For example, it is known that finely porous and highly weathered, or otherwise altered coarse aggregate particles can be especially susceptible to freeze-thaw damage.  This damage is either manifested through fracturing of the particle or the surrounding paste resulting in what is commonly called D-cracking of the pavement. Finely porous aggregate near the pavement surface can also suffer freeze-thaw damage in the form of popouts.  For pavements being constructed in regions experiencing freezing and thawing cycles, petrography can be used to identify potentially susceptible materials within the coarse aggregate sources.  Thus, this standard can provide an initial screening in the aggregate selection process.  Furthermore, the petrographic properties of aggregates with known performance records can be compared with new sources to make an initial assessment of the new aggregate’s suitability for use in concrete pavement construction.

It is recommended by some that aggregates with high absorption (greater than 2 percent in 24 hours) should not be used in a freeze-thaw environment (Pigeon and Plateau 1995).  This is because even though the aggregates may maintain their integrity under freezing and thawing, they may rapidly expel water that can either fracture the paste or dissolve soluble paste components.  Therefore, aggregates having questionable petrographic or absorption characteristics should be tested in a confined state (i.e., embedded in concrete) using test methods such as ASTM C 666 Method A or C 682 (which are described later).

Chemically unstable minerals, such as soluble sulfates and unstable sulfides, or volumetrically unstable materials such as smectites can be readily identified using petrographic means.  If present in sufficient quantity, these minerals can have a deleterious effect on hardened concrete.  Soluble sulfates can lead to internal sulfate attack, whereas unstable sulfides can form sulfuric acid, resulting in acid attack. Volumetrically unstable materials will shrink and swell under moisture cycles resulting in particle and/or paste degradation.

In addition, identification of many alkali–silica reactive constituents can be accomplished using this standard.  Commonly recognized alkali–silica reactive constituents include opal, cristobalite, tridymite, siliceous and some intermediate volcanic glass, chert, glassy to cryptocrystalline acid volcanic rocks, synthetic siliceous glass, some argillites, phyllites, metamorphic graywackes, schists, gneisses, gneissic granites, vein quartz, quartzite, and sandstone. It is noted that this list does not identify all constituents of North American aggregates that are potentially reactive (PCA 1995). An excellent summary of the petrography of alkali–silica reactive aggregates is provided in ACI 221.1R-98 (ACI 1998).  If more than the following quantities of constituents are observed in the fine and coarse aggregate, it shall be considered potentially reactive (PCA 1995):

• Optically strained, microfractured, or microcrystalline quartz exceeding 5.0 percent (a common constituent of granite and granite gneiss).
• Chert or chalcedony exceeding 3.0 percent.
• Tridymite or cristobalite exceeding 1.0 percent.
• Opal exceeding 0.5 percent.
• Natural volcanic glass in volcanic rocks exceeding 3.0 percent.

If ASTM C 295 indicates that the aggregate is potentially reactive according to the above criteria, or if potential for alkali–silica reactivity is suspected, ASTM C 1260 and/or C 1293 should be conducted. It is noted that a petrographic examination may not detect small amounts of reactive material and that the results should therefore be confirmed by expansion tests as will be discussed later (ACI 1998).

Petrographic evaluation of aggregate can also be used to identify potential alkali–carbonate reactive aggregate.  These are usually calcareous dolomites or dolomitic limestones with clayey insoluble residues.  Some alkali–carbonate reactions can occur in dolomites and very fine-grained limestones that are essentially free of clay, but it is uncertain whether these reactions are deleterious.  The structure and texture of reactive carbonate aggregates is rather unique and identifiable (Dolar-Mantuani 1982).  Identification of dolomite in a fine-grained carbonate rock makes it desirable to run ASTM C 586 and/or ASTM C 1105 (ACI 1998).

ASTM C 586: Test Method for Potential Alkali Reactivity of Carbonate Rocks for Concrete Aggregates (Rock Cylinder Method)

This test method is used to determine the expansive characteristics of carbonate rocks immersed in a NaOH solution. This test should be conducted if the potential for ACR is suspected or if ASTM C 295 indicates that potentially reactive constituents are present.  The test is fairly rapid and is an effective tool for screening aggregate sources (ACI 221.1R-98).  Small cylinders (35 mm long by 9 mm diameter) of the suspect rock are immersed in a 1 N NaOH solution at room temperature.  The change in length of the specimen is measured at 7, 14, 21, and 28 days of immersion, and at 4-week intervals thereafter. If the test is continued beyond 1 year, measurements should then be made at 12-week intervals.  Generally, expansive tendencies are observed during the first month (Farny and Kosmatka 1997).  A 28-day expansion equal to or exceeding 0.10 percent indicates the potential for deleterious expansion in the field.  Results of this test can be used qualitatively to predict expansion of concrete in the field, but quantitative predictions of concrete expansion are not possible. This test should therefore not be used alone, but instead in conjunction with other tests such as ASTM C 295 and C 1105 to predict whether an aggregate will likely be susceptible to ACR.

ASTM C 666: Test Method for Resistance of Concrete to Freezing and Thawing

This test method is used to assess the resistance of concrete specimens to rapidly repeated cycles of freezing and thawing. It is considered by many to be the best available test method for evaluating freeze-thaw resistance of aggregate, but it is not without its critics (Pigeon and Plateau 1995; ACI 1992). In this test method, concrete beams are prepared with the aggregate under evaluation and subjected to rapid freezing and thawing cycles. In Procedure A, the specimens are frozen and thawed in water whereas in Procedure B freezing occurs in air while thawing is done in water.  Procedure A is the preferred method.  Many SHAs have modified this procedure to address their specific needs and observations.

Deterioration is measured through the reduction in the dynamic modulus of elasticity of the concrete, the linear expansion of the specimen, and/or through the weight loss incurred.  According to the PCA (1992), a number of SHAs use an expansion failure criterion of 0.035 percent in 350 freeze-thaw cycles or less to help indicate whether or not an aggregate is susceptible to freeze-thaw deterioration.

Criticism of the test method primarily centers on the fact that it is not representative of actual field conditions.  The concrete is saturated and then subjected to rapid freezing and thawing, which is unlikely to occur in the field.  Thus, although the test is able to rank aggregate from excellent to poor, it cannot be used reliably to predict the field performance of marginal aggregate (ACI 1992).  Because the test is more severe than actual field conditions, aggregates that pass this test are generally going to perform well in the field. But it may reject aggregate that has established good field performance.

As noted in the TRB Circular entitled Durability of Concrete, only Procedure A in the standard should be used; Procedure B should never be used (TRB 1999).

ASTM C 1105: Test Method for Length Change of Concrete Due to Alkali–Carbonate Rock Reaction

If ASTM C 295 and C 586 indicate that an aggregate is potentially alkali–carbonate reactive, ASTM C 1105 should be conducted. It is recognized as the best indicator of potentially deleterious expansion due to carbonate aggregate in concrete (ACI 1998). This test calls for six concrete specimens to be fabricated using the aggregate under evaluation and the job cement, if possible.  The specimens are kept in moist storage and their change in length is measured at 7, 28, and 56 days, and at 3, 6, 9, and 12 months.  In the appendix of the ASTM standard, average expansions equal to or greater than 0.015 percent at 3 months, 0.025 percent at 6 months, and 0.030 percent at 1 year are considered indicative of a potentially deleteriously aggregate source.  The appendix in the standard suggests that if the aggregate is judged potentially reactive, the following measures can be employed to control the effects of the reaction:

• Use the smallest nominal maximum coarse aggregate practical.
• Use a maximum of 20 percent reactive rock in the coarse aggregate, a maximum of 20 percent reactive rock in the fine aggregate, or a total of 15 percent reactive rock in the coarse and fine aggregate combined.
• Use cement having an alkali content below that of cement for which the results indicate that the aggregate is nonreactive.  However, some have noted that this is not very effective in controlling ACR (Farny and Kosmatka 1997).

Prevention of ACR is discussed in detail later in this chapter.

ASTM C 1260: Test Method for Potential Alkali Reactivity of Aggregates (Mortar Bar Method)

This test method is specifically designed to establish the aggregates’ potential for alkali–silica reactivity.  In this test, the aggregate is sized and cast into mortar bars at a water to cementitious material ratio (w/cm) of 0.47.  Specimens undergo 2 days of moist room and water cure, and are then submerged in a 1 N NaOH solution for 14 days.  Expansion is measured at 1, 3, 7, 10, and 14 days.  The total duration of the test is 16 days.

The following 14-day expansion criteria are presented in the appendix of C 1260 to judge potential reactivity:

• Expansion greater than 0.20 percent is indicative of potentially deleterious expansive reactions.
• Expansions between 0.10 and 0.20 percent have been observed for aggregates with variable field performance.  Supplemental information must be developed.
• Expansions less than 0.10 percent are indicative of innocuous behavior in most cases.

Some granitic gneisses and metabasalts have been found to be deleterious in the field even though expansion in ASTM C 1260 has been less than 0.10 percent. Thus, unless good field performance can be demonstrated for these types of aggregate, the aggregate in question should be considered potentially reactive.

Because the specimens are submerged in 1 N NaOH solution, the alkali content of the cement is relatively unimportant.  Also, the test will reveal potential reactivity, but will not predict field performance because mixture alkalinity is not considered. But the rapidity of this test and its ability to detect reactivity in slowly reactive aggregates makes it very attractive. In combination with ASTM C 295 and evaluation of field performance records, most alkali-silica reactivity problems can be anticipated and mitigated (Stark 1994).   If the expansion falls between 0.10 and 0.20, the aggregate should be tested using ASTM C 1293.

It is noted in the standard that specimens undergoing apparently deleterious expansion should undergo further analysis to determine if the expansion is indeed a result of ASR.  ASTM C 856, Practice for Petrographic Examination of Hardened Concrete, is one recommended procedure to identify the alkali-silica gel in the test specimen.

In closing, ASTM C 1260 is considered to be a quick and reliable test for characterizing the potential for reactivity of slowly as well as rapidly reactive rock types (ACI 221.1R-98).  But it is a severe test that may reject acceptable aggregates, possibly necessitating long-term testing using ASTM C 1293.

ASTM C 1293: Standard Test Method for Concrete Aggregates by Determination of Length Change of Concrete Due to Alkali–Silica Reaction

In this test procedure, concrete prisms made of the aggregate under evaluation are constructed with a Type I high alkali cement plus sufficient reagent quality NaOH to simulate a 1.25 percent Na₂O equivalent cement. The concrete proportions are strictly defined with a cement content of 420 ± 10 kg/m³ and a water-to-cement ratio (w/c) of 0.42 to 0.45 being specified.  The specimens are sealed in containers and are suspended above water at 38^oC. Length measurements are made at 7, 28, and 56 days and at 3, 6, 9, and 12 months and every 6 months thereafter for as long as desired.  These are compared to the 1-day length and used to determine if expansion due to ASR is occurring.  Aggregates in specimens having expansion in excess of 0.04 percent after 1 year are considered potentially reactive. 

As with ASTM C 1260, it is recommended that if deleterious expansion is observed, ASTM C 856 should be used to confirm that the cause of expansion is ASR. ASTM C 1293 can be used to assess both fine and coarse aggregate.  Due to its long duration, it is also more effective than ASTM C 1260 for assessing slowly reactive aggregates.

AASHTO T 103: Standard Method of Test for Soundness of Aggregates by Freezing and Thawing

This test method is conducted on unconfined coarse aggregate to evaluate freeze-thaw susceptibility.  It is used to initially screen aggregate. Aggregates that fail this test are not permitted for use in concrete.  Those that pass this test are subjected to further testing.  In figure III-2, this would be ASTM C 666 (Method A).

Each SHA would need to establish failure criteria based on experience. For example, in the Kansas Department of Transportation version of this test, a coarse aggregate specimen is prepared containing material passing the 25-mm sieve and retained on the 2.36-mm sieve.  The gradation is carefully selected and recorded.  The specimen is then saturated and cycled between freezing and thawing for 25 cycles.  After 25 cycles, the cumulative percentage of material retained on each of the four sieves (19, 9.5, 4.75, and 2.36 mm) is determined.  A freeze-thaw loss ratio is calculated by dividing the sum of cumulative percentages of aggregate retained on each sieve after freeze-thaw cycling to that for the original aggregate gradation.  If the calculated freeze-thaw loss ratio is less than 0.85, the aggregate is rejected.

of Aggregate Selection Process

This section provided a methodology based on standards and test methods that can be used to evaluate the potential durability of aggregate for use in concrete pavements.  It is assumed that the aggregates under evaluation pass the requirements of ASTM C 33.  In evaluating an aggregate source for durability, it is important that ASTM C 295 first be used to characterize the physical and chemical properties of the aggregate. Supplemental testing is then conducted to further define the physical and chemical stability of aggregate. Demonstrated field performance is a crucial element of any study of aggregate durability, especially in light of the limitations of available test methods.  Agencies are therefore exhorted to maintain detailed records of all aggregate investigations and attempt to correlate these investigations with field performance.

Cementitious and Pozzolanic Materials

The selection of cementitious and pozzolanic material is an extremely important element in designing durable concrete pavements. In many applications, the use of a standard ASTM C 150 Type I cement will provide satisfactory results. But the designer should not take durability for granted, and therefore must carefully consider the properties of the cement and additional cementitious/pozzolanic materials in the context of the long-term physical and chemical stability of the concrete pavement. This section describes important factors to consider when selecting cementitious/pozzolanic materials for use in a pavement project. Cements identified through ASTM C 150 and C 595M are described, as well as fly ash, slag, and silica fume additives with the emphasis on enhancing concrete durability.

Cement

The standard specification for portland cements used in the United States is presented in ASTM C 150. The following cement classifications are presented in the ASTM standard:

• Type I–For use when the special properties specified for any other type are not required.
• Type IA–Air-entraining cement for the same uses as Type I, where air-entrainment is desired.
• Type II–For general use, particularly when moderate sulfate resistance or moderate heat of hydration is desired.
• Type IIA–Air-entraining cement for the same uses as Type II, where air-entrainment is desired.
• Type III–For use when high early strength is desired.

• Type IIIA–Air-entraining cement for the same uses as Type III, where air-entrainment is desired.

• Type IV–For use when low heat of hydration is desired.
• Type V–For use when sulfate resistance is desired.

The most common cement employed in pavement construction in the United States is Type I. The required chemical properties of Type III cements are similar to Type I, but Type III cements are ground finer to promote the development of higher early strength.  Type III cements are gaining more widespread use, particularly in applications where high early strength is needed such as for rapid-setting repairs and “fast track” construction.  In areas where external sulfate attack is a problem, Type V cements are used.  Air-entrained cement, designated with an “A,” has small quantities of air-entraining material interground with the clinker during manufacture.

In addition to these cements, ASTM C 595M provides standard specifications for blended hydraulic cements. These specifications recognize the following five classes of blended cements:

• Type IS–Portland blast-furnace slag cement
• Type IP and Type P–Portland-pozzolan cement
• Type I (PM)–Pozzolan modified portland cement
• Type S–Slag cement
• Type I (SM)–Slag modified portland cement

These cements are formed by intimately blending portland cement with fine materials such as GGBFS, fly ash or other pozzolans, hydrated lime, and pre-blended cement combinations of these materials (Kosmatka et al. 2002).  In the past, these have not been commonly used on paving projects in the United States, but they may see increased use to produce durable concrete pavements, particularly Type IS, Type IP, and Type I (PM). Each of these three can be further designated as air-entraining (A), moderate sulfate resistant (MS), and/or moderate heat of hydration (MH).   For example, an air-entraining, moderate sulfate resistant pozzolan modified portland cement would be designated as Type I (PM)-A(MS).  There are many references available that provide detailed descriptions of the physical and chemical characteristics of cements. One such reference is Design and Control of Concrete Mixtures by the PCA (Kosmatka et al. 2002).

It is known that variations in the physical and chemical characteristics of cement can have an impact on the initiation and progression of MRD in concrete pavements. A recently completed study by the PCA (1996) confirmed that slight changes in cement properties have occurred since the 1950’s. Cement produced in the 1990’s achieves strength more quickly in the first 7 days, but has reduced long-term strength gain than its 1950’s counterpart.  This is partially a result of the fineness of modern cements, which in all cases are more finely and uniformly ground than cements from the 1950’s. But chemistry also plays a role. Modern cements have greater amounts of C₃S (3 to 10 percent more) and less C₂S (5 to 14 percent less) than cements from the 1950s.  This contributes to faster early strength gain at the expense of long-term strength development. Also, sulfate content for modern cements is 0.5 to 1.2 percent higher and the total alkali content has increased by an average of 0.08 percent (Na equivalent) for Type I cement. The increase in sulfate can be partially accounted for by the increased demand for gypsum necessitated by finer grinding of the cement, but changes in processing and raw materials and the use of waste fuels are also contributing factors (Gress 1997).  Although the results of the PCA report do not show dramatic changes in cement properties, some have voiced concern that the combination of small changes in cement characteristics may negatively affect the durability of concrete.

As a result, some researchers are suggesting that the movement toward quicker setting cement and faster construction has been detrimental to long-term concrete durability. Mehta (1997), for example, argues that for concrete to remain durable, it must remain relatively impermeable. He proposes that durable concrete should be slow-hardening with high creep and low elastic modulus at early ages to resist drying shrinkage and thermal cracking.  Although he acknowledges that this can be accomplished through the use of coarser ground cement with low C₃S content, he doesn’t directly advocate a return to the cement of the past.  He instead suggests that there is an overemphasis on speed of construction that has resulted in the unnecessary use of high early-strength mixtures. To address this, instead of using Type I or Type III portland cement, blended cements containing slag and/or fly ash could be used.  The lower heat of hydration and denser, less soluble microstructure will produce less permeable, more durable concrete.  This would necessitate a major change in the way concrete quality is judged, shifting strength requirements to 56 days or 90 days instead of the 28 days commonly used today.  The use of mixtures having slower strength gain of this type may be practical for many paving projects, although they are not practical in situations where early opening is required.  It must be realized that speed of construction and long-term concrete durability both need to be considered.  Thus, when selecting cementitious materials, the designer should not focus exclusively on 28-day strength gain as the measure of concrete quality. And unless the constraints of the project demand high early strength (i.e., high user costs dictate early opening), the use of blended cements or pozzolanic replacement should be considered.

Cement properties related to the development of specific MRD types are discussed later in this section of this guideline, which also provides detailed information regarding mix design and construction practices that can be exercised to control specific MRD in concrete pavements.

Fly Ash

Fly ash is an industrial byproduct produced from the burning of coal. It is primarily silica glass containing silica, alumina, iron, and calcium, with magnesium, sulfur, sodium, potassium, and carbon as minor constituents (Kosmatka et al. 2002). Fly ash is classified according to ASTM C 618 as either Class C or Class F, the primary difference being the minimum percentage of silicon dioxide, aluminum oxide, and iron oxide present (combined total of 50 and 70 percent, respectively).  The burning of bituminous or anthracite coals tends to produce low calcium fly ash (containing less than 10 percent analytic CaO) that has historically been classified as Class F.  Alternatively, Class C fly ash is typically high in calcium (containing 15 to 35 percent CaO), being the product of the combustion of subbituminous or lignite coals (Mehta and Monteiro 1993; Dewey et al. 1996).

Mehta and Monteiro (1993) provide an excellent description of the differences between high and low calcium fly ash. High calcium fly ash (typically ASTM C 618 Class C) is both cementitious and pozzolanic.  It consists mostly of silicate glass containing calcium, magnesium, aluminum, and alkalies, with small amounts of crystalline material present in the form of quartz and C₃A.  Free lime and periclase may also be present, as may C and C₄A₃ if high sulfur coals were used.  The particle size distribution has 10 to 15 percent larger than 45mm with a Blaine fineness of 300 to 400 m²/kg.  Most particles are solid spheres having diameters less than 20 mm.  The particle surfaces are generally smooth but are not as clean as that found on low calcium fly ash, possibly due to the deposition of alkali-sulfates on the surface (Mehta and Monteiro 1993).

Low calcium fly ash (typically classified as ASTM C 618 Class F) is mostly silicate glass containing aluminum, iron, and alkalis (Mehta and Monteiro 1993).  Small quantities of crystalline material exist in the form of quartz, mullite, sillimanite, hematite, and magnetite.  The particle size distribution is typically slightly coarser than that of the high calcium fly ashes, with 15 to 30 percent larger than 45 mm, a Blaine fineness of 200 to 300 m²/kg, and having an average diameter of 20 mm. Most of the particles are solid, but cenospheres (completely hollow, empty spheres) and plerospheres (hollow spheres packed with numerous small spheres) may be present (Mehta and Monteiro 1993).

Fly ash can be incorporated into concrete either through blending or intergrinding with portland cement and classified under ASTM C 595M as Type IP, Type P, or Type I(PM) or introduced during mixing either as an addition to the portland cement or as a partial replacement.

It is widely recognized that concrete can benefit greatly from the inclusion of fly ash, either as an addition to or replacement of portland cement. Because of its smooth spherical shape and broader size distribution than ASTM Type I portland cement, fly ash acts as a “microaggregate,” packing into spaces that would normally be left empty. This improves workability and reduces the water demand of the fresh concrete, and generally results in less bleeding and segregation and improved finishability (Kosmatka et al. 2002; Mehta and Monteiro1993).

The use of fly ash can also produce a denser, less permeable microstructure that is less susceptible to chemical attack. In addition to the physical contribution of improved particle packing, the hydration of the pozzolan will further reduce permeability by filling remaining pore space with hydration products. The pozzolanic reaction that occurs with low calcium fly ash can significantly reduce ASR by converting calcium hydroxide (CH) to calcium silicate hydrate (CSH) and chemically tying up the alkalis in the concrete (Kosmatka et al. 2002).  Similarly, sulfate attack can also be reduced through the addition of low calcium fly ash.  Once again, the pozzolanic reaction will combine the silica in the fly ash with CH and alkalis, reducing the potential for deleterious gypsum formation (ACI 1992).

However, the benefit offered by a specific fly ash is highly dependent on its individual characteristics.  It is widely acknowledged that the current ASTM C 618 classification of coal fly ash is inadequate to describe whether the addition of a fly ash will be beneficial to the performance of concrete.  Not only is specific knowledge regarding chemical and physical characteristics important, but also the fly ash’s mineralogical structure must be identified to understand its reactivity (Dewey et al. 1996). Additionally, the amount of CaO present has a significant effect on the ability of the fly ash to mitigate the initiation and progression of certain MRD such as ASR (Gress 1997). And, in some instances, sulfates and alkalis present in fly ash may actually be detrimental to concrete durability.

It is therefore necessary to carefully select fly ash for a specific concrete application and demonstrate through testing that it will have a beneficial impact on performance.  Many studies have found that different volumes of similarly classified fly ashes are required to gain the same desired enhancement, and that an optimal volume for one particular fly ash may actually be detrimental to concrete durability for another fly ash.

Class C fly ashes have been particularly problematic as far as concrete durability is concerned.  Because of their higher CaO content, they do not react with CH and alkalis to form CSH to the same degree as lower calcium Class F fly ashes. Thus, they are widely recognized as not being as effective in controlling ASR or sulfate attack, and in some cases, may be detrimental (ACI 1992; Kosmatka et al. 2002; Farny and Kosmatka 1997). Also, Class C fly ashes contain more reactive crystalline compounds such as C₃A, C , and C₄A₃ and more alkalis and sulfates than do low calcium Class F fly ashes (Mehta and Monteiro 1993).  Finally, both Class C and F fly ashes also commonly contain small percentages of unburned carbon. If present in quantities greater than 5 percent, this unburned carbon will result in increased water demand to maintain a given consistency and increased quantity of air-entraining admixture to obtain a desired air content (Mehta and Monteiro 1993).

Recommendations regarding the addition of fly ash to concrete to mitigate specific MRD in concrete pavements are reserved for later sections of this guideline.  In general, it is evident that fly ash can be a powerful and cost-effective ally in producing durable concrete through physical and chemical modification to its microstructure.  The addition of low calcium fly ash will reduce concrete permeability while converting soluble CH into less soluble, more stable CSH. It is important to demonstrate the ability of the fly ash to mitigate ASR and increase sulfate resistance through standard tests such as ASTM C 227 and C 1293 for ASR and ASTM C 1012 for sulfate resistance.

The current state of knowledge suggests that the use of Class C fly ash should be discouraged when attempting to mitigate potential MRD problems such as ASR or sulfate attack.  Also, the fly ash’s contribution of sulfates, aluminum, and alkalis to the concrete must be considered.  In all cases, concrete mixtures containing fly ash should be tested to ensure that adequate durability exists. In areas subjected to deicer applications, special care must be taken to ensure that a high cementitious material content and low w/cm are used. Sufficient curing time to account for the slower hydration rate of concrete containing fly ash is required prior to deicer application to prevent scaling.

Other Cementitious/Pozzolanic Materials

Two other industrial byproducts that may be added to concrete as supplemental cementitious/pozzolanic materials are GGBFS and silica fume. Each may be beneficial in preventing MRD in some situations.

Ground Granulated Blast Furnace Slag

GGBFS is a byproduct of the production of cast iron, in which the molten slag (at a temperature of 1500 ^oC) is rapidly chilled by quenching in water.  This forms a glassy, sand-like material that is then finely ground to less than 45 mm, having a surface area of 400 to 600 m²/kg Blaine (Kosmatka et al. 2002). GGBFS is nonmetallic, consisting mostly of silicate glass containing calcium, magnesium, aluminum, and silicate with potentially small quantities of crystalline compounds of melilite present (Mehta and Monteiro 1993). This rough textured material is cementitious in nature, meaning that it possesses hydraulic cementing properties. When combined with portland cement, the NaOH or CaOH activates the GGBFS, which hydrates and sets in a manner similar to portland cement (Kosmatka et al. 2002). Specifications for GGBFS for use in concrete are provided in ASTM C 989.

GGBFS may be blended or interground with the cement and classified according to ASTM C 595M as Type IS, Type S, or Type I(SM). Alternatively, it may be added during batching as supplemental cementing materials or as a partial replacement for portland cement.

In producing durable concrete, the addition of GGBFS has had positive effects. Like fly ash, it reduces the permeability of concrete and therefore should have a beneficial effect on durability by limiting the penetration of harmful agents, decreasing the solubility of the paste, and preventing the rapid movement of pore solution within the concrete.  GGBFS has been used effectively to control both ASR and external sulfate attack.

Silica Fume

Silica fume is the byproduct of induction arc furnaces used in the silicon metal and ferrosilicon alloy industries where quartz is reduced to silicon at temperatures up to 2000 ^oC (Mehta and Monteiro 1993).  This produces SiO vapors that oxidize and condense to minute spherical, noncrystalline silica.  The particle size is roughly two orders of magnitude smaller than fly ash, having an average diameter of 0.1 mm and a surface area of 20,000 to 25,000 m²/kg, which is twice that of tobacco smoke (Kosmatka et al. 2002).  It is highly pozzolanic, but its incredibly high surface area makes it difficult to handle, significantly increasing the water demand of the concrete unless water-reducing admixtures are used.   Standard specifications for silica fume are provided in ASTM C 1240.

Due to its fine size, silica fume is able to pack very tightly into void spaces between cement and aggregate particles.  It can thus be used to significantly reduce concrete permeability and therefore block the ingress of chlorides.  For this reason, silica fume concrete has seen extensive use on bridge decks. Its highly pozzolanic nature also converts alkalis and CH into CSH, making the concrete less susceptible to chemical attack. Thus, it can be used to mitigate ASR and sulfate attack (ACI 1992; Kosmatka et al. 2002), although its effectiveness must be established through testing.

Silica fume has not been widely used in pavements because of problems encountered in handling and special concerns related to curing.  It is also costly and is therefore unlikely to find widespread use in pavement applications, except under special circumstances.

Admixtures

A number of noncementitious admixtures can be added during proportioning or mixing to enhance the properties of freshly mixed and/or hardened concrete.  Admixtures used in pavement construction in North America include air entrainers, water reducers, accelerators, retarders, corrosion inhibitors, noncementitious minerals, and ASR inhibitors.  An excellent description of these various admixtures can be found in a number of sources (Kosmatka et al. 2002; Mehta and Monteiro 1993; Mindess and Young 1981). The following is a brief discussion specifically focused on the impact various admixtures have on concrete pavement durability.

Air-Entraining Admixtures

Air-entraining admixtures are specified and tested under ASTM C 260 and C 233, respectively.  Air-entraining admixtures are added just prior to or during concrete mixing and can be used in place of or in conjunction with air-entraining cement. The entrained air voids protect the hardened concrete against freeze-thaw damage and deicer scaling.  They also improve the workability of the fresh concrete, significantly reducing segregation and bleeding.

Air-entraining admixtures function by stabilizing bubbles formed in the fresh paste during mixing.  This is accomplished through surface-active agents that concentrate at the interface between air and water, reducing the surface tension so the bubbles are stable.  These surface-active agents are composed of molecules that are hydrophilic (water loving) at one end and hydrophobic (water fearing) at the other. Thus they align at the interface with their hydrophilic ends in the water and the hydrophobic ends in the air (Mindess and Young 1981). Typical compounds used as air entrainers include salts of wood resins (Vinsol resins), salts of sulfonated lignin, salts of petroleum acids, alkylbenzene sulfonates, and salts of sulfonated hydrocarbons among others.

The dosage rate for air-entraining admixtures is usually very small, on the order of 0.005 to 0.05 percent active ingredients by weight of cement. Thus they are normally diluted to assist in accurate batching (Mindess and Young 1981). The amount of entrained air required to protect concrete depends on the exposure level and the nominal maximum aggregate size. Recommended air contents for freeze-thaw resistant concrete from ACI (1992) are reproduced in table III-9. Concrete pavements subject to deicer application are considered to be in severe exposure conditions.

The air content of fresh concrete can be determined using ASTM C 173 or C 231.  It is noted that air content alone does not ensure the adequacy of the air-void system, but that a good correlation between air content and freeze-thaw distress exists for air entrained concrete. The complete air-void system in hardened concrete can be assessed microscopically using procedures described in ASTM C 457.

Water-Reducing Admixtures

Water-reducing admixtures are added to reduce the quantity of mixing water required to produce concrete of a given consistency. This allows for a reduction in the w/cm while maintaining a desired slump, and thus has the beneficial effect of increasing strength and reducing permeability. A reduction in water content by 5 to 10 percent is obtainable through the use of

Table III-9. Recommended air contents for freeze-thaw distress-resistant concrete (ACI 1992).

¹   A reasonable tolerance for air content in field construction is ± 1.5 percent.

²   Outdoor exposure in a cold climate where the concrete will be only occasionally exposed to moisture prior to freezing, and where no deicing salts will be used.  Examples are certain exterior walls, beams, girders, and slabs not in direct contact with soil.

³    Outdoor exposure in a cold climate where the concrete may be in almost continuous contact with moisture prior to freezing, or where deicing salts are used.  Examples are pavements, bridge decks, sidewalks, and water tanks.

⁴   These air contents apply to the whole as for the preceding aggregate sizes. When testing these concretes, however, aggregate larger than 37.5 mm (1-1/2 in) is removed by handpicking or sieving and the air content is determined on the minus 37.5 mm (1-1/2 in) fraction of the mixture.  (The field tolerance applies to this value.)  From this the air content of the whole mixture is computed.

conventional water reducers that are specified under ASTM C 494 Type A.  This class of water reducer typically will retard set, and accelerators are often added to offset this effect.  If it is desired that the water reducer also acts as a retarder or accelerator, it can be specified under ASTM C 494 Type D and Type E, respectively. ASTM C 494 water reducers are typically composed of lignosulfates, hydroxylated carboxylic acids, or carbohydrates.

The effect of water reducers on the fresh concrete properties varies with the chemical composition of the admixture, the concrete temperature, cement composition and fineness, cement content, and the presence of other admixtures (Kosmatka et al. 2002). The effect on the air-void structure is unclear, with some sources reporting an improvement (Kosmatka et al. 2002) while others reporting possible adverse affect (Pigeon and Plateau 1995).  Thus, the fresh and hardened concrete properties of mixtures containing water reducers should be thoroughly evaluated during design to determine the extent of detrimental interactions that may occur.

High range water reducers (ASTM C 494 Type F and G) can reduce water content by 12 to 30 percent, but their use is primarily restricted to applications where “flowing” concrete is desired ((Kosmatka et al. 2002). Also, the entrained air voids produced in concrete made with superplasticizers are often large, increasing the spacing factor, and on occasion instability in the air-void system may occur (Pigeon and Plateau 1995).

Accelerators

Accelerators, specified under ASTM C 494 Type C, are used to increase the rate of concrete strength development. Their use in concrete paving is generally restricted to fast track projects or cold weather construction where the rapid liberation of heat helps prevent freezing during early stages of hydration. Accelerating strength development can have negative effects on concrete durability, and thus should only be done when critical time constraints exist.

The most commonly used accelerator is calcium chloride as specified under ASTM D 98. This admixture is a very effective accelerator and is relatively inexpensive.  But its use can contribute to MRD by increasing the corrosion of reinforcing steel, drying shrinkage stresses, scaling potential, potential for alkali–aggregate reactions, and susceptibility to sulfate attack.  Therefore, the use of a calcium chloride as an accelerator in pavement applications is discouraged.

Other nonchloride, noncorrosive accelerators are available that contain compounds such as triethanolamine, sodium thiocyanate, calcium formate, calcium nitrite, and calcium nitrate.  Their impact on concrete durability is not well documented, but it is very unlikely that durability will be improved (except for corrosion resistance for those that inhibit corrosion of embedded steel). It has been generally observed that concrete microstructure produced in rapidly setting concrete is coarser and composed of more soluble hydration products that are prone to chemical attack. It is therefore emphasized that accelerators should only be used when absolutely necessary, and that the designer/engineer understand that their use will likely have a negative impact on the long-term durability of the concrete.

Retarders

Retarders, specified under ASTM C 494 Type B, are used to delay the setting time of concrete.  Common compounds used as retarders are lignin, borax, sugars, tartic acids, and salts. Most retarders also act as water reducers, and may also have some air entraining properties (Kosmatka et al. 2002). As with all admixtures, it is critical that the effect of the retarder on the fresh and hardened concrete properties be investigated during the mix design stage.

Corrosion Inhibitors

Corrosion inhibitors are admixtures specifically added to concrete to inhibit the corrosion of steel.  Their use in pavement construction is quite limited, as alternative methods of protecting steel (e.g., the use of high quality, low permeability concrete, adequate cover, and epoxy coatings) have been used with relatively good success. But in pavement structures containing significant amounts of steel that will be subjected to frequent chloride applications, the use of a corrosion inhibitor might be considered.

The most common corrosion inhibitor is calcium nitrite, which also acts as an accelerator.  A number of studies have verified its ability to prevent corrosion by reinforcing and stabilizing the passive film that protects steel (Kosmatka et al. 2002; ACI 1992). The dosage rate varies, but usually falls in the range of 26 to 110 ml/kg cement. It is unclear what other effects the addition of calcium nitrite has on concrete durability, but it is likely to result in an increase in CH formation in the paste and higher alkali levels, which in turn may adversely affect concrete durability.  Other corrosion inhibitors include compounds such as sodium nitrite, sodium benxoate, certain phospates, flousilicates, or fluoaluminates. Corrosion inhibitors are relatively expensive, increasing the cost of a concrete structure by approximately 8 percent (Mehta 1997).

Noncementitious Mineral Admixtures

A number of noncementitious mineral admixtures could be added to concrete.  Pulverized limestone is one that may improve durability of pavements, having some benefit in reducing ASR. It is noted that it is not as effective as pozzolans.  Typically noncementitous mineral admixtures are added to improve workability caused by lack of fines. As such, the addition of the mineral admixture for this purpose is considered a mix design issue.

ASR Inhibitors

Lithium compounds are the best known ASR inhibitors.  Their effectiveness in preventing deleterious ASR depends on the lithium compound used, addition rate, aggregate reactivity and cement alkalinity (TRB 1999). The most effective and easiest to handle lithium compound is lithium nitrate (LiNO₃).  Other lithium compounds that have been investigated include lithium hydroxide (LiOH), lithium hydroxide monhydrate (LiNO₃·H₂O) and lithium carbonate (Li₂CO₃) (AASHTO 2000). When lithium hydroxide is added to concrete, it forms minimally expansive lithium-bearing ASR gel, which is generally not damaging to the concrete (Farny and Kosmatka 1997).  It is noted that ASTM C 1260 cannot be used to assess the effectiveness of lithium compounds (Farny and Kosmatka 1997). The recommended use of lithium compounds for mitigation of ASR is presented in the AASHTO Guide Specifications (AASHTO 2000).

Water

Almost any potable water that has no pronounced taste or odor can be used to make concrete.  ASTM C 94 provides guidance regarding development of compressive strength and time of set as reproduced below in table III-10. It is noted that the criteria have no provision regarding long-term durability of the concrete.  It is therefore recommended that if the water is of questionable quality, the chemical requirements provided in ASTM C 94 for wash water be applied. These are reproduced in table III-11. From a concrete durability perspective, alkali, chloride, and sulfate contents must be considered when selecting mixing water.

Considerations in Proportioning and Mix Design for Prevention of MRD in Concrete Pavements

To this point, the guideline has considered the properties of the individual constituent materials of a concrete mixture. Once the materials are selected, they must be economically combined to form concrete that is workable and easy to consolidate, develops adequate strength, and has long-term durability.  The PCA (Kosmatka et al. 2002) and the ACI (1991) both present acceptable methods for proportioning concrete mixtures. The discussion in the following section focuses exclusively on elements of the proportioning and mix design process that have a direct bearing on enhancing concrete durability.

The key to durable concrete is the use of high quality materials arranged in a dense, relatively impermeable matrix.  The selected aggregates should be strong and not be reactive or susceptible to freeze-thaw or moisture damage within the environment in which they serve. If this condition is met, the hydrated cement paste is the weak link in the matrix, and thus the goal of proportioning aggregates is to ensure that they occupy as high a percentage of the concrete volume as possible.  The hardened paste microstructure should be free of microcracking, possess

Table III-10. ASTM C 94: Acceptance criteria for questionable water supplies.

¹  Comparisons should be based on fixed proportions and the same volume of test water compared to control mix using city water or distilled water.

Table III-11. ASTM C 94: Optional chemical limits for wash water.

¹   Wash water reused as mixing water in concrete can exceed the listed concentrations of chloride and sulfate if it can be shown that the concentrations calculated in the total mixing water, including mixing water on the aggregates and other sources, does not exceed the stated limits.

²   Other test methods that have been demonstrated to yield comparable results can be used.

³   For conditions allowing the use of CaCl₂ accelerator as an admixture, the chloride limitation may be waived by the purchaser.

a fine pore structure with an adequate air-void system to protect it from freeze-thaw damage, and be composed of stable and relatively insoluble hydration products.  Steel must be protected with adequate cover, and the concrete must be properly cured and free of drying shrinkage cracking.

Specific requirements for the selection of concrete mixture characteristics, aggregate grading, and tests for fresh and hardened concrete are discussed below.

Selection of Concrete Mixture Characteristics

The first step in the proportioning process is to select mixture characteristics consistent with the production of durable pavement concrete. The w/cm is probably the most important factor in this regard. It is recommended that the w/cm not exceed 0.45 for pavements that will experience freezing and thawing in a moist environment and will be exposed to deicer applications.  In the same environment, reinforced concrete pavements should be constructed with a maximum w/cm of 0.40 unless 12.5 mm of additional concrete cover is used (ACI 1991). If external sources of sulfate are present, the recommendations presented in table III-12 should be followed (ACI 1992). When the environment is such that durability is not a controlling factor, the w/cm can be selected based on strength requirements (TRB 1999), but it is generally not recommended that a w/cm exceeding 0.50 be used in pavement concrete.

Table III-12. Recommendations for normal weight concrete subject to sulfate attack (ACI 1992).

1. Sulfate expressed as SO₄ is related to sulfate expressed as SO₃ as in reports of chemical analysis of cement as SO₃ x 1.2 = SO₄.
2. When chlorides or other depassivating agents are present in addition to sulfate, a lower water-cement ratio may be necessary to reduce corrosion potential of embedded items.
3. Or a blend of Type I cement and a ground granulated blast furnace slag or a pozzolan that has been determined by tests to give equivalent sulfate resistance.
4. Or a blend of Type II cement and a ground granulated blast furnace slag or a pozzolan that has been determined by tests to give equivalent sulfate resistance.
5. Use a pozzolan or slag that has been determined by tests to improve sulfate resistance when used in concrete containing Type V cement.

In addition to the w/cm, it is sometimes recommended that a minimum cement content of 335 kg/m³ be used in environments subjected to severe freeze-thaw cycling, deicer applications, or severe sulfate exposure (Kosmatka et al. 2002). The TRB Circular entitled Durability of Concrete, indicates that the rationale for such requirements may be invalid, advocating the use of performance-based specifications instead (TRB 1999).

Commonly, a 28-day compressive strength or modulus of rupture is specified for paving concrete as a general measure of concrete quality. As addressed previously, strength at 56 days or 90 days might be considered for the production of durable pavement concrete, as it better allows for the use of fly ash and GGBFS as an addition or replacement for portland cement.

Slump requirements for slip-formed paving concrete typically fall in the range of 25 to 75 mm.  Concrete consistency should be such that the mix is not difficult to consolidate and is not prone to segregation under the action of internal vibration. It also must be stiff enough that sloughing of the paving edges is not a problem.   Table III-9 should be used to select an appropriate entrained-air content based on the exposure condition and the nominal maximum aggregate size.

As discussed in the Transportation Research Circular (TRB 1999), to ensure that an economical, high quality concrete mixture is obtained, the amount of cementitious material used should be minimized while maintaining the required w/cm.  This can only be accomplished through a reduction in water content. Steps that can be used to achieve this outcome include the use of (TRB 1999):

• The stiffest practical mixture.
• The largest practical nominal aggregate size.
• The optimum ratio of fine-to-coarse aggregate.
• A uniform distribution of aggregate to minimize paste content.
  

Aggregate Grading

The aggregate characteristics that have the largest impact on proportioning are the particle size distribution (grading) and the nature of the particles (shape, angularity, porosity, and surface texture) (Kosmatka et al. 2002).  These two parameters affect both the handling of fresh concrete and the properties of hardened concrete. 

Traditional grading methods presented in standard mix design procedures are based on the blending of a coarse aggregate and a fine aggregate grading as presented in ASTM C 33.  Both the ACI and PCA procedures recommend that the largest practical aggregate size be used for job conditions.  This will maximize economy as the increased aggregate size will minimize void space and thus minimize paste requirements. Larger size aggregates also have the benefit of increasing aggregate interlock across cracks and joints, thus improving pavement structural performance, while reducing drying shrinkage. In each design procedure, the maximum aggregate size for the coarse aggregate is selected based on the slab thickness and the spacing between reinforcing bars. For pavements, both ACI and PCA recommend that the maximum aggregate size should be less than one-third the slab thickness and three-quarters the free space between reinforcing bars or reinforcing bars and formwork. For example, a 200-mm-thick plain jointed concrete pavement (JCP) could, under these guidelines, use a maximum aggregate size up to 66 mm. 

The maximum aggregate size typically used in modern pavement construction is significantly smaller than the maximum permissible under the standard mix design practices. In addition, the maximum size of the aggregate used in paving concrete has decreased in recent years, particularly in the northern midwestern States, where aggregate freeze-thaw deterioration has been a problem.  For example, the Michigan Department of Transportation (MDOT) commonly uses a 19-mm maximum aggregate size in pavement concrete, whereas in the mid-1970’s it was not unusual to use 37.5- to 50-mm maximum coarse aggregate size.  The movement toward smaller maximum coarse aggregate size was in response to poor freeze-thaw durability of some aggregate sources. But over time, the practice of using smaller aggregate has become standard practice for all pavement concrete regardless of the aggregate’s durability.  MDOT and others are currently evaluating the use of larger aggregate mixtures to improve aggregate interlock and minimize drying shrinkage. 

Using a larger maximum aggregate size is one method to increase the proportion of aggregate in the mixture. A second method is to achieve a more continuous aggregate grading than is commonly used in traditional mix design methods.  Strategic Highway Research Program (SHRP) research recommended such a modification to ASTM C 33 because there is considerable variation within the packing densities of aggregate specified according to the standard (Roy et al. 1993b). These variations could lead to noticeably different packing densities for aggregate combinations having the same maximum aggregate size. Also, it was noticed that some gradations that were acceptable under ASTM C 33 produce mixes with little tolerance for aggregate proportioning during construction.

The U.S. Air Force believes that aggregate gradation has been a major factor in joint spalling, with poorly proportioned mixtures being more difficult to construct and exhibiting early distress manifestations (Muszynski and LaFrenz 1996; Muszynski et al. 1997).  Consequently, they have changed their specification for grading aggregates in concrete mixtures, considering both the coarseness and workability of the gradation. Aggregate is no longer just considered to be either fine or coarse, but instead is viewed as a single combined aggregate blend.  This approach is echoed in the TRB Circular on concrete durability, which states that a lack of mid-sized aggregate (around 9.5 mm size) results in concrete with high shrinkage, high water demand, and poor workability (TRB 1999). Blending of multiple aggregate sources can be used to address this problem (Shilstone 1990).

Mehta (1997) states that it is desirable from the perspective of durability to maximize the aggregate content and minimize the paste content to promote the watertightness of the concrete. It is believed that the use of the largest maximum aggregate size (considering freeze-thaw durability of the aggregate) with a continuous grading may provide an opportunity to meet these goals.

Tests for Hardened Concrete Relating to Durability

A number of standard tests are available that can be used to establish the susceptibility of concrete to certain MRDs.  Unfortunately, many of these standards are designed to test only a single mix constituent, such as the cement or the aggregate.  For example, ASTM C 1293 Test Method for Concrete Aggregates by Determination of Length Change of Concrete Due to Alkali-Silica Reaction is designed to only test the aggregate under consideration for ASR susceptibility, and not how the cementitious materials or other admixtures affect the outcome. Likewise, a test such as ASTM C 1012 Test Method for Length Change of Hydraulic-Cement Mortars Exposed to a Sulfate Solution will test the sulfate resistance of the cementitious material, but not in combination with the other mix components. Additionally, many of these tests do not adequately model expected field exposure conditions, and thus may not reflect actual field performance. Finally, many of the tests used to evaluate MRD are long term and are thus not ideally suited to mix design.

Even with these limitations, a few standard test methods may be useful in assessing the potential durability of pavement concrete.  It is assumed that the constituent materials have been screened as described in this guideline. This is particularly important for aggregate screening, as it is used to prevent such MRD types as aggregate freeze-thaw deterioration, ASR, and ACR. It is also assumed that common tests on fresh concrete, such as slump, air content, and strength testing of hardened concrete, are conducted.  When appropriate, the actual job mix formula must be used with the mix constituents including aggregate, cement, admixtures, and water.  The duration of the mixing cycle in the laboratory should be consistent with that anticipated during construction.

AASHTO PTP 34-99: Proposed Standard Method of Test for Restrained Drying Shrinkage

There is a standard test method currently under review for the determination of drying shrinkage. The test, specified in AASHTO TP 34-99, helps distinguish concrete that will undergo excessive drying shrinkage, which in turn can result in cracking and loss of watertightness.  In this ring test, specimens are molded in two layers surrounding a steel ring. After 24 h the outer mold is removed, exposing the concrete surface.  The top and bottom faces of the concrete rings are covered with silicone rubber sealer to prevent any moisture loss other than through the circumferential area. A steel ring inside the concrete specimen restrains the concrete specimen as it tries to shrink.  This restraint will develop internal tangential tensile stresses, which will cause the concrete to crack once its tensile strength is exceeded (Kraai 1985).  The time to cracking and the width and length (area) of these cracks represent the concrete’s resistance to drying shrinkage cracking.

ASTM C 457: Practice for Microscopical Determination of Parameters of the Air-Void System in Hardened Concrete

This test is commonly used to assess the air-void distribution in hardened concrete.  Most often used as a diagnostic test, ASTM C 457 could be applied during the mix design process to assess the adequacy of the air-void system.  It can also be conducted during construction on field cores. This type of evaluation will assist in determining if interactions between the various constituents has led to changes in the anticipated air-void system.  Test results could also be used as a baseline to determine if field construction is producing an air-void system consistently different than that produced in the lab. Finally, results of ASTM C 457 could be used to assess the potential of secondary ettringite formation (SEF) by monitoring the filling of voids with ettringite over time, as it provides a baseline measurement of the unfilled air-void system.

ASTM C 666: Test Method for Resistance of Concrete to Rapid Freezing and Thawing

This test has been previously described in this guideline. In the evaluation of aggregate, it is typical to make a standard mix (specified cement type, w/c, and air content) to compare aggregate performance. But it is known that the full interaction of constituent materials can not be easily anticipated, and thus freeze-thaw testing of concrete specimens made with the job mix formula and materials may be useful in identifying potential problems. Unfortunately, this is a long-term test, requiring upwards of 3 months to conduct.  It therefore should be reserved for use when marginal aggregates are being considered.

ASTM C 672: Test Method for Scaling Resistance of Concrete Surfaces Exposed to Deicing Chemicals

ASTM C 672 is the most common test used to investigate the scaling potential of concrete.  In this test, a CaCl₂ solution (other chemical deicers may be used) is ponded on the surface of rectangular concrete specimens that are then subjected to freeze-thaw cycling. The specimens are placed in a freezer (-17.8^oC) for 16 to 18 h and then manually removed to a thawing environment for 6 to 8 h. A surface layer of the water/salt solution is maintained at all times.  A visual inspection is made at 5, 10, 15, 20, 25, and 50 cycles. The concrete is rated on a scale of 0 to 5 (0 is for concrete surfaces showing no sign of scaling and 5 is for a surface that is severely scaled with coarse aggregate visible over the entire surface). The subjectivity inherent in the rating scale is one problem with this test.  As a result, it is becoming common for researchers to measure the mass of scaled material to gain a more objective measure of scaling resistance.

This test is most applicable when the concrete pavement will be subjected to deicer applications and supplementary cementitious materials are being used.  Concrete made with fly ash and GGBFS have exhibited variable deicer scaling resistance, and therefore mixtures containing these materials should be lab tested if field experience does not exist.  Control samples of mixes with known scaling resistance should also be tested, as the results of this test are qualitative.

ASTM C 1202: Electrical Indication of Concrete’s Ability to Resist Chloride Ion Penetration

This test provides excellent correlation to the results obtained by AASHTO T 259, Standard Method of Test for Resistance of Concrete to Chloride Ion Penetration, for a wide variety of concrete types and qualities (Perenchio 1994).  One exception is concrete made with silica fume because the active pozzolanic nature of the material makes it seem more impermeable to chloride ion penetration than it really is.  Another exception is concrete made with certain admixtures that affect the electrical conductivity of the mixture, including calcium chloride or calcium nitrite.  Regardless, the rapidity, ease of use, and reliability make this test very attractive when investigating not only the chloride ion permeability characteristics of concrete, but also to assess permeability in general. Table III-13 can be used to make a general assessment of the chloride ion permeability of the concrete tested using this test method (ASTM C 1202).  It is noted that the assessment is not specific, but instead the chloride permeability is assigned a qualitative rating. Pavement concrete should have a qualitative permeability not greater than low, and in aggressive environments (exposure to chemical deicers, sulfates, etc.), a qualitative rating of very low or less is desirable.

Table III-13. Chloride ion penetrability based on charge passed (AASHTO T 277).

Although this test has been embraced by many SHAs due to its ease of use, it suffers some limitations that make it impractical when evaluating some mixtures.   The three main limitations are: 1) the current passed is related to all ions in the pore solution and not just chloride ions, 2) the measurements are made before a steady-state migration is achieved, and 3) the temperature of the specimen increases due to the applied voltage (Stanish et al. 2000).  The first limitation is most problematic for the assessment of concrete permeability in mixtures containing various admixtures (e.g., accelerators, etc.) that will affect the ion concentration of the pore solution.

ASTM C 642: Test Method for Specific Gravity, Absorption, and Voids in Hardened Concrete

Porosity is a measure of the proportion of the total volume of pore space in concrete irrespective of the interconnectivity of the pores (Neville 1996).  Absorption tests can be used to measure porosity, but the degree of interconnectivity between the pores influences the measured absorptivity.  Thus, although porosity and absorptivity are commonly correlated, there is not a direct relationship. A variety of techniques are used for determining the absorption rate of concrete.  One common test is ASTM C 642, Test Method for Specific Gravity, Absorption, and Voids in Hardened Concrete, which entails drying a concrete specimen at 100 to 110^oC and then immersing it in water at 21^oC for at least 48 h. This type of test is commonly used as a quality control test for precast members. 

Sorptivity Test

As a result of shortcomings in common absorption testing procedures (e.g., ASTM C 642), there has been movement toward a type of testing known as sorptivity testing.  Sorptivity testing measures the rate of absorption by capillary suction of water into the concrete (Neville 1996).  Generally, it is too difficult to mathematically model this flow in all but a single direction, and thus sorptivity tests are configured to establish one directional flow into the specimen (Stanish et al. 2000).  Sorptivity tests typically require that the sample be at a standard moisture content before testing is begun.  The benefits of sorptivity testing are reduced time, low cost of equipment, and simplicity of procedure.  The proposed ASTM standard test for sorptivity requires only a scale, a stopwatch, and a shallow pan of water (Stanish et al. 2000). One attractive feature of this approach is that the sample is conditioned for 7 days, with the temperature never exceeding 50 ^oC.  This is important since damage to the concrete microstructure can result at the higher drying temperatures (100 ^oC or higher) recommended in other test methods, thus biasing the test results.  Data reduction includes plotting the gain in mass per unit area over the density of water versus the square root of elapsed time, with the slope of the best-fit line being reported as sorptivity.

Testing Needs

As mentioned, a limitation of many of the standard tests is that they are designed to test only one element of the mix design.  For example, in ASTM C 1260 or C 1293, aggregate is tested under controlled, severe exposure conditions.  Aggregates that pass either of these tests will likely not incur ASR in the field, but for marginal aggregate, the results are less certain.  If marginal aggregates are to be used, it is recommended that low alkali cement and/or a low calcium fly ash or GGBFS be added. Unfortunately, there is no rapid standard test method available that can be used to judge the effectiveness of these admixtures, primarily due to the nature of the ASR reaction, which in some cases can take years to appear.

Efforts are under way to address this limitation.  For example, the proposed AASHTO Guide Specification for ASR (AASHTO 2000) recommends using both ASTM C 1260 and C 1293, as well as ASTM C 441 (Test Method for Effectiveness of Mineral Admixtures in Preventing Excessive Expansion of Concrete Due to Alkali-Aggregate Reaction) to test the effectiveness of proposed mitigation.  But for other MRD types, the implementation of new test methods is not imminent.  Gress (1997) laments the fact that a rapid standard test does not exist for the evaluation of internal sulfate attack in field-proportioned concrete (although the prevalence of this distress in paving concrete has yet to be determined). Currently, test methods employed by researchers primarily focus on the expansion of the cementitious materials and can take years to complete.  Considerable research is needed in this area before a standard test will be available. Until then, screening constituent materials offers the best alternative to ensure concrete durability.

Construction Considerations for Preventing MRD

Construction can have a very important impact on the occurrence of MRD. As stated, durable concrete must be relatively watertight and nonreactive.  From a construction perspective, this means that it must be properly batched, mixed, consolidated, and cured under conditions that will produce little microcracking and stable hydration products. Construction practices that have a direct impact on batching, mixing, consolidation, and curing include ambient and seasonal conditions, batching, duration of the mixing cycle, method of consolidation, steel placement, finishing, and curing.  The following discussion focuses exclusively on those elements of the construction sequence that have a direct bearing on concrete durability. It is assumed that the concrete is well designed and workable, and that standard concrete construction practices are being followed.

Ambient and Seasonal Conditions

Ambient and seasonal conditions can play an important role in the durability of concrete through their influence on the development of drying shrinkage cracking, excessive heat of hydration, and inadequate curing prior to deicer application. Hot-weather and cold-weather construction practice as described in Design and Control of Concrete Mixtures (Kosmatka et al. 2002) and elsewhere must be followed. It is imperative that plastic and drying shrinkage cracking be avoided to maintain the integrity of the concrete surface.  This can only be accomplished if an awareness of the relationship between relative humidity, temperature, and wind exists and is acted upon by the construction team. Curing practices should be established that are flexible enough to respond to changes in climatic conditions. A tool that is available to assist in predicting early-age climatic effects is the software package HIPERPAV (McCullough and Rasmusssen 1999).

Hot ambient temperatures potentially may affect the development of MRD. In addition to increased problems with plastic and drying shrinkage cracking, air loss, and workability, it is known that relatively high concrete temperatures can occur during curing if a high cement content mix is placed in the late morning on a very hot day. Gress (1997) cites a number of studies that found pavement temperatures exceeded 45^oC during curing. It is feasible that if precautions are not taken, the internal temperature of a thick slab made with a rich, Type III cement mixture placed during a hot summer day may approach 60^oC, a temperature at which DEF may occur.  It is therefore necessary to control the mixture temperature during paving to avoid detrimental heat build-up within the first 24 h.

Batching

Batching is obviously a very important construction factor that contributes to the durability of the concrete.  If the mixture is improperly batched, either through improper addition rate or the order in which materials are added, undesirable consequences may occur. One of the most important aspects is the control of water, since it directly impacts the w/cm.  The water content of the aggregates must be monitored and accounted for, and provisions must be made for the water remaining in the mixer after washing (TRC 1999).  Also, the accurate batching of admixtures can be critical with respect to durability.

Mixing Cycle

As speed of construction has increased, the use of shorter mixing cycles has been investigated by a number of SHAs.  This shorter mixing cycle addresses the increased material demand resulting from the use of larger, more powerful pavers. Unfortunately, a decrease in the length of the mixing cycle affects the homogeneity of the resulting concrete, with balling, clumping, and sand streaking becoming more prevalent (Gress 1997). In addition, air-entraining admixtures require a certain minimum amount of mixing to disperse the air voids uniformly throughout the mixture.  If this does not happen, freeze-thaw damage in the paste may occur. Thus, any shortening of the mixing cycle should be carefully investigated as it may have a detrimental impact on the durability of the concrete. 

Consolidation

The goal of consolidation of paving concrete is the removal of entrapped air while avoiding segregation and disruption of the entrained air system. The concrete mixture properties are key to good consolidation.  Assuming that a high quality, properly graded, workable mixture is used, it can be consolidated through mechanical means including both internal and external vibration.  Internal vibration is applied through immersion-type vibrators typically located after the strike-off in the paver.  Important factors to consider are the frequency of vibration, amplitude, and the speed of paving. Frequency is typically set between 7,000 and 12,000 vpm, although a recent study recommends fixing the frequency at 10,000 vpm (Gress 1997).  This same study recommends weekly calibration checks on the vibrators, a negative (down from horizontal) vibrator slope of 30^o or to one-half the pavement depth, and a fixed, uniform paver speed.  Other sources state that 8,000 vpm is more typical. In any event, it is critical that the vibration is set to consolidate the concrete without segregation or disruption to the entrained air-void system (TRB 1999).

It is emphasized that under-vibration can result in poor consolidation and over-vibration may lead to segregation and the elimination of the air-void system in the immediate vicinity of the internal vibrator. It has been noted on a number of MRD affected projects that the vibrator trails are visible on the pavement surface and appear to be an initiation point for cracking, spalling, and scaling. In some instances, petrographic analysis has revealed that the air content in the path of vibration is significantly less than that between the vibrators.

When constructing concrete pavements, considerable attention must be directed to produce a mix that has consistent consolidation characteristics. Early in the construction of the project, a paving rate and vibrator frequency must be established that adequately consolidates the concrete without harmful segregation or disruption to the air-void system.  If the mix is found to be harsh and difficult to place, construction should be stopped until the problem is corrected.  Too often, difficulties in placement are not discovered until a pavement begins to suffer early signs of deterioration at which point little can be done.

Steel Placement

The placement of embedded steel in concrete pavement has a direct bearing on the potential for corrosion.  Steel used at joints, such as dowel and tie bars, are typically placed at mid-depth, thus concrete cover is typically not an issue. But in these applications, the joint provides a direct route for chloride ion and moisture ingress, and thus these bars must be protected with a durable coating to avoid corrosion.  Epoxy is the most commonly used protective coating, although the cladding of dowels with plastic and stainless steel has also been used by some agencies.

For reinforcing steel used in jointed reinforced concrete pavement (JRCP) or in CRCP, adequate concrete cover is essential to prevent corrosion. At least 50 mm of quality concrete cover is required in environments where chloride-based deicers are used. To ensure that at least 90 to 95 percent of the reinforcing steel is adequately protected, it is recommended that a design cover of 65 mm be used.  During construction, great care must be exercised to avoid areas of “high steel” that become particularly susceptible to corrosion.

Finishing

In slipform paving, mechanical floating is commonly used immediately following the vibrating pan to embed large aggregate particles, correct small surface imperfections, and “close the surface.”  If the entire slipform paving operation proceeded smoothly, no other finishing other than texturing may be required.  But in most cases, additional finishing is needed to correct surface imperfections.

After paving, the surface is checked using a 3- to 4-m straight edge. Surface imperfections are commonly corrected using a hand-operated float.  It is common that little bleed water occurs with stiff, slipformed concrete, but under certain conditions, it may be present. It is emphasized that finishing should not be conducted when bleed water is present, as there is a danger that either bleed water will become trapped beneath the concrete surface, creating a plane of weakness, or air content in the surface layer will be reduced. In either case, scaling of the surface may occur due to freeze-thaw damage, particularly if chemical deicers are used.

Finishers commonly desire to add more water to the surface to assist in the finishing operation.  This tendency must be resisted as working water into the concrete surface decreases the w/cm and a