Guide for Curing Portland Cement Concrete Pavements, II

CHAPTER 4: INFORMATION GATHERED TO COMPLETE THE GUIDELINES

This chapter describes the work that was done to add information to the guide or to revise information already in the guide. The rationale for this needed work is in the last part of chapter 3. Information in this chapter is taken from available literature and/or guidance. Additional information obtained from laboratory work or from original analysis is contained in chapter 5.

Guidelines for Selecting Concrete Materials and Mixture Designs

Choice of materials and mixture proportions can have a significant effect on curing requirements, although choices of materials and mixture proportions are normally based on the expected properties of the hardened concrete and not on curing considerations. However, there are some details to be brought to the designer's attention, as discussed in chapter 2. The information that follows is an expanded discussion.

Hydraulic Cement

The major curing issue in choosing cements is the rate of strength gain and hence the length of curing time required to attain adequate physical properties. However, there are several other pertinent properties, as discussed below. Hydraulic cement is covered by five ASTM specifications and three AASHTO specifications.

The diversity in ASTM specifications reflects trends within the cement industry. ASTM C 150 is the first standard cement specification (1940) still in use.⁽⁴⁴⁾ ASTM C 595 was first published in 1967 to meet the needs of the emerging blended cement industry.⁽⁶¹⁾ This specification has not been widely used, mostly because of user preference and not because of any negative quality issues. ASTM C 1157 was first introduced in 1992 as a performance-based specification for blended cements in an effort to overcome some of the user preference problems.⁽⁵³⁾ The scope has since been broadened to include all hydraulic cements, including portland cement (now specified also under C 150).⁽⁴⁴⁾ There is considerable interest in the cement manufacturing industry to move all cement specifications to this standard. The performance-based format of C 1157 allows for more freedom to use innovative materials and processes.⁽⁵³⁾ A major factor driving this trend is pressure to reduce carbon dioxide (CO₂) emissions.

Partially for economic reasons, it is popular to purchase pozzolan separately and replace part of the cement with it. This is only common practice, however, when using portland cement as the primary cementitious material. There is concern that cements furnished under other specifications (e.g., C 595⁽⁶¹⁾and C 1157⁽⁵³⁾) may perform badly when blended with pozzolan because most of them already contain a fraction of pozzolan or slag and the effects of adding a third cementitious material are largely unknown. It is plausible that this kind of three-way cementitious mixture could be successfully used, but there is currently little or no guidance on this practice.

ASTM C 150⁽⁴⁴⁾/AASHTO M 85.⁽⁶²⁾ Some ACI guidance (e.g., ACI 308⁽³¹⁾) makes a significant distinction about the length of moist curing required when using the various cements covered by ASTM C 150:⁽⁴⁴⁾

• Type I-7 days.
• Type II-14 days.
• Type III-3 days.
• Type IV-no recommendation.
• Type V-no recommendation.

These recommendations assume a curing temperature of ≥ 10 °C.

The shorter curing times recommended for Type III cements is probably still valid. The distinction between Type I and Type II cement is, for all practical purposes, no longer valid. Until about 1980, Type II cements were commonly made with lower tricalcium silicate (C₃S) contents than Type I cements, hence were slower to reach required strengths and needed extra curing time. With changes in some details of C 150 requirements, the Type II cements now can be manufactured to gain strength at the same rate as Type I cements.⁽⁴⁴⁾ Most are now formulated for this higher early strength property. An exception would be a Type II cement meeting the optional C 150⁽⁴⁴⁾ heat of hydration (HH) requirement. Cements meeting this requirement are rare except as a special purchase item. Type IV cement is more rare than a Type II HH cement and normally used only in mass concrete. Type V cement is specified in high sulfate environments. Early strength gain of these cements is typically a little slower than Type I and II cements, but faster than the Type II HH cement. A prescriptive curing time of 10 days would be a reasonable interpolation of the above recommendations.

AASHTO M 85⁽⁶¹⁾ is identical to ASTM C 150⁽⁴⁴⁾ except for the requirement of £55 percent C₃S for Type II cements. Most ASTM Type I and II cements meet this requirement, but it does exclude a few cements with unusually high level of C₃S. This difference does not make a significant change in length of curing requirements developed around the ASTM types.

ASTM C 1157⁽⁵³⁾ (no AASHTO equivalent). No guidance was found that makes recommendations about curing times required for cements meeting this specification. The following equivalency in strength gain rate with C 150⁽⁴⁴⁾ cements is reasonably valid:

• Type GU = Type I.
• Type MH = Type II HH-not normally used with pavements.
• Type HE = Type III.
• Type MS = Type II without the HH requirement.
• Type LH = Type IV-not normally used with pavements.
• Type HS = Type V.

ASTM C 595⁽⁶⁰⁾/AASHTO M 240.⁽⁶³⁾ This is the original specification for blended cements. The ASTM and AASHTO specifications are identical except for some requirements on sulfate resistance and MgO levels. The differences are irrelevant to curing. These specifications contain a relatively complicated system of classification based on composition. The types designated as I(SM), IS, I(PM), IP is the close approximation to a Type I with respect to strength-gain properties and is most likely to be used for paving.

ASTM C 845⁽⁶⁴⁾ (no AASHTO equivalent). This specification covers expansive cement. The only product currently marketed is Type K. Cements meeting this specification were marketed aggressively for many years in the late 1980s and early 1990s. Availability has been greatly reduced in recent years. This cement is rarely used for large paving projects because of cost, but has been used in smaller paving and flat work because the expansive property helps avoid shrinkage. The chemical mechanism that causes the expansive property requires quite a lot of water, so that the maximum benefit of this property of the cement is obtained if wet curing is applied.

ASTM C 989⁽⁶⁵⁾/AASHTO M 302.⁽⁶⁶⁾ Ground granulated blast-furnace slag is actually a hydraulic cement, but it is often referenced as a supplemental cementing material (SCM). Its specification and test methods (C 989)⁽⁶⁵⁾ are managed by ASTM Committee C 9 (on concrete and concrete aggregates), along with pozzolans and other admixtures. This is a result of past internal politics in ASTM and not based on technical considerations. ASTM recognizes the term "ground granulated blast furnace slag" as representing the 100 percent slag product (C 989)⁽⁶⁵⁾, and the term "slag cement" as representing a portland cement-slag blend (C 595).⁽⁶¹⁾ Some 100 percent slag products are marketed and labeled as "Slag Cement." The confusion in this terminology is currently unresolved. However, the user can recognize the 100 percent slag product by the "C 989" designation and the blended cement product by the "C 595" designation.

Slag is sold under three designations based on strength development potential: Grade 80, Grade 100, and Grade 120. The grade designations are based on the approximate expected strengths of portland cement-slag mixtures, expressed as a percentage of controls at 28 days. Grades 100 and 120 would be suitable for paving. Grade 80 is more suitable for mass concrete applications and is less commonly available than grades 100 and 120.

Pozzolans

Pozzolans are specified by AASHTO M 295⁽⁶⁷⁾ and ASTM C 618⁽⁶⁸⁾. These two specifications have the same requirements except for the limits on loss on ignition and sulfate resistance. Neither of these has curing implications. Three classes are described:

• Class N-natural pozzolan, little marketed in recent years.
• Class F-fly ash from bituminous and anthracite coals, widely available.
• Class C-fly ash from subbituminous and lignite coals, widely available.

Pozzolans typically have two major effects on concrete that pertain to curing. These are time of setting and strength development.

Most fly ashes retard time of setting of cement pozzolan blends. This retardation varies from less than an hour to as much as several hours. There is no requirement in C 618⁽⁶⁸⁾/M 295⁽⁶⁷⁾ for effect on time of setting and, in practice, there is no selection that is normally made on the basis of this property. In paving this retardation affects timing of the texturing operation and application of final curing. Delay of these operations means that there is a longer period of time when the concrete may need to be protected from evaporation to prevent plastic shrinkage cracking.

Most pozzolans also retard the rate of strength gain in concrete, affecting the length of time they must be cured to get the desired physical properties. An exception to this generalization is when pozzolans are interground with portland cement during manufacture of blended cements. The grinding operation tends to effectively increase the surface area of the pozzolan so that there is essentially no strength-gain penalty.

Strength-gain retardation is particularly acute with Class F fly ash. Some research has recommended 10 to 14 days curing to get desired strength properties, while 7 days is recommended for portland cement. Class C fly ash tends give better early strength gain. This is because these materials are only partially pozzolanic and usually contain a significant fraction as cementitious compounds. Under proper circumstances, some Class C fly ashes can be used as the basis for hydraulic cement, although these cements are currently used only for patching.

The hydraulically active nature of Class C fly ash sometimes has a negative effect on early workability. Some materials cause the concrete to lose workability rapidly. This property is not covered by requirements in C 618⁽⁶⁸⁾/M 295.⁽⁶⁷⁾ The reactions that cause this tend to consume a considerable amount of water and cause the concrete to take on the appearance of early setting. This is not, however, the traditional setting reaction in concrete that involves the hydration of C₃S in the cement. This reaction is usually retarded by Class C fly ash, so the common manifestation of this effect is that the concrete loses workability within a few minutes, but then the time of setting is delayed. It has not been demonstrated, but this pattern of behavior may contribute to susceptibility to plastic shrinkage cracking.

Other Considerations on Cements and Pozzolans

Heat of Hydration. The HH requirements in C 150⁽⁴⁴⁾, C 1157⁽⁵³⁾, and C 595⁽⁶¹⁾are intended principally to meet the needs of mass concrete construction. However, at least one State (Florida) now specifies a HH limit on cement used in summer paving. The limit is for 335 kilojoules per kilogram (kJ/kg) at 7 days for concrete containing no pozzolan and 375 kJ/kg (88 cal/g) for concrete containing pozzolan. These limits do not conform to any standard ASTM requirements. Heat of hydration for Type I/II cements at 7 days is commonly near 335 kJ/kg, and most sources would probably meet this limit, but some companies make Type I and II cements that evolve as much as 395 kJ/kg (95 cal/g) at 7 days. The AASHTO M 85⁽⁶²⁾ limit of 55 percent on C₃S would probably effectively exclude these.

Fineness of Cement and Pozzolan. Very finely divided cements and pozzolans are noted for contributing to increased susceptibility to plastic shrinkage. Typical values of fineness for cements are 350 to 400 m²/Mg (ASTM C 204), and for pozzolans are 10-30 percent retained on a 45 mm sieve.⁽⁵³⁾ Values much finer than this may contribute to plastic shrinkage cracking. Type III portland cements typically have a fineness of 500 m²/kg or more. Silica fume is approximately an order of magnitude higher still. Fineness of slag is usually not measured, but typically it is ground somewhat finer than ordinary Type I/II cements. Grade 120 may reach 500 m²/Mg.

Admixtures

Water-reducing and air-entraining admixtures are commonly used in pavement concrete. There are some implications on curing with the water-reducing admixtures (WRAs). Some WRAs tend to retard the time of setting of cementitious materials significantly. This retardation is typically in addition to the retardation caused by use of fly ash. Unless the retardation is truly needed to overcome the accelerating effects of high concrete temperatures then, as discussed above under the section on pozzolan retardation, this retardation can cause an extension of the time when excessive evaporation must be avoided and possibly the timing of the texturing operation.

Mixture Proportions

Three mixture proportion properties have relevance to curing: water-cementitious materials ratio, cement content, and percentage of pozzolan.

Water-Cement Ratio. Paving mixtures are typically proportioned to a low w/c, but curing related problems may develop if water-cement ratios are too low. Sometimes water-cement ratios less than 0.40 are used to get high early strength development and to get desired cohesiveness for slip-form paving. A negative side to this practice is that low water-cement ratio concrete may have little or no bleed water. In moderation, bleed water is useful for buffering against the effects of high evaporation rates. If evaporation rates exceed bleeding rates, then conditions easily develop that favor plastic shrinkage cracking. Another negative feature of low water-cement ratios is internal desiccation. At water-cement ratios less than about 0.40, essentially all of the mixing water is consumed by cement hydration and gel water adsorption. This leaves the capillary pores in a condition exactly analogous to evaporative drying. This results in drying shrinkage and the potential for cracking unless water is added to the concrete during curing operations.

Cement Content. Some agency specifications set minimum levels of cement content to around 325-355 kg/m³. Other specifications are based around a minimum flexural strength limit. It is common practice to use higher cement contents if there is difficulty in getting the minimum strength or if there is uncertainty with respect to acceptance testing issues surrounding flexural strength testing. The precision of this test method is so poor that owner-contractor disputes over strength results are common. A negative side to using more rather than less cement, other than cost, is that it results in a concrete with a high paste content. It is the paste content that dominates the long-term drying shrinkage behavior of the concrete, particularly if the cement in the paste is largely hydrated. Ideally, cement contents should be just high enough to obtain the necessary strength (with enough extra to protect against testing uncertainty), but not higher than that.

Percent of Pozzolan. As noted above, portland-pozzolan mixtures hydrate slower than portland cement, so strength gain is slower and setting times are often extended. The hydration rate affects the length of the curing period required to get adequate strength. In USACE airfield construction, strength development problems normally limit pozzolan contents to about 20 percent for Class F and 25 percent for Class C fly ashes. In this type of construction, common practice is to slip form alternate lanes. Then, when there is sufficient strength to support the paving machine on the slip-formed concrete, the alternate lanes can be paved using conventional paving techniques. It is the rate of strength development of the slip-formed lanes that limits pozzolan contents in this type of construction.

Guidelines for Protection Between Placing and Curing

The time between placing concrete and application of curing has been termed the "Initial Curing Period" (ICP) in ACI 308R.⁽⁶⁾ This is a critical time for pavements under severe drying conditions because of the relatively large areas of unprotected concrete that are often left exposed and because of the relatively large effects of loss of water on plastic concrete. This condition cannot always be effectively remedied by early application of final curing measures because of the possibility of either damage to the concrete surface or poor membrane formation (in the case of use of curing compounds).

The conventional guidance (ACI 308⁽³¹⁾, caption to figure 1) on curing practices during the ICP is that evaporation rates exceeding 1.0 kg/m²/h are severe enough that protection of concrete surfaces from evaporative water loss should be required, and it cautions that evaporation rates more than 0.5 kg/m²/h may also result in damaging shrinkage.

These drying rate limits are apparently derived from typical bleeding rates of concrete (Al-Fadhala and Hover, 2001).⁽³²⁾ If evaporation rates of bleed water exceed bleeding rates, then the near-surface zone of the concrete will start to dry and shrink, potentially resulting in plastic shrinkage cracking. Paving mixtures, particularly those used with slip-form paving, tend to have low water contents and high content of fine materials, which tend to reduce the amount of bleeding. It is commonly believed that slip-form paving mixtures do not bleed at all. Therefore, for paving mixtures, it becomes important to know what the bleeding rates are for a particular placement so that critical evaporation conditions can be determined.

ASTM C 232 describes a test method for measuring bleeding in concrete.⁽⁶⁹⁾ The conditions of the test are tightly prescribed, which is necessary for comparing mixtures and materials, but this tends to make the method insensitive to details of actual placements. The rate and duration of bleeding depend on the thickness of the placement and to some degree on the nature of the base on which the concrete is placed (some bleed water will drain into a porous base).

A useful field test is a modification of C 232 in which the depth of the specimen is adjusted to simulate the depth of the actual placement.⁽⁶⁹⁾ A standard 150 mm by 300-mm mold is useful for this. If the concrete is to be placed on a drainable base, a layer of sand can be placed in the bottom. A more detailed protocol is described in the revised guide (Poole 2005).⁽⁵⁾ A time-dependent bleeding pattern can be determined for a concrete mixture.

An approximation of evaporation rates can be anticipated from the weather forecast. From this, an approximation can be made of whether or not a critical situation will develop. Then during placing, the ACI evaporation rates can be estimated with simple testing equipment. Temperature-humidity measuring devices and wind meters can be obtained for modest cost, so that onsite conditions can be determined periodically. Alternatively, evaporation rate can be determined directly by measuring mass loss from an open container of water located near the construction site. If this kind of testing shows that evaporation rates are likely to exceed bleeding rates, then protection measures can be implemented. Use of evaporation reducers is becoming a common practice in paving.

In reviewing State DOT practices, it was found that several States have a maximum time limit of 30 min during the initial curing period that the concrete can be left in an unprotected condition. This limit was discussed with a paving engineer who has had some experience with plastic shrinkage cracking in a California desert situation. The engineer thought such a limit might have been adequate to prevent the cracking he had observed.

There is not much information on which to base the length of such a limit, other than perhaps such practical experience. Plastic shrinkage cracking is such a multivariable problem that it is very difficult to develop simple guidance on length of exposure. However, some hypothetical calculations can provide at least an approximate frame of reference.

Holt (2000) published a water-loss versus shrinkage curve and a tensile strain capacity versus time curve that suggests that a water loss of about 1 kg/m² during the initial curing period might be sufficient to cause cracking in the concrete under investigation.⁽⁷⁰⁾ Thus at evaporation rates of 1 to 2 kg/m²/h, which is relatively high, evaporation would exceed the 1 kg/m² limit in 30 to 60 min. However, as shown above, paving mixtures are likely to have a low bleed rate, perhaps averaging considerably less than these values. A laboratory investigation of bleed rates of paving mixtures is reported in chapter 5.

Nomograph

The "estimated evaporation rate" used in much guidance is usually a calculation based on the ACI nomograph found in ACI 308.⁽³¹⁾ The nomograph only represents the evaporation rate of a free-water surface, such as exists with bleed water on the surface of concrete, and does not represent evaporation of water from fresh or hardened concrete when the free-water surface disappears below the exposed surface of the concrete. Representing this process in form of an equation allows development of practical software tools for translating site conditions to evaporation rates. The nomograph was translated into a calculation formula using Menzel's equation, which is the basis of the nomograph. The equation was found in the literature in non-SI (International System) units and the conversion to SI units is not obvious, therefore it is presented here in non-SI units (the final form of the equation later in this section is in SI units).

ER = 0.041(e_o - e_a)(0.253 + 0.215V)     (1)

where,

ER is the estimated evaporation rate (lb/ft²/h),

e_o is the vapor pressure of water at the evaporating surface (psi),

e_a is the vapor pressure of water is air (psi), and

V is the wind velocity (mph).

Vapor pressure of water in air varies with temperature, so that these components of the equation can be replaced by the vapor pressure/temperature relationship, which was determined by fitting an exponential curve to handbook data. The resulting equation is as follows:

ER = 4.88{0.1113+0.04224(^WS/0.447)} (0.00443)(e^{0.0302(CT1.8)+32}) - (^RH/100) e^{0.0302(AT1.8)+32}     (2)

where,

ER is evaporation rate (kg/m²/h)

WS is the windspeed meters per second (m/s),

CT is concrete temperature (°C),

AT is air temperature (°C), and

RH is relative humidity (%).

This calculation assumes that the concrete temperature represents the temperature at the evaporating surface. Since this is not strictly true because of the surface cooling that occurs during evaporation, it seems that a direct measurement of the concrete surface temperature would be the best value to use in the calculation.

Effect of Temperature on Time of Setting

Exponential functions have been found to well represent the effect of temperature on chemical reactions. The well-known Arrhenius equation, which forms the basis for the maturity method, is an example. The effect of temperature on time of setting should also be approximated by such a relationship. The following function has a form similar to the Arrhenius equation and will be inserted into the guide.

    (3)

where,

TOS is the time of initial setting (h) under field conditions,

TOS_stdtemp is the time of initial setting (h) under standard laboratory temperatures,

CT is the concrete temperature in the field placement (K), and

LabTemp is the temperature of the laboratory test (K).

K is a constant that varies with the cementitious materials, commonly called the activation energy.

A value of the constant K of 5000 Kelvins has been found to be a good default value for cementitious systems in general.

Time of setting is typically measured on concrete using ASTM C 403 at laboratory temperatures of about 23 °C, but this can actually be any temperature.⁽⁷¹⁾ In field conditions, concrete temperatures are usually allowed to vary between approximately 10 and 35 °C, so in place time of setting could vary significantly from the laboratory-determined value.

This equation was evaluated using data from several publications that describe the effects of temperature on time of setting. The data were presented graphically in these publications, so numbers used in this analysis were taken off of the graphs. The data in these publications represent time of setting under three similar temperature conditions: 10-15 °C, 23-25 °C, and 35-38 °C. The time of setting at the 23-25 °C condition was taken as the standard determination, and time of setting at the colder and hotter conditions was calculated using the above equation. Table 1 summarizes the data, the observed time of setting, and the time of setting predicted using this equation. A value of K of 5000 K was first used. A value of 4500 K was found to give a better fit of the data.

The average error is −0.1 h (−4.1 percent) for the low-temperature conditions and 0.6 h (−11 percent) for the high temperature predictions.

Curing Compounds

Product literature on curing compounds from many suppliers presents a large diversity of products. On first reading there seems to be quite a lot of overlap in properties among the various produces offered by each manufacturer. Quite naturally manufacturer's product literature lists all of the positive features of the material, so that the differences among the individual products tend to get lost in the long list of common features. The following is a reasonably complete list of properties and options. Combining the properties would result in hundreds of different products.

1. Type 1 versus 2. Type 1 materials are clear and used for architectural applications. Type 2 materials are white pigmented for use in hot weather or where the white pigment can assist in evaluating application quality. Type 1 may be supplied with a fugitive dye to assist in assessing application. The dye fades within a week of exposure to sunlight. Paving applications normally require Type 2.
2. Class A versus B. Class A is unrestricted, active ingredients often being wax-based or a linseed oil-based material because of cost. The active ingredient of Class B materials are resins. These are usually held by the manufacturer to be a somewhat higher performance class of materials and may be more expensive than Class A. There have been no implications that Class B materials do not meet the stated water retention requirements, but they may be messy to work with. Both Class A and B are normally allowed in paving.
3. Water Retention. Water retention (ASTM C 156) levels are set by various specifications.⁽¹⁸⁾ There are two different ASTM specifications (one of which is duplicated as an AASHTO specification) each with a different water retention requirement, the USACE has a different requirement for airfield paving, and State DOTs often have individual requirements. A large part of the diversity in product offerings is attributable to the variation in user's performance requirements for this property.
4. Drying Time. Drying time requirements in ASTM and AASHTO specifications are constant at £4 h, but there is variation among agency specifications. Some agencies require 30 min or less. This property also contributes to product diversity.
5. Compatibility With Coatings. Many curing compounds serve as effective bond breakers with materials applied over them, such as paints and adhesives.
6. Dissipating versus Nondissipating Curing Compounds. For some applications, it is necessary that the curing compound be removed. Dissipating curing compounds tend to break up and flake off when exposed to sunlight for a few days. Non-dissipating curing compounds are durable to the effects of sunlight.
7. VOC Compliance. Recent U.S. Environmental Protection Agency regulations require that some types of construction use low VOC materials, defined as less than 350 grams per liter (g/L) of volatile compounds. This requirement is often used in specification for indoor concreting where accumulation of volatile organics can present a health hazard. Other applications are not restricted.
8. Vertical Surfaces. Application to vertical surfaces requires that the viscosity of the compound be high enough to prevent it from sagging when applied to vertical surfaces at the recommended dosage. This property may also be important for curing deeply textured horizontal surfaces. These materials are likely to be of a higher viscosity than those designed for application to horizontal surfaces, and may not be suitable for spray-on application with the same equipment normally used in paving.
9. Special Properties. Some curing compounds are advertised to possess other special properties, such as acting as sealers after the curing period is over.

Application rates and practices vary considerably among agencies. Application rates vary from about 2.5 m²/L to about 6 m²/L. Some agencies require application in two coats or power spray equipment-even giving details on the pressures required, type of spraying patterns, type of agitation equipment, etc. These are detailed in Volume I.⁽⁵⁾

Tined Surfaces

Grooving and tining have the effect of increasing the surface area of the concrete beyond the simple surface area calculated from the length and width measurements of the structure. The amount of the increase depends on the grooving pattern. On simple analysis, it would appear that the application rate of curing compounds to ungrooved pavements should be increased in proportion to the increase in surface area caused by the grooving, as represented by the following equation:

The grooved surface area is calculated from the distance between grooves and the width and depth of the grooves. This equation was put into the guide as a planning tool. It was evaluated using data from Shariat and Pant (1984).⁽⁷⁴⁾

Shariat and Pant⁽⁷⁴⁾ showed that moisture loss for a grooved specimen tested according to ASTM C 156⁽¹⁸⁾ increased in proportion to the extra surface area created by the grooving. In their tests, surface area of the grooved to the nongrooved surface increased by a factor ranging from 128 to 145 percent. In another part of the study, they investigated the relationship between curing compound application rate and water retention of grooved specimens in which the grooved surface to nongrooved surface-area ratio was 1.35 (0.74 inverted, as expressed in the above equation). There was a fairly large scatter in these data, but on the average, the amount of curing compound needed to achieve the water retention of the grooved mortar was 3.9 m²/L. The application rate to the non-grooved mortar was 5.0 m²/L. The above calculation would result in a calculated application rate of 3.7 m²/L, indicating that the calculation correctly accounts for the increased surface area.

Evaporation Reducers

Little information was found on evaporation reducers (now the preferred terminology-also sometimes called evaporation retarders) except for manufacturer's information sheets. There are currently no specifications for these products.

Holt (2000) included one such product in an investigation of plastic shrinkage cracking and showed significant benefit.⁽⁷⁰⁾ At least one manufacturer claims an 80 percent reduction in evaporation rates. Laboratory work performed in conjunction with this investigation showed about a 50 percent reduction.

The action principle of these products is that they form an oil-like film on the surface of the concrete or bleed water, which then physically retards the rate that water molecules can evaporate. The active ingredient is polyvinyl alcohol.

Application rates are about 5 m²/L, the same as for curing compound. Given that most of the carrier is water, this liquid provides something of a buffer from evaporation for a few minutes to an hour or more, depending on the drying conditions, independently of the action of the alcohol film.

Effectiveness of Water-Added Curing in Low Water-Cement Ratio Concrete

It is recognized that low water-cement ratio concrete can consume all of its mixing water in hydration, creating the potential for autogenous shrinkage cracking and that added water during curing should be an effective tool for preventing this. However, the caution often expressed is that water does not effectively penetrate low water-cement ratio concrete. Some additional literature was found on this subject that can be used as at least a partial evaluation of this subject.

The continuity among pores and with the surface of concrete tends to decrease rapidly in low water-cement ratio concretes, so that introducing appreciable water into the concrete may be difficult (Meeks and Carino, 1999⁽⁴⁾; Cather, 1994⁽⁷⁵⁾). The onset of capillary discontinuity varies with water-cement ratio. Table 2 gives times to capillary discontinuity from three different references. The numbers are similar but not in exact agreement.

For purposes of curing concrete pavements, the important point is that capillary discontinuity of 0.40 water-cement ratio concrete develops in 2 to 3 days and that any attempt to introduce water into the concrete through wet curing methods will decrease sharply in efficiency after this time. It is not uncommon in modern slip form concrete mixtures (and bridge deck concretes) for water cement ratio to be below 0.40 and as low as about 0.35. The information cited above was apparently assembled before such concretes were as common in practice.

Test Methods for Determining Amount of Curing Compound Application

Several methods for estimating the application rate of in place curing compound, after the fact, were mentioned in the interim report. These included indirect methods based on flow rates through the spray nozzles along with rate of movement of the curing machine, infrared images of the pavement, direct measurement with sampling coupons, and visual estimation. Another potential method is reflectance.

The Minnesota DOT wrote a rather extensive report on curing compound application along with the method of estimation based on flow rates.⁽⁷⁹⁾ This has been extracted in parts and put into the guide. Some data on the infrared approach was developed and that method rejected as being too sensitive to a number of variables to be of practical use. These data are presented in chapter 5. Methods for direct measurement with sampling coupons for use of visual estimation and for use of reflectance have been developed and put into the guide.

Length of Curing

Fixed time interval curing specifications tend to oversimplify the problem. Development rates of the physical properties of concrete vary widely, depending on materials, proportions, and temperature. The advantage of fixed time interval requirements is that interpretation and enforcement are relatively simple. Because specifications tend to oversimplify the problem, they are somewhat conservative. Project specifications should allow more complicated, performance-based requirements as an option.

The prescriptive approach to specifying length of required curing has been largely based on strength-gain rates of the cements used to make the concrete. AASHTO guidance is that curing should be at least 3 days at temperatures 10 °C and above. No mention is made of the effects of pozzolan or slag. ACI guidance varies, but approximately converges on 7 days for concrete containing Type I (or equivalent, presumably), and 14 days for Type II, also at temperatures ≥ 10 °C. The latter guidance was developed in a time when Type II cement actually gained strength slower than Type I cement. In general, this is no longer true. The ACI guidance is also silent on pozzolans and slag. Most of States surveyed reported prescriptive curing times of 3 days. A smaller number reported 7 days, and a few reported 4 days.

The AASHTO guidance may be based on development of flexural strength and on the nature of paving concrete mixtures. Paving mixtures are typically proportioned with relatively high cement contents and relatively low water-cement ratios. As a result, 3‑day strengths are relatively higher than at the same age when using leaner, higher water-cement ratio concretes. Also, one set of USACE data showed that flexural strength in airfield paving mixtures developed even faster than compressive strength. A discussion of these data is presented in chapter 5.

Three days is probably adequate time for most paving mixtures to develop strength sufficient for service when temperatures are around 20 °C or higher. However, rates at 10 °C may be only about half this rate and it appears that additional time would be of benefit. Perhaps an approach similar to that used in the UK ⁽⁷⁾ would be appropriate. This guidance bases lengths of curing required on maturity concepts, resulting in a set of curing times that graduate upwards as temperatures drop. Such an approach will be developed for this project.

On the question of rate of development of durability properties in the near-surface zone of the concrete, examination of the literature suggests that these properties develop at about the same

rate as compressive strength development. The length of curing for mixtures containing pozzolan or slag seems to converge on 7-10 days at about 20 °C.

Test Methods for Determining Efficiency of Curing of the Near-Surface Zone

As mentioned in the preliminary report, strength is a good measure of the overall state of curing of concrete, but does not represent well the effects of moisture loss in the near-surface zone, usually thought to be about 50 mm thick. A wide variety of test methods have been developed for determining the degree of physical property development in the near-surface zone as a result of curing. Many of these require specialized equipment and are not easily applied in the field due to lack of control of moisture content of the concrete. Several were thought to be plausible and some work was done to check this plausibility.

Ultrasonic Pulse Velocity (UPV). UPV has been reported to be useful as a field tool for measuring strength development. One drawback is that the measurement path does not appear to be necessarily confined to the near-surface zone of the concrete, which is the zone of interest. A simple evaluation of this concept is described in chapter 5.

Rebound Hammer. The rebound hammer (ASTM C 805) is often criticized as an unreliable method for measuring strength development in concrete because of its sensitivity to surface irregularities.⁽⁷⁹⁾ However, some preliminary work suggested that rebound numbers could possibly reflect differences in curing. An evaluation of this method is described in chapter 5.

Water Absorption. Water absorption has been demonstrated as a good method for distinguishing degree of curing in mortars. It was the basis of ASTM C 1151 for evaluating curing compounds.⁽²⁷⁾ It has since been withdrawn for lack of use in the curing compound industry, but in principle it may still be a useful method, although there is some uncertainty about its use in concrete. It is one of the methods sensitive to the moisture content of the concrete; hence, it is difficult to apply in the field. However, it can be easily applied to cores without elaborate gear and since coring is such a common way to obtain samples to evaluate field concrete. An evaluation of this approach is described in chapter 5.

Abrasion Resistance. Abrasion resistance is the basis for another test method basis that is often mentioned in the literature as being heavily dependent on degree of curing. The effect of poor curing on the abrasion resistance of lab specimens is fairly obvious when simple nonquantitative approaches like scratching or grinding the surface are employed. Past experience in USACE has shown that putting this on a quantitative basis reproducible among laboratories would not be straightforward. Therefore, this method was not pursued further.

Chemically Combined Water. Measuring the amount of chemically combined water is one of the oldest methods for measuring the amount of hydration in cement pastes. It is not field applicable, but like water absorption, it can be executed on cores. Preliminary work suggested that the method might be more sensitive to details when applied to concrete, given the relatively small amount of cement in concrete and the relatively small amount of water consumed in early hydration reactions. Consequently, this method was not pursued further.

Criticality of Thermal Gradient (Short-Term Thermal Shock)

Guidance exists and is cited in the guide on limits to temperature gradients that can develop between the surface of concrete and the interior due to rapid cooling of the surface, either by evaporation or by application of cold curing water. Additional guidance will be developed on the level of evaporation expected to result in sufficient cooling that these limits might exceed.