group, 2016: pp. 37–74.
https://doi.org/10.1201/b19074.
[86] V. Mechtcherine, A. Gram, K. Krenzer, J.H. Schwabe, S. Shyshko, N.
Roussel, Simulation of fresh concrete flow using Discrete Element Method

81
 (DEM): Theory and applications, Mater. Struct. 47 (2014) 615–630.
https://doi.org/10.1617/s11527-013-0084-7.
[87] ASTM C157/C157M, Standard Test Method for Length Change of
Hardened Hydraulic-Cement Mortar and Concrete, ASTM International,
West Conshohocken, PA, 2017. https://doi.org/10.1520/C0157.
[88] M. Zajac, A. Rossberg, G. Le Saout, B. Lothenbach, Influence of limestone
and anhydrite on the hydration of Portland cements, Cem. Concr. Compos.
46 (2014) 99–108. https://doi.org/10.1016/j.cemconcomp.2013.11.007.
[89] ACI Committee 201, Guide to Durable Concrete (ACI 201.2R-16),
Farmington Hills, MI, 2016.
[90] ASTM C143/C143, Standard Test Method for Slump of Hydraulic-Cement
Concrete, ASTM International, West Conshohocken, PA, 2020.
https://doi.org/10.1520/C0143.
[91] ASTM C231/C231M, Standard Test Method for Air Content of Freshly
Mixed Concrete by the Pressure Method, in: ASTM International, West
Conshohocken, PA, 2017: pp. 1–10.
[92] ASTM C403, Standard Test Method for Time of Setting of Concrete
Mixtures by Penetration Resistance, ASTM International, West
Conshohocken, PA, 2016. https://doi.org/10.1520/C0403.
[93] W. Lerch, Plastic Shrinkage, J. Am. Concr. Inst. 28 (1957) 797–802.
https://doi.org/10.14359/11555.
[94] ASTM C39/C39M, Standard Test Method for Compressive Strength of
Cylindrical Concrete Specimens, ASTM International, West
Conshohocken, PA, 2020. https://doi.org/10.1520/C0039.
[95] ASTM C1202, Standard Test Method for Electrical Indication of
Concrete’s Ability to Resist Chloride Ion Penetration, ASTM International,
West Conshohocken, PA, 2019. https://doi.org/10.1520/C1202-19.2.
[96] M. Thomas, Supplementary cementing materials in concrete, Taylor &
Francis Group, 2013. https://doi.org/10.1201/b14493.
[97] ASTM C1876, Standard Test Method for Bulk Electrical Resistivity or
Bulk Conductivity of Concrete, ASTM International, West Conshohocken,
PA, 2019. https://doi.org/10.1520/C1876-19.
[98] A. Gholizadeh Vayghan, J.R. Wright, F. Rajabipour, An extended chemical
index model to predict the fly ash dosage necessary for mitigating alkali-
silica reaction in concrete, Cem. Concr. Res. 82 (2016) 1–10.
https://doi.org/10.1016/j.cemconres.2015.12.014.
[99] J.H. Ideker, A.F. Bentivegna, K.J. Folliard, M.C.G. Juenger, Do Current
Laboratory Test Methods Accurately Predict Alkali-Silica Reactivity?, ACI
Mater. J. 109 (2012) 395–402. https://doi.org/10.14359/51683914.
[100] J. Tanesi, T. Drimalas, K.S.T. Chopperla, M. Beyene, J.H. Ideker, H. Kim,

82
 L. Montanari, A. Ardani, Divergence between Performance in the Field and
Laboratory Test Results for Alkali-Silica Reaction, Transp. Res. Rec. 2674
(2020) 120–134. https://doi.org/10.1177/0361198120913288.
[101] AASHTO T 380, Standard Method of Test for Potential Alkali Reactivity
of Aggregates and Effectiveness of ASR Mitigation Measures (Miniature
Concrete Prism Test, MCPT), American Association of State Highway and
Transportation Officials, Washington, DC, 2019.

83
Appendix A. X-Ray Diffractograms

6000
Z He = Hematite
L = Lime
5000 Mu = Mullite
Q Q = Quartz
Z = Zincite
Intensity (Counts)

4000
Z
3000
Z

2000 Z Z
Z Z
Q
1000 Mu Q Z
Mu L L Mu
He
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.1 X-ray diffraction pattern of F-G

4500
Ah = Anhydrite
Z
4000 He = Hematite
L = Lime
3500 Ma = Magnetite
Mu = Mullite
Intensity (Counts)

3000 Q = Quartz
Z Z = Zincite
2500
Q
2000 Z

1500 Z Z
Z Z
Q
1000 Ah
Ma L Z
Mu Mu Q L Mu He
500
He
Ma HeMa
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.2 X-ray diffraction pattern of F-Z

84
 4500
Z Ah = Anhydrite
4000 C = C3A
He = Hematite
3500 L = Lime
Mu = Mullite
Intensity (Counts)
3000 P = Periclase
Z
Q = Quartz
2500 Q Z = Zincite
Z
2000
Z Z
1500
Z Z
Q C
1000 Ah
Mu Z
L Mu P Q Mu
500
He
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.3 X-ray diffraction pattern of BA-P

6000
Z Al = Albite
Cl = Clinoptilolite-Na
5000 Cr = Cristobalite
Mu = Mullite
Q = Quartz
Intensity (Counts)

4000 Z Z = Zincite
Cl
3000 Q Z
Cr

2000 Z Z
Cl Z
Al Z
Q
Cl Cl
1000 Z
Cl Mu Mu Q Mu
Cl
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.4 X-ray diffraction pattern of BA-S

85
 4500
Z Ah = Anhydrite
4000 C = C3A
L = Lime
3500 Mu = Mullite
Q P = Periclase
Intensity (Counts)
3000 Q = Quartz
Z
Z = Zincite
2500
Z
2000
Z
1500
Z Z
Q Z
1000 C
Ah P Q Z
Mu L Mu Mu
500

0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.5 X-ray diffraction pattern of BA-V

6000
He = Hematite
Q Z Ma = Magnetite
5000 Mu = Mullite
Q = Quartz
Z = Zincite
Intensity (Counts)

4000
Z

3000
Z

2000 Z
Q Z Z
Z
1000 Q Z
Mu Mu Mu
He Ma
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.6 X-ray diffraction pattern of BC-B

86
 5000
Q
Z C = C3A
4500 Mu = Mullite
P = Periclase
4000
Q = Quartz
3500 Z = Zincite
Intensity (Counts)

3000 Z

2500
Z
2000
Q Z Z
1500 Z
Z
1000 C Q
Z
Mu Mu P Mu
500
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.7 X-ray diffraction pattern of BC-M

6000
Z Ks = Kalsilite
Mu = Mullite
5000 Pl = Portlandite
Q = Quartz
Q Z = Zincite
Intensity (Counts)

4000
Z

3000 Z

2000 Z Z
Pl Q Z
Z
Ks Ks
1000 Mu Mu Z
Q Mu
Pl
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.8 X-ray diffraction pattern of RM-S3

87
 6000
C = C3A
Z L = Lime
5000 Mu = Mullite
P = Periclase
Intensity (Counts) Q = Quartz
4000 Z = Zincite
Z
3000 Q Z

2000 Z Z Z
Z
Q C
1000 Mu Mu P Z
L Q Mu

0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.9 X-ray diffraction pattern of RM-S9

6000
He = Hematite
Ma = Magnetite
5000 Z Mu = Mullite
Q = Quartz
Z = Zincite
Intensity (Counts)

4000
Z
Q
3000
Z

2000 Z Z Z
Q Z
1000 Z
Mu Mu Q Mu
He Ma
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.10 X-ray diffraction pattern of I-S

88
 4500
Z Ah = Anhydrite
4000 C = C3A
G = Gehlenite
3500 L = Lime
Intensity (Counts) Mu = Mullite
3000 P = Periclase
Z
Q Q = Quartz
2500 Z = Zincite
Z
2000

1500 Z
Z Z
Z
1000 Q G C
Ah P
L Mu Q Z
500 Mu Mu

0 P
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.11 X-ray diffraction pattern of C-H

5000
Z An = Anorthite
4500 Au = Augite
Mu = Mullite
4000
P = Periclase
3500 Q Q = Quartz
Intensity (Counts)

Z Z = Zincite
3000
2500 An Z
2000
Z
1500 Z Z
Z
Q An
1000 Au Z
Q
Mu Mu P Mu
500
Au
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.12 X-ray diffraction pattern of MBA

89
 7000
A = Anatase
Z He = Hematite
6000 Q Mu = Mullite
M = Muscovite
Intensity (Counts) 5000 Q = Quartz
Z = Zincite
4000 Z

3000 Z

Z Z
2000 Q Z
M Z
A
Q Z
1000 Mu Q
M Mu A
He
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.13 X-ray diffraction pattern of M-D

7000
Ac = Anorthoclase
S = Sanidine
6000 Z
Q = Quartz
Z = Zincite
5000
Intensity (Counts)

4000 Q Z
Ac
3000 Z

Z
2000 Z Z
QSS Z

1000 Z
Ac Ac Q Q

0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.14 X-ray diffraction pattern of P-O

90
 7000
S = Sanidine
Z Q = Quartz
6000
Z = Zincite

Intensity (Counts) 5000

4000 Z

Z
3000

S Z
2000 Z Z
S S Q Z

1000 Z
S Q

0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.15 X-ray diffraction pattern of P-P

4000
Z Me = Merwinite
3500 Z = Zincite

3000 Z
Intensity (Counts)

2500
Z
2000

1500 Z
Z
Z Z
1000
Me Z
Me
500 Me

0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.16 X-ray diffraction pattern of S

91
 6000
Z Ac = Anorthoclase
Q Al = Albite
5000 Q = Quartz
Z = Zincite
Intensity (Counts)
4000
Z

3000
Z
Ac
Z
2000 Q Al Z
Al Z
Z
1000 Q Z
Ac Ac Q
Al
0
0 10 20 30 40 50 60 70 80
2-Theta (Degrees)

Figure A.17 X-ray diffraction pattern of T-P

92
Appendix B. Scanning Electron Microscopy
Images

Figure B.1 SEM image of F-G

Figure B.2 SEM image of F-Z

93
Figure B.3 SEM image of BA-P

Figure B.4 SEM image of BA-S

94
Figure B.5 SEM image of BA-V

Figure B.6 SEM image of BC-B

95
Figure B.7 SEM image of BC-M

Figure B.8 SEM image of RM-S3

96
Figure B.9 SEM image of RM-S9

Figure B.10 SEM image of I-S

97
Figure B.11 SEM image of C-H

Figure B.12 SEM image of MBA

98