U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations

Report
This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-04-122
Date: February 2005

Computer-Based Guidelines for Concrete Pavements Volume II

Previous | Table of Contents | Next

CHAPTER 2. EARLY-AGE PAVEMENT BEHAVIOR

2.1 GENERAL

The behavior of concrete pavements at early ages is influenced by a number of factors including, but not limited to:

During the first hours after construction, the interactions of these factors result in volume changes in the concrete, primarily due to changes in temperature and moisture. During the hardening state, stresses in the concrete start to buildup due, in part, to slab curling and warping and restraint to axial movements at the slab-subbase interface. As concrete is weak in tension, undesirable situations may occur if the stresses in the concrete exceed the developing strength.

In this context, early age is understood as the first 72 hours after pavement placement. Experience demonstrates that significant stresses in the pavement develop during this curing period. These stresses may lead to undesirable situations if not properly controlled.

2.1.1 Temperature and Moisture Changes

The following sections describe the various factors that influence the temperature and moisture state of early-age concrete.

2.1.1.1 Heat of Hydration

A number of properties of concrete are governed by the heat of hydration of the cementitious materials. These properties include concrete set time, strength development, and modulus of elasticity development. In addition, the heat of hydration contributes to the temperature increase in the concrete during the first hours after placement.

The combination of cementitious materials and water results in a variety of chemical reactions in which hydration products are formed. The hydration of the cementitious materials is exothermic, and the degree of hydration is related to the amount of heat liberated at any point during the hydration stage. The rate of hydration and total heat liberated during hydration depends on factors such as the type of cement, admixtures, water-to-cementitious materials ratio (w/cm), fineness of the cement, and curing temperature.

2.1.1.2 Concrete Temperature

In the early ages, concrete temperature is a function of the heat of hydration and climatic conditions. The heat generated due to hydration results in a temperature rise in the concrete as a function of the thermal conductivity and specific heat of the paste and aggregate. On the other hand, climatic conditions such as air temperature, solar radiation, cloud cover, and convection due to windspeed affect the amount of heat lost or gained through the surface of the pavement. This heat loss or gain is transported through the depth of the slab, as a function of the concrete thermal conductivity and specific heat. Heat conduction to or from the subbase also affects the temperature of the concrete.

An intrinsic relationship exists between heat of hydration and concrete temperature. Because the heat of hydration has a direct correlation with the development of mechanical properties such as strength and modulus of elasticity, the temperature at which the concrete is cured will affect its mechanical properties.

2.1.1.3 Moisture in Concrete

Moisture in the concrete may be classified into two types. The first type is structural water, chemically bound within the cement paste. The second type is water contained in the pore structure. The sum of these portions equals the total water content in the paste. In addition to temperature development, the moisture state in the pore structure of hydrating concrete is another important factor affecting its early-age behavior. It has been recognized that the level of moisture at which a concrete structure is cured can have a significant effect on its strength, and also on the stress development due to shrinkage.

The moisture profile in a pavement structure is a function of climatic conditions such as relative humidity, temperature, and precipitation. The moisture state of the underlying subbase also affects this moisture profile. In addition to the external moisture conditions at which the pavement structure is exposed, the moisture gradients in the pavement are also a function of the thermodynamic equilibrium conditions of moisture flow and the random geometrical nature of the pores in the concrete structure. In general, differences of energy due to capillary forces, gravity, and thermal gradients are the primary factors governing the moisture transport in concrete. The flow of moisture is also a function of the concrete moisture diffusivity and the characteristics of the porous media.(4)

2.1.2 Development of Concrete Properties

This section describes the development of various early-age concrete properties.

Concrete Set

Concrete set is defined as the transition of concrete from a plastic to a solid state. Concrete set also marks the start of development of significant mechanical properties. It also is common to associate the set time with the moment where the stresses in the slab begin to develop due to volumetric changes coupled with various restraints.

Strength and Modulus of Elasticity Development

Tensile strength development in concrete begins to increase after the concrete sets. It develops primarily as a function of the water-to-cement (w/c) ratio, cement, admixtures, and aggregate characteristics and content. It is also a function of the energy of compaction (consolidation), the curing temperature, and moisture state. The strength of the concrete depends on the strength of the cement paste, the strength of the aggregates, and the bond strength of the cement/aggregate interface. The rate of strength development will be a function of the cement properties such as the cement fineness, cement compounds, and admixtures used.

Cements with a higher fineness commonly will hydrate faster, developing a high early strength, although the strength development at later ages may be lower as compared to coarser cements. Furthermore, it has been observed that fine ground cements are associated with durability problems.(5) A similar effect on the strength development occurs when the concrete is cured at high temperatures. A rapid strength increase is observed at early ages, and a less steep strength increase is noted with the long-term strength.

The effect of moisture on concrete strength also should be emphasized. In concrete pavements, pronounced moisture profiles may occur as a consequence of climatic conditions. It has been observed that whenever the internal relative humidity in the concrete drops below 80 percent, the development of strength can be significantly affected.(6) Similar to strength, the hardening process of concrete also contributes to its stiffness or modulus of elasticity. The concrete modulus of elasticity is directly related to the concrete strength, and also depends on the type of aggregates and its volume in the concrete mix.

2.1.3 Development of Stresses

Stresses in concrete pavements develop due to a number of factors. The following sections discuss some of these.

2.1.3.1 Axial Expansion and Contraction

As the temperature changes after set, the concrete tends to expand or contract as a function of its CTE. The concrete CTE is a function of the CTE of the paste and the CTE of the aggregates. Depending on the concrete-making materials and specifically on the type of coarse aggregate selected, the concrete CTE may be higher or lower. Typically, calcareous aggregates such as limestone and dolomite have a low CTE (4 to 9' 10-6 m/m/°C) while siliceous aggregates, such as quartzite, possess a high CTE (10 to 12.5'ยด 10-6 m/m/°C). Also, because the CTE of the hardened cement paste is greater than that of the aggregate, as the percentage of aggregate increases, the overall concrete CTE decreases. CTE has been found to be one of the most influential factors on the performance of concrete pavements. Stresses in the concrete also develop due to restraints imposed at the slab-subbase interface that resist the expansion and contraction of the slab. Stresses develop in the concrete due to this restraint coupled with the temperature changes from day to night. Expansive movements will lead to compressive stresses, while contractive movements will lead to tensile stresses.

2.1.3.2 Plastic Shrinkage

As the name implies, plastic shrinkage occurs in the plastic state of concrete, before it sets. Plastic shrinkage is a function of the evaporation of water from the surface of the concrete. Evaporation is a function of air temperature, windspeed, ambient relative humidity, concrete temperature, and many other factors. If the water at the surface evaporates faster than the water that migrates from the interior of the slab, plastic shrinkage can occur. Plastic shrinkage can also occur at the bottom of the slab by suction due to a dry subbase.(7) Plastic shrinkage commonly takes the form of random cracks at the surface of the concrete.

2.1.3.3 Autogenous Shrinkage

Autogenous shrinkage is related to the reduction in volume of the cement paste as a result of self-desiccation. Self-desiccation is a process where physical and chemical changes in the cement components during hydration lead to drying of the capillary pores in the concrete. This reduction in volume is significant after initial setting and is more pronounced in the case of high-strength concrete where low w/c ratios are used. Autogenous shrinkage has been observed to be less significant for high w/c ratios.(8)

Autogenous shrinkage is not associated with moisture transport, temperature changes, or external restraints. Shrinkage reducing admixtures are available to minimize autogenous shrinkage. However, autogenous shrinkage can also be minimized or avoided by maintaining water available during the cement hydration. This can be accomplished by water curing or by including saturated porous aggregate in the mix (commonly termed internal curing).(8) When enough water is available during hydration, self-desiccation is avoided, and some minor swelling of the concrete may occur instead.(9)

2.1.3.4 Drying Shrinkage

Drying shrinkage of concrete occurs due to drying when concrete is exposed to unsaturated air. The normal wetting and drying cycles to which the concrete is subjected lead to changes in the moisture state of the concrete. Although some of the water movement may be replaced by subsequent wetting periods, some of the water lost to the environment is irreversible and leads to permanent concrete shrinkage. Drying shrinkage is a function of a number of factors, including the w/c ratio, cement type, cement content, admixtures used, type and amount of aggregates, and climatic conditions.

2.1.3.5 Curling and Warping

Temperature and moisture transport in the concrete result in gradients that can lead to curling and warping movements and stresses. The concrete slab will curl up or down depending on whether the top of the slab is cooler or warmer than the bottom. If the top of the slab is cooler than the bottom, the upper surface will tend to contract, and the bottom will tend to expand, causing the slab to curl up. A downward curling usually is observed if the opposite temperature differential occurs. Due to its exposure to the environment, the top of the slab typically experiences higher drying shrinkage than the bottom. This can cause an upward curvature of the slab, commonly known as warping. Warping has been observed to change as a function of the climate, including the relative humidity and rainfall. Slab curling and warping are, in turn, restrained by the weight of the slab as well as subbase restraints. Compressive and tensile stresses are generated in the pavement as a function of this restrained movement.

2.1.3.6 Creep/Relaxation

It has been observed that creep and relaxation effects can be significant during the early age of concrete.(10) When concrete is subjected to a constant load, both an instantaneous and a delayed elastic strain occur. Strain due to plastic deformation of the concrete develops and increases during the loading period. If the load is removed, the elastic strain is totally recovered, while the strain due to plastic deformation is not. This viscous behavior is commonly known as creep. Conversely, when concrete is subjected to a constant strain level, relaxation of stresses occurs.

During the early age, concrete behaves more like a viscoelastic material than it does later in its hardened state, when its behavior is more purely elastic. During this early-age period, concrete is more susceptible to creep and relaxation effects. Because most of the axial, shrinkage, and curling movements in the pavement are restrained by external factors (such as the slab weight and subbase restraints), the pavement is constantly subjected to tensile and compressive strains. The corresponding level of stress that develops is a function of the creep-relaxation characteristics of the concrete.

Figures 1 and 2 illustrate the typical relaxation effects in concrete as a consequence of temperature loading. After the concrete is placed, the concrete temperature increases due to the heat of hydration (figure 1). Immediately after the concrete sets, strength begins to develop. As the concrete is restrained to move in the axial direction, the concrete tries to expand due to the temperature increase, and compressive stresses develop (figure 2). Because of the creep characteristics of the concrete at such an early age, the compressive stresses are relaxed significantly. After the maximum temperature is reached and the concrete starts a cooling period, the concrete begins to contract. Due to the continued relaxation, the concrete is subjected to tensile stresses even before it cools down to the set temperature at which the compressive stresses initially started. Eventually, as the concrete keeps contracting, increased tensile stresses are generated. When the concrete stresses exceed the tensile strength of the concrete, cracks form. The creep and relaxation properties of concrete are a function of a number of factors, including the moisture state, temperature, concrete properties, stress level, duration of load, and concrete age.(10)

Figure 1.  Graph.  Conceptual representation of temperature development in a concrete element with time.  Graph depicts relationship between Temperature (Y-axis) versus Time (X-axis) represented by a sine-shaped curve.  PCC sets at about halfway between minimum and maximum temperatures.

Figure 1. Conceptual representation of temperature development in a concrete element with time.

Figure 2.  Graph.  Conceptual effect of creep/relaxation on concrete stresses.  Graph depicts relationships between Stress (Y-axis) and Time (X-axis) with positive Y being Compression and negative Y being Tension.  The solid sine-shaped curve represents actual stress with relaxation effect.  The dashed sine-shaped curve represents theoretical stress with no relaxation effect.  The dashed curved line (below the X-axis) decreasing from left to right represents the tensile strength.  The point in which the tensile strength line meets the actual stress line represents the formation of cracks.

Figure 2. Conceptual effect of creep/relaxation on concrete stresses.(11)

2.1.4 Thermal Cracking

The combination of axial and curling thermal stresses commonly leads to the development of significant compressive and tensile stresses. Because the concrete is weaker in tension than in compression, any condition leading to a decrease in concrete temperature, thermal gradients in the slab, and/or the continued drying shrinkage originated from moisture changes may result in thermal cracking whenever the stresses that develop exceed the tensile strength of the concrete.

2.2 EARLY-AGE INDICATORS OF LONG-TERM PERFORMANCE OF JPCP

The size of the joint opening in JPCP is one of the primary early-age indicators of the pavement's long-term performance. This opening controls the load transfer efficiency across the joint, which controls how well the JPCP will perform over time. JPCP can either have doweled or nondoweled joints. In this section, the physical mechanisms that influence joint movement are described. In addition, the influence of joint opening on pavement distresses is discussed.

2.2.1 Physical Mechanisms Governing Joint Opening at Early Ages

The joint opening at early ages is primarily controlled by the effects of temperature and moisture changes on pavement and subbase properties such as concrete CTE, drying shrinkage, and subbase restraint. Before the concrete's final set, a common assumption is that the pavement is free of stress and strain. After this time, stress develops in the pavement due to phenomena such as climatic conditions, hydration, creep, and shrinkage.

As an example, if the pavement is subjected to an increase in temperature (+pressureT), the pavement will expand about its centerline, and the joint will close. During a decrease in temperature (-pressureT), the pavement will contract about its centerline, and the joint will open (see figure 3).

Figure 3.  Sketch.  Schematic of joint opening due to temperature drop (negative delta T).  Sketch shows two slabs contracting due to a temperature drop, thus causing the joint to open.

Figure 3. Schematic of joint opening due to temperature drop (-pressureT).

It has been observed experimentally that a change in temperature is the most influential parameter affecting joint opening. Joint opening varies with the daily temperature cycles due to changes in the pavement's thermal conditions.(12) CTE, drying shrinkage, and pavement length also affect the joint opening. To control the joint opening to acceptable levels, proper curing of the pavement at the time of placement is of paramount importance. If the concrete's temperature differential can be minimized, the size of the joint opening will decrease as a result. Likewise, drying shrinkage should also be minimized if the joint opening is to stay tight. This can also be accomplished with adequate curing methods.

2.2.2 Effect of Joint Opening on Load Transfer Efficiency—JPCP without Dowels

In JPCP without dowels at the joints, the size of the joint opening governs the load transfer efficiency across the joint (see figure 4). The aggregate interlock transfers the load, and its efficiency is a function of factors such as aggregate type, aggregate size, angularity, and abrasion characteristics.) However, if the joint opening is too wide (0.6 millimeters (mm) or greater), the aggregate cannot provide load transfer.(14) In addition, because of the wide opening, water can infiltrate the subbase from the surface of the pavement. The water will infiltrate faster as the joint opening increases. The size of the joint opening also determines how quickly incompressibles can fill the joint over time. Incompressibles apply pressure on the pavement edges and can cause spalling, as described in section 2.2.4.

Figure 4.  Sketch.  Schematic of joint in JPCP without dowels.  Sketch shows a slab with a joint that has an opening, which allows water and incompressibles to infiltrates, thus causing loss of aggregate interlock/load transfer and a saturated subbase.

Figure 4. Schematic of joint in JPCP without dowels.

The loss of load transfer across the joint can result in the development of significant deflections, particularly for the case of granular subbases, as shown in figure 5. This eventually can lead to faulting and transverse cracks, so it is desired that the pavement maintain good load transfer efficiency. This phenomenon will be discussed in more detail in section 2.2.4.

Figure 5.  Sketch.  Schematic of JPCP after loss of aggregate interlock/load transfer at the joint.  Sketch shows the slab on the right to be completely separated from the slab on the left and is deflected to be lower than the slab on the left.  The slabs sit on a saturated subbase.

Figure 5. Schematic of JPCP after loss of aggregate interlock/load transfer at the joint.

2.2.3 Effect of Joint Opening on Load Transfer Efficiency—JPCP with Dowels

For JPCP with dowels across the joint, the dowels act to partially restrain vertical movement. As figure 6 illustrates, the adjoining slabs deflect a similar amount. This minimizes faulting, while still allowing horizontal movement due to temperature fluctuations. The size of the joint opening does not significantly affect the load transfer efficiency component due to the dowels.

Figure 6.  Sketch.  Schematic of doweled joint in JPCP.  Sketch shows the slab on the right deflecting, but it pulls down the slab on the left and does not become separated or lowered because of the dowel bar.  The slabs sit on a saturated subbase.

Figure 6. Schematic of doweled joint in JPCP.

2.2.4 Distresses Influenced by Joint Opening—JPCP without Dowels

As described in the previous sections, the size of the joint opening can impact significantly the behavior of the pavement in the long term. Three possible pavement distresses that can result due to adverse behavior are spalling, faulting, and transverse cracking.

Spalling. Spalling is one of the pavement distresses that can occur at a concrete pavement joint. Incompressibles that enter the joint exert pressure on the pavement edges. This pressure can induce significant stresses in the concrete that can eventually cause the concrete to spall. This form of distress may be delayed or prevented if the joint opening is tight enough to prevent infilling of incompressibles. More details on spalling are provided in sections 4.5 and 4.6.

Faulting. When aggregate interlock at the joint does not provide adequate load transfer, the pavement profile will not be continuous (section 2.2.1). Faulting can occur due to this mismatch in slab deflections at the joint. A larger joint opening can also increase the amount of water infiltration into the joint. Faulting can initiate when aggregate interlock at the joint is no longer effective. More details on faulting are provided in sections 4.1 and 4.2.

Transverse Cracking. Transverse cracks can form in JPCP due to a number of different mechanisms. The two mechanisms discussed here are both the result of a large joint opening. The first form of transverse cracking is termed top-down cracking; it is caused by the erosion of the subbase support. Severe erosion of the granular subbase material from under the leading edge of the pavement can cause tensile stresses to form at the pavement surface. This extreme loss of support can cause a top-down transverse crack to form. For a nonerodable stabilized subbase, transverse cracking can still occur, but the mechanism is different. Incompressibles can become trapped beneath the pavement and the subbase. This leads to a slab lift-up, and a bottom-up transverse crack can result. Both bottom-up and top-down transverse cracking can be avoided or delayed if water seepage into the pavement joint is prevented. This is possible if a tight joint opening can be maintained. Transverse cracking is discussed in greater detail in section 4.3.

2.2.5 Distresses Influenced by Joint Opening—JPCP with Dowels

In doweled JPCP, pavement distresses are related predominately to the bearing stress placed on the concrete at the dowel-concrete interface. If this pressure is too great, spalling of the overlying concrete is possible. Likewise, faulting is possible if the bearing stress is excessive. The dowels will lose their ability to transfer load if their bearing stress causes the concrete around them to fail. Over time, faulting will be apparent. Yet, if the bearing stress is kept below the crushing strength of the concrete, faulting can be maintained at an acceptable level.(15) Dowel looseness also is possible if the dowel-concrete stress is very high. The concrete near the dowel can crush, and voids can develop underneath as the crushed concrete particles are removed. For more detail on dowel looseness, refer to section 5.1.7.

2.3 EARLY-AGE INDICATORS OF LONG-TERM PERFORMANCE OF CRCP

As its name implies, CRCP refers to concrete pavement constructed with no transverse contraction joints, and is reinforced with steel. For this type of pavement, concrete is allowed to crack randomly as a consequence of volume changes resulting from temperature and moisture variations that are restrained by steel and subbase friction.

Concrete and steel tend to contract and expand with changes in temperature. Because the CTE of steel and concrete are usually different, the expansion and contraction movements are not uniform, and stresses can develop in both elements. Concrete drying shrinkage, slab curling and warping, and subbase restraint each contribute uniquely to stress development. Since concrete is weak in tension, whenever the stresses that develop are higher than the tensile strength of the concrete, transverse cracks form to relieve the stresses. Because the steel has a high yield strength, it keeps cracks together. Keeping the cracks tight is essential in maintaining load transfer through aggregate interlock, and to avoid water infiltration and intrusion of incompressibles through the cracks. Subsequent drops in temperature and loss of moisture in the concrete can reduce the transverse crack spacing further. Later in the life of the pavement, externally induced stresses due to wheel loads and seasonal changes in climatic conditions can reduce the initial crack spacing even further. It has been observed that the crack spacing decreases rapidly during the early age of the pavement, then remains fairly constant from the initial year until pavement wearout, as illustrated in figure 7.(16)

Figure 7.  Graph.  Conceptual reduction in mean crack spacing over time.  Graph depicts relationship between Mean Crack Spacing (Y-axis) versus Time (X-axis).  The relationship as a whole is represented with a decreasing trend that can be broken into three phases: an initial phase, represented by the linear portion decreasing sharply from left to right; a stable phase, represented by a linear portion decreasing very little from left to right; and a fatigue phase, represented by a linear portion decreasing moderately from left to right.

Figure 7. Conceptual reduction in mean crack spacing over time.(16)

The crack spacing typically is greater near the ends of the pavement than at the central section. This is because the pavement is more restrained to movement at the center of the slab than at the ends. Therefore, most of the longitudinal movement due to volume changes occurs at the ends of the slab, and higher stresses are generated at the central section of the pavement.

Previous experience has shown that primary early-age pavement indicators of performance on CRCP include crack spacing, crack width, and steel stress. The following sections describe these behavioral indicators in more detail.

2.3.1 Crack Spacing

CRCP slabs usually distribute traffic loads in both the longitudinal and transverse directions. However, in the case of short crack spacings, the slab acts more as a beam, with its longer dimension on the transverse direction. As a result, the load is distributed primarily in the transverse direction, and significant transverse tensile stresses may develop. Due to this condition, longitudinal cracks may form as a function of the magnitude of the stresses, tensile strength, fatigue characteristics of the concrete, and cumulative load applications. The longitudinal cracks lead to a distress condition that is commonly known as a punchout. Punchouts are a typical distress manifestation in CRCP where a block of the concrete slab is separated from the rest of the pavement by cracks in the transverse and longitudinal directions. The severity and progression of punchout distresses is a function of the traffic loads, support conditions, and load transfer between cracks. On the other hand, larger crack spacings commonly result in wider cracks that may lead to spalling problems.

McCullough et al. demonstrated that the crack spacing could be properly controlled to fall within certain limits to minimize such undesirable conditions. He proposed a crack spacing of 1.7 to 2.4 meters (m) for this purpose.(2) The American Association of State Highway and Transportation Officials (AASHTO) guidelines later recommended a similar crack spacing of 1.1 to 2.4 m,(17) although, in some instances, crack spacings of less than 0.6 m have performed well under very good soil-support conditions.(18) Crack spacing in CRCP is also affected due to the variability of materials and construction procedures. Therefore, it is a recommended practice to evaluate crack spacing in terms of its average value as well as in terms of its distribution. For a given crack spacing distribution, the percentage of crack spacings that do not fall within the recommended range of 1.1 to 2.4 m will lead more likely to distress during the life of the pavement.

2.3.2 Crack Width

Crack widths affect CRCP performance in several ways. For example, excessive crack widths may lead to undesirable conditions (such as infiltration of water) that later result in corrosion of the reinforcing steel. Incompressibles can also enter the cracks. Because the pavement is subject to contraction and expansion movements as well as deflections due to traffic loadings, this can lead to excessive stresses at the cracks that eventually lead to spalling. In addition, for wider cracks, there is less contact between the surfaces of the cracks, resulting in poor aggregate interlock. This results in increased slab deflections and stresses. Higher stresses in the concrete, in turn, lead to spalling, faulting, additional cracking, and punchouts.

Recent investigations have found that cracks that form within the first few days after pavement construction tend to be wider than cracks that form later during the life of the pavement.(16) One possible explanation for this behavior is due to the bond strength development at the concrete-steel interface. At earlier ages, the bond strength is weaker than at a later age. Therefore, the bond strength does not restrain the movement of the crack at earlier ages as it does later when this bond is stronger, and the restraint to movement is higher. This phenomenon results in tighter cracks. Furthermore, concrete drying shrinkage also increases with time after placement. Because drying shrinkage is one of the factors that govern the contraction of the concrete, the increase of drying shrinkage with time will also affect crack width. It has been observed that cracks that develop during the first few days are more affected by the remaining drying shrinkage than cracks that occur later when the remaining drying shrinkage is not as large.

A crack width of 0.6 mm has been found to be effective in reducing water percolation.(2) The AASHTO guidelines limit the crack width to 1 mm to avoid spalling and to limit water penetration.(17) For temperatures below freezing, crack width is not as much of a consideration, because frozen conditions do not permit the penetration of water.(16) A strong correlation between crack spacing and crack width has been found in some references.(2) It is believed that this correlation is due primarily to the fact that many of the mechanisms that influence crack spacing also influence crack width.(2, 19)

2.3.3 Steel Stress

The level of stress at which the concrete and the steel are subjected in a CRCP will influence its performance in the long term. As stated before, due to volume changes in the concrete, the reinforcement acts as a restraining system to keep cracks together. Consequently, significant stresses develop in the steel at the crack locations. The design of the steel reinforcement must consider the possible steel fracture and/or excessive plastic deformation. Usually, the stress at which the steel is subjected is limited to a reasonable percentage of the ultimate tensile strength to avoid steel fracture, and allowing only a small amount of plastic deformation.( 17, 20)

2.3.4 Factors Affecting Crack Spacing, Crack Width, and Steel Stress

It has been shown that crack spacing, crack width, and steel stress on CRCP are a function of the concrete tensile strength and the level of stresses that result due to the restraint to volume changes. Therefore, any factor affecting the concrete tensile strength, the pavement restraint, or factors contributing to volume changes will affect the cracking characteristics of CRCP, and therefore can influence its long-term performance.

Primary factors known to affect crack spacing and crack width in the early age include:

2.4 EARLY-AGE INDICATORS OF LONG-TERM PERFORMANCE COMMON TO JPCP AND CRCP

The early-age behavior of JPCP and CRCP is different in many respects. However, because both pavement types are made with portland cement concrete (PCC), they share some mechanisms that later lead to long-term distresses. Some of the primary common phenomena occurring in the early age of the pavement include pavement delamination and built-in curling. These two conditions are known to be early-age indicators of the future performance of the pavement in the long term, and are presented in the following section.

2.4.1 Delamination

2.4.1.1 Moisture Loss

Evaporation of bleed water at the concrete surface is governed by the concrete temperature and climatic conditions such as windspeed, relative humidity, and air temperature. However, at any point during the early age of the pavement, the loss of moisture in concrete pavements will also be a function of the moisture transport characteristics of the concrete, the water available in the mix, and in particular, the water that has not been used for hydration. Moisture gradients usually develop in concrete pavements due to the loss of moisture to the environment. Under conditions where significant moisture is lost to the environment, critical moisture gradients may develop in the slab. Typical moisture gradients resulting from excessive moisture loss usually present a sharp drop in moisture close to the pavement surface, as illustrated in figure below.

Figure 8.  Graph.  Moisture gradient resulting from excessive moisture loss.  Graph depicts relationship between Slab Depth (Y-axis, negative direction) and Relative Humidity (X-axis, from 0 to 100 percent), which is represented by a nonlinear trend decreasing from left to right.  The slab thickness is the distance from the X-axis to a dashed horizontal line.  Relative Humidity decreases sharply at the slab top.

Figure 8. Moisture gradient resulting from excessive moisture loss.

2.4.1.2 Strength Development

Because water available in the concrete mix is a determinant factor for the continuing hydration of the concrete, this hydration may be affected if excessive loss of moisture occurs. In addition, since the strength development is a function of hydration, the strength of the cement paste also may be affected due to moisture loss. Previous experience has indicated that the loss in moisture leads to reduced concrete strength, specifically at the surface. This reduction in concrete strength occurs because of the loss of strength gain in the cement paste, and in particular, at the interface between the paste and the aggregate. This interruption is due to the lack of water available for hydration.

2.4.1.3 Vertical Stress Development

Undesirable situations resulting from moisture loss are typically observed in the form of plastic shrinkage cracking, but also and most importantly, due to a radical change in the gradient of the moisture profile. Vertical tensile stresses and shear stresses develop as a consequence of the difference in shrinkage at different depths of the slab, as well as the pavement restraint to movement.

2.4.1.4 Formation of Delamination

In theory, whenever the tensile or shear stresses exceed the concrete strength, horizontal cracks tend to develop. The horizontal cracking that usually occurs close to joints or cracks and has typical depths of 13 to 76 mm is known commonly as delamination.(16) The depth of the delamination will be primarily a function of the evaporation rate, the type of curing, and the time to application of curing. Climatic conditions leading to high evaporation rates and poor curing procedures will also lead to significant reduction in strength and increased tensile and shear stresses in the concrete, producing deeper delaminations. On the other hand, climatic conditions leading to low evaporation rates and/or good curing procedures will typically result in shallower delaminations, and may even prevent delamination from occurring.

Other concrete delamination mechanisms commonly are cited in the literature. Delaminations at the steel depth have been observed in some CRCP in the past. This type of delamination is attributed to several factors. Some projects were paved in a two-stage construction process, placing steel on top of the first layer of concrete in plastic state, then constructing the second layer immediately following this. In some cases, delamination was observed due to a delay in placement of the second layer.(21) However, delamination of plain concrete pavements with no reinforcement also has occurred at middepth. In these cases, concrete placement has occurred during cold temperatures. This type of delamination has been attributed to sharp temperature drops, leading to excessive temperatures gradients in the slab, compounded with the restraint imposed by reinforcement/dowels at that location and traffic loads.

2.4.1.5 Spalling

Usually, the formation of delaminations is not evident in the early age; this depends on the subsequent climatic conditions, traffic loading, and depth. Delaminations may result in spalling distresses that will influence the performance of the pavement in the long term (see section 4.5). Delamination spalling is characterized by small pieces of concrete near cracks or joints that become loose from the pavement as a result of shear forces imposed by traffic loads, temperature fluctuations, and concrete fatigue characteristics. The amount of spalling distress and severity usually increases with time.

2.4.2 Built-In Curling

2.4.2.1 Temperature Gradients at Set

The temperature gradients during the early age are a function of the environmental conditions and the heat of hydration of the concrete. The initial thermal gradient in the slab at set will influence the curling shape of the slab.

2.4.2.2 Built-In Curling

Built-in curling is a term used for the curling state that develops at set and that later influences the curled shape of the slab as the thermal gradient is subsequently modified by the hydration process and climatic conditions.(22) The time-dependant relaxation properties of the concrete also affect the curled shape of the slab. As illustrated in figure 9, if the temperature throughout the slab depth was constant (thermal gradient at set = 0), the curling and warping state of the slab would be a function of the subsequent thermal gradient and the drying shrinkage of the slab. If the thermal gradient at time t is zero, the shape of the slab will only be a function of the drying shrinkage and the restraint conditions of the slab as imposed by its own weight and the subbase (figure 9). For this case, the slab would tend to warp up slightly due to the drying shrinkage. However, if the thermal gradient is positive (hotter at the top than at the bottom), then the curling of that positive thermal gradient would tend to counteract the warping due to the drying shrinkage (figure 10), possibly resulting in a flat shape. On the other hand, if the thermal gradient is negative, then the curling of the slab toward the top will be increased by the drying shrinkage (figure 11), resulting in a moderately upward shape.

 

Figure 9.  Diagram.  Effect of drying shrinkage and thermal gradient on curling and warping

Figure 9. Effect of drying shrinkage and thermal gradient on curling and warping
(thermal gradient at set = 0 and thermal gradient at time t = 0).

 

Figure 10.  Diagram.  Effect of drying shrinkage and thermal gradient on curling and warping

Figure 10. Effect of drying shrinkage and thermal gradient on curling and
warping (thermal gradient at set = 0 and thermal gradient at time t is positive).

 

Figure 11.  Diagram.  Effect of drying shrinkage and thermal gradient on curling and warping

Figure 11. Effect of drying shrinkage and thermal gradient on curling and warping
(thermal gradient at set = 0 and thermal gradient at time t is negative).

In general, when concrete sets, the temperature through the slab generally is not uniform, but rather is a function of the climatic conditions, the heat of hydration, and curing methods. Because there is a thermal gradient at set, the curling and warping of the slab at any time will be a function of that initial thermal gradient, the drying shrinkage, and the current thermal gradient of the slab. This is illustrated in figures 12-14 for a positive thermal gradient at set and in figures 15-17 for a negative thermal gradient at set.

Figure 12.  Diagram.  Effect of positive thermal gradient at set on curling and warping (thermal gradient at set is positive and thermal gradient at time lowercase T equals zero).  Diagram shows a text box (Thermal Gradient at Set is Positive) which flows to a slab with flat shape at set.  It curves slightly upward with the thermal gradient at time lowercase T equals 0; this plus the slightly upward shape due to drying shrinkage result in a moderately upward shape.

Figure 12. Effect of positive thermal gradient at set on curling and
warping (thermal gradient at set is positive and thermal gradient at time t = 0).

Figure 13.  Diagram.  Effect of positive thermal gradient at set on curling and warping (thermal gradient at set is positive and thermal gradient at time lowercase T is positive).  Diagram shows a text box (Thermal Gradient at Set is Positive) which flows to a slab with flat shape at set.  It stays flat with the thermal gradient at time lowercase T is positive; this plus the slightly upward shape due to drying shrinkage result in a slightly upward shape.

Figure 13. Effect of positive thermal gradient at set on curling and warping
(thermal gradient at set is positive and thermal gradient at time t is positive).

Figure 14.  Diagram.  Effect of positive thermal gradient at set on curling and warping (thermal gradient at set is positive and thermal gradient at time lowercase T is negative).  Diagram shows a text box (Thermal Gradient at Set is Positive) which flows to a slab with flat shape at set.  It curves significantly upward with the thermal gradient at time lowercase T is negative; this plus the slightly upward shape due to drying shrinkage result in a significantly upward shape.

Figure 14. Effect of positive thermal gradient at set on curling and warping
(thermal gradient at set is positive and thermal gradient at time t is negative).

For a positive thermal gradient, considering average temperature equal to the average temperature at set, whenever the subsequent thermal gradient becomes zero (figure 12), the pavement will tend to curl up due to the initial thermal gradient at set. This occurs due to the fact that the top fibers cool down from the initial temperature at set, and the bottom concrete fibers will warm up. This effect is accentuated by the warping generated due to the drying shrinkage. However, no curling is observed if the current thermal gradient is positive and similar to the thermal gradient generated at set (figure 13). For this case, the shape of the slab will only be a function of the drying shrinkage. The most critical case would be observed if the thermal gradient of the slab is negative, as illustrated in figure 14. For this case, the difference in temperature from the positive gradient at set will generate a significant amount of upward curling of the slab, which will be further accentuated by drying shrinkage.

A similar situation would occur for the case of negative thermal gradient at set as illustrated in figures 15-17. For a negative thermal gradient, the worst case would occur with a positive thermal gradient at time t as represented in figure 16.

Figure 15.  Diagram.  Effect of negative thermal gradient at set on curling and warping (thermal gradient at set is negative and thermal gradient at time lowercase T equals zero).  Diagram shows a text box (Thermal Gradient at Set is Negative) which flows to a slab with flat shape at set.  It curves slightly downward with the thermal gradient at time lowercase T equals 0; this plus the slightly upward shape due to drying shrinkage result in a possible flat shape.

Figure 15. Effect of negative thermal gradient at set on curling and warping
(thermal gradient at set is negative and thermal gradient at time t = 0).

Figure 16.  Diagram.  Effect of negative thermal gradient at set on curling and warping (thermal gradient at set is negative and thermal gradient at time lowercase T is positive).  Diagram shows a text box (Thermal Gradient at Set is Negative) which flows to a slab with flat shape at set.  It curves significantly upward with the thermal gradient at time lowercase T is positive; this plus the slightly upward shape due to drying shrinkage result in a slightly downward shape.

Figure 16. Effect of negative thermal gradient at set on curling and warping
(thermal gradient at set is negative and thermal gradient at time t is positive).

Figure 17.  Diagram.  Effect of negative thermal gradient at set on curling and warping (thermal gradient at set is negative and thermal gradient at time lowercase T is negative).  Diagram shows a text box (Thermal Gradient at Set is Negative) which flows to a slab with flat shape at set.  It stays flat with the thermal gradient at time lowercase T is negative; this plus the slightly upward shape due to drying shrinkage result in a slightly upward shape.

Figure 17. Effect of negative thermal gradient at set on curling and warping
(thermal gradient at set is negative and thermal gradient at time t is negative).

2.4.2.3 Pavement Smoothness, Faulting, and Cracking

It has been observed that built-in curling may translate later into faulting and cracking problems as the level of stresses developed in the pavement slabs is increased as a function of traffic loads and long-term daily and seasonal climatic conditions. This will affect further the curling shape and the support conditions under the pavement. For positive thermal gradient at set, it is believed that the pavement is more susceptible to corner cracks (figure 14), while for negative thermal gradient at set, the pavement would be more susceptible to midslab cracking (figure 16). The effect of drying shrinkage suggests that built-in curling would be more critical for stresses due to corner cracking than for midslab cracking. In addition, under certain circumstances, built-in curling could lead to a more pronounced curling shape of the slab, which may influence the initial pavement smoothness and, therefore, its riding quality during the service life.

 

Previous | Table of Contents | Next

Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000
Turner-Fairbank Highway Research Center | 6300 Georgetown Pike | McLean, VA | 22101