DEVELOPING PREDICTIVE TOOLS AND GUIDANCE TO INCREASE THE DURABILITY OF CONCRETE INFRASTRUCTURE BY REDUCING ALKALI-SILICA REACTION (ASR) AND EARLY-AGE CRACKING 

DEVELOPING PREDICTIVE TOOLS AND GUIDANCE TO INCREASE THE DURABILITY OF CONCRETE INFRASTRUCTURE BY REDUCING ALKALI-SILICA REACTION (ASR) AND EARLY-AGE CRACKING 

ABSTRACT

Concrete degradation modes that reduce concrete durability are typically summarized as the following: cracking, physical attack, and chemical attack. Examples of cracking that reduces concrete durability are those that are caused by restrained volume reduction of concrete and those due to service loads exceeding the tensile strength of concrete. An example of physical attack is stresses induced by freezing and thawing cycles. Examples of chemical attack are alkali-silica reaction (ASR) and corrosion of embedded reinforcing steel. This dissertation develops predictive statistical tools and guidance to reduce concrete degradation and increase the durability of concrete infrastructure by reducing 1) cracking (specifically at early-ages) and 2) alkali-silica reaction (ASR). Therefore, this dissertation is separated into two (2) paths.

Research Path 1: Identify potential causes of early-age cracking and provide best practices guidance to mitigate future cracking.

The main objective of this research path is to identify the causes of longitudinal early-age cracking in concrete deck segments placed adjacent to newly replaced bridge deck expansion joints in Pennsylvania. The work consists of: 1) a literature review of the causes of early-age cracking on bridge decks; 2) a review of past and active bridge deck rehabilitation projects; and 3) an experimental evaluation of the most commonly used PennDOT bridge deck concrete mixtures to revise/update PennDOT specifications. Overall, an integrated approach to ensure proper selection and design of concrete materials, proper structural design of the deck (including the repair section), and proper construction and curing methods is needed to minimize early-age cracking of the concrete deck and repair sections. Main conclusions are material properties and construction practices play a greater role in determining the early-age concrete cracking characteristics for traditionally designed bridge decks. Results may be used as a best practices outline for reducing earlyage concrete cracking.

Research Path 2: Develop predictive statistical tools and guidance to reduce ASR in concrete infrastructure. 

It is important to limit the unnecessary waste of high quality pulverized coal fly ash because it is estimated that the supply of ASTM C 618 compliant fly ash (or ‘concrete grade’ fly ash) in the U.S. in 2030 will be approximately 14 million tons while the demand by the concrete industry will exceed 35 million tons. Accordingly, it is clear that all existing

‘concrete grade’ fly ash must be used efficiently and that the future and present needs of concrete producers include the availability of new supplemental cementitious materials (SCMs) and/or chemical admixtures effective at mitigating ASR. Therefore, this dissertation develops predictive statistical models to efficiently utilize the existing supply of ‘concrete grade’ fly ash and determines the effectiveness of novel SCMs and/or chemical admixtures effective at mitigating ASR for use in concrete as fly ash use sunsets.  Task 1: Develop predictive statistical models to assist in mix proportioning of concrete tomitigate ASR.

Three predictive statistical models are developed as part of this dissertation. The 1^st statistical model modifies and evaluates an existing model to determine fly ash dosages necessary to mitigate ASR in concrete mixtures containing highly reactive recycled glass sand according to the accelerated ASTM C 1567 test. For the first time, this statistical model provides a blueprint for material producers to attain the proper fly ash dosage necessary to mitigate ASR when using very highly reactive recycled glass sand aggregates. The 2^nd statistical model modifies and evaluates an existing model to determine fly ash dosages necessary to mitigate ASR in concrete mixtures according to ASTM C 1293. Finally, a 3^rd statistical model is summarized, which is a novel statistical model which also determines fly ash dosages necessary to mitigate ASR in concrete mixtures according to ASTM C 1293.

Task 2: Develop guidance for deploying next generation supplemental cementitious materials (SCMs) and chemical admixtures effective at mitigating ASR

This dissertation provides a summary of the performance of recycled soda-lime glass powder (SLGP) and Al(OH)₃ as supplemental cementitious materials (SCMs) at mitigating ASR. It is determined that both admixtures undergo the pozzolanic reaction and reduce ASR. However, results indicate SLGP is a net contributor to the total alkalis of the cementitious pore solution, thus exacerbating ASR over time. With respect to Al(OH)₃, for the first time, it is proven that Portland cement admixed with Al(OH)₃ will mitigate ASR according to long-term ASTM C 1293 testing. The three (3) main ASR reducing mechanisms are described within.

Table of Contents

List of Figures……………………………………………………………………………………………………… ix

List of Tables……………………………………………………………………………………………………… xiii

 

Acknowledgments ………………………………………………………………………………………………….. xiv

1. Chapter 1: Introduction/Organization. ………………………………………………………………. 1

1.1.Concrete Constituents and Durability………………………………………………………….. 1

• Literature Review: Causes of Early-Age Concrete Cracking ……………………… 2
• Literature Review: Alkali-Silica Reaction (ASR) ………………………………………. 13
• Research Needs: Early-Age Concrete Cracking …………………………………………. 26
• Research Needs: Alkali-Silica Reaction …………………………………………………….. 27
• Dissertation Organization ………………………………………………………………………….. 28
• References …………………………………………………………………………………………………. 34

2. Chapter 2: Causes of Early-Age Cracking in Concrete Bridge Deck Expansion

Joint Repair Sections ………………………………………………………………………………………… 39

• Introduction ……………………………………………………………………………………………….. 40
• Research Objectives …………………………………………………………………………………… 41
• Literature Review ………………………………………………………………………………………. 42
• Review of Past and Active Deck Rehabilitation Projects ……………………………. 46
• Experimental Evaluation of Bridge Deck Concrete Mixtures …………………….. 49
• Conclusions ……………………………………………………………………………………………….. 59
• References …………………………………………………………………………………………………. 62

3. Chapter 3: Reliability of Chemical Index Model in Determining Fly Ash Effectiveness against Alkali-Silica Reaction Induced by Highly Reaction

Glass Aggregates ……………………………………………………………………………………………… 67

• Introduction and Research Significance ……………………………………………………… 68
• The Chemical Index Model of Malvar and Lenke ………………………………………. 70
• Materials and Testing Protocol…………………………………………………………………… 72
• Results ………………………………………………………………………………………………………. 75
• Discussion ………………………………………………………………………………………………… 80
• Conclusions ……………………………………………………………………………………………… 83
• References ………………………………………………………………………………………………… 84

4. Chapter 4: Predicting Fly Ash Dosage to Prevent ASR by Introducing the

Concrete Prism Test (CPT) Chemical Index Model ………………………………………….. 86

• Introduction and Research Significance ……………………………………………………… 87
• AMBT Chemical Index Model…………………………………………………………………… 89
• Modifications to Create the CPT Chemical Index Model …………………………… 91
• Historical Concrete Prism Test Results for Model Development ………………. 93
• Model Predictions Based on Concrete Prism Test Data …………………………….. 99
• Discussion ………………………………………………………………………………………………… 101
• Conclusions …………………………………………………………………………………………….. 104

4.6. References ………………………………………………………………………………………………… 108

5. Chapter 5: Summary of Other Scientific Contributions …………………………………….. 111
   • Introduction ………………………………………………………………………………………………. 111
   • Section 1 – Summary: Performance of Recylced Soda-Lime Glass

Powder in Mitigating Alkali-Silica Reaction……………………………………………………. 112

• Background …………………………………………………………………………………………….. 113
• Conclusions …………………………………………………………………………………………… 114
• Cementitious Materials and Aggregates …………………………………………………. 115
• Results and Discussion …………………………………………………………………………… 117

• Section 2 – Summary: Mitigation of Alkali-Silica Reaction by Hydrated

Alumina ………………………………………………………………………………………………………….. 131

• Background …………………………………………………………………………………………… 132
• Conclusions …………………………………………………………………………………………… 133
• Materials ……………………………………………………………………………………………….. 134
• Results and Discussion ………………………………………………………………………….. 135

• Section 3 – Summary: An Extended Chemical Index Model to Predict Fly Ash Dosage Necessary for Mitigating Alkali-Silica Reaction in Concrete …. 146
  • Background ………………………………………………………………………………………….. 147
  • Conclusions …………………………………………………………………………………………… 148
  • Historical Concrete Prism Test Results and Model Development …………… 148
  • Extended Chemical Index Model …………………………………………………………… 149
• References ………………………………………………………………………………………………… 162

 

Chapter 6: Main Conclusions and Recommended Follow-on Work ………………………… 165

• Research Path 1: Task 1 – Identify Potential Causes of Early-Age

Concrete Cracking  ………………………………………………………………………………………….. 165

• Research Path 1: Task 2 – Develop Best Practices Guidance Through

Field Observations and Experimental Evaluations of Concrete Mixes

Commonly used on PennDOT Bridge Decks and Revise their Construction

Practices and Specifications …………………………………………………………………………….. 166

• Research Path 2: Task 1- Develop Models to Assist in Mix Proportioning

Concrete to Mitigate ASR ……………………………………………………………………………….. 167

• Research Path 2: Task 2 – Develop Guidance for Deploying Next Generation Supplemental Cementitious Materials (SCMs) and Chemical

Admixtures Effective at Mitigating ASR  ………………………………………………………… 167

• Recommended To-Go Work …………………………………………………………………….. 168

Chapter 1

 

Introduction/Organization

1.1.      Concrete Constituents and Durability

Portland cement concrete is the most abundantly used construction material in the world. Concrete consists of a heterogeneous mixture of Portland cement, supplementary cementitious materials (SCMs; e.g. pulverized coal fly ash (FA), ground granulated blast furnace slag (slag), etc.), water, fine aggregate (also known as sand; particles well graded between 4.75mm and 75μm), and coarse aggregate (particles well graded between 25mm to 4.75mm). These constituents are batched, mixed, placed, and cured to produce a hardened concrete that achieves a target uniaxial compressive strength, internal microstructure, air void network, and capillary porosity for optimal durability. The aggregates (both fine and coarse) are present to take up interstitial volume within the concrete system, reduce shrinkage, and provide potential load bearing capacity at early ages while the water is hydrating the cementitious materials into a binder. Moreover, aggregates aide in producing economically feasible concrete by assuming ~60-70% of the concrete system’s volume, thus reducing the use of expensive Portland cement.

The cement industry accounts for approximately 5-7% of the anthropogenic carbon emissions in the United States (¹). A great way to ensure a concrete structure is sustainable is to design the structure to be durable enough to meet or exceed its expected service lifetime. Therefore, one of the best definitions of concrete durability can be stated as follows: ‘…every concrete structure should continue to perform its intended functions, which is maintain its required strength and serviceability, during the specified or traditionally expected service life (²).’

Concrete degradation modes that reduce concrete durability are typically summarized as the following: cracking, physical attack, and chemical attack. Examples of cracking that reduce concrete durability are those that are caused by restrained volume reduction of concrete (i.e. due to drying shrinkage and/or thermal contraction) and those that are due to service loads exceeding the tensile strength of concrete. An example of physical attack is stresses induced by freezing and thawing cycles. Examples of chemical attack are alkali-silica reaction (ASR), chemical sulfate attack, and corrosion of embedded reinforcing steel. This dissertation will focus on increasing the durability of concrete infrastructure by reducing early-age cracking and ASR. Below, please find a literature review of the causes of early-age concrete cracking and alkali-silica reaction (ASR).

 

1.2.      Literature Review: Causes of Early-Age Concrete Cracking

Concrete cracking can be a primary cause of early infrastructure deterioration and it is known to considerably decrease the durability and service-life of the built environment. These cracks facilitate penetration of chlorides and moisture and therefore accelerate the corrosion of reinforcing steel. Aside from structural damage, cracking is also unsightly and the resulting distresses will decrease the ride quality in pavements and bridges (see Figure 1-1). Several state DOTs have performed or funded studies over the last few decades to identify the causes and effective mitigation practices for this problem. These studies evaluated typical causes of early-age cracking and the contribution of concrete material properties, construction practices, and structural design factors to this problem (these three contributors to early-age concrete cracking will be examined further in Chapter 2).

Cracking in concrete bridge decks results when the net internal tensile stress is greater than the tensile strength of concrete. Often, tensile stresses are caused as a result of the restrained shrinkage or thermal contraction of concrete; although cracking may also occur due to mechanical loading (e.g., early opening of bridge to traffic, overloading, and fatigue at a later age). Figure 1-2 shows how tensile stresses develop as a result of restrained shrinkage and thermal contraction of a newly constructed concrete repair patch. The tensile stresses increase with time as concrete experiences more shrinkage until these stresses exceed the tensile strength of concrete, at which time the concrete cracks (³). In addition to stresses developed as a result of an external restraint, moisture and temperature gradients in concrete (due to preferential drying or cooling at surfaces) can cause a non-uniform shrinkage strain profile which results in self-restraint and stress formation within concrete.

There are several causes or types of shrinkage in concrete: plastic shrinkage, chemical and autogenous shrinkage, drying shrinkage, and carbonation shrinkage. These types of shrinkage will be discussed below.

 

Figure 1-2: Tensile Stress development (³)

1.2.1.Plastic Shrinkage

When concrete is in a fresh or plastic state, plastic shrinkage cracking can occur if the rate of evaporation exceeds the rate at which the bleed water reaches the surface of concrete (⁴). As a result, tensile stresses develop at the concrete surface, which can result in cracking due to the very low tensile strength capacity of fresh concrete. In addition, differential settlement over rebar or where a change in the member’s cross section occurs can lead to plastic cracking (Figure 1-1). Plastic shrinkage cracking is especially common in high w/cm (water to cementitious material ratio) concretes (⁵). Plastic shrinkage can be reduced with proper moist curing, reducing evaporation rates, and installing wind breaks, so the surface of concrete never dries (discussed further in Chapter 2).

1.2.2.Chemical and Autogenous Shrinkage

When Portland cement hydrates, the net volume of hydration products (e.g., C-SH, portlandite (Ca(OH)₂), and other products) is less than the volume of the reactants (e.g., cement and water). This volume reduction is known as chemical shrinkage and is approximately equal to 64ml per 1 kg of Portland cement (1.77in³/lbs. of cement) for neat cement paste (⁶). As long as concrete is in the fresh state, this chemical shrinkage results in an overall settlement of the upper surface of fresh concrete but no tensile stresses are developed. After concrete sets, further chemical shrinkage serves as a driving force for autogenous shrinkage which increases the risk of cracking of restrained concrete members. It should be noted that chemical shrinkage is an intrinsic property of concrete and does not lead to cracking unless it results in large autogenous shrinkage.

After concrete sets, chemical shrinkage can no longer be accommodated by settlement of the upper surface of fresh concrete. As such, air-filled capillary voids form in the interior of the concrete as the water is consumed by hydration reactions. This phenomenon is commonly known as self-desiccation of concrete (⁷), resulting in a uniform drying of the entire cross section of the concrete member. This phenomenon is fundamentally different than drying shrinkage in which drying occurs at exposed surfaces of concrete while the interior bulk concrete remains near water saturation.

As a result of self-desiccation and the formation of many small air-filled voids that are uniformly distributed throughout the concrete member, capillary stresses develop which causes a uniform volume reduction of concrete.  This volume reduction that resulted from self-desiccation is known as autogenous shrinkage. If the concrete member is restrained, even in the absence of any external drying, autogenous shrinkage can cause tensile stresses and cracking (⁸).  Autogenous shrinkage is inversely related to the size of capillary pores inside concrete. As such, this type of shrinkage can be especially problematic in low w/cm (e.g., <0.36) and high cement content concretes (⁹). It also increases at higher temperatures (¹⁰). To combat this phenomenon, internal curing has been developed which entrains concrete with fine and well-distributed water reservoirs (e.g., saturated lightweight fine aggregates or super-absorbing polymers) which can gradually release this water to the interior of concrete to prevent self-desiccation (¹¹, ¹², ¹³).

1.2.3.Drying Shrinkage

Drying shrinkage occurs when moisture is lost from the surface of hardened concrete.  Moisture evaporation results in development of capillary stresses which reduces the volume of concrete. If this contraction is restrained (Figure 1-3) tensile stresses and cracking can result. In the absence of restraint, no visible cracks would form. The top image of Figure 1-4 is a schematic illustration showing that without a restraining friction, concrete slabs would shrink but would not crack. However, in a more realistic scenario (illustrated by the bottom image of Figure 1-4) the restraint provided by the sub-base results in cracking of the slab as it undergoes drying shrinkage. For concrete bridge decks, in most applications, external restraint can be caused by girders and stay-in-place (SIP) forms.  Aggregate and reinforcing steel can provide internal restraint which may lead to microcracking.

 

Figure 1-3:  Cracking of restrained concrete due to drying shrinkage (¹³)

 

 

Figure 1-4: Restrained drying shrinkage resulting in cracking of concrete slab (¹⁴)

 

Drying shrinkage of concrete is dependent on its aggregate content, w/cm, and the relative humidity of the ambient conditions. Aggregates do not shrink (or shrink very little) comparing to cement paste. As such, the presence of aggregates provides an internal resistant and considerably reduces the shrinkage of concrete. Pickett (¹⁵) suggested the following equation (Equation 1-1) that relates the shrinkage of concrete (є_con) (μm/m) to the shrinkage of cement paste (є_p) (μm/m) and the aggregate volume fraction (V_agg):

con =p(1−Vagg)n                                     Eq. 1-1

 

where n is a parameter that ranges from 1.2 to 1.7, depending on the stiffness of the aggregates and the paste (¹⁶).  This correlation is shown graphically in Figure 1-5a. In addition, the drying shrinkage of concrete depends on the stiffness of aggregates (Figure 1-5b). Aggregates with low absorption and high modulus of elasticity are the most effective at reducing concrete shrinkage.

Figure 1-5: (a) Concrete shrinkage as a function of aggregate volume fraction; (b) Effect of aggregate stiffness on shrinkage of concrete (⁴)

 

Moisture evaporation from concrete is a function of temperature and relative humidity of the ambient. Concrete surfaces dry faster and shrink more in hot and dry ambient conditions. Figure 1-6a and b shows the approximate relationship between relative humidity and mass loss and drying shrinkage of concrete.

In addition, w/cm has been shown to affect drying shrinkage; mainly due to faster drying in high w/cm high porosity concrete and a lower stiffness of these materials. Also, specimen geometry affects the magnitude of drying shrinkage. Specimens with small surface to volume ratio (e.g., large square and cylindrical sections) dry slowly, comparing with thin specimens with high surface area (e.g., slabs, decks, overlays). As such the former shows a slower drying shrinkage over time; although the ultimate shrinkage may be similar.

Figure 1-6: Relationship between ambient relative humidity (%RH) and (a) weight loss,  and (b) drying shrinkage of concrete (⁴)

1.2.4.Carbonation Shrinkage

In addition to drying shrinkage caused by moisture evaporation from concrete, the atmospheric carbon dioxide (CO₂) can chemically react with hardened cement paste and cause an irreversible shrinkage, known as carbonation shrinkage. The magnitude of carbonation shrinkage is a function of relative humidity and temperature (⁴). Since atmospheric CO₂ is always present (except in very controlled laboratory chambers), carbonation shrinkage always occurs simultaneously with drying shrinkage. Unless in controlled laboratory settings, the majority of drying shrinkage measurements performed in laboratories and all field measurements result in reporting shrinkage values that are a combination of drying and carbonation shrinkage.

1.2.5. Thermal Contraction

Another source of volume instability and potential for concrete cracking is thermal contraction which is especially a concern for concrete at early-ages. The heat of hydration causes the temperature of fresh concrete to rise. Often concrete sets near its peak temperature; and afterwards, as concrete cools to ambient temperature, it contracts (⁸). If this thermal contraction is restrained by adjacent members (for example, bridge girders, abutments, approach slabs, adjacent existing deck slabs and longitudinal rebar, and/or metal deck pan), tensile stresses develop inside the concrete which can result in cracking. Thermal cracking can occur from both externally applied temperature gradients as well as gradients formed internally. The temperature difference between peak concrete temperature and temperature of supports (e.g., steel forms or girders) provides a source of external temperature gradient. The supports also act as a restraint as concrete cools resulting in residual tensile stresses and possibly cracking. Internal temperature gradients form when concrete does not cool at the same rate throughout. This occurs typically when the concrete surface is exposed to ambient air temperatures (cools or heats quickly) while the interior of the concrete remains at a different temperature.

Other factors contributing to thermal cracking are aggregate content, cement content, and w/cm. Low aggregate concretes with low w/cm are prone to considerable heat of hydration development which can subsequently result in thermal contraction cracking

(⁸). Often, to control temperature development of concrete, supplemental cementitious materials (e.g., fly ash or slag) are used. Other means to control thermal contraction include cooling the concrete ingredients (water, aggregate) prior to mixing.

1.2.6.Effect of Mechanical Loads

Previous research (¹⁷,¹⁸,¹⁹,²⁰) shows tensile stresses caused by mechanical loading of bridges are far smaller than the stresses generated by restrained shrinkage, unless specifications are ignored and traffic or heavy construction equipment are allowed access prior to the concrete gaining required minimum strength. In repair applications, the loads caused by adjacent traffic lanes remaining open during construction may contribute to early-age cracking. Issa (²¹) showed that vibrations due to adjacent traffic lanes will only contribute to plastic cracking when concrete is under-vibrated or has too high of slump. In order to reduce moments in bridge decks causing tensile strain, it is recommended that concrete be placed first in the center of continuous bridge deck spans at the positive moment regions before the negative moment regions (^21,22). Other factors that may result in cracking of concrete in the long term are creep and fatigue. Repeated loading (e.g., traffic) can cause graduation formation and propagation of microcracks which leads to surface macrocracks after many cycles of loading.

With respect to early-age cracking, another phenomenon influencing the magnitude of tensile stresses inside concrete is stress relaxation (¹³). Relaxation is an alternative form of creep that is caused by the viscoelastic nature of concrete. Both the solid microstructure as well as the internal moisture of concrete can move gradually in response to sustained stresses which result in stress relaxation. This means that the actual magnitude of tensile stress in concrete is less than what is predicted by Hook’s law from the magnitude of shrinkage strain. This is shown in Figure 1-7 (²³).  Research has shown that stress relaxation is lower for concrete with higher elastic modulus; as such, high strength/stiffness concrete is even more prone to shrinkage cracking (¹⁸,²⁴).

 

Figure 1-7: Comparison between elastic and relaxed stresses in a restrained concrete slab undergoing

shrinkage (²³)

 

1.2.7.Causes of Early-Age Concrete Cracking

The factors that affect early-age cracking on concrete bridge decks are divided into three categories: (1) concrete material properties; (2) construction practices; and (3) structural design factors. Chapter 2 provides a further literature review of each category, identifies the primary and secondary contributors to early-age concrete cracking on bridge decks, and performs a suite of fresh, mechanical, and durability tests to develop a concrete mix design robust against early-age concrete cracking.

 

 

 

 

1.3.      Literature Review: Alkali-Silica Reaction (ASR)

Alkali-silica reaction (ASR) is a major concrete durability issue caused by an expansive series of reactions between the cementitious pore solution, silicate phases found in certain aggregates, and external moisture. ASR damages major infrastructure including parking structure (Figure 1-8), pavements, bridges, retaining walls, dams and hydraulic structures, and nuclear power plants. ASR causes tensile stress development and cracking within the concrete and on the concrete’s surface (²⁵). Since the aggregate is well dispersed within the concrete matrix, ASR cracking is widespread within the concrete volume and is typically manifest on the surface by many interconnected cracks (map-cracking) that exude a gel (Figure 1-8). Crack widths are typically between 0.1 mm to 10 mm (²⁵). Therefore, ASR induced cracking, by itself, mostly affects the appearance and serviceability of concrete rather than the structural integrity. Still, the abundance of cracks induced by ASR can facilitate the ingress of harmful chloride ions which will contribute to corrosion of embedded reinforcing steel.

The four (4) major variables needed to form ASR within concrete are the following: 1) aggregates with reactive forms of silica, 2) soluble calcium, 3) a high pH cementitious pore solution, and 4) an external source of moisture. These four (4) variables will dictate the extent which ASR degrades the structure and will control the three (3) distinct mechanisms which cause ASR: 1) silica dissolution, 2) silica gelation, and 3) swelling of the gel.

 

Figure 1-8: Concrete infrastructure affected with ASR (Photo taken by author)

 

1.3.1.Four (4) Major Variables Leading to ASR

Below, please find a review of the four (4) major variables leading to ASR and why each is present in the Portland cement concrete system.

1.3.1.1.ASR Variable 1): Reactive aggregates

Aggregates assume ~60-70% of the concrete system’s volume. Meta-stable silica minerals found in aggregates which are recognized to be susceptible to ASR are classified as amorphous, micro-crystalline, or strained. Specific minerals include, but are not limited to, opal, chalcedony, christobalite, tridymite, and cryptocrystalline and strained quartz (²⁵). These minerals are often found in the following aggregates: greywacke, phyllites, schists, gneisses, sandstones, glassy or cryptocrystalline rhyolites, chert, flint, shale, and artificial glass (²⁵,²⁶). The different mineral forms of silica can be determined by petrographic analyses and/or X-ray diffraction. In the Commonwealth of Pennsylvania, approximately 35% of all aggregate are reactive and therefore use of reactive aggregate cannot always be avoided.

1.3.1.2.ASR Variable 2): Soluble Calcium

The following are the typical participants in the hydration of Portland cement (²⁷):

• Alite (C₃S, tricalcium silicate doped with foreign ions)
• Belite (C₂S, dicalcium silicate doped with foreign ions)
• Tricalcium aluminate (C₃A)
• Calcium aluminate ferrite phases (C₄AF)
• Free lime (CaO)
• Alkali sulfates ((Na₂/K₂)SO₄)
• Calcium carbonate (CC)
• Gypsum (CaSO₄) and other calcium sulfate phases – Batch water

Calcium is clearly abundantly available in concrete. The dissolution of gypsum

(CaSO₄) and free lime (CaO), along with hydration of Portland cement clinker phases such C₃S and C₃A, release Ca²⁺ ions into the fresh concrete. Due to the continuous release of

Ca²⁺ ions from Portland cement clinker phases and the abundance of solid portlandite (Ca(OH)₂); concrete quickly reaches portlandite saturation (²⁷,²⁸,²⁹). A saturation factor of between 2 and 3 is reached within 1-3 hours. In later phases of hydration, Ca²⁺ ion concentration declines steadily due to further precipitation of hydration phases (see Figure 1-9 for Ca²⁺ ion concentration in concrete pore solution over the first 24-hours of hydration).

Solid portlandite in hardened concrete is the second most abundant product of

Portland cement hydration (first are calcium-silicate-hydrate phases (C-S-H)). Hardened Portland cement pastes can contain as much as 20-22% portlandite by mass (²⁷). Portlandite is found exclusively in imperfect crystalline and cryptocrystalline phases. Although initially found as small crystals in young Portland cement pastes, massive portlandite crystals (~10mm+) of indeterminate shape are observed in older pastes (²⁷).

A pure solution saturated with Ca(OH)₂ at 20°C contains 0.0211 mol/L Ca(OH)₂. Therefore, at portlandite saturation, the Ca²⁺ ion concentration is 0.0211 mol/L and the OH^– ion concentration is 0.0422mol/L (yielding a pH ~ 14 – log₁₀(1/0.0422) = 12.6) (³⁰). Since ~20-22% of hardened concrete is portlandite, an almost inexhaustible source of potential aqueous calcium ions are available to be released into the concrete pore solution upon consumption of ionic calcium into hydration phases. The availability of portlandite also acts a buffer against the reduction of the pH below 12.6.

However, the pH of normal cements range between 13.5 and 13.9 (([OH^–] = 0.320.79mol/L) (²⁵). Thus, the alkali sulfate impurities from the clinker play a role in increasing the pH of concrete another order of magnitude beyond a pH of 12.6.

1.3.1.3.ASR Variable 3): High pH cementitious pore solution

The most influential impurities in Portland cement clinker are alkali oxides (Na₂O and K₂O) which are introduced into the Portland cement pore solution as soluble alkali sulfates (Na₂SO₄ and K₂SO₄). These compounds may be attached to clinker raw feed (clay, limestone, shale, etc.) or to the coal used to burn the raw feed in the cement kilns (Hobbs 1988). On a mass basis, Na₂O and K₂O constitute ~ 0.60% of low alkali cement and ~1.25% of high alkali cement; although 0.90% is typical for cements created in the 21^st Century. Once mixed with concrete batch water, these alkali sulfates will dissolve rapidly into the pore solution. After dissolution, the sulfate anion will quickly enter low solubility products (ettringite, monosulfate, or C-S-H). These reactions will release an equivalent amounts of hydroxyl ions [OH^–] into the pore solution for charge balancing purposes which will bond with available alkali (Na⁺ and K⁺) cations producing alkali hydroxides (NaOH and KOH). As seen in Figure 1-9, hydroxyl ion concentration increases as sulfate ions decrease in the pore solution (consumed into hydration products). Therefore, alkalis are viewed negatively because their hydroxides are soluble and allow hydroxyl ions to enter and remain in the pore solution. It is typically understood that alkalis from the clinker will yield an equivalent amount of hydroxyl ions unless bound in a hydration product (²⁵,²⁷,³¹). K⁺, Na⁺, and OH^– concentrations typical approach there maximum limits within 28 to 90-days of curing (²⁸).

 

Figure 1-9: Pore solution ionic composition (⁴)

The pH of low alkali cements ranges between 12.7 and 13.1 ([OH^–] = 0.05-

0.12mol/L) and the pH of normal cements range between 13.5 and 13.9 (([OH^–] = 0.320.79mol/L) (²⁵). Literature suggests (³²,³³,³⁴) a potential correlation exists such that the hydroxyl ion concentration of the concrete pore solution needs to be greater than 0.20-0.30 mol/L (pH = 13.3-13.5) for aggregate attack to be ample enough for ASR to proceed, but this is not proven.

The majority of the alkalis which contribute to ASR are associated with the alkali sulfates in Portland cement. However, alkali contributions are also connected to aggregates (particularly artificial aggregates like ground soda-lime silicate glass), supplemental cementitious materials (SCMs), and external sources such as seawater, deicers, and anti-

icers (³⁵).

1.3.1.4.ASR Variable 4): Water

For cracking and expansion to result from ASR, an external source of water is necessary. The vast majority of concrete infrastructure is directly exposed to water for its entire service lifetime. Expansion typically only occurs when the external and internal relative humidity is between 85%-90% (²⁵,³⁶). Therefore, the concrete must be maintained moist for expansion to proceed and will cease if moisture is removed.

Poor drainage will allow water to pond/accumulate on the top surface or side faces of concrete. If the concrete mix proportions are such that the w/cm and water content are sufficient high enough to facilitate the ingress of water, adequate water will be available to cause ASR expansion.

1.3.2.Three (3) ASR Mechanisms

As mentioned previously, if adequate amounts of the four (4) major variables described above are available, ASR expansion is likely. These four (4) variables will dictate the extent which ASR proceeds and will control the three (3) distinct mechanisms which cause ASR: 1) silica dissolution, 2) silica gelation, and 3) swelling of the gel. These mechanisms are described below.

 

 

1.3.2.1.ASR Step 1: Silica Dissolution

Silica dissolution in concrete aggregates proceeds when tetrahedral coordinated silica (≡Si-O-Si≡(s) where ‘≡’ represents three (3) separate bridging oxygen bonds) enters into the high pH (>13.4) concrete pore solution and is immediately attacked by hydroxyl ions [OH^–]. This nucleophilic attack causes network dissolution of ≡Si-O-Si≡(s) by disrupting each bridging oxygen bond until all are broken and Si(OH)_4(aq) (mono-silicic acid) dissolves into the bulk solution (³⁷). Since the concrete pore solution pH is sufficiently high, upon release into the solution, stepwise hydrolysis ionizes Si(OH)_4(aq) by deprotonation from low solubility Si(OH)_4(aq) into complexes with much higher solubility such as H3SiO−4 (aq) (or SiO(OH)−3 (aq))  and H2SiO24−(aq) (or SiO2(OH)22−(aq)) (37,35). The deprotonation process will release small amounts of H⁺ ions into the pore solution. However, since the pore solution pH is typically between 13.5-13.9 and is sufficient buffered by a saturation of portlandite which prevents the pH from dropping below 12.6, no long-term widespread decrease in pH is observed. In concrete, it is the dissolution of silica on the surface of or within aggregates that leads to ASR. See Figure 1-10 for conceptualized stepwise silica network dissolution and hydrolysis (from (³⁷)). See Figure

1-11 for increasing solubility limits for amorphous silicates as a function of pH (from (³⁸)). Equations 1-2 through 1-4 offer the simplified stepwise network dissolution and hydrolysis of silica into negatively charged polymeric silicate species (³⁵).

 

Figure 1-10: Stepwise silica network dissolution and hydrolysis (note Si(OH)₅^–_aq = SiO(OH)₃^–_aq+ H₂O) (³⁷)

 

𝑆𝑖𝑂2(𝑠) + 2𝐻2𝑂 ↔ 𝑆𝑖(𝑂𝐻)4(𝑎𝑞) LogKsp = -2.92                        Eq.1-2

𝑆𝑖(𝑂𝐻)4(𝑎𝑞) + 𝐻2𝑂 ↔ (𝑆𝑖𝑂(𝑂𝐻)3−)𝑎𝑞 + 𝐻3𝑂+ LogKI = -9.47            Eq.1-3

         𝑆𝑖(𝑂𝐻)4(𝑎𝑞) + 2𝐻2𝑂 ↔ (𝑆𝑖𝑂2(𝑂𝐻)22−)𝑎𝑞 + 2𝐻3𝑂+ LogKII = -22.12       Eq.1-4

 

 

 

Figure 1-11: Solubility limit of amorphous silica in aqueous solution vs. pH at 25C (³⁸)

 

1.3.2.2.Step 2: Silica Gelation

The next step in the development of ASR is the condensation of silica sols via charge balancing between the exposed ionized silicate anion (silicate di-, tri-, and tetrameric ions formed during hydrolysis) and neighboring metal cations (Ca²⁺, Na⁺, K⁺, Al³⁺,

Fe^2/3+, etc…) in the high pH pore solution. These sols undergo aggregation into what is known as an ‘ASR gel’. These ‘ASR gels’ are often formed at the surface of concrete aggregates or within its cracks where the hydrolyzed silica is most abundant (³¹,³⁵).

Calcium may play an outsized role in the gelation properties of ‘ASR gels’ due to its omnipresence in concrete either dissolved in the pore fluid or as solid portlandite (which maintains the potential to dissolve into the pore fluid). Ca²⁺ may exchange with Na⁺ and/or

K⁺ in mature ‘ASR gels’ recycling alkalis back into the pore solution to potentially participate in further ASR gelation activities. The dissolution of Ca(OH)₂ to replenish Ca²⁺ ions also maintains the high pH in the pore solution which is necessary to sustain the dissolution of silica (³⁵).

1.3.2.3.Step 3: Silica Swelling

Depending upon the gel composition and intermediary cation exchange, the ‘ASR gel’ will absorb moisture and swell, leading to tensile stress development and cracking in concrete. Without moisture, the gels would remain innocuous and not swell. The mineralogical composition of these ‘ASR gels’ vary greatly depending on aggregate type, aggregate porosity, silica type, binder type, and structure location (²⁶ and references therein). It is established that the composition of ‘ASR gel’ greatly affects its hygroscopic nature and expansion behavior (³⁹,⁴⁰). Calcium also modifies the viscosity and yield strengths of ‘ASR gels’, potentially altering its migration potential and internal stress development (⁴¹,³⁵ and reference therein).

 

 

1.3.3.Mitigation of ASR

ASR is clearly a sophisticated concrete degradation mode. However, methods of mitigation are available (as summarized by ³⁵). For new structures, mitigation methods are typically broken down into the following categories: 1) use of non-reactive aggregates, 2) use of supplemental cementitious materials (SCMs), 3) use of low alkali cement, and 4) use of lithium based admixtures. For existing structures, mitigation methods are usually limited to restricting external moisture ingress (noted as “Mitigation Method’ 5 below).

Recently, ASTM C 1778-20 and AASHTO R80-17 were developed as tools for use by material producers and specifiers to determine appropriate testing methods to define aggregate reactivity potential and provide performance based and prescriptive based approaches to mitigate ASR in new structures. These documents assign ASR risk and prevention ‘levels’ associated with aggregate reactivity, structure importance, and specific oxides in the Portland cement and SCM. These standards are valuable for material producers and suppliers and provide positive forward momentum towards helping the concrete community eradicate ASR in new structures.

 

Figure 1-12: ASTM C 1260/1567 (AMBT) test setup

 

1.3.3.1.Mitigation Method 1: Non-reactive aggregate

For new structures, ASR can be avoided by eliminating reactive aggregates that are identified by standard tests such as ASTM C 1260 (~14-day test) (Figure 1-12) or ASTM C 1293 (~1-2 year test) (Figure 1-13). Since the use of reactive aggregates cannot be always avoided (e.g., 35% of utilized aggregate sources in Pennsylvania are reactive, as stated previously), this option is not always logistically and/or economically feasible.

 

 

1.3.3.2.Mitigation Method 2: Use of Supplemental Cementitious Materials (SCMs)

Supplemental cementitious materials (SCMs) such as pulverized coal fly ash (FA) and ground granulated blast furnace slag (slag) are effective at mitigating ASR due to the pozzolanic reaction. The primary pozzolanic reaction is an attack on the SiO₂ or Al₂O₃SiO₂ framework of the SCM by hydroxyl ions in the pore solution. The hydroxyl ions attach to the silica and break the bridging oxygen bonds (similar to silica dissolution described previously) eventually detaching the silica which can either remain in place or migrate into the solution depending on the pore network (²⁷,³¹). The negatively charged silica anions can charge balance easily with H⁺, Na/K⁺, and Ca²⁺ ions available in the pore solution to form low solubility pozzolanic C-S-H. This C-S-H typically has a low Ca/Si ratio which indicates substitution of Ca²⁺ with Na/K⁺ cations. Over time, this process reduces the mass % of solid portlandite from 20-22% in Portland cement concrete to as low as 3-6% for Portland cement-SCM binary concretes. The reduction of portandite occurs because 1) calcium ions in the pore solution consumed in the pozzolanic reaction are readily replaced by more calcium ions by dissolving portlandite (note: pore solution pH is typically above pH = 12.6, which drives portlandite saturation of the concrete pore solution) and 2) solid portlandite can react with ionic silica within the pore solution (²⁷,³¹).

The mechanisms by which SCMs such as FA mitigate ASR are the following: 1) reduce the alkalinity and pH of the pore solution by binding alkalis within pozzolanic reaction products, 2) lower the portlandite quantity through consumption of calcium into pozzolanic reaction products, 3) lower the Ca/Si ratio of C-S-H, 4) and lower the permeability and ion mobility through refining the pore network (²⁷,³⁵).

1.3.3.3.Mitigation Method 3: Use low-alkali cement

As mentioned previously, low alkali cement (~0.60% Na₂O_eq by mass) can limit the pH of the Portland cement concrete pore solution to between 12.7-13.1. If no alkalis are contributed from the aggregates and external sources (³⁵), reducing the pH to this range may be adequate to mitigate ASR (recall literature suggests [OH^–] = 0.2-0.3 mol/L, or pH < 13.3-13.5, may be effective to stop ASR). However, low alkali cements provide no mechanism for consuming portlandite or binding alkalis from aggregates or external sources. Therefore, alkali and calcium ions are still free within the system and not bound up in reaction products of the secondary pozzolanic reaction. These reasons call into question the effectiveness of low alkali cement at mitigating ASR over the long-term.

1.3.3.4.Mitigation Method 4: Lithium admixtures

Main lithium compounds used as admixtures to mitigate ASR are LiOH, LiCO₃, and LiNO₃. The mechanisms which these admixture mitigate ASR are not fully understood but can be summarized as follows (³⁵): 1) reduce dissolution rate of silica, 2) impede ASR gel formation, 3) reduce repulsive forces between the colloidal gel particles of ASR to reduce swelling, and 4) incorporate into ASR gels and alter properties and expansion. Lithium based admixtures can be added during batching with the other concrete constituents at the ready-mix facility to mitigate ASR.

1.3.3.5.Mitigation Method 5: Restrict external moisture

Expansion typically only occurs when the external and internal relative humidity is between 85%-90% (²⁵,³⁶). Therefore, ASR susceptible concrete must remain free of water to prevent ASR. The first option to keep the concrete free of water is to redirect water to a different location. If this proves impossible, penetrating silane sealers and/or surface applied urethane and epoxy coatings can be effective at reducing water infiltration and internal relative humidity (²⁵,³⁵).

Penetrating silane sealers are broken into two categories: 1) pore blockers and 2) water repellants. Silane blockers will penetrate into concrete and block pores while silane water repellants (~10-15 angstroms in size) will penetrate into concrete pores, react with surface hydroxyl groups, coat the pore walls, and render the pores hydrophobic (⁴²). The silane water repellents are typically applied to the surface of concrete in a water or isopropyl alcohol solvent which helps the silane penetrate into the pores. The organic groups bound to the silane produce a coating that is non-polar, has a low surface energy

(high water contact angle), and is hydrophobic (⁴³). Use of water repellants are favored over pore blockers because the former prevents moisture ingress while allowing water vapor transfer.

 

1.4.      Research Needs: Early-Age Concrete Cracking

Transverse cracking in newly constructed concrete bridge decks is a common problem reported by many state departments of transportation (DOTs) and several cities. Several state DOTs have performed or funded studies over the last few decades to identify the causes and effective mitigation practices for this problem. These studies evaluate typical causes of early-age cracking and the contribution of concrete material properties, construction practices, and structural design factors to this problem. Although many studies have been performed since the 1980s to identify the causes and effective mitigation practices for early-age cracking on concrete bridge decks, very few studies have focused on cracking in repair sections, especially next to rehabilitated deck expansion dams.

Cracking (Figure 1-14) of newly placed Portland cement concrete adjacent to bridge deck expansion joint dam replacements is observed on several newly rehabilitated sections of bridge decks according to the Pennsylvania Department of Transportation (PennDOT). PennDOT recognizes the problem and requests Penn State to diagnose the causes and recommend potential solutions. In response, this dissertation outlines two (2) research needs: 1) identify the potential causes of early-age concrete cracking in the newly rehabilitated bridge deck dams and 2) develop best practices guidance through field observations and experimental evaluations of concrete mixtures commonly used on PennDOT bridge decks to revise/supplement specifications to prevent early-age concrete cracking on future projects. Chapter 2 addresses these needs.

1.5.      Research Needs: Alkali-Silica Reaction

State and federal agencies as well as private sector companies recognize the need for basic and applied research on ASR in concrete structures. The Federal Highway

Administration (FHWA) allocated $10M to establish ‘funding for furthering the development and deployment of techniques to prevent and mitigate alkali silica reactivity.’ This dissertation examines the following ASR research needs: 1) development of models to assist in mix proportioning of ASR-mitigated concrete, and 2) development of guidance for deploying next generation supplemental cementitious materials (SCMs) and chemical admixtures which are effective at mitigating ASR.

1.5.1.ASR Research Need 1

This dissertation will develop statistical models for use by concrete practitioners to determine the dosage of fly ash needed to mitigate ASR. These statistical models have the potential to provide accurate fly ash dosages specific to each concrete mixture, thus reducing the lengthy wait for an expensive 2-year ASTM C 1293 test and reducing unnecessary waste of high quality fly ash. It is important to limit the waste of high quality fly ash because it is estimated that the supply of ASTM C 618 compliant fly ash (or ‘concrete grade’ fly ash) in the U.S. in 2030 will be approximately 14 million tons while the demand by the concrete industry will exceed 35 million tons. These statistical models are intended to be used as supplements and/or alternates to ASTM C 1778-20 and

AASHTO R80-17. Chapters 3 and 4 will address this research need. Section 3 of summary

Chapter 5 will also address this research need.

1.5.2.ASR Research Need 2

Summarized in this dissertation is guidance for deploying aluminum-rich SCMs and ground glass pozzolans as next generation ASR reducing SCMs and chemical admixtures through material characterizations and performance evaluations. The intention of ASR research need 1 is to efficiently use the current supply of ‘concrete grade’ fly ash. However, based on the scarcity and expected limited future supply of ‘concrete grade’ fly ash, it is clear the future and present needs of concrete producers include the availability of new supplemental cementitious materials and/or chemical admixtures effective at mitigating ASR. Lithium-based ASR reducing admixtures can be effective at mitigating ASR but cannot fill this gap as they are in short supply (~less than 0.002% of the Earth’s crust consists of lithium), expensive (potentially adding over 75% to the cost of the concrete), and are in high demand by competing industries such as the ever growing battery market. Literature has shown SCMs containing alumina (Al₂O₃) are more effective at mitigating ASR than SCMs of pure silica (³⁵,⁴⁴ and reference therein). Also, soda-lime glass powder (SLGP), known as ground glass pozzolan, has been suggested to be useful at mitigating ASR in concrete containing reactive aggregates due to a substantial pozzolanic performance. Summary Chapters 5 will address this research need.

 

 

1.6.      Dissertation Organization

The objective of this dissertation is to improve the durability and increase the service lifetime of concrete infrastructure through: 1) the development of best practices guidance to reduce early-age cracking and 2) the development of predictive tools and guidance to reduce alkali-silica reaction (ASR). This dissertation is therefore separated into six (6) chapters as summarized in Table 1-1. Chapters 2, 3, and 4 are uniquely written as separate journal manuscripts – complete with an abstract, literature review, research significance, and materials/methods sections. These chapters are each published as individual journal papers, with this dissertation’s author as 1^st author. Chapter 5 summarizes additional published work of this dissertation’s author as 2^nd author.

Chapter 1 provides a literature review of the causes of early-age cracking and alkalisilica reaction (ASR) in concrete infrastructure. Moreover, Chapter 1 lays out the research needs, objectives, and contributions of this dissertation. Chapter 6 summarizes dissertation findings/conclusions and suggests future work. Below, please find abstracts for Chapters 2-5.

Chapter 2(⁴⁵) investigates the causes of early-age concrete cracking by assessing the concrete mixtures specified for bridge deck rehabilitation projects, as well as reviewing the structural design and the construction and curing methods implemented by the contractors. The work consists of: 1) a comprehensive literature review of the causes of cracking on bridge decks; 2) a review of previous bridge deck rehabilitation projects that experienced early-age cracking along with construction observations of active deck rehabilitation projects; and 3) an experimental evaluation of the two most commonly used bridge deck concrete mixtures. Based on the literature review, the causes of concrete bridge deck cracking can be classified into three categories: concrete material properties, construction practices, and structural design factors. The most likely causes of the observed early-age cracking were found to be inadequate curing and failure to properly eliminate the risk of plastic shrinkage cracking. These results underscore the significance of proper moist curing methods for concrete bridge decks, including repair sections. This document also provides a blueprint for future researchers to investigate early-age cracking of concrete structures.

Table 1-1: Dissertation Chapters and Objectives

Chapter # and Title	Objectives
2: Causes of Early-Age Cracking in Concrete Bridge Deck Expansion Joint Repair Sections	Identify the causes of longitudinal early-age cracking in concrete deck segments placed adjacent to newly replaced bridge deck expansion joints in Pennsylvania.
3: Reliability of Chemical Index Model in Determining Fly Ash Effectiveness against Alkali-Silica Reaction Induced by Highly Reactive Glass Aggregates	Evaluate and modify a previously developed chemical index model to assist material designers in determining fly ash dosages necessary to mitigate ASR in concrete mixtures containing highly reactive recycled glass sand.
4: Predicting Fly Ash Dosages to Prevent ASR by Introducing the Concrete Prism Test (CPT) Chemical Index Model	Develop and use a new database of ASTM C 1293 (CPT) results to modify an existing statistical model to create the CPT chemical index model to predict the dosage of fly ash necessary to reduce ASR expansion below CPT threshold values.
5: Summary of Other Scientific Contributions	1) Does soda-lime glass powder mitigate ASR and if so, by what mechanism? 2) Does aluminum hydroxide mitigate ASR and if so, by what mechanism?, and 3) Develop a novel CPT extended chemical index model to more accurately predict fly ash (both Class F and Class C) dosages necessary to mitigate ASR.

Chapter 3(⁴⁶) evaluates and modifies a previously developed statistical chemical index model (Malvar and Lenke (⁴⁷)) to assist material designers in determining fly ash dosages necessary to mitigate ASR in concrete mixtures containing highly reactive recycled glass sand. The Malvar and Lenke (⁴⁷) statistical model considers the fly ash chemical composition, the cement chemical composition, and aggregate reactivity (AMBT expansion results at 14-year) as input parameters and predicts the required fly ash dosage as output. This chapter independently evaluates and modifies the chemical index model using six (6) new fly ashes, one (1) new OPC, and one (1) new highly reactive aggregate

(glass sand) to predict the dosage of fly ash necessary to reduce ASR expansion below AMBT acceptance threshold values. Based on the chemical oxide composition of the cement and fly ash and their corresponding AMBT expansion values, this model provides empirical nomographs that can be used by concrete suppliers as well as government transportation agencies to create a binary (cement-fly ash) concrete mixture that can be durable against ASR when using glass sand aggregates. Results show the predictions using existing Malvar and Lenke model parameters are conservative. Once the regression parameters are revised based on the AMBT results specific to glass sand, the model is determined to be accurate for low lime fly ashes (Class F) while still conservative for high lime fly ashes (Class C). For the first time, this chapter provides a blueprint for use by concrete material producers and specifiers to attain fly ash dosages necessary to mitigate ASR when using recycled glass sand aggregates.

Chapter 4(⁴⁸) modifies the existing statistical model of Malvar and Lenke (⁴⁷) based on CPT factors including: modifying expansion threshold limits to 0.04%, comparing 1year expansion results of 100% OPC mixtures to 2-year binary OPC-fly ash mixtures, implementing the new database of CPT results, and new regression fit parameters. Based on the chemical oxide composition of the new cement and fly ash and CPT results (obtained via extensive literature review) at 1-year and 2-years, the CPT chemical index model provides empirical nomographs that can be used by concrete suppliers as well as government transportation agencies to create a binary (cement-fly ash) concrete mixture that can be durable against ASR. Using selective data points for validation, the CPT chemical index model is slightly conservative at predicting Class F fly ash dosages to mitigate ASR and conservative at predicting Class C fly ash dosages to mitigate ASR (predictions similar or less conservative compared to ASTM C 1778). Acquiring more CPT results using Class C fly ash may aid in reducing the error and creating a formidable ASR prediction tool. Chapter may be used by concrete material producers and specifiers to recommend Class F fly ash dosages necessary to prevent ASR.

Chapter 5(⁴⁹,⁵⁰,⁵¹) summarizes the other major contributions of the author. Specifically, this chapter summarizes a study which evaluates the alkali-silica reaction (ASR) mitigation performance of soda-lime glass powder (SLGP). This study determines that a sufficient dosage of SLGP can mitigate ASR according to the accelerated mortar bar test (ASTM C1567, AMBT); however, the 2-year concrete prism test (ASTM C 1293, CPT) shows that the same type, fineness, and dosage of SLGP which mitigates ASR according AMBT does not sufficiently reduce ASR below the CPT expansion failure threshold. Pore fluid composition monitoring reveals that SLGP serves as a net contributor to the pore solution alkalinity over time. Therefore, it is recommended that long-term ASR testing, as opposed to the accelerated ASTM C 1567 test, is necessary to determine if and at what dosage SLGP may be able to mitigate ASR. Also summarized in this chapter is a study which proves Al(OH)₃ is effective at mitigating ASR. For the first time, long-term concrete prism tests (ASTM C 1293) prove the partial replacement of Portland cement with Al(OH)₃ will mitigate ASR expansion. Cementitious pore solution analyses, TGA analyses, and microstructural analyses show the main ASR reducing mechanisms of Al(OH)₃ are the following: 1) reduction of pH and alkalis in pore solution, 2) consumption of portlandite and soluble calcium, and 3) reduction of silica dissolution and aggregate damage. Finally, summarized in this chapter is a novel multiple nonlinear regression extended chemical index model developed for predicting fly ash dosages necessary to mitigate ASR per CPT. The model uses the oxide compositions of Portland cement and fly ash as well as aggregate reactivity to predict fly ash dosage levels to replace Portland cement and control ASR. The model successfully predicts the fly ash dosages required to mitigate ASR for different aggregates and cement-fly ash combinations. The prediction errors in most cases meet ASTM C 1293 multi-laboratory precision criterion.

 

             

1.7.      References

 

• PCA Cement Manufacturing Fact Sheet (2012), accessed 2012,

<http://www.cement.org/briefingkit/pdf_files/ManufacturingFactSheet.pdf>

 

• Neville, A.M., Properties of Concrete, Fourth Edition, 1995, Pearson Education, Limited. Harlow, Essex, England.

 

• Radlinska , A., Pease, B., and Weiss, J. (2007). A Preliminary Numerical Investigation on the Influence of Material Variability in the Early-age Cracking Behavior of Restrained Concrete. Materials and Structures. 40(4) 375-386

 

• Mindess, S., Young, J.F., and Darwin, D. (2003). Concrete. 2^nd Pearson Education Inc. Upper Saddle River. NJ.

 

• Cohen, M.D., Olek, J., and Dolch, W.L. (1990). Mechanism of Plastic Shrinkage in Portland Cement and Portland Cement-Silica Fume Paste and Motar. Cement and Concrete Research, 20(1), 103-119.

 

• Jensen, O., and Hansen, P. (2001). Water-Entrained Cement-based Materials I. Principles and Theoretical Background. Cement and Concrete Research, 31, 647654.

 

• Radlinska, A., Rajabipour, F., Bucher, B., Henkensiefken, R., Sant, G., and Weiss, J. (2008). Shrinkage Mitigation Stategies in Cementitious Systems: A Closer Look at Differences in Sealed and Unsealed Behavior. Transportation Research Record. 2070. 59-67.

 

• ACI 231-10: (2010). Early-Age Cracking: Causes, Measurement, and Mitigation. American Concrete Institute. Farmington Hills, MI.

 

• Jensen, O., and Hansen, P. (1996). Autogenous Deformation and Change of the Relative Humidity in Silica Fume-Modified Cement Paste. ACI Materials Journal. 93(6) 539-543.

 

• Jensen, O., and Hansen, P. (1999). Influence of Temperature on Autogenous Deformation and Relative Humidity Change in Hardening Cement Paste. Cement and Concrete Research. 29(4) 567-575.

 

• Bentz, D.P., and Jensen, O. (2004). Mitigation Strategies for Autogenous Shrinkage Cracking. Cement and Concrete Composites, 26(6), 677-685.

 

• Jensen, O., and Hansen, P. (2002). Water-Entrained Cement-based Materials II.

Experiemental Observations. Cement and Concrete Research, 32, 973-978.

 

 

• ACI 224R-01: (2001). Control of Cracking in Concrete Structures. American Concrete Institute. Farmington Hills, MI.

 

• Portland Cement Association. (2011). Design and Control of Concrete Mixtures. 15^th Resources. Skokie. IL.

 

• Pickett, G. (1956). Effect of Aggregate on Shrinkage of Concrete and Hypothesis Concerning Shrinkage, Journal of ACI, Vol. 52, pp. 581-590

 

• L’Hermite, R.G. (1960). Volume Changes of Concrete, Fourth International Symposium on the Chemistry of Cement, Washington, D.C., pp. 659-694.

 

• Schmitt, T.R., and Darwin, D. (1995). Cracking in Concrete Bridge Decks. Report No. K-TRAN:KU-94-1, Kansas Department of Transportation. Topeka. KS.

 

• Krauss, P., and Rogalla, E. (1996). Transverse Cracking in Newly Constructed Bridge Decks. NCHRP Report No. 380, Transportation Research Board. Washington, D.C.

 

• Frosch, R., Blackman, D.T., and Radabaugh, R.D. (2003). Investigation of Bridge Deck Cracking in Various Bridge Superstucture Systems. Purdue University. West Layfayette, IN.

 

• Hadidi, R., and Saadeghvaziri. M.A. (2005). Transverse Cracking of Concrete Bridge Decks: State of the Art. Journal of Bridge Engineering, 10(5) 503-510.

 

• Issa, M. (1999). Investigation of Cracking in Concrete Bridge Decks at Early Ages. Journal of Bridge Engineering, 4(2) 116-124.

 

• Babaei, K., and Hawkins, N. (1987). Evaluation of Bridge Deck Protective Strategies. Transportation Research Record, NCHRP Report 297, Washington, D.C.

 

• Weiss, J., Yang, W., and Shah, P. (1998). Shrinkage Cracking of Restrained Concrete Slabs. Journal of Engineering Mechanics, 124(7), 765-774.

 

• Darwin, D., Browning, J., and Lindquist, W.D. (2004). Control of Cracking in Bridge Decks: Observations from the Field, Cement Concrete and Aggregates. 26, 148– 154.

 

• Hobbs, D.W., (1988) Alkali-Silica Reaction in Concrete, Thomas Telford, London.

 

• Broekmans, M., A.T.M., (2012) “Deleterious Reactions in Aggregate with Alkalis in Concrete,” Reviews in Mineralogy & Geochemistry, Vol. 74, 279-364.

 

• Hewlett, P.C., Lea’s Chemistry of Cement and Concrete, 1998, John Wiley & Sons Inc, New York, NJ.

 

• Benstad, J., and Barnes, P., Structure and Performance of Cements, 2^nd Edition, 2002, Taylor & Francis Group, New York, NY.

 

• Rajabipour, F., Sant, G., and Weiss, J. (2008). Interactions Between Shrinkage

Reducing Admixtures (SRA) and Cement Paste’s Pore Solution. Cement and Concrete Research, 38, 606-615.

 

• Diamond, S. (1975). Long-Term Status of Calcium Hydroxide Saturation of Pore Solutions in Hardened Cements. Cement and Concrete Research, 5, 607-616.

 

• Taylor, H.F.W., (1997) Cement Chemistry, 2nd Edition, Thomas Telford Publishing, Heron Quay, London E144JD.

 

• Diamond, S., Barneyback., R.S., Struble, L. (1981). On the Physics and Chemisty of Alkali-Silica Reactions. Proceedings of the 5^th International Conference on Alklai-Aggregate Reaction, Cape Town, 1981, p. S252/22.

 

• Kollek, J., Varma, S., and Zaris, C. (1986). Measurements of OH^– Concentrations of Pore Fluids and Expansioion due Alkali-Silcia Reaction in Composite Cement Mortars. Proceedings of the 8^th International Congress of the Chemistry of Cement, Vol 4, pp. 183-189.

 

• Shehata, M., and Thomas, M. (2006). Alkali Release Characteristics of Blended Cements. Cement and Concrete Research, 36, 1166-1175.

 

• Rajabipour, F., et al., (2015) ‘Alkali-Silica Reaction: Current Understanding of the

Reaction Mechanisms and the Knowledge Gaps,’ Cement and Concrete Research, 76, 130-146.

 

• Burgher, B., Thibonnier, A., Folliard, K.J., Ley, T., and Thomas, M. (2008). Investigation of the Internal Stresses Caused by Delayed Ettringite Formation in Concrete. Report No. FHWA/TX-09/0-5218-1, Texas Department of Transportation, Austin, TX, 114 pp.
• Iler, R., (1979) The Chemistry of Silica: Solubility, Polymerization, Collois, and Surface Properties, and Biochemistry, John Wiley & Sons, Inc., United States of America.

 

• Maraghechi, H., Rajabipour, F., Pantano, C., and Burgos, W. (2016). Effect of Calcium on Dissolution and Precipitation Reactions of Amorphous Silica at High Alkalinity. Cement and Concrete Research, 87, 1-13.

 

• Gholizadeh-Vayghan, A., Rajabipoour, F. The Influence of Alkali-Silica Reaction (ASR) Gel Composition on its Hydrophilic Properties and Free Swelling in contact with Water Vapor. Cement and Concrete Research, 94, 2017, 49-58.

 

• Gholizadeh-Vayghan, A., Rajabipour, F. Quantifying the Swelling Properties of Alkali-Silica Reaction (ASR) Gels as a Function of Their Composition. Journal of the American Ceramic Society, 2017, 1-18.

 

• Gholizadeh-Vayghan, A., Rajabipour, F., and Rosenberger, J.L., CompositionRheology Relationships in Alkali-Silica Reaction Gels and the Impact of the

Gel’s Deleterious Behavior. Cement and Concrete Research, 83, 2016, 45-56.

 

• Attanayake, U., Liang X., Ng, S., and Aktan, H. (2006) Penetrating Sealants for Concrete Bridge Decks – Selection Procedure. ASCE Journal of Bridge Engineering, 11(5), 533-540.

 

• Sudbrink, B., Moradllo, M., Hu, Q., Ley, T., Davis, J., Materer, N., and Apblett, A.

(2017). Imaging the Presence of Silane Coating in Concrete with Micro X-Ray Fluorescence. Cement and Concrete Research, 92, 121-127

 

• Thomas, M.D.A., The effect of supplementary cementing materials on alkali-silica reaction: A review, Cement and Concrete Research, 2011, 41, 209-216.

 

• Wright, J.R., Rajabipour, F., Laman, J., and Radlińska, A. (2014), Causes of Early-Age Cracking in Concrete Bridge Deck Expansion Joint Repair Sections. Advances in Civil Engineering. Article ID 103421, 10 pages.

 

• Wright, J.R., Shafaatian, S., and Rajabipour, F. (2014). Reliability of the Chemical Index Model in Determining Fly Ash Effectiveness against Alkali-Silica Reaction Induced by Highly Reactive Glass Aggregates. Construction and Building Materials, Volume 64, 166-171.

 

• Malvar, L.J., and Lenke, L. (2006). Efficiency of Fly Ash in Mitigating Alkali-Silica Reaction Based on Chemical Composition. ACI Materials Journal, 103(5), 319-326.

 

• Wright, J.R., Gholizadeh, A., and Rajabipour, F. (2017). Predicting Fly Ash Dosages to

Prevent ASR by Introducing the Concrete Prism Test (CPT) Chemical Index Model,’ International Journal of Advances in Engineering Sciences and Applied Mathematics, 1-10, DOI: 10.1007/s12572-017-0188-3.

 

• Gholizadeh, A., Wright, J.R., Rajabipour, F. (2016). An Extended Chemical Index Model to Predict the Fly Ash Dosage Necessary for Mitigating Alkali-Silica Reaction in Concrete. Cement and Concrete Research, 82, 1-10.

 

 

 

• Shafaatian, S., Wright, J.R., and Rajabipour, F. (2019). Performance of Recycled SodaLime Glass Powder in Mitigating Alkali-Silica Reaction. ICE Green Materials Journal, themed issue of Durable Infrastructure, Institution of Civil Engineers (ICE), 7, 1, 2839.

 

• Szeles, T., Wright, J.R., Rajabipour, F., Stoffels, S. (2017). Mitigation of Alkali-Silica Reaction by Hydrated Alumina. Transportation Research Record, 2629, 15-23.

DEVELOPING PREDICTIVE TOOLS AND GUIDANCE TO INCREASE THE DURABILITY OF CONCRETE INFRASTRUCTURE BY REDUCING ALKALI-SILICA REACTION (ASR) AND EARLY-AGE CRACKING 

Sharing is caring!