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

Skip to content
Facebook iconYouTube iconTwitter iconFlickr iconLinkedInInstagram

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

This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-HRT-04-150
Date: July 2006

Chapter 4. Cracks

Previous | Table of Contents | Next


Cracks in HCC may have several causes (see ACI 201.1R and ACI 224.1R): plastic shrinkage, settlement, drying shrinkage, thermal stresses, chemical reactions, weathering (freezing and thawing, wetting and drying, heating and cooling), corrosion of reinforcement, poor construction practices (e.g., retempering), construction overloads, errors in design and detailing,and externally applied loads. For our purposes, nine types of cracks are discussed:

  1. Microcracks.
  2. Crazing.
  3. Scaling.
  4. Cracks caused by insufficient air-void content.
  5. Cracks caused by delamination at the reinforcement.
  6. Cracks caused by a chemical reaction.
  7. Cracks caused by drying shrinkage.
  8. Cracks caused by thermal volume change.
  9. Cracks caused by plastic shrinkage.


4.2.1 Microcracks

The term microcracks includes all very fine cracks, in any direction, at the surface or within the mass of the concrete that are not easily visible with the unaided eye, but may be seen with a magnifying glass or microscope. Microcracks are often extremely difficult to observe on the textured surface of a placement. Chapter 8 provides instructions on the observation and recordingof microcracks.

4.2.2 Crazing

Crazing is a fine, very shallow pattern of cracking that occurs in the exposed surface of a concrete placement. Usually it cannot be seen while the observer is in motion in a vehicle. It occurs when the thin surface layer shrinks significantly relative to the interior mass of the concrete and may result from excessive paste or water content in the surface layer or rapiddrying. Crazing usually has a very fine pattern, with the individual uncracked central portions usually no more than 50 mm across. The depth of craze cracking should be recorded. Paste-rich surface layers can also impede the rise of bleed water, trapping it just beneath the surface. If the surface layer flakes off, usually to a depth of about 3 mm, it may be called paper scaling. The underside of any loose flakes of concrete should be examined for clean sand grains and calciumhydroxide that would indicate trapping of bleed water and the casts of ice crystals that would indicate freezing before hardening.

4.2.3 Scaling

Scaling is a local flaking or peeling of the mortar at the surface and can be classified as light (without exposure of coarse aggregate), medium (5 to 10 mm deep), severe (11 to 20 mm deep),or very severe (> 20 mm deep) (ACI 201.1R). Scaling is usually attributable to a weak surface layer exposed to the physical effects of freezing and thawing cycles or deicing chemicals.

4.2.4 Cracks Caused by Cycles of Freezing and Thawing While Concrete Is Critically Saturated

Concrete that lacks either sufficient strength (maturity) or an adequate air-void system will develop laminar cracking if critically saturated and exposed to freezing and thawing cycles (see figures 26 and 80). The deterioration will usually proceed from the exterior inward as progressive scaling.

4.2.5 Cracks Caused by Reinforcement Corrosion

Corrosion and the resulting expansion of the reinforcement will cause lateral cracking in the plane in which the reinforcement is situated. Normally, the high pH of concrete protects the reinforcing steel from oxidation. The passivating effect of concrete on steel can be negated by the intrusion of chloride ions or by carbonation of the paste surrounding the reinforcement. The concrete between the reinforcement and the outer surface of the element thus serves as a barrier to the ingress of chloride ions or carbon dioxide. Construction plans should specify the concrete cover thickness, usually around 50 to 68 mm. Cracks extending from the exposed surfaces toward the level of the reinforcement may significantly decrease the effectiveness of the cover concrete as a barrier, especially if the cracking is over an extensive area. The concrete is considered to have only the depth of protection that exists from the bottom of the cracks to the reinforcement (see figures 27 and 28).

4.2.6 Cracks Caused by a Chemical Reaction

Chemical reactions may take place between concrete constituents (the aggregates and the paste) and ionic species in solution within the concrete mass. Cracking resulting from AAR is discussed in chapter 10.

4.2.7 Cracks Caused by Drying Shrinkage

The term drying shrinkage is commonly used in connection with shrinkage cracking that occurs after the HCC has attained final set and a particular degree of bonding has developed between the aggregate and the paste. After the concrete has reached final set, the paste behaves as a brittle material.

Most HCC shows evidence of drying shrinkage. Cracking can be minimized by good workmanship, proper proportioning of the mixture, and sufficient jointing performed soon after hardening. In jointed concrete, uncontrolled cracks may form if the joints were not formed early enough, are not working properly, or the shrinkage in the hardened state is excessive. In continuously reinforced HCC, very narrow, fairly regularly spaced transverse cracks are expected to form. The cracks allow for the changes in volume of the concrete with drying and varying temperatures.

4.2.8 Cracks Caused by Thermal Volume Change

Thermal effects on concrete volume can cause cracking with a disposition similar to that caused by drying shrinkage and, in fact, thermal and drying effects will often occur in concert. The coefficient of thermal expansion (CTE ) of concrete is a proportional function of the CTE of its constituent materials. Since common aggregate materials differ considerably in their CTE , they consequently exert considerable influence on the concrete CTE (Scanlon and McDonald, 1994; Lane, 1994).

Thermal cracking is a predominant concern in mass concrete placements if the heat of hydration is not controlled and large temperature differences are allowed to develop between the inner core and cooling outer skin of the placement. In thin members, such as pavements and bridge decks, thermal cracking is most likely to present problems when the concrete undergoes large temperature swings during the first several days after placement. Such effects are most pronounced in heavily reinforced structures such as continuously reinforced concrete pavementand decks; however, they can present serious problems for plain jointed pavements if midslab cracks result. HIPERPAV software, published by FHWA, provides a means to predict potential thermal cracking problems for pavements (FHWA, 2003).

4.2.9 Cracks Caused by Plastic Shrinkage

Plastic shrinkage is a form of drying shrinkage that occurs while the HCC is unhardened and malleable, and the bond between the components of the material is very weak. Plastic shrinkage cracking is caused by excessive evaporation of the water at the concrete surface because of improper curing procedures for a concrete under the climatic conditions existing at the time of placement (see appendix B and Lerch, 1957; Price, 1982; and Scott, Lane, and Weyers, 1997). Plastic shrinkage occurs in the first few hours after placement before final set (when the rate of evaporation exceeds the rate of bleeding). At this stage, the HCC has some properties of a solid, but no appreciable bond exists between the aggregate particles and the cement paste. This sort of cracking is also called early cracking or morning cracking, the latter because it is often the concrete that was placed in the morning and then exposed to afternoon drying conditions that exhibits this type of cracking (see figures 33 through 40).

Plastic shrinkage cracking often occurs in high-quality HCCs when curing is not promptly or adequately applied and appropriate measures to avoid excessive evaporation have not been taken. The intrinsic quality of the concrete is not necessarily adversely affected by plastic cracking. Provided the concrete was adequately moisture cured, the defect is more cosmetic in nature, except in situations where the concrete is intended to provide protection for the reinforcement.

A form of early cracking where the cracks are located directly over the steel can occur when the depth of cover over the reinforcement is shallow (Price, 1982; Dakhil, et al., 1975). This cracking is directly related to shrinkage and settlement of the concrete over the steel as the bleed water leaves the concrete and the volume of the paste is diminished. It may be accompanied by segregation of the coarse aggregate from the paste. As with HCCs with a lower w/cm, this condition is exacerbated by drying atmospheric conditions. Evidence of this type of fault can be seen on interior surfaces of specimens of these HCCs. This type of cracking is commonly called settlement or subsidence cracking (ACI 224R).

Occasionally, there may be severe bleed channels in HCC that might be confused with plastic shrinkage cracking. However, upon close inspection, such bleed channels show clear evidence of being a waterway, whereas plastic shrinkage cracks show clear evidence of having been pulled open by the shrinkage of the concrete. Bleed channels are trains of water voids caused by the upward movement of mixing water as the fine solids settle. Such trains of voids may occur in HCCs with a high w/cm. Some forms of plastic shrinkage cracking may have zones of such void trains. In HCCs with a low w/cm, incomplete consolidation may cause fold lines and collections of voids. Cracks occurring in HCCs can usually be distinguished by the nature of the associated voids and the appearance of the fracture surfaces (showing signs of either brittle or plastic deformation) (see section 4.3).

The following may occur in unusual and extreme cases of rapid evaporation, causing drying of the HCC before hardening:

These are features of extreme drying and are not usually present in HCCs affected by plastic shrinkage cracking. The absence of these features in HCC exhibiting plastic cracking does not indicate that it was not subjected to deleterious drying conditions during or shortly after placement.


4.3.1 Overview

For transportation departments and other purchasers of concrete placed by a contractor (or other agency), it is frequently important to distinguish between early plastic shrinkage cracking and the brittle cracking that may occur because of structural stress or later drying shrinkage. The contractor is obligated to prevent the dehydration (and consequent cracking) of the fresh concrete that can occur when wind, low humidity, or both promote rapid drying. HCC with a low w/cm (0.40 or less) and latex-modified concretes are more apt to suffer this sort of failure than are concretes with a high w/cm ratio. When it can be shown that the cracking is caused by a failure of the contractor to refrain from placing concrete during unfavorable weather conditions, employ sufficient methods to prevent drying, or both, the contractor may be obligated to provide a new surface or accept a lower payment.

The usual rule of thumb in the literature is: If the cracks go through the aggregate particles and cause them to break, the cracking should not be considered plastic cracking. Many observers call any crack that goes around the aggregate particles a plastic/early shrinkage crack. This can be in error. There may be other reasons for the crack to go around the aggregate particles.

Some aggregates are more fragile than others and some may crack during final placing and finishing. It is possible to judge a crack to be a later crack on the criterion of broken aggregate when, in reality, the aggregate particle just happened to have a zone of weakness in the crack plane and the crack was an early crack. The specimen examined (a cross section of a crack) is a very small portion of a crack. The interior of the crack surface is a very small, nonrepresentative portion of the crack.

Cracks that preferentially go around aggregate particles indicate that the bond between the aggregate and paste was a weak point at the time the cracking occurred. The bond may be weak because of any of the following:

Thus, it is necessary that the criteria for deciding that a specific crack is a plastic shrinkage crack be more than the fact that the crack skirts the aggregate particles.

4.3.2 Analogy With Clay Materials

The difference between drying shrinkage cracking and plastic shrinkage cracking can be explained further by use of an analogy using clay materials.

Hard brittle materials (drying shrinkage cracking): Consider a broken ceramic object (pottery) or rock. If all the pieces can be found and fitted back together, the material will solidly fill the same space as did the original object. Any internal voids will almost invariably be recognizable by either their shape or the nature of the interior surface (different from a fracture surface). If all the chips are available, the expression of the crack on the surface of a hard brittle material will be a thin sharp line.

If the material is sandy clay that was fired in a kiln ("fired" is analogous to the hardening of concrete), all the cracked surfaces will fit back together if all the fragments are preserved and there are no air pockets present. If the bond and the tensile strength of the ceramic are as strong as the tensile strength of the sand, then the crack will fracture the sand and the crack surfaces will neatly and completely fit back together. If the sand has much greater tensile strength than the fired ceramic, any cracks occurring in the material are likely to detour around the sand grains.

This logic may be directly extrapolated to cracks in hardened HCCs, hardened latex-modified HCC, and many other highway materials. In the case of these materials, the general type of void and the nature and luster of the interior of the voids should be carefully studied so that they can be recognized in the path of any crack under study.

It may be that a macrocrack occurring on the riding surface of HCC that was originally a thin sharp line was worn wider by the abrasive action of traffic. The course of the crack below the surface expression should be examined. Care must be taken to extrapolate the evidence in light of all relevant facts, including the age of the placement, the amount of traffic, the pertinent weather conditions, and the strength and general condition of the concrete placement.

Malleable materials (plastic shrinkage cracking): Now, consider a crack caused by the pulling apart of a piece of modeling clay or other such material. Because of the plastic nature of the material, there may be small "bridges" of the material spanning the crack, there will be deformation of the sides of the crack, and the two sides of the crack will not be able to be fitted back together without reshaping of the crack walls. A crack of this nature starts at the exposed surface. In the case of unhardened HCC, this is the driest portion. Here, the crack is the widest because the surface is the origin and tension is greatest as there is no adhesion above to resist the pulling apart. The edges of the crack are often rounded back. If the material is ceramic clay and the material was baked in a kiln in the cracked condition, it would be obvious in the finished piece that the two interior surfaces of the crack would not fit back together. If the material is malleable sandy clay, then the crack will go around the sand and the crack surfaces will show that deformation occurred while the material was plastic.

This is entirely analogous to the situation in plastic shrinkage cracking in HCC. The crack occurs while the material is plastic and is then "baked" (the shape is preserved) by the continuing hydration of the cement and the complete hardening of the HCC.

4.3.3 Procedures

Distinguishing between drying shrinkage cracking and plastic shrinkage cracking is a five-step procedure, as shown in table 12.

Table 12. Procedure for distinguishing between plastic and drying shrinkage cracking.
  1. Study the intersection of the sidewalls of the specimen with the wearing surface.
  2. Study the wearing surface of the specimen.
  3. Study the expression of the crack on all cut, cored, and lapped surfaces, and focus down into the crack.
  4. Examine the interior surface of the crack.
  5. If necessary, study the types of voids present in other areas of the HCC and compare their surface with the various types of surfaces within the crack.
  1. Study the intersection of the sidewalls of the specimen with the wearing surface: Plastic shrinkage cracking may occur before or just after the surface is screeded and the wearing surface texture is created. In these cases, it may happen that one or more of the specimens submitted may show that mortar was worked into the top of the crack. Sometimes portions of the plug of mortar will have fallen out or been chipped off the specimen during coring and handling. It is clear that the plug is a definite indication of early plastic shrinkage cracking (see figure 33).

    If the cracking occurred just before texturing, the creation of the wearing surface would naturally work mortar into the crack (see figure 34). If cracking occurred after texturing and was observed by the workers, a deliberate effort may have been made to correct the flaw.

    Figure 33. Plastic shrinkage cracking was covered up by mortar filling over it on top of a 100-mm core.

    Plastic shrinkage cracking was covered up by mortar filling over it on top of a 100-millimeter core.  The top section of a broken concrete core is shown with areas of finer mortar material noticeable near the surface. The mortar was worked into surface plastic shrinkage cracks while the concrete was still plastic. These weaker areas of the hardened concrete cracked again. A near-vertical crack is shown with a crack branching off at a lower angle. Chunks of the top surface have broken off.

    Mortar was worked into the top of this set of cracks before the concrete hardened. Later, the concrete cracked in the same area. Because of the lack of interlock, the new crack follows the mortar boundaries. At the right side of the picture is a depression in the surface where a mortar plug was lost before the photograph was taken. (This specimen had to be glued together to enable the photograph to be made.)

    Figure 34. Plastic shrinkage cracking occurred before a surface texture was formed.

    The top section of a concrete core is shown with a crack and void noticeable below a surface layer of mortar leaving a void about 10 millimeters below the finished concrete surface.

    The defect became hidden by the finishing procedures that pushed unhardened mortar over the crack.

  2. Study the wearing surface of the specimen: With the microscope, look for features that will define the crack. Check for any features that will indicate the nature of the cracks. If photographs can be obtained of the wearing surface just after cracking but before the surface is worn by any traffic, small bridges of material may be seen stretching across the crack, which will indicate plastic shrinkage cracking (see figures 35 and 36).
  3. Study the expression of the crack on all cut, cored, and lapped surfaces, and focus down into the crack: If you see any little bridges of paste that tie the two sides together or if any uncracked aggregate particle bridges the crack, the crack plane existing all around the particle, and the aggregate firmly attached to the hardened paste on either side, you can conclude that the crack was caused by plastic shrinkage (see figures 37 through 39). Bridges of paste or aggregate are not found in all plastic shrinkage cracks. The defining feature is the fact that the sides of the cracks have been distorted by the tensile forces because of shrinkage that occurred while the HCC was still malleable.

    The greater the difference between the width of the crack at or near (beneath any mortar plug) the wearing surface and the width of the crack at its deep end, and the more quickly it tapers to nearly nothing, the greater the likelihood of the crack having been caused by plastic shrinkage. The V-shape of the crack is part of the distortion of the edges of the crack caused by the tensile forces. Commonly, the crack is a plastic shrinkage crack near the surface; however, with depth, the crack is straighter and the zone of weakness has been extended after the final set by other forms of drying shrinkage or by structuralstress, sometimes completely through the slab.

    The path of a plastic shrinkage crack on any vertical concrete surface is usually quite tortuous because it is deflected by each piece of aggregate. Typically, the crack is wider at the top than at the bottom; however, even at the bottom, it is not the sort of crack found in a brittle ceramic (see figure 40).

    Do not be confused or alarmed by the presence of abundant road dirt or mud created by the drilling of the cores. This mud may appear layered and flaky, but the structures will be fragile.

    Figure 35. Plastic shrinkage cracking occurred in a latex-modified concrete overlay (top view of core (small threads of latex span the crack)).

     The top view of a core section taken through the exposed top surface of the concrete showing an irregular plastic shrinkage crack with small threads of latex spanning the crack

    Rules are graduated in millimeters.

    The crack shows the distortion that accompanies tension cracking in an unhardened medium. This specimen was collected before traffic destroyed these delicate structures.

    Figure 36. Plastic shrinkage cracking: Small bridges of latex paste connect sides of crack.

     Vertical cross section cut down through the core shows the irregular crack going down about 25 millimeters below the surface of the concrete.

    Rules are graduated in millimeters.

    Figure 37. Small bridge of paste across a crack caused by shrinkage that took place while the ordinary portland cement concrete was not hardened (100-mm core, road surface at right).

     Photo shows a crack on the surface of a 100-millimeter core. The crack formed during the concrete’s plastic stage.

    Rules are graduated in inches.

    These little bridges may occasionally be found on broken, sawed, cored, or lapped surfaces.

    Figure 38. Bridge of paste across plastic shrinkage crack: Closeup of bridge.

      Closeup of bridge. A closeup of the crack shows a bridge of paste between its two sides.

    Rules are graduated in inches.

    Figure 39. Tortuous path of a plastic shrinkage crack in a concrete that has not completely separated.

       Photo shows a section through hardened concrete with the tortuous path of a plastic shrinkage crack. In this case the crack has not completely separated, showing how bridges of material form as the plastic concrete is stretched

    It is easy to envision how little bridges of concrete form when a crack system of this nature is stretched further. Narrow microcracks are emphasized with ink. The specimen is approximately 100 mm top to bottom.

  4. Examine the interior surface of the crack: Delay any destructive procedures until the total surface has been studied and any air-void determinations have been made. Then, break the HCC with a hammer blow in an area away from the crack and examine the surface produced. In the most common case of a crack produced in an area not including a plastic shrinkage crack, the interior surface will be made up of a multitude of tiny angular planes with the exception of the interior of the air voids. There may be an occasional rounded surface of an aggregate particle or of the cast of an aggregate particle; however, the major portion of the surface will be a collection of planes.

    Break the HCC on the crack and examine the interior surface. Usually the interior surface of a crack produced by plastic shrinkage is made up of very tiny globules of cement paste. The surface of the paste inside the crack does not look like a cracked surface. Rather, the paste appears, as one might imagine it would, as if, when the sides of the crack were no longer in contact, minute stringers of paste had momentarily bridged the crack and then, as the tensile strength of the stringers was exceeded, the stringers coalesced into tiny globules. Each globule appears as if a paste stringer shrunk in on itself. If it is water-worn sand, none of the sand grains will be broken and they will present a naturally rounded surface. A few fragments of the coarse aggregate may be broken, as will occur in all concrete construction. If the aggregate is shaley or naturally fragile and contains a plane of weakness approximately parallel with the plane of the plastic shrinkage crack, the crack may preferentially traverse this aggregate.

    Figure 40. Plastic shrinkage crack on a lapped surface (wearing surface is at the top of the photograph).

    Photo shows a typical plastic shrinkage crack that is wider at the concrete surface and has an irregular shape as the crack becomes narrower toward the bottom of the vertical cross section shown
  5. If necessary, study the types of voids present in other areas of the HCC and compare their surface with the various types of surfaces within the crack.


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