Concrete Cracks: ¿A Shrinking Problem?

By Scott M. Tarr

While often overlooked, the shrinkage potential of a concrete mixture is perhaps the most important consideration for concrete used to construct industrial floors on ground. All concrete, even shrinkage-compensating concrete, shrinks. Because of a loss in volume, concrete shrinkage can lead to cracking when base friction or other restraint occurs. Shrinkage also causes curling/warping which can lead to a variety of slab issues including decreased load-carrying capacity (structural cracking) and joint stability problems such as spalling. Shrinkage-induced cracking and curling/warping can also contribute to decreased vehicle ride (a potential health and safety issue) and floor covering failures (buckling caused by reverse warping). In short, shrinkage is a factor in most concrete slab-on-ground performance issues. To increase the performance of concrete floors, shrinkage must be better understood and addressed.

Material Handling Equipment has Changed

Concrete shrinkage has become an increasingly important design issue with industrial floors. One of the main reasons for this is the evolution of the material distribution industry. As the industry has evolved and the market has become increasingly competitive, the efficiency of the operation has become an important consideration. Material handling equipment has been developed to move more product at a faster, yet still safe, pace. Racking equipment can safely hold more load and is increasing in height. However, the size of baseplates has not increased substantially, which results in greater contact pressures and flexural stresses. 

Another feature related to the evolution in material handling which influences concrete floors is the wheel design for lift truck vehicles. Wheels have evolved from large pneumatic tires to small solid castors which increase the vehicle stability. However, these wheels apply significantly greater contact pressures on smaller footprints which increases the importance of joint stability in order to maintain the effectiveness of joint fillers to resist joint deterioration. Industrial slab design must become more sophisticated in order to service these modern efficient operations.

Concrete Mixtures have Changed

In addition to the changes in the material handling equipment, the shrinkage of concrete mixtures typically used for slabs on ground has increased over the past couple decades. There are many factors which contribute to this increase such as high paste content and the increasing use of admixtures. Also, the availability of good quality aggregates is diminishing which has forced the use of aggregates with properties which promote higher shrinkage concrete mixtures. Another significant factor is the increasing demand for fast-track construction. Many mixtures with rapid setting and strength gain performance have an increased shrinkage potential.

Causes of Concrete's Volume Changes

The shortening of concrete slabs can be caused by temperature decreases or moisture loss. These two causes are also related to curling and warping of slabs, respectively. Curling is the deformation of the slab due to a difference in temperature between the surface and the bottom of the slab (temperature gradient). Like most materials, concrete expands and contracts with a change in temperature. If the slab surface is cooler than the slab bottom, the surface contracts causing the slab edges to curl upward. Slab "warping" is the deformation of the slab surface profile due to a difference in moisture between the surface and bottom of the slab (moisture gradient). As with a sponge, if the slab surface is allowed to dry and the bottom is kept moist, the edges will tend to warp upward. Exterior pavement slabs typically have a permanent upward edge warp and experience curling on a daily basis due to surface warming and cooling cycles related to exposure to the sun. In general, the edges of interior concrete floor slab panels warp upward due to a moisture gradient through the slab depth.

Temperature Contraction. Concrete has a coefficient of thermal expansion and contraction of about 0.0000010 mm/mm/°C (0.0000055 in./in./°F). After hardening, concrete will contract as a result of cooling after the peak heat of hydration (typically coinciding with drop in ambient temperature at night). A 22°C (40°F) drop in temperature between day and night can cause about 0.8 mm (1/32 in.) of contraction in a 3 m (10 ft) length of concrete. This magnitude of shortening is sufficient to cause cracking, especially at early ages when the strength of the concrete is low. Therefore, it important to consider the potential for early age temperature drops and install contraction joints properly to control the location of cracking. Keeping concrete placement temperatures as low as possible and close to the ultimate operating temperature of the facility [target 26°C (80°F)] helps to decrease the magnitude of the temperature drop.

Drying Shrinkage. After hardening, concrete begins to shrink as water not consumed by cement hydration leaves the system. This is known as drying shrinkage. Water above that necessary to hydrate cement is required for proper workability and finishability – the water is called "water of convenience." In general, the higher the additional water content, the higher the shrinkage potential. For small, unrestrained concrete specimens (prisms), a low ultimate shrinkage (strain) is considered to be less than 520 millionths (at 50% relative humidity and 23°C [73°F]). Typical concrete shrinkage has been measured at 520 to 780 millionths. However, for some mixtures, shrinkage exceeding 1100 millionths has been documented. Using concrete with a higher drying shrinkage increases the risk of problems with the floor performance.

Testing. Concrete drying shrinkage can be measured in the laboratory. The American Concrete Institute (ACI) 302 Guide for Concrete Floor and Slab Construction recommends following ASTM C 157Standard Test Method for Length Change of Hardened Hydraulic-Cement Mortar and Concrete. However, as drying shrinkage requires a period of time to complete (typically over 6 months for lab specimens and 12 to 18 months for field slabs), it is unlikely that the testing can be completed within typical pre-construction schedules. ACI 360 Design of Slabs-on-Groundsuggests following the procedure in ACI 209 Prediction of Creep, Shrinkage, and Temperature Effects in Concrete Structures which includes a method of predicting the ultimate drying shrinkage from early-age data measured at, for example, 14 or 28 days. While the accuracy of the prediction improves with later-age test data, the test can be used to compare potential mixtures as well as confirm that the design of the slab and joint spacing are appropriate for the concrete available. The design may need to include mechanical load transfer devices in joints, steel reinforcement, or, in cases where only high shrinkage concrete is available, the use of post-tensioning or shrinkage-compensating concrete.

As with any comparative test, the procedure for measuring shrinkage in the laboratory must be duplicated exactly in order to compare the results. Exact mixture proportions and resulting slump should be reported for the batch used to fabricate test specimens. For instance, if the project specifications allow a 125 mm (5 in.) maximum slump, the test specimens should not be fabricated using concrete with a 75 mm (3 in.) slump. The specimens should also be the appropriate size as dictated by the maximum size of the coarse aggregate. Some additional items to consider when analyzing lab results include:

• method of batching and mixing concrete


• method of consolidation


• method of curing


• length of curing (0, 7, 28 days)


• storage conditions subsequent to curing


• method of measuring beams


• point of "zero" measurement (length of mold, initial beam length, length after soak).

Considering the potential differences between laboratories, it is an obvious advantage to have the same lab perform all testing on the trial mixtures being considered. That way, even if the standard test method is modified, the same procedure is still followed for comparative mixtures. For slabs on ground, reasonable modifications to ASTM C 157 test may include altering the curing conditions from a 28-day lime bath soak to a 7-day soak or moist cure since this may better reflect actual construction conditions. However, since the specimens have a relatively large surface area in comparison to their size, care should be taken to prevent drying during the curing period to accurately simulate anticipated construction conditions. After curing, the specimens should be stored in air in a strictly controlled drying environment as required by the standard procedure.

Laboratory shrinkage test results can be used to consider the potential for joint/crack widening of the floor. However, concrete in a slab on ground shrinks less than that measured in the laboratory. There are several reasons why this occurs. First, unlike the laboratory samples, a slab is restrained by the subgrade. Subgrade drag coefficients are well-documented for different types of subbase material. Another reason slabs shrink less than lab samples is because a floor in contact with the ground does not dry the same amount as small beams exposed to 50% RH on all surfaces. Typically, slab moisture is very high (95% RH or greater) at the bottom. Even when ambient air conditions are substantially below 50% RH, the slab interior remains at a significantly higher moisture condition, especially when a vapor retarder is not located immediately beneath the slab. The surface may dry but the bottom remains moist which is why warping occurs. 

Joint/crack widening is related to this moisture gradient as well. Therefore, joints and cracks are typically much wider at the floor surface than at the bottom. As a result, predicting, or even measuring joint width at the slab surface is an ineffective method of determining joint load transfer capability. In general, considering sub-slab frictional restraint and actual in-place moisture condition, the shrinkage of typical concrete (520 to 780 millionths) is actually closer to 100 to 300 millionths (or even less at the slab bottom). Experience with the performance of locally-available concrete mixtures in floor slabs can provide good information to consider anticipated joint performance.

Effect of Concrete Ingredients on Shrinkage

Many researchers¹ have studied the factors associated with the shrinkage of concrete mixtures. The most influential factor is the type of coarse aggregate used. Hard, dense aggregate is able to restrain the shrinkage of the cement paste. In contrast, using aggregate with a higher compressibility can increase the shrinkage of the concrete mixture by about 120 to 150 percent. Therefore, locally-available materials play a critical role in the shrinkage behavior of the concrete. The properties of aggregate from various quarries should be considered if shrinkage is to be minimized. Some recommendations include using a large topsize aggregate and optimizing the gradation of the aggregate and combining aggregate sources to minimize gap-grading and corresponding paste content of the concrete. However, the overall benefit of these suggestions is dependent on the aggregate properties used. If the aggregate is of poor inherent quality, maximizing the size, gradation, and content may have little effect on the concrete shrinkage. Likewise, blending a large aggregate with poor qualities to a mid-size aggregate with good properties may increase the resulting shrinkage behavior of the concrete mixture.

Other factors that have been found to have a significant impact on the shrinkage of concrete mixtures includes the use of shrinkage-promoting admixtures (such as accelerators), the use of dirty aggregate which increases water demand and using a cement with high shrinkage characteristics. The cumulative effect of these factors has been found to be multiplicative and not additive. So, combined factors can easily increase concrete shrinkage by several hundred percent. Therefore, the specific impact of any set of materials should be determined by laboratory testing.

The shrinkage of a concrete mixture can have a significant impact on the performance of floors on ground. With the increasing demand for structural load-carrying capability and corresponding floor performance, shrinkage has become a growing issue. As repairs and maintenance can be costly, good joint performance is essential for industrial concrete floors. Therefore, it is important for slabs to remain in contact with the supporting base (minimal warping) and the joints to have minimal widening. The shrinkage potential of the concrete mixture must be well-understood so proper design and construction methods can result in the expected long-term serviceability for the owner. Even for commercial floors where heavy loading does not occur, concrete shrinkage can result in warping relaxation subsequent to installation of floor coverings or coatings. When surfaces are reprofiled in preparation for installation of floor finishes, slab distortion can result in delamination and buckling of the flooring. Knowledge of potential concrete shrinkage can help minimize such problems. 

The information above and further discussion of this topic is included in the book Concrete Floors on Ground(EB075) by the Portland Cement Association.

 

1. Tarr, Scott M., and Farny, James A.; Concrete Floors on Ground, EB075, Fourth Edition, Portland Cement Association, Skokie, Illinois, USA, 2008, 256 pages.