Click to expand
Click to expand
  • Concrete shrinkage is often misunderstood, even among those with substantial expertise in the design and construction of concrete slabs on ground. The drying shrinkage potential of concrete, though, is a major factor in the most common kinds of distress in concrete floors.

    Problems directly stemming from concrete shrinkage include cracking, warping (or curling), and joint spalling. Since these are the most common complaints with industrial concrete floors, ACI 302.1R “Guide for Concrete Floor and Slab Construction,” includes the following advice:

    Because minimizing shrinkage is of prime importance, special attention should be given to selecting the best possible concrete mixture proportions. The shrinkage characteristics of a concrete mixture can be determined by ASTM C 157. Should it be necessary to determine if a proposed mixture has other than normal shrinkage, the proposed mixture should be compared to the specified or a reference mixture using ASTM C 157.

    PCA’s Concrete Floors on Ground includes the following guidance:

    Too often, specifications for concrete floors only include requirements for compressive strength and slump. However, many different concrete mixtures with widely variable performance can satisfy given strength and slump requirements. Other characteristics should be considered when specifying requirements or submitting a mixture for approval. Factors such as workability, finishability, and shrinkage performance are extremely important to consider when anticipating the long-term serviceability of concrete slabs on ground.

    Workability and finishability are improving through efforts to optimize the aggregate gradation by blending several aggregates. But this does not necessarily result in low shrinkage concrete. Although industry guidelines stress the importance of the shrinkage potential, there is confusion about why it’s necessary to measure, who it’s helpful for, and how it can be done efficiently prior to construction. A common misconception is that the test takes too long to get useful information prior to placement. Another common question is “What can be done once we have the information?” This article is intended to bring some clarity to the test procedure and its usefulness to the design and construction of concrete floors on ground.

    The impact of shrinkage

    Shrinkage is a factor in most common problems with concrete slabs. Many times, slab problems have a substantial negative impact on the efficiency of the facility operation, which increases the cost of running the facility.

    Cracking. Most people can readily recognize “shrinkage cracks,” which are due to the restraint of concrete shrinkage. The restraint is typically provided by friction between the bottom of the slab and the base, and the amount of shrinkage is controlled by the concrete mixture and ambient conditions. Sawcut joints installed at the proper depth, time, and spacing encourage cracks to form beneath the sawcuts and out of sight. These joints should be cut to a depth of ¼ the slab thickness within 8 hours after final finishing.

    But what about spacing? ACI Committee 360, Design of Concrete Slabs on Ground, includes a chart (See Fig. 1) that helps determine the maximum joint spacing. Joints are typically recommended on column lines to relieve the restraint caused by column penetration. The chart helps determine how many joints are needed between columns. This chart is based on the shrinkage potential of the mixture and the spacing recommendations are for concrete with low, typical, or high shrinkage. The concrete shrinkage potential should be known in order to specify the appropriate joint spacing. If unknown, the shortest joint spacing should be used.

    But using the lowest joint spacing isn’t always the best choice. Adding joints between columns increases the cost of construction and maintenance. Often, the decision comes down to being slightly greater than the maximum spacing recommended or being well below what the chart suggests. Knowing the concrete shrinkage potential can allow the designer to specify the necessary joint spacing to minimize the risk of random cracking.

    Warping. Cracking can also be the result of differential drying shrinkage between the top and bottom of the slab. Contractors can meet a specified floor flatness and levelness measured immediately after finishing. But, depending on the shrinkage potential, the slab may not stay flat. Because the concrete at the exposed surface dries more than the concrete lower in the slab, the top shrinks more than the bottom causing the slab edges at joints to warp (or curl) upward. The difference in shrinkage depends on the ambient conditions and on the shrinkage potential of the mixture. If the shrinkage is low, the relative difference and corresponding warping will be less than if it’s high. The use of reinforcement or dowels has very little impact on warping magnitude.

    The magnitude of warping is important since large warping results in no support at the slab edges along joints, which can result in excessive internal and load-induced stress. The stress can result in cracking and corner breaks. Even when cracking doesn’t occur, excessive warping can impact ride quality and joint performance.

    Joint Spalling. Perhaps the most common distress of industrial concrete slabs is joint spalling, which greatly accelerates wear of the material handling equipment in the building. Most people don’t understand the connection between joint spalling and shrinkage.

    Joint spalling occurs when hard-wheeled lift trucks cross joints that are not supported by semi-rigid joint filler. This filler is typically installed about 90 days after slab placement. Unfortunately, concrete continues to dry for 12 to 18 months. As drying continues the concrete shrinks and the joints widen. The filler cannot stretch within the widening joint and separates from the joint wall. Once separation occurs, the filler can no longer support the joint edges against spalling under hard wheels. I recommend that the joints be refilled a year after placement to restore the support of the joint. In fact, since this is an anticipated occurrence, ACI Committee 301, Specifications for Structural Concrete, has a new section on industrial floor slabs, which requires the second joint filling to be included in the original bid, allowing the contractor to get paid up front.

    Joint Stability. Another cause of filler separation is poor joint stability. As joints widen, the effectiveness of aggregate interlock decreases causing the adjacent slabs on each side of the joint to deflect independently of one another when lift trucks cross. This differential deflection is called joint stability. ACI says that the required stability of joints subjected to hard-wheeled traffic is 0.010 inches. Anything higher than that can result in filler separation leading to joint spalling. A conservative recommendation is to install dowels at all joints that will be subjected to hard-wheeled traffic. This may be unnecessary if low shrinkage concrete is available. So, to determine if dowels should be used in joints on a specific project, the concrete shrinkage potential should be evaluated.

    Evaluating concrete shrinkage

    Concrete shrinkage can significantly impact the performance of slabs. Countless slab investigations for nationwide retail distribution firms have shown that a single mix or slab design does not provide consistent performance in different regions of the country. Concrete is made from locally-available materials and these materials determine the shrinkage potential. The design of concrete slabs, therefore, must be based on specific concrete mixtures. Specifying compressive strength and slump is not sufficient—mixes with the same strength and slump can shrink very differently in different regions.

    To measure the drying shrinkage potential of concrete the test used is ASTM C 157. The specific mixture, including admixture dosage range, should be evaluated. For slabs, the majority of the test should be followed, but PCA and ACI recommend some significant modifications and clarifications:

    The beam size is based on aggregate top size.

  • Once molded, the beams should be kept moist for the first 24 hours.
  • Section 10.2 of ASTM C157 should be followed to get the initial baseline length measurement.
  • Instead of soaking the samples in a lime-saturated bath for 28 days, the specimens should be moist cured after the initial measurement until they reach an age of 7 days.
  • Since we are measuring the drying shrinkage, the specimens should be stored in air in accordance with Section 11.1.2 (73 ± 3° F and 50 ± 4% RH). The samples must be stored on drying racks so that air circulates around them.
Click to expand
Click to expand

During the air-drying period, the samples are measured and the reduction in length is reported as drying shrinkage (%). The rate of shrinkage varies depending on the mixture.

The drying shrinkage potential referenced by PCA and ACI is the ultimate drying shrinkage. To measure the ultimate drying shrinkage, the test must be continued until the samples no longer show a length change. The standard requires measurements at 4, 7, 14, 21, and 28 days of drying and additional measurements at 8, 16, 32, and 64 weeks.

A common complaint is that the testing period far exceeds any construction schedule. But the ultimate shrinkage potential can be predicted from early-age measurements using ACI 209R, “Prediction of Creep, Shrinkage, and Temperature Effects in Concrete Structures.” This report includes the following equation:

This equation can be used to predict the shrinkage from data measured at any age, although the earlier the data, the less accurate the prediction. To be sure that the appropriate test age is used, the predictions can be plotted. The graph typically appears like that shown in Figure 2. At early ages, the predictions are unstable and the graph has a steep slope. Once an adequate age is reached, the predictions stabilize and remain constant. This usually occurs around 21 to 28 days, but additional measurements may be necessary between 28 and 56 days to verify the prediction has stabilized.

Ultimate drying shrinkage ranges dramatically depending on the materials used in the mixture. PCA and ACI include discussions on the impact of the different materials used but the most influential factor is the coarse aggregate.

As part of the chart on joint spacing, ACI defines “typical” concrete shrinkage to be in the range of 0.052 to 0.078% which is about what PCA has published for decades. Low shrinkage concrete falls below this range and high shrinkage concrete falls above. The shrinkage assumed is important to the project. If low shrinkage is assumed, but it’s actually high, it may have a serious impact on the slab.

Applying the results to the project

So what is the contractor supposed to do with the shrinkage results? Simply give them to the designer. It is unfair to include a maximum shrinkage requirement in the project specifications since the contractor does not have an opportunity to evaluate shrinkage prior to submitting the project bid. If locally-available materials don’t satisfy the requirement, the cost of decreasing the shrinkage by hauling in better aggregate or using a shrinkage-reducing admixture (SRA) would not have been accounted for in the bid. The designer is responsible for determining if the slab design is appropriate. The specifications can include common requirements for optimizing the aggregate gradation but this doesn’t assure low shrinkage. The best time to make this determination is before the slab is constructed.

If the shrinkage potential is higher than expected, there are a number of design features that can be used. Sometimes, bringing in aggregates with lower compressibility and higher dimensional stability is feasible. And some projects have effectively used SRAs.

If these modifications are not possible, there are other options. Depending on the shrinkage potential, the joint spacing can be reduced, the use of dowels or reinforcement can be incorporated, or the design can be upgraded to a post-tensioned or shrinkage-compensated system. The designer must explain these options and their associated cost to establish the appropriate expectations. Once the final design is established, the contractor can build it. But, while the contractor can meet or exceed initial floor flatness/levelness, the concrete shrinkage and the selected slab design will determine how flat the slab remains. The relationship between these factors must be fully understood and explained by the designer so that the owner either receives the desired slab or accepts and understands the alternative.

Formerly with CTLGroup and Concrete Engineering Specialists, Scott Tarr is a consulting engineer and President of North S.Tarr Concrete Consulting, P.C. in Dover, N.H.