Guidelines for Concrete Mixtures Containing Supplementary Cementitious Materials to Enhance Durability of Bridge Decks (2007)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

45 Introduction The raw materials for use in concrete must be selected based on availability and potential for durability. There are certain deterioration mechanisms that are directly related to the raw materials, such as ASR of the aggregate or unsoundness in the cement. Careful review of the raw materials at this early stage of the mixture development is an important part of optimiz- ing the concrete performance, so that the likelihood of these mechanisms affecting the bridge deck durability is minimized. The raw materials for bridge deck concrete include cement, coarse aggregate, fine aggregate, supplementary cementitious materials, and chemical admixtures. Choices regarding potential raw materials may be based on information already available from the suppliers in mill reports, test reports, or other formats. However, additional testing may be necessary to provide reasonable confidence in the quality of the raw materials required. The objectives of Step 2 of the methodology are (1) the generation of a list of locally available materials consistent with the performance objectives of the project and (2) the identification of those materials that will be considered for selection as variables in the test matrix. Raw Materials Selection Process Figure S2.1 provides a procedure to organize and evaluate the information about raw materials that potentially may be used in designing the concrete mixture. As with Figure S1.1, addi- tional background information for each of the decisions or tasks to be performed in this flowchart was collected (and is pre- sented in “Guidance on Raw Materials Selection”). The code shown in brackets, e.g., [C], for each topic in the flowchart refers to the subsection discussing that topic. The first box in Figure S2.1 directs the user to list all avail- able raw materials on Worksheet S2.1 before their evaluation and selection. Listing these materials will help the user to review all possibilities, so that a wide range of materials can be considered. As the user moves through the remainder of the flowchart, the most durable sources identified based on the guidance or the user’s own experience should be selected from each category and circled on Worksheet S2.1. Likewise, if a material was determined to be undesirable it should be ex- cluded and have an “X” marked through it on Worksheet S2.1. When making these choices for each type of material, the user may want to include sources with a range of properties so that the effect of the properties in that range can be assessed. Next, the user is directed to list all cement sources on Worksheet S2.2. Current mill certificates of potential sources of cement should be gathered and reviewed. Testing may be performed on the cements for properties that are not rou- tinely published on the mill certificates such as early stiffen- ing. For cements that are not pure portland cement (such as blended cements), the user is directed to obtain the compo- sition of the cement (i.e., the percentage of portland cement, and the type and percentage of integral SCM components) from the manufacturer. Worksheets S2.3 and S2.4 are provided to help organize test information regarding the aggregates; both fine and coarse aggregates should be verified to meet AASHTO M 6, Standard Specification for Fine Aggregate for Portland Ce- ment Concrete and AASHTO M 80, Standard Specification for Coarse Aggregate for Portland Cement Concrete (or ASTM C 33, Standard Specification for Concrete Aggregates), respectively. As a part of these specifications, the mineralog- ical composition of the aggregates should be determined, if not already available, by performing a petrographic analysis of the aggregate according to ASTM C 295, Standard Guide for Petrographic Examination of Aggregates for Concrete. From this analysis, the quantities of all the constituents in the aggregate, including the potentially alkali-silica reactive material, are tabulated. If the amount of material that is potentially deleteriously reactive with alkalis indicates a po- tential for harmful expansion, further work, such as an S T E P 2 Select Durable Raw Materials

46 assessment of the service record of the aggregate (discussed in “Guidance on Raw Materials Selection” of this chapter), is re- quired. If records are not available, then ASTM C 1260, Test Method for Potential Alkali Reactivity of Aggregates (Mortar- Bar Method), must be performed to assess ASR potential. Depending on the results, either the mixture must be ad- justed according to the ASR mitigation guidelines in this chapter and tested according to either ASTM C 1567, Stan- dard Test Method for Determining the Potential Alkali-Silica Reactivity of Combinations of Cementitious Materials and Aggregate (Accelerated Mortar-Bar Method), or a modified 2-year version of ASTM C 1293, Test Method for Determi- nation of Length Change of Concrete Due to Alkali Silica Reaction, or the aggregates can be tested without mitigation in a 1-year ASTM C 1293 test. Figure S2.1 together with the background in the following section provides recommenda- tions on the assessment of the aggregates and the potential use of SCMs or low alkali cement for ASR mitigation. The SCMs should also be screened; the compositions of each type of material should be reviewed from current cer- tificates of analysis; and the data compiled in Worksheets S2.5 through S2.9. Lastly, the chemical admixtures should be reviewed for compliance with the respective AASHTO or ASTM specifica- tions. Letters of compliance should be requested from the admixture producers. Other additional testing on the admix- tures is usually limited; however, the effect of chemical admixtures on setting time and other properties of concrete can be important and should be compiled. Worksheet S2.10 has been provided to help with this process. After all the information regarding the raw materials has been reviewed, Worksheet S2.1 will have some sources circled and some sources with an “X” marked through them to reflect which are viewed as likely and unlikely choices, respectively. Suggested ranges of SCMs to be tested for ASR mitigation should be placed on Worksheet S1.1 and compared with the ranges suggested for optimizing the other concrete properties. When ASR is a concern, the minimum levels of SCMs given in the ASR Mitigation Guidelines override the minimum values for all other properties when summarizing each SCM column. The sources and ranges of SCMs for experimental testing will be selected from Worksheets S1.1 and S2.1 in Step 3. There- fore, while working through Step 2, some thought should be given to which sources would be of interest for inclusion in the experimental program. Guidance on Raw Materials Selection The following subsections contain background discus- sions regarding concrete performance requirements, which accompany Figure S2.1. Cement [C] Three categories of specifications govern cement types: (1) AASHTO M 85 (ASTM C 150), Standard Specification for Portland Cement; (2) AASHTO M 240 (ASTM C 595), Stan- dard Specification for Blended Hydraulic Cements; and (3) ASTM C 1157, Standard Performance Specification for Hydraulic Cement. There are slight differences between the AASHTO and ASTM specification requirements. Each cate- gory will be described separately. AASHTO M 85 (ASTM C 150) Portland Cement Worksheet S2.2 can be used to organize the information about various cement sources that may be available for a proj- ect. Selected relevant data from AASHTO M 85 (ASTM C 150) specifications have been listed, with notes regarding their importance. The specifications have more requirements than those listed in Worksheet S2.2, and the user should refer to the specifications if needed. Further background informa- tion regarding basic information about portland cement is presented in the following subsection. Chemistry. The main components of portland cement are calcium silicates (i.e., C2S and C3S, where C and S repre- sent CaO and SiO2, respectively), calcium aluminates (i.e., C3A and C4AF, where A and F stand for Al2O3 and Fe2O3, respec- tively), and calcium sulfate (i.e., CS-, where S- stands for SO3) (11). Calcium sulfate (mostly gypsum, calcium sulfate dihy- drate) is included to control the setting of C3A, which may cause rapid stiffening (flash setting) of the concrete if not adequately controlled. Types. Cement is classified by AASHTO M 85 (ASTM C 150) into five types: Type I for normal usage, Type II for mod- erate sulfate resistance and moderate heat of hydration, Type III for high early strength, Type IV for low heat of hydration, and Type V for high sulfate resistance. Types I, II, and possi- bly III are used in bridge deck construction. These types of ce- ment are produced by controlling the chemical composition and fineness of the cement. The properties of the cement and its chemistry are regularly tested by the manufacturer and re- ported on a mill report, which also certifies the cement’s con- formance to the applicable ASTM or AASHTO standards. Rate of Hydration. The rate of hydration of cement— and thus the rate at which the concrete produced with the cement sets, generates heat, and gains strength—is governed by cement chemistry and fineness. C2S, C3S, and C3A make up the bulk of cement, and the relative portions of these com- pounds influence reaction rates and setting: C3A reacts very quickly with gypsum in a reaction that generates heat and some stiffening. Then the reactions that are responsible for

strength begin, and the compounds that participate in this, in order from fastest to slowest, are C3S, C3A (influenced by the gypsum), and C2S (21, 11). The more C3S and C3A present, the more quickly hydration can be expected to occur. Cements with finer size distributions can also be expected to hydrate more quickly because of the greater surface area exposed to water during mixing. Cement fineness, the surface area for a given mass, is esti- mated using AASHTO T 153 (ASTM C 204), Test Method for Fineness of Hydraulic Cement by Air Permeability Appara- tus, which measures the Blaine fineness, and AASHTO T 98 (ASTM C 115), Test Method for Fineness of Portland Cement by the Turbidimeter, which measures Wagner fine- ness. Limits for cement properties obtained using these meth- ods are provided in AASHTO M 85, although Blaine fineness is easiest to measure and most commonly reported. Typical Blaine finenesses range from 350 to 425 m2/kg for Type I or II cement and 550 m2/kg for Type III. Fineness measurements give an indication of the relative size of the average cement particle but do not describe the size distribution that may sig- nificantly influence the reactivity of the cement and the rhe- ology of the concrete (55). The setting time of cement is measured using AASHTO T 131 (ASTM C 191), Test Method for Time of Setting of Hydraulic Cement by Vicat Needle, and AASHTO T 154 (ASTM C 266), Test Method for Time of Setting of Hydraulic Cement Paste by Gillmore Needles. In the Vicat test, initial and final setting is defined by the time required for a 1-mm diameter needle to penetrate a paste specimen of given con- sistency to a maximum depth of 15 mm. For the Gillmore test, initial and final set is defined as when needles under 0.25- and 1-lb weights, respectively, can be supported by paste without producing indentations. ASTM C 150 specifies that initial and final setting time for all types of cement be greater than 60 and less than 600 minutes, respectively, when using the Gillmore apparatus and greater than 45 and less than 375 minutes, re- spectively, when using the Vicat apparatus. Although both tests are used, the correlation between setting time of paste measured using either method and of concrete determined using AASHTO T 197 (ASTM C 403), Standard Test Method for Time of Setting of Concrete Mixtures by Penetration Re- sistance, is not always consistent, perhaps because of the arbi- trary definitions used for initial and final setting (55). As stated, calcium sulfate is used to control the reaction of the C3A in the cement. The content and form of the calcium sulfate must be balanced with the reactivity of the C3A, which is influenced by the fineness and other properties of the ce- ment. When not properly balanced, false set (when too much calcium sulfate reacts initially) or flash set (when not enough calcium sulfate reacts with the C3A) may occur. Stiffening due to false set can be reversed by additional mixing while flash set is irreversible. This balance and the likelihood of early stiffening can be evaluated by testing according to ASTM C 359, Standard Test Method for Early Stiffening of Hydraulic Cement (Mortar Method). Durability Aspects. The chemistry of the cement also has implications for the long-term durability of the concrete. Certain reactions—such as that of free lime (CaO) and peri- clase (MgO) which form calcium and magnesium hydroxide, respectively—may occur after the cement has set, resulting in expansive products that may cause concrete cracking. AASHTO M 85 places limits on the amount of MgO that may be present in the cement. In addition, AASHTO T 107 (ASTM C 151), Standard Test Method for Autoclave Expan- sion of Portland Cement, tests the propensity of the cement for developing expansion; AASHTO M 85 limits acceptable expansion to less than 0.8%. Sulfate (SO3) may contribute to a reaction that produces expansive ettringite (DEF or internal sulfate attack); limits for this compound that vary depending on the type of the cement and the amount of C3A present are defined in AASHTO M 85. While DEF is not fully understood, if tem- peratures above 150°F (65°C) are developed during initial curing, ettringite that forms from C3A and sulfate com- pounds immediately after water and cement are combined may break down. The components that make up ettringite remain in the hardened concrete and reform ettringite over time in the presence of moisture (21). Currently, only long- term (2-year or longer) test procedures on heat-cured sam- ples are available to evaluate the risk of DEF. Keeping bridge deck concrete below 150°F (65°C) during the first 4 days of age is recommended in lieu of testing. Internal sulfate attack occurs when there is sufficient sulfate to cause such reaction without heat curing. If internal sulfate attack is a possible deterioration mechanism, the cement can be evaluated according to ASTM C 1038, Standard Test Method for Expansion of Hydraulic Cement Mortar Bars Stored in Water. This method measures the expansion of 1×1×11.25-in. (25×25×286-mm) mortar bars stored in lime-saturated water for 14 days. The amount of alkali present in the cement is also listed on the mill report in terms of equivalent or total alkali, which represents the contribution of both Na2O and K2O. Specific limits on the equivalent alkali (0.6%) are given only if the cement is to be labeled low alkali. The amount of alkali pres- ent is significant because a harmful reaction, ASR, occurs between alkali in cement and reactive silica that may be pres- ent in the aggregate. The product of this reaction is a gel that expands when wet and can cause damage to the concrete over time. The ASR requires both cement of sufficient alkalinity and aggregate that is reactive, and it can be avoided if either of these conditions is not met. The significance of cement alkalinity should be assessed based on testing of the candidate 47

48 aggregates, which is a time-consuming process. A procedure for aggregate evaluation is outlined in “Aggregates [A1].” The amount of alkali present in the cement can also affect the reaction rates of the phases of the cement. AASHTO M 240 (ASTM C 595) Blended Hydraulic Cements AASHTO M 240 (ASTM C 595) covers the requirements for blended cements. These cements are composed of port- land cements pre-blended with GGBFS and/or pozzolans. There are six classes of blended cements: • Type IS, portland blast furnace slag cement. GGBFS con- stitutes 25% to 70% by mass of the blended cement. • Type IP, portland-pozzolan cement. This blend consists of either portland or portland blast furnace slag cement and fine pozzolan. The pozzolan constitutes 15% to 40% by mass of the blended cement. • Type P, a portland-pozzolan cement. This blend is for use when higher strengths at early ages are not required. • Type I (PM), pozzolan-modified portland cement. This blend consists of either portland or portland blast furnace slag cement and fine pozzolan. The pozzolan constitutes less than 15% by mass of the blended cement. • Type I (SM), slag-modified portland cement. GGBFS con- stitutes less than 25% by mass of a mixture of GGBFS with portland cement. • Type S, slag cement. GGBFS constitutes at least 70% by mass of a mixture of GGBFS with portland cement. AASHTO M 240 should be consulted as to the various chemical and physical requirements of blended cements. It is important to obtain from the manufacturer information on both the blend used in the cement (actual quantities of port- land cement and types and quantities of SCMs) and, if possi- ble, the chemical and physical composition of the portland cement itself. This information is needed when setting up the test matrix so that duplicate additions of an SCM do not occur inadvertently. ASTM C 1157 Hydraulic Cement ASTM C 1157 is a performance specification, and there are no chemical requirements for cements that are manufactured according to this specification (they can include SCMs). The types of cements include the following: • Type GU for general use • Type HE for high early strength • Type MS for moderate sulfate resistance • Type HS for high sulfate resistance • Type MH for moderate heat of hydration • Type LH for low heat of hydration • Option R for low reactivity with alkali-reactive aggregates ASTM C 1157 should be consulted for the various physical requirements of these types of hydraulic cements. It is important to obtain from the manufacturer information on both the blend used in the cement (actual quantities of port- land cement and types and quantities of SCMs) and, if possi- ble, the chemical and physical composition of the portland cement itself and of the blend. Supplementary Cementitious Materials Class C fly ash, Class F fly ash, GGBFS, silica fume, and Class N natural pozzolans are the most common types of SCM used in bridge deck concrete; each is discussed in this section. Fly Ash [FA] Fly ash is the finely divided residue created from the com- bustion of ground or powdered coal in coal-fired electric power–generating plants. When coal is ignited to 2700°F (1500°C), any non-combustible materials melt and form droplets. These droplets are rapidly cooled and then collected from the flue gases. The cooled droplets maintain their spher- ical shape and can be solid or hollow ranging in size from less than 1 μm to greater than 100 μm with a median size of 5 to 20 μm. The composition of the fly ash depends on the coal source (4, 21). AASHTO M 295, Standard Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use in Concrete, and ASTM C 618, Standard Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use in Concrete, separate fly ash into two classes, Class C and Class F. Class C Fly Ash Composition. Class C fly ash originates from subbitumi- nous and some lignite coals. It is generally composed of 50% to 90% calcium aluminosilicate glass manifested as solid and hollow spheres. The crystalline phases include relatively chemically inactive phases such as quartz, mullite, ferrite spinel, and hematite. Some of the more chemically active phases include calcium sulfate, alkali sulfate, C2S, C3A, and others. The SiO2, Al2O3, and Fe2O3 contents of Class C fly ash are higher than those of Type I portland cement; the CaO content is lower. The CaO content of Class C fly ash is higher than Class F fly ash (typically 10% to 30% versus 0.7% to 7.5%). The carbon contents of Class C fly ash are generally less than 2% and the Blaine fineness is typically about 300 to

450 m2/kg (comparable to Type I portland cement); its rela- tive density is around 2.5 to 2.7. Hydration. Class C fly ashes are considered hydraulic materials because they will harden on their own. They also undergo some pozzolanic reactions with the calcium hy- droxide product of portland cement hydration. As for port- land cement, calcium silicate hydrate is the main hydration product of Class C fly ash. When Class C fly ash is combined with portland cement, the hydration of the fly ash is assisted by the heat and hydroxyl ions generated by the hydrating portland cement, which facilitates the breakdown of the glassy structure of the fly ash. The glassy structure reacts with calcium hydroxide forming calcium silicate hydrate C-S-H. Specifications. The specifications that govern Class C fly ash are AASHTO M 295 and ASTM C 618. The minimum amount of SiO2 + Al2O3 + Fe2O3 in Class C fly ash is 50%. There are also chemical requirements for maximum sulfur trioxide, maximum moisture content, and maximum loss on ignition. Mandatory physical requirements include a limit on the amount of material retained when wet sieved on a 45-μm (No. 325) sieve, 7- and 28-day minimum percentage of con- trol strength (strength activity index), maximum water requirement as a percentage of control, maximum autoclave expansion or contraction (soundness), and uniformity. AASHTO M 295 (but not the most recent version of ASTM C 618) gives an optional chemical requirement for the amount of available alkalis that should be specified if the aggregate is potentially reactive. ASTM C 618 suggests that a mortar-bar expansion test be conducted to assess reactivity instead and provides criteria for evaluating the results. There are some optional physical requirements, including a maxi- mum increase in drying shrinkage, maximum difference of AEA required to produce a given air content in a single batch compared to 10 preceding batches, and effectiveness in con- tributing to sulfate resistance. Class F Fly Ash Composition. Class F fly ash originates primarily from burning anthracite and bituminous coals. Like Class C fly ash, it is generally composed of 50% to 90% aluminosilicate glass manifested as solid and hollow spheres. The crystalline phases include relatively chemically inactive phases such as quartz, mullite, ferrite spinel, and hematite. The SiO2, Al2O3, and Fe2O3 contents of Class F fly ash are higher than those of Type I portland cement; the CaO content is lower. The CaO content of Class F fly ash is lower than Class C fly ash (0.7% to 7.5% versus 10% to 30%). The carbon contents of Class F fly ash are generally less than 5%, although sometimes it may be as high as 10% and may adversely affect air entrainment. The Blaine fineness of Class F fly ash is on the order of 300 to 450 m2/kg (comparable to Type I portland cement); its rela- tive density is around 2.3 to 2.4. Hydration. Class F fly ashes are considered pozzolanic materials that react with the calcium hydroxide from the hydrating portland cement to form additional C-S-H. When Class F fly ash is combined with portland cement, the hydra- tion of the fly ash is assisted by the heat and hydroxyl ions generated by the hydrating portland cement, which facilitates the breakdown of the glassy fly ash structure, allowing the fly ash to react. Class F fly ash hydration products are believed to be more fluid than those of portland cement, thereby filling pores and making the paste more dense and less permeable. Specifications. The specifications that govern Class F fly ash are AASHTO M 295 and ASTM C 618. The minimum amount of SiO2 + Al2O3 + Fe2O3 for Class C fly ash is 70%. There are also chemical requirements for maximum sulfur trioxide, maximum moisture content, and maximum loss on ignition. Mandatory physical requirements include a limit on the amount of material retained when wet sieved on a 45-μm (No. 325) sieve, 7- and 28-day minimum percentage of con- trol strength (strength activity index), maximum water requirement as a percentage of control, maximum autoclave expansion or contraction (soundness), and uniformity. As for Class F fly ash, there is an optional chemical requirement on the amount of available alkalis in AASHTO M 295 and a maximum limit on mortar-bar expansion given in ASTM C 618 intended to assess the effectiveness at controlling ASR. There are some optional physical requirements, including a maximum increase in drying shrinkage, maximum difference of AEA required from a control, and effectiveness in con- tributing to sulfate resistance. Ground Granulated Blast Furnace Slag [S] GGBFS, also called slag cement or just slag, is an indus- trial by-product of the iron-making process in a blast fur- nace. It occurs when the iron oxides (such as iron ore) are heated to around 2700°F (1500°C). In this process the liq- uid, metallic iron is separated from the molten slag oxides that float on its surface. The slag is rapidly quenched in water, thereby forming a glassy material. This material is dried and then ground finer than Type I portland cement so that it has a surface area between 400 and 600 m2/kg as measured by AASHTO T 153 (ASTM C 204); the particles have an angular shape. Because GGBFS is produced simul- taneously with a manufactured product, its variability within a source is minimal; however, different plants produce different products, such that the variability between sources can be higher (particularly when compar- ing GGBFS sources outside North America). 49

50 Composition. GGBFS is composed primarily (90% to 95%) of glassy calcium silicates and calcium aluminosilicates. The composition of the calcium silicates is in the same region of the CaO-SiO2 phase diagram as C2S. The SiO2, Al2O3, and MgO contents of GGBFS are higher than those of typical Type I cement; Fe2O3 and CaO contents are lower. Sulfide sulfur is also a component of GGBFS. The relative density of GGBFS is 2.94. Hydration. Many GGBFSs will hydrate by themselves, albeit slowly. Some require activators such as alkalis or lime. The principal reaction product of GGBFS is essentially the same as that of portland cement, which is calcium silicate hydrate (6). When GGBFS is combined with portland cement, the hydroxyl ions released during portland cement hydration break down the glassy structure of the GGBFS and allow its hydration. GGBFS hydrates are more fluid than those of portland cement, which reduces pore sizes and increases the denseness of the paste. Specifications. The specification that governs GGBFS is AASHTO M 302 (ASTM C 989), Standard Specification for Ground Granulated Blast-Furnace Slag for Use in Concrete and Mortars. In this specification, three grades of slag (80, 100, and 120) are defined based on their performance in slag activity tests. These tests compare the compressive strength of 50% slag/portland cement blended mortars to 100% portland cement mortars (specific requirements are stipulated for total alkali content and 28-day compressive strength of the portland cement). The grades refer to the percentage of strength of the slag blended mortars to the ref- erence mortar at 28 days. Grades 100 and 120 are by far the most commonly used for bridge deck concrete. Other physical requirements in the specification include a limit on the amount of material retained on a 45-μm sieve when wet screened, specific surface by air permeability, and air content of slag mortar. There are chemical requirements for the maximum amount of sulfide sulfur as well as sulfate ion reported as SO3. While not specified for slag, the potential for possible future deleterious reactions can be evaluated by testing the slag in combination with the cement at a similar mixture ratio as that expected to be used (as is specified with fly ash) for soundness (autoclave expansion). This test would be con- ducted in accordance with ASTM C 151, Test Method for Autoclave Expansion of Portland Cement. An appropriate criterion for soundness is that the percentage of change in length should not be more than 0.8%. Silica Fume [SF] Silica fume is an industrial by-product, which results from the production of silicon-metal or ferrosilicon alloys at 3632°F (2000°C). It is the material that is condensed from the flue gases of electric arc furnaces that are used to reduce high- purity quartz with coal or coke and wood chips (7). Composition. Silica fume consists of at least 85% silicon dioxide, in an amorphous (non-crystalline) form. Because sil- ica fume is derived from flue gases (similarly to fly ash), the particles are spherical; however, they are extremely small with an average diameter of 0.1 μm or 1/100th of a cement particle. The relative density is 2.2. Silica fume is supplied in a con- densed powder form or as slurry in water. Special handling and mixing procedures are needed to assure uniform distribution and to minimize nodules or lumps of silica fume and settling. Hydration. Silica fume is very reactive because of its small particle size and, when mixed with portland cement, reacts with calcium hydroxide to form C-S-H. It increases the density of the paste because of particle packing as well as its reaction products. It may increase the bond between the paste and aggregate, as well as the density of the interfacial transi- tion zone. The C-S-H based on silica fume has been reported to have a lower calcium/silicon ratio than that made from portland cement alone, which allows it to incorporate ion substitutions into its structure such as alkalis. Specifications. The specifications that govern silica fume in concrete are AASHTO M 307 and ASTM C 1240, Standard Specification for Silica Fume Used in Cementitious Mixtures. The chemical requirements of this specification restrict the minimum SiO2 content (85%), the maximum moisture content (3.0%), and the maximum loss on ignition (6.0%). The physical requirements are for oversize particles retained on a 45-μm (No. 325) sieve, accelerated pozzolanic strength activity index (minimum percentage of control is 105% at 7 days), and a minimum specific surface (15 m2/g). There are optional physical requirements regarding unifor- mity of the quantity of air-entraining admixture required to produce an air content of 18%, reactivity with cement alka- lis, and sulfate resistance. Natural Pozzolans (Class N) Types. Diatomaceous earth, high-reactivity metakaolin, and calcined clay are a few of the types of SCMs that are cat- egorized as natural pozzolans. Diatomaceous Earth. Diatomaceous earth consists of deposits of microscopic algae skeletons, which are composed of silica and have open frameworks. These skeletons are found mostly in western states and are relatively uncommon. High-Reactivity Metakaolin. High-reactivity metakaolin is formed by heating kaolinite clay (Al2O3⋅2SiO2⋅2H2O) to between 1100°F to 1650°F (600°C and 900°C) such that the chemically combined water is driven off, resulting in an

amorphous structure of metastable aluminosilicate (Al2O3⋅ 2SiO2). It is not a by-product material; the source is natural. Metakaolin reacts with calcium hydroxide to form calcium sil- icate and calcium aluminate hydrates (14). Its relative density is 2.5; the average particle size ranges from 0.5 to 20 μm (4). Calcined Clay. The origin of calcined clay is kaolinite mix- tures, e.g., 85% kaolinite, 10% quartz, and 5% other clays (10). The mixtures are heated to 550–750°C to release chemically bound water, resulting in an amorphous aluminosilicate phase. The composition is 54% SiO2, 38% Al2O3, and 2% Fe2O3. The particles are not as fine as metakaolin and can be used with silica fume to reduce permeability. When hydrated, C-S-H is formed as well as ettringite and possibly stratlingite (C2ASH8). Specifications. The specifications that govern Class N pozzolans are AASHTO M 295 and ASTM C 618. The mini- mum amount of SiO2 + Al2O3 + Fe2O3 for Class N pozzolan is 70%. There are also chemical requirements for maximum sulfur trioxide, maximum moisture content, and maximum loss on ignition. There is an optional available alkali limit. Mandatory physical requirements include maximum per- centage of material retained when wet sieved on a 45-μm (No. 325) sieve, 7- and 28-day minimum percentage of con- trol strength (strength activity index), maximum water requirement as a percentage of control, maximum autoclave expansion or contraction (soundness), and uniformity. There are some optional physical requirements, including a maximum increase in drying shrinkage, maximum difference of AEA required from a control, effectiveness in controlling ASR, and effectiveness in contributing to sulfate resistance. Aggregates [A1] AASHTO M 6/80 (ASTM C 33) AASHTO M 6, Standard Specification for Fine Aggregate for Portland Cement Concrete; AASHTO M 80, Standard Specifi- cation for Coarse Aggregate for Portland Cement Concrete; or ASTM C 33, Standard Specification for Concrete Aggregates, defines the requirements for grading and quality of normal weight aggregates. Worksheets S2.3 and S2.4 are provided to help organize test data of the fine and coarse aggregates. Grading. The grading requirements for coarse aggregate are given in Table 2 of AASHTO M 80. Typically, the largest size aggregate considered appropriate for the application should be used to reduce the concrete paste volume and lower shrinkage and cost. Grading is a measure of the particle size distribution of an aggregate as determined by a sieve analysis. The grading and grading limits are expressed as the percentage of material passing each sieve. Grading of the aggregate affects aggregate and cement proportions and water requirements. It can also affect the workability of the concrete. In general, aggregates with a smooth grading curve will produce the best results. For highway construction, AASHTO M 43 (ASTM D 448), Standard Classification for Sizes of Aggregate for Road and Bridge Construction, lists the same 13 sieve size numbers as in AASHTO M 80 plus an additional six more coarse aggregate sizes. Fine aggregate has only one range of particle sizes given in AASHTO M 6 (ASTM C 33) as shown in Table S2.1. In leaner mixtures or mixtures with small-size coarse aggregate, the grading that is close to the maximum recommended per- centage passing each sieve is most desirable for workability. AASHTO allows the amounts passing the No. 50 (300 μm) and No. 100 (150 μm) sieves to be reduced to 5% and 0%, respectively, if • The concrete is air entrained and contains at least 400 lbs of cement per cubic yard (237 kg/m3), or • The concrete contains more than 500 lbs of cement per cubic yard (297 kg/m3), or • An approved SCM is used to supply this deficiency in the material passing these two sieves. Other requirements for the fine aggregate are • Less than or equal to 45% must be retained between any two consecutive sieves and • The fineness modulus (FM) as calculated by adding the cu- mulative percentage retained on each designated sieve and dividing by 100 must be between 2.3 and 3.1. The higher the FM, the coarser the aggregate. The FM of the fine ag- gregate is used to estimate the proportions of the coarse and fine aggregate in the mixture. AASHTO M 80 (ASTM C 33) allows a wide range in grad- ing. Generally, the amount of cement and water required in a mixture decreases as the maximum size of the coarse aggre- gate increases because of the decrease in the total aggregate surface area. The maximum size of the aggregate is the size of the small- est sieve that all of the aggregate will pass through. The nom- inal maximum size of the aggregate is the size of the smallest sieve through which the major portion of the aggregate will pass. The nominal maximum size sieve may retain 5% to 15% of the aggregate (depending on the size number). For exam- ple, the maximum size of a No. 67 aggregate is 1 in. (25 mm) and the nominal maximum size is 0.75 in. (19 mm), because 90% to 100% of this aggregate must pass the 0.75 in. (19 mm) sieve, and 100% must pass the 1 in. sieve. The maximum size of the aggregate should not exceed • One-fifth of the narrowest dimension of a concrete member, • Three-quarters of the spacing between reinforcing bars and three-quarters of the spacing between the reinforce- ment and forms, and • One-third the depth of a slab. 51

52 Combined Aggregate Grading. An assessment of the combined grading of the coarse and fine aggregate can provide a better indication of the aggregate performance in the con- crete. For example, inadequate quantities of mid-size aggregate particles (e.g., 0.375 in. [9 mm]) can result in concrete with poor workability (including poor pumpability), higher water demand, and consequently higher shrinkage characteristics. A smooth distribution of aggregate sizes, as illustrated in Figure 5-10 of Kosmatka et al. (21), represents an optimum type of gradation. Shilstone (56) provides suggested options for optimizing grading of aggregate and describes the bene- fits of a combined aggregate analysis. Particle Surface Texture. Smooth, rounded aggregate particles are harder to bond to than rough, angular particles. Bond of the cement paste to the aggregate particles is a par- ticularly important consideration for concrete with relatively high flexural strength. Some SCMs, such as metakaolin and silica fume, improve the transition zone and the bond between cement paste and certain aggregates. Improvement of the transition can help to reduce the permeability of some concretes. Particle Shape. Because flat and elongated particles re- quire an increase in mixing water that can result in lower strength if the w/c is not adjusted by increasing the cement content, such particles should be limited to 15% by weight of total aggregate. The determination of flat and elongated par- ticles is given in ASTM D 4791, Standard Test Method for Flat Particles, Elongated Particles, and Flat and Elongated Parti- cles in Coarse Aggregate, and a method for providing an index of aggregate particle shape and texture is given in ASTM D 3398, Standard Test Method for Index of Aggregate Particle Shape and Texture. Soundness. The durability of an aggregate can be assessed by the soundness test. In this test, the aggregate is soaked in 10% sodium sulfate solution or 15% magnesium sulfate solu- tion and then dried. This process is repeated for five cycles to create a salt crystallization (salt hydration) pressure that sim- ulates pressures generated by water freezing in the pores of the aggregate. Although this test has been found to be appropri- ate for stratified rocks with porous layers or weak bedding planes, it also has been known to give erroneous results for some aggregates. The aggregate can be accepted if it has given satisfactory service when exposed to weathering similar to that to be encountered, or if it gives satisfactory performance when tested in concrete with ASTM C 666, Standard Test Method of Resistance of Concrete to Rapid Freezing and Thawing. Deleterious Substances Coarse Aggregate. Deleterious substances described in AASHTO M 80 (ASTM C 33) include clay lumps and friable particles, chert (with a relative density of less than 2.40), material finer than the 75-μm (No. 200) sieve, and coal and lignite. Maximum allowable amounts of these substances are presented in Table 3 of the standard. The allowable amounts of deleterious materials for bridge deck concrete vary de- pending on the weathering region (exposure) of the structure as shown in Table S2.2. Fine Aggregate. Table S2.3 lists the recommended limits for deleterious substances in fine aggregate. AASHTO M 6 and M 80 (ASTM C 33) have limits on dele- teriously reactive materials, which will be discussed in the following section on ASR. Alkali Silica Reactivity [ASR] According to AASHTO M 80 (ASTM C 33), the aggregate shall not contain any materials that are deleteriously reactive with the alkalis in cement in an amount sufficient to cause ex- cessive expansion of the concrete. Use of the potentially reactive aggregate is not prohibited when used with a cement containing less than 0.60% alkalis, calculated as sodium oxide equivalent (Na2O + 0.658 K2O) or with the addition of a material that has been shown to prevent harmful expansion due to the alkali- aggregate reaction. However, many reactive aggregates may react at later ages when used with low-alkali cement. Laboratory tests for determining whether an aggregate is potentially reactive include ASTM C 295, Standard Guide for Petrographic Examination of Aggregates for Concrete; ASTM C 1260, Test Method for Potential Alkali Reactivity of Aggre- gates (Mortar-Bar Method); ASTM C 1293, Test Method for Determination of Length Change of Concrete Due to Alkali Silica Reaction; ASTM C 289, Test Method for Potential Alkali- Silica Reactivity of Aggregates (Chemical Method); ASTM C 227, Test Method for Potential Alkali Reactivity of Cement- Aggregate Combinations (Mortar-Bar Method); ASTM C 441, Test Method for Effectiveness of Mineral Admixtures or Ground Blast Furnace Slag in Preventing Excessive Expansion of Concrete Due to the Alkali-Silica Reaction; and ASTM C 1567, Test Method for Determining the Potential Alkali-Silica Reactivity of Combinations on Cementitious Materials and Aggregate (Accelerated Mortar-Bar Method). ASTM C 295 [A2]. ASTM C 295 should be used to deter- mine the type(s) and amount(s) of reactive components. Fine and coarse aggregate containing more than the quantities of the constituents listed in Table S2.4 are considered potentially reactive (57). Materials known to be deleteriously alkali-carbonate reac- tive include calcareous rocks that contain substantial amounts of relatively large crystals of dolomite scattered in a finer grained matrix of calcite and clay. The acid-insoluble residue of these rocks typically contains a significant amount

of clay. Alkali-carbonate test procedures will be discussed in the following section. Service Record Evaluation [A3]. Before any laboratory tests beyond the petrographic analysis are performed, the his- torical performance of the aggregate should be reviewed. Valid, comparable concrete service record data, if available, can take precedence over laboratory test results. However, several significant criteria must be met for a truly valid com- parable service record, including a satisfactory performance record for at least 10 to 20 years, although longer periods of documented service are often required. For a valid compari- son, the composition of the historical concrete mixtures must be similar to the composition of the proposed concrete mix- tures with respect to the following: • Alkali contents of the cements • Cement content of the concrete • Water-cement ratio • Presence and amount of SCMs • Geological composition of aggregate (preferably as deter- mined by petrographic reports of both historical and current aggregate) If the similarity of composition cannot be documented to this extent, the service record should not be used as a basis for accepting the aggregate. Laboratory Testing. If the designer chooses to forgo pre- screening, the chosen ASR test performance can be evaluated as part of the full experimental design process. In this case, the chosen mitigation tests must be included as the responses in the experimental design. ASTM C 1260 and ASTM C 1293 are standardized tests, and certain variables such as cement con- tent and w/cm are fixed. Because the results of these stan- dardized tests are independent of cement content and w/cm, keeping these values constant as required will not cause any problems in the analysis of the entire test program defined in the experimental design. Therefore, cement content and w/cm can still be chosen as variables in the experimental design and varied in all tests except those related to ASR. ASTM C 1260 [A4]. ASTM C 1260 is an accelerated screening test for aggregate that develop deleterious expan- sions slowly over a long period of time. This test is generally considered overly conservative and some aggregates that perform well in the field have been shown to fail this test. However, such an innocuous result gives good confidence that the aggregate is acceptable for use. Generally accepted ex- pansion limits for this test are as listed in Table S2.5. Figure S2.2 shows a desirability function for ASTM C 1260 gener- ated from the ASTM criteria. This function eliminates mix- tures with aggregates that expand beyond the recommended limits and only gives full credit (desirability equal to 1) if the expansion is well below the limits. Aggregates with expansion greater than 0.10% should be either • Tested using ASTM C 1293, • Evaluated based on historical performance (see subsection “Service Record Evaluation”), or • Reevaluated with mitigating SCMs using ASTM C 1567, Standard Test Method for Determining the Potential Alkali- Silica Reactivity of Combinations of Cementitious Materi- als and Aggregate (Accelerated Mortar-Bar Method). ASTM C 1293 [A5]. ASTM C 1293 uses an elevated alkali content and the exposure condition of ASTM C 227 (storage over water in a closed container maintained at 100.4 ± 3.6°F [38.0 ± 2°C]). Aggregates with expansions equal to or greater than 0.04% at 1 year are considered potentially deleteriously reactive. ASTM C 1293 is considered to be the most reliable test method for assessing ASR of an aggregate. Its main dis- advantage is the length of time (1 year) needed for the testing. Figure S2.3 shows a desirability function for ASTM C 1293 generated from the ASTM criteria. This function eliminates mixtures with aggregates that expand beyond the recom- mended limits and only gives full credit (desirability equal to 1) if the expansion is well below the limits. According to the Appendix of ASTM C 33, “When inter- preting expansion of laboratory specimens, consideration should be given not only to expansion values at specific ages, but also to the shape of the expansion curve, which may indicate whether the expansion is leveling off or continuing at a constant or accelerating rate.” This statement applies when using ASTM C 1260, ASTM C 1293, or any of the other methods discussed here. Other Laboratory Testing for ASR ASTM C 289 (Chemical Method). ASTM C 289 is a rela- tively rapid screening test (48 hours) that measures quantities of dissolved silica and reduction in alkalinity; it provides help- ful information except for some slowly reactive rocks, such as some granite gneiss and quartzite. In addition, “[r]esults may not be correct for aggregates containing carbonates or mag- nesium silicates, such as antigorite (serpentine), or con- stituents with late-slow reactivity.” This test is somewhat unreliable and should not be used as the sole determination of aggregate reactivity. If the results indicate deleterious or potentially deleterious reactivity, the aggregate should be tested by using ASTM C 227, ASTM C 1260, or ASTM C 1293. ASTM C 227 (Mortar-Bar Method for Cement Aggregate Combinations). Expansions measured in this test are generally considered excessive if they exceed 0.10% at 6 months—a not conservative limit for some slowly reactive 53

54 aggregates. Thus, this method is not suitable for slowly reac- tive aggregates typically containing strained or microgranu- lated quartz. Aggregates that may be slowly reactive should be evaluated using ASTM C 1260 or ASTM C 1293. ASTM C 441. ASTM C 441, Standard Test Method for Effectiveness of Mineral Admixtures or Ground Blast Furnace Slag in Preventing Excessive Expansion of Concrete Due to the Alkali-Silica Reaction, measures the effectiveness of a mineral admixture or SCM for mitigating excessive expansion due to a potentially reactive aggregate. This test method uses mortar bars as in ASTM C 227. ASTM C 618 provides additional criteria for its use with pozzolans. ASTM C 989, Standard Specification for Ground Granulated Blast- Furnace Slag for Use in Concrete and Mortars, Appendix X3, describes the use of ASTM C 441 for GGBFS including guid- ance on interpretation of the results. Determining the Risk of ASR. A risk-evaluation process for making decisions regarding potentially reactive aggregates has been proposed by Fournier et al. (58). This process is based on the assessment of the following factors: • The degree of reactivity of the particular aggregate • The size of the concrete element and the environmental conditions it will face • The expected service life of the structure For the proposed process, the level of risk for concrete exposed to humid air is rated on a scale of 1 to 4 as shown in Table S2.6. Reactivity Level. The degree of reactivity of the aggregate is based on ASTM C 1293 and C 1260 testing data as shown in Table S2.7. Level of Prevention. Based on the risk level and the in- tended service life, a Level of Prevention was proposed as shown in Table S2.8. Preventative Measures. Depending on the level of pre- vention needed, the preventative measures listed in Table S2.9 are recommended. Essentially, these recommendations limit the alkali contributed by the cement, require use of SCMs to control expansion, or both. Recommended Levels of SCMs for Mitigation. Fournier et al. (58) suggest types and quantities of SCMs for ASR mit- igation based on experience and previous laboratory inves- tigations. These guidelines consider the level of prevention needed and the composition of the SCM. They suggest that laboratory studies can be used to evaluate the effectiveness of SCMs that do not meet recommended compositional requirements or to evaluate the effectiveness of lower con- centrations of the SCMs. Examples of some of their recom- mendations are listed in Table S2.10. Summary of ASR Mitigation Guidelines [A6]. If an aggregate is judged to be potentially deleteriously reactive by ASTM C 1260 or C 1293 or by historical performance, the fol- lowing measures can be taken to mitigate this reaction (these guidelines should be incorporated with the other durability guidelines listed in Worksheet S1.1): • Use a low-alkali cement (<0.60% Na2Oeq) and/or limit the total alkalinity of the concrete to 1.8 to 3.0 kg/m3 (3.0 to 5.0 lb/yd3) based on the risk level and desired service life as previously described. • Use a blended cement (ASTM C 595 with Table 2 optional mortar expansion requirement) known and tested to mit- igate ASR. • Use a cement meeting performance specification ASTM C 1157 with Option R. • Use a pozzolanic material shown to be effective in control- ling ASR (a minimum of 25% to 30% [by weight of cement] Class F fly ash has been shown to be effective for mitigating ASR). • Test and use 5% to 8% silica fume by weight of cement (combinations of SCMs must be used if these levels are not effective). • Test a GGBFS shown to be effective for preventing exces- sive expansion per ASTM C 989, Appendix X3. ASTM C 989 states a minimum of 40% GGBFS alone “will generally prevent excessive expansion with cements having alkali contents up to 1.0%.…” • Test combinations of SCMs per footnote “c” in Table S2.10. • Test 10% to 15% replacement of cement by metakaolin (59). The type(s) and quantities of the above SCMs or combina- tions of SCMs should be tested to evaluate their ability to ad- equately mitigate deleterious expansion using either test method ASTM C 1567, described in the following section, or ASTM C 1293 with mitigating measures included and per- formed for 2 years. To be considered effective in the 2-year ASTM C 1293 test, the expansion should be less than 0.04%. ASTM C 1567 [A7]. ASTM C 1567, Standard Test Method for Determining the Potential Alkali Reactivity of Combinations of Cementitious Materials and Aggregate (Ac- celerated Mortar-Bar Method), is based on ASTM C 1260, but allows the addition of SCMs to be tested in an accelerated manner. According to the Appendix of this method, “[c]om- binations of cement, pozzolan or ground granulated blast- furnace slag, and aggregate that expand less than 0.10% at

16 days after casting are likely to produce acceptable expan- sions when tested in concrete (that is, Test Method C 1293) and to have a low risk of deleterious expansion when used in concrete under field conditions.” Furthermore, “[c]ombina- tions of cement, pozzolan or ground granulated blast-furnace slag, and aggregate that expand more than 0.10% at 16 days after casting are indicative of potentially deleterious expan- sion. However, the potential for deleterious reaction should be confirmed by testing the same combination of materials in concrete (that is, Test Method C 1293). The expansion may be reduced by retesting the material combination using the pozzolan or ground granulated blast-furnace slag at a higher replacement level.” Alkali-Carbonate Rock Reaction Alkali-carbonate rock reaction is a mechanism where the alkali from the cement reacts with carbonate forms of aggre- gate. ASTM C 586, Standard Test Method for Potential Alkali Reactivity of Carbonate Rocks as Concrete Aggregates (Rock- Cylinder Method), is a preliminary screening test to indicate the potential for deleterious expansion of carbonate rocks. Such rock is relatively infrequent. Another test, ASTM C 1105, Standard Test Method for Length Change of Concrete Due to Alkali-Carbonate Rock Reaction, is used to evaluate specific combinations of materials. A cement aggregate com- bination might be classified as potentially reactive if the aver- age expansion of six specimens is greater than 0.015% at 3 months, 0.025% at 6 months, or 0.300% at 1 year. ASTM C 1293 (CSA A23.2-26A) is the most suitable test for identifying potentially reactive alkali-carbonate rock ag- gregate when an unknown cement is to be used. Concrete expansion exceeding 0.04% at 1 year is considered as poten- tially alkali-carbonate reactive. Pozzolans generally have not been found to control alkali- carbonate reaction. Mitigation measures include the following: • Selective quarrying • Diluting the reactive rock to less than 20% of the aggregate in the concrete • Using smaller maximum size of the aggregate • Using very low alkali cement. An alkali limit of 0.60% Na2Oeq may not be adequate to prevent excessive expansion Air-Entraining Admixtures [AEA] Air-entraining admixtures are chemicals that, upon being mixed in concrete, stabilize air bubbles into the paste fraction of the concrete. The purpose of air entrainment is for freez- ing and thawing durability (see Step 1), although entrained air also can improve workability and reduce segregation. The materials typically used for this purpose are Vinsol (wood) resins or combinations of materials such as sulfonated hydrocarbons, wood rosins, and tall oil fatty acid soaps. The specification governing air-entraining admixture ma- terials is AASHTO M 154 (ASTM C 260), Standard Specifi- cation for Air-Entraining Admixtures for Concrete. The physical requirements for the concrete containing the air- entraining admixture include deviation from initial and final setting time of not more than 1 hour 15 minutes (1:15) ear- lier nor 1:15 later than the control concrete. The compressive strength is required to be at least 90% of the control at 3, 7, and 28 days. The purchaser may also require that the flexural strength at 3, 7, and 28 days be a minimum of 90% of the con- trol. The minimum relative durability factor as measured according to ASTM C 666 cyclic freezing tests is required to be 80%. The bleeding as a fraction of the net amount of mix- ing water has a limit of 2% over the control. Chemical Admixtures [CH] The specification for chemical admixtures is AASHTO M 194 (ASTM C 494), Standard Specification for Chemical Admixtures for Concrete. There are seven types of chemical admixtures: • Type A, water-reducing • Type B, retarding • Type C, accelerating • Type D, water-reducing and retarding • Type E, water-reducing and accelerating • Type F, water-reducing, high range • Type G, water-reducing, high range, and retarding For each type of admixture, there are requirements re- garding the performance of the concrete in which they are added. These requirements include a maximum deviation of the initial and final times of setting from the control; a mini- mum percentage of compressive strength of the control at 1, 3, 7, and 28 days, 6 months, and 1 year; a minimum percent- age of flexural strength of control at 3, 7, and 28 days; a max- imum shrinkage in terms of percentage of control or increase over control; and a minimum relative durability factor. However, admixtures meeting ASTM M 194 can cause unan- ticipated changes to setting times and shrinkage. A certificate of compliance to AASHTO M 194 (ASTM C 494) can be requested from the manufacturer. The behavior of the chemical admixture in a specific cementitious blend, or with other chemical admixtures, and at varying dosages can be complicated and is best tested in trial batches. It is recommended that the manufacturer provide data on admixture compatibility with SCMs and other admixtures and review the admixture’s effect on strength, setting, and shrinkage. 55

56 Example from Hypothetical Case Study To demonstrate how the worksheets in this step may be used, examples of several completed worksheets are pre- sented in Tables S2.11 through S2.15. According to the pro- cedure outlined in the previous section, Table S2.11 lists all the potential materials considered. Then the properties of the cement sources that were identified, namely “Cemsource 1” and “Cemsource 2,” are listed in Table S2.12 (completed Worksheet S2.2). Once all the data were collected for the cements, the sources were compared, and a selection of that material type was made. In this case, “Cemsource 2” was selected because of the comparatively lower alkali content; this selection was denoted by a box drawn around the Source in Table S2.11. A similar process was followed for the fine aggregate using Worksheet S2.3 (see Table S2.13) and “Fineagg manufacturer 2” was selected because of the higher fineness modulus and the better soundness test results; this selection was recorded in Table S2.11. Completed versions of Worksheet S2.5 (Table S2.14) and Worksheet S2.10 (Table S2.15) can be used to perform the same functions for Class C fly ash sources and air-entraining and chemical admixtures. If the user would be interested in testing which Class C fly ash would give the best performance (i.e., using a Class C fly ash as a type factor in the experiment), two Class C fly ash sources could be selected by having boxes drawn around them in Worksheet S2.1 (Table S2.11).

Worksheets for Step 2 Raw Material Source 1 Source 2 Source 3 Source 4 tnemeC etagergga eniF etagergga esraoC hsa ylf C ssalC hsa ylf F ssalC gals ecanruf tsalb detalunarg dnuorG emuf aciliS MCS rehtO erutximda gniniartne-riA erutximda lacimehC erutximda lacimehC :rehtO Worksheet S2.1. List of available raw materials.

Test/Property AASHTO Limit Cement 1 Cement 2 Cement 3 Cement 4 Manufacturer Plant location Mill report date AASHTO M 85 (ASTM C 150) Cements Type C3S (%)1 ≤ 58 for Type II C2S (%)2 C3A (%)3 ≤ 8 for Type II Total alkalis (Na2Oeq ) (%)4 ≤ 0.60 for low alkali optional requirement SO3 (%) 3.0 (unless C3A > 8%, then 3.5 for Type I)5 MgO (%)6 ≤ 6.0 Rapid stiffening (y/n)7 AASHTO M 240, ASTM C 595, or C 1157 Cements Type Portland cement, % Second constituent, % Third constituent, % Fourth constituent, % 1Relates to early-age strength gain. 2Higher contents indicate slower early-age strength gain, but may have higher ultimate strength. 3C3A reacts with sulfate to form ettringite; higher values indicate less resistance to external sulfate attack. 4This value is important if potentially reactive aggregates are being used in the mixture. 5These limits are for Type I and II cements; if SO3 exceeds these limits, request ASTM C 1038 backup data. The expansion in water according to ASTM C 1038 should not exceed 0.020% at 14 days. Type III cement has different limits; see ASTM C 150 for details. 6Excessive amounts of MgO (periclase) can result in unsoundness (deleterious expansion). 7Prescreening cements by ASTM C 359, Standard Test Method for Early Stiffening of Portland Cement (Mortar Method), may be desirable to test for flash or false set or high water demand. The needle penetration at 11 minutes or on remix should be greater than 35 mm. Worksheet S2.2. Cement data.

Test/Property AASHTO M 6 Class A Limit Local Requirements Fine Agg. 1 Fine Agg. 2 Fine Agg. 3 rerutcafunaM noitacol tiP Date of last ASTM C 295 petrographic examination ygolareniM yramirP )DSS( ytivarg cificepS )%( yticapac noitprosbA Clay lumps and friable particles ≤ 3.0% max ≤ 2.0% max, concrete subject to abrasion Material finer than 75-μm (No. 200) sieve ≤ 3.0% max, all other concrete Coal and lignite, concrete where surface appearance is not important ≤ 0.25%, max noitadarg dradnats steem kcehC 1.3-3.2 suludom sseneniF Organic impurities Lighter than color standard Soundness Weighted average loss ≤10%* stnemeriuqer lacoL secnatsbus suoireteled rehtO Types and amounts (%) of particles deleteriously reactive with alkalis ASTM C 1260 Expansion <0.10%† ASTM C 1293 Expansion <0.04%† * When sodium sulfate is used; 15% when magnesium sulfate is used † ASTM C 33 requirements Worksheet S2.3. Fine aggregate data.

Test/Property AASHTO M 80 Class A Requirements† Local Requirements Coarse Agg. 1 Coarse Agg. 2 Coarse Agg. 3 Manufacturer Pit location Check meets standard gradation Date of last ASTM C 295 petrographic examination Primary Mineralogy Grading size number Specific gravity (SSD) Absorption capacity (%) Clay lumps and friable particles ≤ 2.0% max. Chert* ≤ 3.0% max. Sum of clay lumps, friable particles, and chert* ≤ 3.0% max. Material finer than 75-μm (No. 200) sieve ≤ 1.0% max. Coal and lignite ≤ 0.5% max. Abrasion ≤ 50% max. Sodium sulfate soundness, 5 cycles ≤ 12% max. ** Types and amounts (%) of particles deleteriously reactive with alkalis -- ASTM C 1260 Expansion <0.10%‡ ASTM C 1293 Expansion <0.04%‡ * Less than 2.40 relative density SSD ** 18% max. if magnesium sulfate is used. † These are the most stringent AASHTO M 80 values. ‡ ASTM C 33 recommendations Worksheet S2.4. Coarse aggregate data.

Test/Property AASHTO M 295 Requirement Fly Ash 1 Fly Ash 2 Fly Ash 3 Fly Ash 4 rerutcafunaM noitacol tnalp/ecruoS SiO2+Al2O3+Fe2O3, % ≥ 50.0 % ,OaC SO3, % ≤ 5.0 Moisture content, % ≤ 3.0 Loss on ignition, % ≤ 5.0 Amt. retained when wet-sieved on 45 μm (No. 325) sieve, % ≤ 34 Strength activity index, 7-day, % of control ≥ 75 Strength activity index, 28-day, % of control ≥ 75 Water requirement, % of control ≤ 105 Soundness: autoclave expansion or contraction, % ≤ 0.8 Density, variation from average, % ≤ 5 Percent retained on 45-μm (No. 325) sieve, percentage points from average ≤ 5 of variation Available alkalis, % ≤ 1.5 Worksheet S2.5. Class C fly ash data.

Test/Property AASHTO M 295 Requirement Fly Ash 1 Fly Ash 2 Fly Ash 3 Fly Ash 4 Manufacturer Source/plant location SiO2+Al2O3+Fe2O3, % ≥ 70.0 CaO, % SO3, % ≤ 5.0 Moisture content, % ≤ 3.0 Loss on ignition, % ≤ 5.0 Amt. retained when wet-sieved on 45 μm (No. 325) sieve, % ≤ 34 Strength activity index, 7-day, % of control ≥ 75 Strength activity index, 28-day, % of control ≥ 75 Water requirement, % of control ≤ 105 Soundness: autoclave expansion or contraction, % ≤ 0.8 Density, variation from average, % ≤ 5 Percent retained on 45-μm (No. 325) sieve, variation, percentage points from average ≤ 5 Available alkalis, % ≤ 1.5 Worksheet S2.6. Class F fly ash data.

Test/Property AASHTO M 295 Requirement Natural Pozzolan 1 Natural Pozzolan 2 Natural Pozzolan 3 Natural Pozzolan 4 Manufacturer Source/plant location SiO2+Al2O3+Fe2O3, % ≥ 70 CaO, % SO3, % ≤ 4.0 Moisture content, % ≤ 3.0 Loss on ignition, % ≤ 5.0 Amt. retained when wet-sieved on 45 μm (No. 325) sieve, % ≤ 34 Strength activity index, 7-day, % of control ≥ 75 Strength activity index, 28-day, % of control ≥ 75 Water requirement, % of control ≤ 115 Soundness: autoclave expansion or contraction, % ≤ 0.8 Density, variation from average, % ≤ 5 Percent retained on 45-μm (No. 325) sieve, percentage points from average ≤ 5 Available alkalis, % ≤ 1.5 Worksheet S2.7. Class N natural pozzolan data.

Test/Property AASHTO M 302 Value GGBFS 1 GGBFS 2 GGBFS 3 GGBFS 4 Manufacturer Source/plant location Grade Amt. retained when wet-sieved on 45-μm (No. 325) sieve, % ≤ 20 Specific surface by air permeability (Method C 204) Air content of slag mortar, % ≤ 12 Grade 100: ≥ 75 7-day slag activity index, %* Grade 120: ≥ 95 Grade 80: ≥ 75 Grade 100: ≥ 95 28-day slag activity index, %* Grade 120: ≥ 115 Sulfide sulfur (S), % ≤ 2.5 Sulfate ion reported as SO3, % ≤ 4.0 *Any individual sample Worksheet S2.8. Ground granulated blast furnace slag (GGBFS) data.

Test/Property AASHTO M 307 Value Silica Fume 1 Silica Fume 2 Silica Fume 3 Silica Fume 4 Manufacturer Source/plant location SiO2, % ≥ 85.0 Moisture content, % ≤ 3.0 Loss on ignition, % ≤ 6.0 Optional: moisture content of dry microsilica, % ≤ 3.0 Optional: available alkalis as Na2O, % ≤ 1.5 Strength activity index: With portland cement at 7 and 28 days, min. percent of control ≥ 100 Worksheet S2.9. Silica fume data.

Test/Property AASHTO M 154* or M 194** Value AEA 1 AEA 2 Chemical Admixture 1 Chemical Admixture 2 Chemical Admixture 3 Brand Name -- Manufacturer -- Chemistry -- AEA: Initial time of setting, allowable deviation from control, not more than (hr:min) 1:15 earlier nor 1:15 later Final time of setting, allowable deviation from control, not more than (hr:min) 1:15 earlier nor 1:15 later Compressive strength, % of control at 3, 7 and 28 days ≥ 90 Chemical admixtures: Type -- Setting time and other requirements See Table 1 of ASTM C 494 * Equivalent to ASTM C 260 ** Equivalent to ASTM C 494 Worksheet S2.10. Air-entraining agent (AEA) and chemical admixture data.

67 Figures for Step 2 Are service records available (see Background Information)? [A3] List Available Materials and Sources (Worksheet S2.1) For each cement source, fill in Worksheet A-S2.2 from Mill Certificate. Verify each meets AASHTO M 85 or M 240 (ASTM C 150, C 595 or C 1157) [C] For each coal fly ash, raw or calcined natural pozzolan source, verify that it meets AASHTO M 295 (ASTM C 618); fill in Worksheets S2.5, S2.6 and/or S2.7 [FA] For each ground granulated blast-furnace slag source, verify that it meets AASHTO M 302 (ASTM C 989); fill in Worksheet S2.8 [S] For each silica fume source, verify that it meets AASHTO M 307 (ASTM C 1240); fill in Worksheet S2.9 [SF] For each chemical admixture, verify that it meets AASHTO M 194 (ASTM C 494); for each air-entraining admixture, verify that it meets AASHTO M 154 (ASTM C 260); fill in Worksheet S2.10 [AEA and CH] For each fine and coarse aggregate, verify it meets AASHTO M 6 and M 80 (ASTM C 33); fill in Worksheets S2.3 and S2.4 [A1] Characterize each aggregate with ASTM C 295. Does the % reactive material indicate potential for deleterious expansion? [A2] Aggregate is candidate for use Perform ASTM C 1260 [A4] Is the 16-day expansion 0.1% or less? Perform ASTM C 1293 [A5] Is the 1-year expansion greater than 0.04%? Test mitigation efficacy of SCMs using ASTM C 1567 [A7] Is the 14- day expansion less than 0.1%? Aggregate is candidate for experimental design with no less than the dosages and combinations of SCMs tested Aggregate source unacceptable for use with the tested mixtures ASR is a concern (Figure S1.1) For each combination of candidate cement and fly ash, perform ASTM C 151 autoclave soundness test. Reject combinations where expansion is greater than 0.8% Circle candidate concrete raw materials (Worksheet S2.1) No No No No Yes Yes Yes Yes Adjust proposed mixes using the ASR Mitigation Guidelines [A6] Do the records indicate good performance in this environment? No Yes Test mitigation efficacy of SCMs with 2-yr ASTM C 1293 Or Is the 2-year expansion greater than 0.04%? Yes No Yes Or No Water: potable or per C 194 Figure S2.1. Selecting durable raw materials.

68 0 0.2 0.4 0.6 0.8 1 0 0.05 0.1 0.15 0.2 Expansion at 16 days (%) De s ira bi lit y Figure S2.2. Desirability function for ASTM C 1260. 0 0.2 0.4 0.6 0.8 1 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 Expansion at 1 year (%) De s ira bi lit y Figure S2.3. Desirability function for ASTM C 1293. Tables for Step 2 Sieve Size % passing by mass 9.5 mm (3/8 in.) 100 4.75 mm (No. 4) 95 to 100 2.36 mm (No. 8) 80 to 100 1.18 mm (No. 16) 50 to 85 0.6 mm (No. 30) 25 to 60 0.3 mm (No. 50) 5 to 30 (AASHTO 10 to 30) 0.15 mm (No. 100) 0 to 10 (AASHTO 2 to 10) Maximum allowable, % (1) Clay Lumps and Friable Particles (2) Chert < 2.4 Specific Gravity Sum of (1) and (2) Material finer than 200 mesh sieve Coal and lignite Weathering Region* C 33 M 80 C 33 M 80 C 33 M 80 C 33 and M 80 C 33 and M 80 Severe 3.0 2.0 5.0 3.0 5.0 3.0 1.0 0.5 Moderate 5.0 3.0 5.0 3.0 7.0 5.0 1.0 0.5 Negligible 5.0 5.0 ** 5.0 ** 7.0 1.0 0.5 * The locations of the weathering regions are given in Figure 1 of ASTM C 33. ** No specifications given. Table S2.1. Grading limits for fine aggregate (AASHTO M 6/ASTM C 33). Table S2.2. Recommended limits (ASTM C 33 and AASHTO M 80) for deleterious substances in coarse aggregates.

69 Max % by mass Deleterious Substances ASTM C 33 AASHTO M 6 Class A Clay lumps and friable particles 3.0 3.00 -200 mesh fraction (for concrete subject to abrasion) 3.0* 2.00 -200 mesh fraction (for all other concrete) 5.0* 3.00 Coal and lignite 1.0 0.25 Other deleterious substances (such as shale, alkali, mica, coated grains, and soft and flaky particles) — Specifier shall insert appropriate limits *For manufactured sand, if -200 mesh (< 74 μm) fraction is free of clay or shale, these limits can be increased to 5% and 7%, respectively—not a category in ASTM C 33. Constituent Amount Optically strained, microfractured, or microcrystalline quartz 5.0% Chert or chalcedony 3.0% Tridymite or cristobalite 1.0% Opal 0.5% Natural volcanic glass in volcanic rocks 3.0% Source: Portland Cement Association (57) Expansion at 14 days* Typical Field Behavior <0.10% Innocuous >0.20% Potentially deleterious 0.10% to 0.20% Includes both innocuous and deleterious behavior * Some aggregates such as granites, gneiss, metabasalts (greenstones) grandodiorites of Grenville age and some horizons of the Potsdam sandstone (upper New York and southwest Quebec) have reacted deleteriously in field concrete, but exhibit less than 0.10% expansion at 14 days. Aggregate Risk Level Non-reactive 1 Moderately reactive 3 Highly reactive 4 Source: Fournier et al. (58) Table S2.3. Recommended limits (ASTM C 33 and AASHTO M 6 Class A) for deleterious substances in fine aggregates. Table S2.4. Quantities of constituents considered potentially reactive in aggregates. Table S2.5. ASTM C 1260 limits on expansion. Table S2.6. Risk levels for concrete exposed to humid air.

70 Reactivity Level ASTM C 1293 % Expansion at 1 year* ASTM C 1260 % Expansion at 14 days** Non-reactive <0.04 <0.15 (0.10 for limestone and certain other aggregates) Moderately reactive 0.04 to 0.12 — Highly reactive >0.12 >0.15 (0.10 for limestone) * Based on a combination of fine and coarse aggregate intended for use in concrete. If combination result is not available, the result for the most expansive of the aggregates shall be used. ** When ASTM C 1260 and ASTM C 1293 data conflict, the results of the ASTM C 1293 tests should be considered definitive. Source: Fournier et al. (58) Table S2.7. Degree of reactivity of aggregates. Measures to Prevent Deleterious ASR* Prevention Level Limit Alkali Contributed by Portland Cement to (kg of Na2Oeq/m3 of concrete)** And/Or SCM Mitigation V — — — W <3.0 Or Sufficient SCM to control expansion X <2.4 Or Sufficient SCM to control expansion Y <1.8 Or Sufficient SCM to control expansion Z <1.8 And Sufficient SCM to control expansion * The first option for conditions W through Z is to reject the proposed aggregate. For condition V, accept for use without any preventative measure. ** lb/yd3 = 1.686 x kg/m3, % Na2Oeq ≈ 0.043 x kg/m3 for normal weight concrete Source: Fournier et al. (58) Cement Replacement Level (% mass)a, c Type of SCM Total Alkali Content of SCM (% Na2Oeq) Composition Requirement Prevention Level W Prevention Level X Prevention Level Y & Z <3.0 <8% CaO 15 20 25 Fly Ash (Class F) 3.0 to 4.5 <8% CaO 20 25 30 GGBFS <1.0 — 25 35d 50d Silica Fume <1.0 SiO2 >85% 2.0 × alkali contentb 2.5 × alkali contentb 3.0 × alkali contentb, e a The maximum alkali content of the cement used in combination with the SCMs should be <1.0% Na2Oeq. b Based on the alkali content of the concrete (expressed as kg/m3 Na2Oeq), except where silica fume is the only SCM when the minimum silica fume content shall be 7.0%. c When two or more SCMs are used together, the sum of the parts of each SCM shall be 1. For example, if silica fume and GGBFS are used together, the silica fume may be reduced to one-third of the recommended amount provided that the GGBFS is at least two-thirds of the minimum GGBFS level recommended. d In some regions, this level of GGBFS is not allowed in deck concrete because of concerns about deicer salt scaling. e Concrete with silica fume additions greater than 5% by weight of cement can be very sensitive to early shrinkage cracking. Source: Fournier et al. (58) Table S2.9. Preventative measures. Table S2.10. Recommended levels of SCMs. Level of Prevention* ASR Risk Level < 5 Years of Service Life (Temporary Element) 5 to 50 Years of Service Life > 50 Years of Service Life 3 V X Y 4 W Y Z *V, W, X, Y, Z represent preventative measures described in Table S2.9. Source: Fournier et al. (58) Table S2.8. Level of prevention.

Raw Material Source 1 Source 2 Source 3 Source 4 Cement Fine aggregate Coarse aggregate Class C fly ash Class F fly ash Ground granulated blast furnace slag Silica fume Other SCM Air entraining admixture Chemical admixture Chemical admixture Other: Fffffff Selected for use Table S2.11. Completed Worksheet S2.1, example list of available raw materials.

Test/Property AASHTO Limit Cement 1 Cement 2 Cement 3 Cement 4 Manufacturer Plant location Mill report date AASHTO M 85 (ASTM C 150) Cements Type C3S (%)1 ≤ 58 for Type II C2S (%)2 C3A (%)3 ≤ 8 for Type II Total alkalis (Na2Oeq ) (%)4 ≤ 0.60 for low alkali optional requirement SO3 (%) 3.0 (unless C3A > 8%, then 3.5 for Type I)5 MgO (%)6 ≤ 6.0 Rapid stiffening (y/n)7 AASHTO M 240, ASTM C 595, or C 1157 Cements Type Portland cement, % Second constituent, % Third constituent, % Fourth constituent, % 1 Relates to early-age strength gain. 2 Higher contents indicate slower early-age strength gain, but may have higher ultimate strength. 3 C3A reacts with sulfate to form ettringite; higher values indicate less resistance to external sulfate attack. 4 This value is important if potentially reactive aggregates are being used in the mixture. 5 These limits are for Type I and II cements; if SO3 exceeds these limits, request ASTM C 1038 backup data. The expansion in water according to ASTM C 1038 should not exceed 0.020% at 14 days. Type III cement has different limits; see ASTM C 150 for details. 6 Excessive amounts of MgO (periclase) can result in unsoundness (deleterious expansion). 7 Prescreening cements by ASTM C 359, Standard Test Method for Early Stiffening of Portland Cement (Mortar Method), may be desirable to test for flash or false set or high water demand. The needle penetration at 11 minutes or on remix should be greater than 35 mm. Table S2.12. Completed Worksheet S2.2, example of cement data.

Test/Property AASHTO M 6 Class A Limit Local Requirements Fine Agg. 1 Fine Agg. 2 Fine Agg. 3 Manufacturer Pit location Date of last ASTM C 295 petrographic examination Primary Mineralogy Specific gravity (SSD) Absorption capacity (%) Clay lumps and friable particles ≤ 3.0% max ≤? 2.0% max, concrete subject to abrasion Material finer than 75-μm (No. 200) sieve ≤ 3.0% max, all other concrete Coal and lignite, concrete where surface appearance is not important ≤? 0.25%, max Check meets standard gradation Fineness modulus 2.3-3.1 Organic impurities Lighter than color standard Soundness Weighted average loss 10%* Other deleterious substances Local requirements Types and amounts (%) of particles deleteriously reactive with alkalis ASTM C 1260 Expansion <0.10%† ASTM C 1293 Expansion <0.04%† * When sodium sulfate is used; 15% when magnesium sulfate is used. † ASTM C 33 requirements. Table S2.13. Completed Worksheet S2.3, example of fine aggregate data.

Test/Property AASHTO M 295 Requirements Fly Ash 1 Fly Ash 2 Fly Ash 3 Fly Ash 4 Manufacturer Source/plant location SiO2+Al2O3+Fe2O3, % ≥ 50.0 CaO, % SO3, % ≤ 5.0 Moisture content, % ≤ 3.0 Loss on ignition, % ≤ 5.0 Amt. retained when wet-sieved on 45 μm (No. 325) sieve, % ≤ 34 Strength activity index, 7-day, % of control ≥ 75 Strength activity index, 28-day, % of control ≥ 75 Water requirement, % of control ≤ 105 Soundness: autoclave expansion or contraction, % ≤ 0.8 Density, variation from average, % ≤ 5 Percent retained on 45-μm (No. 325) sieve, percentage points from average ≤ 5 of variation Available alkalis, % ≤ 1.5 Table S2.14. Completed Worksheet S2.5, example of Class C fly ash data.

Test/Property AASHTO M 154* or M 194** Value AEA 1 AEA 2 Chemical Admixture 1 Chemical Admixture 2 Chemical Admixture 3 Brand Name -- Manufacturer -- Chemistry -- AEA: Initial time of setting, allowable deviation from control, not more than (hr:min) 1:15 earlier nor 1:15 later Final time of setting, allowable deviation from control, not more than (hr:min) 1:15 earlier nor 1:15 later Compressive strength, % of control at 3, 7 and 28 days ≥ 90 Chemical admixtures: Type -- Setting time and other requirements See Table 1 of ASTM C 494 * Equivalent to ASTM C 260 ** Equivalent to ASTM C 494 Table S2.15. Completed Worksheet S2.10, example of air entraining agent (AEA) and chemical admixture data.