Tag Archives: Concrete

Association Cooperation

In the October issue of The Construction Specifier, authors Ward R. Malisch, PhD, PE, and Bruce A. Suprenant, PhD, PE (both of the American Society of Concrete Contractors [ASCC]) wrote our cover story, “Bridging the Specification Gap between Divisions 03 and 09: Concrete and Floorcovering Associations Unite.” The piece looked at how their association teamed up with six other flooring groups to find a solution to a ‘specification gap’ between Divisions 03 and 09 in terms of floor surface flatness requirements.

For space reasons, we had to hold off including a little more background on how these associations collaborated. That ‘missing’ information follows, in the words of Malisch and Suprenant:

The impetus for developing the American Society of Concrete Contractors (ASCC) Position Statements came from a group of contractor members who became aware of a paper published by a national wood flooring organization—not, it should be noted, the National Wood Flooring Association (NWFA)—that stated the organization did not believe in F-numbers and felt they should not be used to measure slabs for gym floors. Rather than trying to decide how they could build a floor that meets unreasonable requirements, ASCC contractors realized they needed to spend their time and resources to educate the industry on the limitations of concrete floors. Thus was born this series, including ASCC Position Statement 6, Division 3 versus Division 9 Floor Flatness Tolerances.

Then, rather than continuing to fight their fellow contractors in the floorcovering industry, ASCC made an effort to get them on board, realizing the greater strength of a united front. ASCC first approached NWFA. With only minor rewriting, that association was eager to endorse the Position Statement.

“For the first time, instead of disagreeing, the two sides have come together to find a common solution to a problem that has cost both groups hundreds of thousands of dollars in rework,” said NWFA president/CEO Michael Martin.

Shortly thereafter, ASCC invited the National Tile Contractors Association (NTCA) to participate in a panel discussion on this topic featuring contractors and technical personnel from both disciplines. Both sides acknowledged the wisdom of a bid allowance to compensate for the incompatibility of the measuring methods, and NTCA became the second flooring association to sign on.

Bart Bettiga, NTCA executive director, commented on the reasons for the document’s usefulness.

“It is our belief this position statement is one of the most important documents we have supported in the past several years,” he said. “This statement accomplishes its goals on many levels. It educates the construction professional about important considerations that must be taken when specifying floorcovering products over concrete substrates.”

“The most important point emphasized in this position statement centers on the disparity related to meeting industry standards in the respective divisions,” Bettiga continued. “Equally important is the call for communication between the related parties and for a proactive approach to be determined prior to the commencement of the work. We strongly support the use of this statement to our members in their communication to the general contractor and architect/specifier on their projects.”

These two organizations were followed by the Flooring Contractors Association. Then, last year, Scott Conwell, director of industry development and technical services for the International Masonry Institute (IMI) contacted the ASCC, asking to add the group’s name, along with those of the Tile Contractors Association of America (TCAA) and the International Union Of Bricklayers and Allied Craftsmen (BAC) to the list of supporters.

“This ASCC Position Statement succinctly brings to light the disparity in requirements for floor flatness between the concrete and the ceramic tile trades,” says Conwell. “The paper effectively brings expectations in line, leading to increased cooperation on the job site to make any corrections to the floor that may be necessary prior to installation of the tile finish.”

Two trades with distinctively different practices and obstacles to overcome but with one goal: to deliver a high-quality product to a satisfied owner.

Don’t Seal Your Fate: Considerations for parking garage surface treatments

All photos © Hoffmann Architects Inc.

All photos © Hoffmann Architects Inc.

by Lawrence E. Keenan, PE, AIA and Robert A. Marsoli Jr., EIT

Elastomeric traffic-bearing membranes have soared in popularity over the past decade. But, what should designers know before specifying one at a parking facility?

It is true parking decks must be protected from the harmful effects of moisture and chlorides, but there is a growing misconception installing a traffic-bearing membrane is a one-way ticket to the garage equivalent of immortality. While a traffic-bearing membrane may be the best option for many situations, it is a big-ticket item, and thorough consideration is necessary to determine whether this costly investment is suited to the garage’s needs.

In order to withstand the punishing abrasion which a parking deck must endure, the traffic-bearing membrane must be hard and durable. At the same time, the membrane must be soft and flexible to bridge over moving cracks and joints without failure. However, traffic-bearing membranes are not perfect. Since hard membranes are generally inflexible, and more pliable membranes do not hold up well to abrasion, choosing the right membrane is a balancing act. Additionally, there are locations where no membrane performs well, such as those areas requiring a flexible membrane, yet are subject to snow plows.

Identifying product properties and applying appropriate selection criteria can guide the specifier in developing a customized system that will provide immediate protection, while also considering future treatment options.

For this parking lot, cracks are routed and sealed as part of a concrete repair project.

For this parking lot, cracks are routed and sealed as part of a concrete repair project.

This test core shows epoxy penetration to the bottom of a crack, as indicated by the arrows on the concrete.

This test core shows epoxy penetration to the
bottom of a crack, as indicated by the arrows
on the concrete.









Sources of deterioration
Since the interior and exterior of parking structures are exposed to the elements, they are more susceptible than other types of buildings to deterioration due to moisture, temperature cycles, and contaminants. Even the best designed and constructed garages need help to survive this onslaught of corrosive forces.

Water is at the heart of most parking deck deterioration. Moisture can facilitate reactions between certain aggregates and alkali hydroxides in the concrete, creating a cycle of expansion, cracking, and further moisture intrusion. Alkali-silica reaction (ASR) is difficult to stop once it has developed. Other minerals, notably sulfates, migrate via penetrating moisture and can lead to formation of gypsum, which can lead to softening and loss of concrete strength, and ettringite, a crystalline mineral the formation of which can result in an increase in solid volume, creating expansive forces that cause cracking and a loss of cohesion and strength in the concrete.

In northern climates, parking decks are subjected to extreme corrosive and deteriorating environments. Moisture, laden with chlorides from de-icing chemicals, tracks into garages and ultimately soaks into the concrete surface. The dissolved chlorides then migrate to embedded steel reinforcement through the pores in the concrete or penetrate through cracks. Once they reach the steel, the salts cause expansive corrosion, ultimately resulting in unsightly, destructive, and costly deterioration.

Moisture’s ability to transport corrosive chlorides is not its only damaging property. Coupled with cold weather, water can damage concrete decks as it expands and contracts during freeze-thaw cycles. Air entrainment, the deliberate incorporation of microscopic air voids in concrete, releases the internal pressure created by freezing water by permitting moisture to flow from void to void. Although this solution to freeze-thaw degradation has been known for years, garages may inadvertently be constructed with insufficient air entrainment, leading to premature concrete breakdown as freezing water destroys it from the inside out.

Applying an epoxy healer/sealer to a concrete deck can be a quick, effective, lowmaintenance option.

Applying an epoxy healer/sealer to a concrete deck can be a quick, effective, low-maintenance option.

Gravity-feeding an epoxy healer/sealer can repair cracks on a concrete deck.

Gravity-feeding an epoxy healer/sealer can repair cracks on a concrete deck.











Deck protection: product types
Technological advances in the chemical industry over the past 30 years have brought concrete sealers a long way from the boiled linseed oil previously used. Today, an industry dedicated solely to concrete protection offers a dizzying array of products to treat concrete before, during, and after production.

Ultimately, the goal of parking garage protection is to stop water from getting into the deck. This may be an over-simplification, but no water means significantly reduced deterioration. The tricky part is water comes in multiple forms. Liquid water is an obvious ‘villain,’ as is the expansive force of ice and snow, but water vapor can be just as damaging.

For example, a chloride-laden deck can actually draw moisture from the air and continue to deteriorate even after the best efforts to keep it dry. In fact, calcium chloride, the most popular and effective of all de-icing chemicals, is commonly used on construction sites for dust control. It is sprinkled onto the dry earth and wets the surface by pulling moisture from the air. Unfortunately, it works equally well at saturating a parking deck.

Remediating the effects of chloride ion attack, freeze-thaw damage, or moisture-driven chemical reactions is both difficult and costly, so preventing any type of water infiltration is a priority. While keeping a garage perfectly dry is an impossible task, through thoughtful product selection, the degree to which moisture can penetrate the parking deck can be limited. For existing parking structures, numerous waterproofing agents that can be applied to the deck’s surface are available.

Penetrating sealers
These liquid-applied treatments, which include silane, siloxane, and silicates, stop water entry by penetrating deep into concrete and forming a barrier that prevents water from entering, limiting chloride ion migration and freeze-thaw damage. These treatments are also vapor-permeable, allowing them to be used at locations where other coatings may be inappropriate, such as slabs-on-grade. Since they are inexpensive and quickly applied, with little or no down-time, penetrating sealers offer a good first line of defense for a parking structure that is in good overall repair. As invisible penetrants working below the surface, these sealers do not affect deck line striping, saving on project duration and cost.

However, these coatings can be short-lived solutions, requiring reapplication every five years or less. They also do not bridge cracks, so they only limit moisture and chloride penetration in intact concrete. Since cracking can be an ongoing process, the ability to bridge new cracks as they form may be important in parking decks that already have evidence of concrete distress.

Methacrylate and epoxy healer/sealers
These coatings both repair cracks and seal pores, so they can be used to restore a deck that has already undergone some deterioration. Low-viscosity methacrylates and epoxies fill the pores in concrete to create a barrier to liquid-water-driven chloride intrusion. They can also be injected or gravity-fed into cracks to structurally heal them. Where desirable, healer/sealers can also limit vapor transmission, although care must be taken not to lock moisture within the deck.

Moderately priced, this class of surface treatments offers a good solution for parking structures starting to show some signs of distress, both to treat deterioration that has already occurred and to prevent continued water-related damage.

Where methacrylate and epoxy healer/sealers fall short is in wet or soiled fractures (to which the materials will not adhere) and moving cracks (which are likely to re-fracture). On parking decks exposed to continuous sunlight, epoxies can degrade quickly under ultraviolet (UV) radiation, so methacrylates should be considered for these areas.

For enclosed parking structures or other areas where fumes might be a problem, offensive odors from methacrylates might prove prohibitive. Unlike the penetrating sealers, healer/sealers are not just ‘coat-and-go;’ surface preparation necessitates shot-blasting, which means increased down time and cost. Also, pavement markings must be reapplied.

Traffic-bearing membranes (elastomeric)
In parking structures with dynamic cracking, shrinkage, or more advanced damage, a traffic-bearing membrane may be the only option to address the ongoing deterioration. Unlike the sealers, these do not penetrate the concrete, but remain on the surface to create a barrier that locks out moisture and chlorides. Most elastomeric membranes have two layers—a base coat that provides the waterproofing protection, and a top coat, which protects the base membrane and provides skid resistance. Together, these yield an attractive, easy-to-clean surface that can give a ‘face lift’ to older, crack-riddled parking decks.

However, a traffic-bearing membrane’s assets are also its downsides. Flexible varieties offer superior crack-bridging, even for moving cracks, but they do not hold up well to abrasion because they are soft and yielding. More rigid varieties, designed to better withstand abrasive forces of heavy traffic, are too stiff to bridge these moving cracks. So while traffic-bearing membranes, as a class of surface treatments may seem to have the ideal combination of properties, in practice no single membrane actually does. Before specifying one of these coatings, the lengthy down-time required for preparation and application, and considerable ongoing maintenance of re-coating or top-coating every five to 10 years, should be considered. Once a traffic-bearing membrane has been installed, it is nearly impossible to return to an uncoated surface in the future.

Applying an impermeable coating to the bottom of an elevated deck traps moisture in the slab, leading to accelerated deterioration.

Applying an impermeable coating to the bottom of an elevated deck traps moisture in the slab, leading to accelerated deterioration.

Unable to evaporate through the coated surface, water entering the slab migrates to the reinforcing steel, leading to corrosion and spalls.

Unable to evaporate through the coated surface, water entering the slab migrates to the reinforcing steel, leading to corrosion and spalls.

Cast-in-place vs. precast
To select the best of the available surface treatments for the parking structure’s characteristics, condition, and situation, designers should consider numerous criteria to determine which products offer the best-performing option for the cost, in terms of both initial investment and long-term maintenance.

Over the years, many different types of parking decks have been developed. For the purpose of investigating surface treatment options, deck types can be simplified into two basic categories: cast-in-place and precast concrete. Usually composed of a single, contiguous, reinforced slab of concrete spanning a concrete or steel frame, cast-in-place decks are constructed onsite.

Due to its nature, concrete shrinks as it cures, which coupled with the external restraint stress from the structure to which it is attached, can lead to crack formation. Cracks are water-borne chlorides’ direct route to reinforcing steel. Once established, these cracks form natural expansion joints that open and close with changing temperature and humidity.

Consequently, protective techniques tend to focus on these moving cracks. If cracks are few and the deck is chloride-free, then routing and sealing, and applying a low-cost sealer, may be appropriate. If the deck is riddled with cracks that cannot be adequately sealed, then elastomeric membranes can begin to look like a good option.

Cast off-site under controlled conditions, precast decks are lifted and welded into place after they have cured and partially dried out. Since the concrete used for this type of construction is typically high-strength and denser than its cast-in-place counterpart, precast decks should rarely experience cracking. However, this manner of construction is favored for fast-track projects, and the end result is rarely defect-free.

As these materials are factory-made and must be lifted into place, precast units do not create a single, contiguous, monolithic structure. Instead, the individual members meet at sealant joints. Extending around each precast unit, these joints add up to miles of sealant that must be maintained and periodically replaced. Even if cracking is not an issue, water migration through failed joints can be just as damaging.

Aside from routine sealant maintenance, surface protection requirements are typically minimal and can usually be addressed with simple low-cost penetrating sealers. Heavily cracked decks may be routed and sealed or treated with rigid epoxies or healer/sealers, since these cracks are typically non-moving. However, the precast deck’s irregular surface does not readily lend itself to flexible membrane-type coatings. The leading edge of each panel quickly becomes a wear point, bumping against automobile tires or catching the tip of a snow plow. Protection techniques that soak into the deck and keep the concrete as the wearing surface are preferred.

Painting the underside of a parking deck may not improve its appearance if moisture causes the coating to bubble and peel.

Painting the underside of a parking deck may not improve its appearance if moisture causes the coating to bubble and peel.

In this case, coating patches were used to repair damage from snowplow blades.

In this case, coating patches were used to
repair damage from snowplow blades.










Concrete quality and condition
Knowing the concrete quality offers insight into the type of deterioration to which it would be most susceptible. This is usually achieved by ordering a petrographic analysis of a test sample. A petrographic analysis is an extraordinarily useful tool in determining what is wrong with concrete or predicting what can go wrong in the future. This analysis can detect most durability issues, so the most appropriate level of protection can be selected.

Chloride content is determined by removing concrete samples from varying depths and analyzing them in a laboratory. If chlorides are moving through the concrete quickly, the deck protection system must be aggressively enhanced to stop further migration. If the chlorides have reached the level of the reinforcement, chances are deterioration has already begun and low-cost sealers are no longer an option. Deck protection that retards water vapor intrusion or effectively inhibits corrosion is now necessary.

While there is nothing inherently wrong with old concrete, the life of a deck does tend to follow a natural progression. Unless design or installation defects are an issue, a new deck can be effectively treated with low-cost sealers that limit the intrusion of chlorides through the concrete. Further along in the life of the deck, a more positive barrier, such as a moderately-priced epoxy sealer, may be necessary to retard moisture entry.

Ultimately, if not properly protected, a deck may require a traffic-bearing membrane to provide the best defense. However, as these membranes are costly and require maintenance and periodic reapplication, waiting to address signs of trouble until there are no other options is not the best course of action. Once a deck has begun to deteriorate, the coating can only retard further deterioration, not stop it.

Evaluating the condition of the concrete slab is an important part of the coating selection process.

Evaluating the condition of the concrete slab is an important part of the coating selection process.

Application of a traffi c-bearing membrane can take several days, and re-coating/top-coating may be required every fi ve to 10 years.

Application of a traffic-bearing membrane can take several days, and re-coating/top-coating may be required every five to 10 years.








Whatever protection system is employed, it must withstand the rigors of its environment. UV degradation may be a problem for some coatings on a top deck. Epoxies, in particular, have difficulties when exposed to direct sunlight. Soft, flexible membranes may not withstand abrasion in high-traffic garages or on a typical turning radius and will fare poorly against snowplows. For example, a coating that looks ‘like new’ after many years in an apartment garage may not withstand a year at an airport or shopping mall.

The damage that can be inflicted by snow removal should not be underestimated. Many coating warranties require snow removal equipment to have rubber tips; others do not cover snowplow damage outright. Unless the garage management operates its own snow removal equipment, coatings at exposed decks will likely encounter a steel plow blade at some point in their service life. There are coating systems tough enough to repel the steel tips, but these super-rigid coatings do not bridge cracks. The best solution depends on finding the right compromise between rigidity and flexibility for a specific situation.

As the adage goes, location is everything. Knowing which surfaces in a parking structure can accept an impermeable coating and which are best left bare is critical to prolonging the life of a garage.

A coating successfully applied to an elevated deck may have disastrous effects in the same garage when applied to a slab-on-grade. As water levels and humidity change, ground moisture seeps up into the concrete slab. Vapor barriers, often installed under slabs-on-grade, are designed to block this moisture from entering the slab. However, in reality, breaches in the barrier or cracks in the slab can still permit water entry. If an impermeable coating is applied to the top surface of the deck, that moisture becomes trapped between two impenetrable surfaces. Unable to escape, the water sits in the slab, leading to chloride and freeze-thaw degradation. Even without a vapor barrier, moisture in the ground rises within the slab and becomes trapped within the deck. Therefore, leaving the slab-on-grade uncoated is the best course of action.

On an elevated deck, that permeability gradient is reversed. Moisture enters the deck from above and migrates through the slab to the underside, where it evaporates. Even with a waterproofing membrane protecting the top surface, the deck is still susceptible to water entering at cracks, joints, and failed coating sections. Coating the bottom of the deck with an impermeable coating invariably leads to trapped moisture and accelerated deterioration. For this reason, the underside of an elevated deck should be similarly treated to a slab-on-grade and left uncoated.

Inappropriately specifi ed or applied coatings can lead to moisturerelated damage that is as detrimental as it is aesthetically unsightly.

Inappropriately specified or applied coatings can lead to moisture-related damage that is as detrimental as it is
aesthetically unsightly.

While the saying ‘you get what you pay for,’ can be applied to surface protection as well as anything else, in terms of quality material selection and skilled application, it is also true lower-cost systems are usually lower-maintenance alternatives. If an inexpensive sealer would suffice, installing a traffic-bearing membrane because it is the high-end option may mean investing in a costly system that may not perform any better in that situation. Additionally, once the membrane is in place, it must be maintained and eventually replaced.

While a simple sealer can help prevent water infiltration, it will not change the parking deck’s appearance. On the other hand, an elastomeric membrane transforms the look of the garage and provides a uniform, fresh-looking surface that is easily cleaned of dirt and stains. For a crack-riddled older garage, this can be a welcome change. In a newer garage, however, the existing concrete surface is likely fine.

Using paint on the underside of a deck to improve its appearance can have problematic effects, since paint is a type of coating. Many of the concrete paints on the market are epoxy-based and relatively impervious to moisture. Even if a vapor-permeable paint is used, successive reapplications increase the coating thickness and so decrease its permeability. Over time, what was once a high-permeability surface can become surprisingly resistant to moisture migration. With the eventuality of peeling paint, spalls, rust, and cracks taken into account, a deck underside painted only for aesthetics begins to lose its appeal, as compared with a simple, uncoated one still intact.

Coating compatibility
Not all surface treatments are compatible. The parking deck protection specified now may limit future options, so both immediate performance goals as well as long-range planning should be considered before committing to a coating. Any applications already in place should also be investigated.

A ‘quick fix’ to get through the winter, for example, might be a less-restrictive sealer that penetrates the slab—rather than one that coats the surface—because various surface treatments can be applied over it in the future. Epoxy healer/sealers can cover such penetrants, and they provide a good base for membrane systems, should one be installed, down the road. However, once a traffic-bearing membrane is installed there is no way to effectively remove it without damaging the slab surface.

Proper application
Even if the right surface treatment is selected for a given project, problems can still result when the application is not executed correctly. Certain coating deterioration issues—such as delamination and blistering—may be avoidable if care is taken in surface preparation and coating techniques.

Before any coating is applied, surface defects must be corrected in order to create a sound substrate for coating application. Any dirt, dust, grease, paint, or other foreign matter should be cleaned, and surrounding areas protected. To prepare concrete for a penetrating sealer, procedures such as power-washing are often used, wherein high-pressure water or steam, sometimes mixed with mild detergents, forces dirt off an exterior surface. Other methods include hand-scrubbing and simple vacuum or broom cleaning. To prepare the deck surface for healer/sealers or traffic-bearing membranes, shot-blasting is required.

It is crucial to wait a minimum of 24 hours following any kind of water washing before applying a coating, in order to allow the deck to dry sufficiently. Cleaned surfaces should be tested for moisture at various sites just prior to application. If excess moisture remains, the coating may trap it inside the parking deck, exacerbating any water-related deterioration. A damp surface can also cause adhesion problems.

In weather conditions such as extreme heat or cold, wind, or rain, the area must be protected and coatings should not be applied. One must also avoid coating in direct heat of sun, as this may result in rapid drying of the material and cause bubbles or wrinkling. It is important to check the specific temperature range recommended by the manufacturer, as these vary from product to product. Also, keep in mind checking the ambient temperature may not be sufficient as surface temperatures may be significantly hotter or colder.

The preferred protection techniques stop deterioration before it begins. If a parking deck is well-maintained from the start, with sealers applied early and cracks promptly addressed, then surface treatment choices can evolve over time as the garage ages and needs change. However, if conditions are such that distress is advanced and progressing rapidly, more immediate and aggressive action must be taken to slow deterioration and minimize its impact.

Before specifying a concrete coating, one should consider the parking deck type, concrete age and quality, and level of exposure to traffic and weather. A surface treatment must not be specified until the garage’s condition has been assessed through investigation, testing, and evaluation. This can help navigate the array of available coatings. The right parking structure protection program should not only protect the deck today, but also anticipate the maintenance needs of tomorrow.

Lawrence E. Keenan, PE, AIA, is the director of engineering with Hoffmann Architects Inc., an architecture and engineering firm specializing in the rehabilitation of building exteriors. He has extensive experience in parking structure rehabilitation, including investigation, repair, and surface treatment consultation. Keenan can be contacted by e-mail at l.keenan@hoffarch.com.

Robert A. Marsoli Jr., EIT, is a project manager at Hoffmann Architects and has developed remediation solutions for a number of parking garages, from design through administration. He also provides preventive treatment consultation services for new construction. Marsoli may be reached at r.marsoli@hoffarch.com.

Bridging the Specification Gap between Divisions 03 and 09: Concrete and floorcovering associations unite

Photo © Michael Marxer (marxerphotography.com). Photo courtesy Mapei

Photo © Michael Marxer (marxerphotography.com). Photo courtesy Mapei

by Ward R. Malisch, PhD, PE, and Bruce A. Suprenant, PhD, PE

Division 03 specifies concrete floor surface flatness requirements to be installed by the concrete contractor. Division 09 specifies the concrete floor surface flatness for the flooring installer that must be met before installing the floorcovering. What does it mean when these requirements are incompatible?

One of the inconsistencies is Division 03 requires the floor flatness to be measured within 72 hours after concrete placement, whereas Division 09 requires the floor flatness to be measured before the floorcovering installation, which may be six to 12 months after the concrete placement. Additionally, Division 03 requires floor flatness to be measured using F-numbers, while Division 09 usually requires floor flatness to be measured as an allowable gap under a 3.1-m (10-ft) straightedge.

Further, Division 03 requires floor flatness not be measured across a construction joint or within 0.6 m (2 ft) of any slab edge, column blockout, or slab penetration. However, Division 09 requires floor flatness to be measured at all these locations. At the same time, Division 09 includes multiple but different floor flatness requirements for carpeting, vinyl, wood, and ceramic tile.

The owner does not want a specification battle; he or she just needs a concrete slab that allows the floorcovering to be installed to achieve a good appearance and obtain the manufacturer’s warranty. Clearly, there must be a cost-effective and efficient solution. Cooperation between the American Society of Concrete Contractors (ASCC) and six associations has led to a solution for bridging the specification gap between Divisions 03 and 09.

Floor fl atness is initially measured within 72 hours after concrete placement using F-numbers to determine contractor’s compliance with Division 03 specifi cations. Flooring installers need a fl oor fl atness metric when they arrive onsite to install fl ooring in compliance with Division 09 specifi cations. However, because concrete fl oor fl atness decreases with time due to curling or defl ection, the initially fl at fl oor placed by the concrete contractor is unlikely to meet the fl oorcovering specifi cation requirements.

Floor flatness is initially measured within 72 hours after concrete placement using F-numbers to determine contractor’s compliance with Division 03 specifications. Flooring installers need a floor flatness metric when they arrive onsite to install flooring in compliance with Division 09 specifications. However, because concrete floor flatness decreases with time due to curling or deflection, the initially fl at floor placed by the concrete contractor is unlikely to meet the floorcovering specification requirements.

The effect of the amount of curling on fl oor fl atness and levelness for a concrete slab with a 4.6-m (15-ft) joint spacing and initially fi nished to a moderately fl at (FF 25), fl at (FF 40), and a very fl at (FF 51) fl oor.

The effect of the amount of curling on floor flatness and levelness for a concrete slab with a 4.6-m (15-ft) joint spacing and initially finished to a moderately fl at (FF 25), fl at (FF 40), and a very fl at (FF 51) floor.













Defining the gap
The gap between floor flatness requirements is illustrated in Figure 1. The concrete contractor produces a floor that meets F-number flatness requirements included in Division 03 and measured shortly after concrete placement. The floorcovering installer arrives onsite far later to start preparation for floor installation. The floor flatness for a concrete slab-on-ground decreases with time due to curling caused by non-uniform concrete drying shrinkage. The floor flatness for an elevated concrete slab decreases with time due to initial deflection caused by the slab’s dead weight and long-term deflection due to creep and shrinkage of the concrete.

The time between the concrete contractor’s work and the flooring installer’s preparation results in a surface change that is the most significant factor in creating the ‘gap.’ Thus, while the specifications may require a suitable concrete surface as placed and finished by the concrete contractor, the resulting changes in surface shape make it unsuitable when the flooring installer arrives onsite. This decrease in flatness often requires flooring installers to do more surface preparation than they originally planned.

It is often impossible to estimate the degree to which floor flatness changes with time, and to determine when the flooring installation might proceed after concrete placement. As will be shown, the gap might be small (e.g. a slight reduction in floor flatness) or significant (e.g. more than a 50 percent reduction in floor flatness based on F-numbers). Thus, it is difficult for flooring installers to decide how much money to put in their bid for surface preparation. It is also difficult for owners to determine how much they need to pay to receive a high-quality final floor finish.

Modeling the effect on fl oor fl atness and levelness of elevated slab defl ection.

Modeling the effect on floor flatness and levelness of elevated slab deflection.








Why the gap exists
Four factors contribute to the gap between the concrete contractor’s finished floor and the flooring installer’s requirements:

  • changes in floor flatness due to curling and deflection;
  • differences in floor flatness measurement methods;
  • differences in floor flatness measurement locations; and
  • dealing with multiple floorcovering requirements.

Changes in floor flatness due to curling
Concrete slabs built flat do not stay flat. The foreword of American Concrete Institute (ACI) 302.1R-04, Guide for Concrete Floor and Slab Construction, states it is completely normal to expect some amount of curling on every project.

Slab curling is caused primarily by differences in moisture content or temperature between the top and bottom of the slab. The slab edges curl upward when the surface is drier and shrinks more, or is cooler and contracts more than the bottom. Curling is most noticeable at construction joints, but it can also occur at saw-cut joints or random cracks. Curling usually results in part of the slab edges and corners losing contact with the underlying base.

There are many factors that influence the amount of curling for a concrete slab-on-ground.1 One of the most important factors is the relative humidity (RH) of the drying environment for the concrete slab. For instance, a concrete slab-on-ground in New Orleans, Louisiana, at 90 percent RH might undergo differential drying shrinkage gradient from the top to bottom surface of as little as 60 x 10-6 in./in. (or mm/mm). While the same concrete slab in Denver, Colorado, exposed to 30 percent RH would undergo a differential shrinkage gradient from the top to the bottom surface as much as 200 x 10-6 in./in.

The magnitude of the shrinkage gradient is three times larger for the slab in Denver versus New Orleans; thus, we would expect a slab to curl more in the former than the latter. Other factors that influence the amount of curling include:

  • potential drying shrinkage magnitude of the concrete mixture;
  • modulus of subgrade reaction;
  • concrete compressive strength and modulus of elasticity;
  • reinforcement ratio;
  • slab thickness; and
  • joint spacing.

Poor curing is often cited as the culprit when a slab curls. ACI 360R-10, Guide to Design of Slabs-on-Ground, states:

Extended curing only delays curling, it does not reduce curling.

In 2003, one of this article’s co-authors reported on F-number floor surface measurements taken at the same location lines at different times on two projects:2

  • a 150-mm (6-in.) thick, 28-MPa (4000-psi) concrete slab containing 19-mm (¾ in.) maximum aggregate size placed directly over a vapor retarder for a gym floor at the University of Maryland; and
  • a 150-mm (6-in.) thick slab with 28-MPa concrete placed on a compactible granular base with saw-cut joints every 4.6 m (15 ft) for an industrial warehouse in Pennsylvania.

Measurements for the gym floor were made 72 hours after concrete placement, and then again seven months later when the flooring installer arrived onsite. The measurements indicated the floor flatness had decreased by 20 percent during the seven months. Similarly, measurements for the industrial slab were taken with 72 hours after concrete placement and then again 12 months later. The measurements indicated floor flatness decreased by 40 percent during that year—this shows the magnitude of floor flatness changes that must be accounted for in bridging the specification gap between Divisions 03 and 09.

The length of lost contact area as a result of curling is about 20 percent of the joint spacing at each end of the slab. For a 4.6-m (15-ft) joint spacing, the slab curl would be expected to change the profile for about 1 m (3 ft) from each end. It is possible to take F-number readings from floors with differing profiles, download those values into a spreadsheet, then add a known amount of curl, and calculate new F-numbers. This allows a comparison of F-numbers before and after the curl.3 Good agreement with this approach was found when compared to the actual F-number measurements for the gym floor and industrial slab.

Figure 2 gives the analytical results showing the effect of the amount of curling on floor flatness and levelness for a concrete slab with a 4.6-m (15-ft) joint spacing and initially finished to a moderately flat (FF 25), flat (FF 40), and a very flat (FF 51) floor as defined by ACI 117-10, Specification for Tolerances for Concrete Construction and Materials.

The amounts of curling considered were 1.6, 3.2, and 6.4 mm (1/16, 1/8 and ¼ in.). A 1/8-in. curl will decrease the floor flatness from an FF 51 to an FF 35, while that same amount of curl will only decrease the floor flatness with an initial FF of 25 to a final FF of 23. As the table shows, the effect of curling is more pronounced on floors with higher initial floor flatness and levelness values. Thus, specifying and paying for higher floor flatness and levelness values in Division 03 may not prove to be a cost-effective solution.

Changes in floor flatness due to deflection
Division 03 requirements state the floor flatness and levelness of elevated slabs must be measured within 72 hours after concrete placement and while the concrete is still supported by the formwork and shoring. However, as soon as the formwork and shoring is removed, the slab deflects downward due to its dead weight. The deflected slab shape changes the floor flatness and levelness, just as curling does.

The concrete industry treats deflection as two parts:

  • initial deflection due to dead weight of the structural members; and
  • long-term deflection due to concrete creep and shrinkage.

ACI 318-11, Building Code Requirements for Structural Concrete, can be used to estimate the additional long-term deflection at 12 months as about 1.4 times the initial deflection. For example, a concrete flexural member spanning 9.1 m (30 ft) was designed for an initial deflection limit of L/360, where L is the span length in inches. Thus the initial deflection would be 30 x 12/360 = 1 in. (about 25 mm). The additional long-term deflection as estimated in accordance with ACI 318 would be 1.4 x 1 = 1.4 in. (about 36 mm) of additional deflection. If the flooring installer arrives one year after the concrete has been placed, he or she could thus expect to see a slab that has deflected about 2.4 in. (about 60 mm).

The effect on floor flatness and levelness of elevated slab deflection can be modeled in the same fashion as the curling effect was for concrete slabs-on-ground. First, initial F-number profiles were simulated, representing varying floor quality, and then superimposed structural deflection values on the profiles. The deflection was assumed to vary with position along the beam as a sine wave, with the initial deflection equal to L/360, L/480 and L/960–deflection values typically used in building code requirements, where L is the length of the span. The deflections were calculated at (1-ft) increments along the beam and added to the simulated F-number readings at the same increment. A 9.1-m (30-ft) span was assumed and the analytical results of this approach are shown in Figure 3.

The analysis shows that for a stiff structure with an FF 25 value, a deflection of L/960 (3/8 in. for a 30-ft span [about 9.5 mm for 9.1 m]), FF decreases by only four percent. Even for an initial profile representing an FF 30 floor, a deflection of L/960 affects the FF value by only about seven percent. However, for an initial FF value of 50, the L/960 deflection causes about a 24 percent decrease in flatness. As is the case with curling deflection—the higher the initial FF value, the greater the effect of dead-load deflection.

A composite overall flatness of FF 35 is the maximum specified value typically used for elevated slabs (ACI 302.1R-04). Based on the analysis, and at this specified value, a deflection of L/960—which indicates a stiff building—will probably result in an FF reduction no greater than about 10 percent. Unfortunately, the same cannot be said for deflection values of L/480 and L/360, which are common for structural steel framing systems supporting concrete slabs placed on metal decking. Since these slabs deflect much more than slabs in reinforced concrete frame buildings, the effect of deflection on FF can also be expected to be greater.

Differences in floor flatness measurement methods
Division 03 floor flatness and levelness requirements are usually specified with the F-number system and thus are measured in accordance with ASTM E1155, Standard Test Method for Determining FF Floor Flatness and FL Floor Levelness Numbers. Division 09 floor flatness is usually specified as a maximum allowable gap measured under a 3.1-m (10-ft) straightedge that rests on two high spots on the concrete surface. It is important to note the straightedge method measures only floor flatness and not levelness. Figure 4 illustrates the two different measurement methods.

If correlations between F-numbers and straightedge gaps are used, it is important to understand the F-number for a given straightedge gap can vary widely. The table in Section 4.5.6 of ACI 117-90 indicates a 3.2-mm (1/8 in.) gap under a 3.1-m (10-ft) unleveled straightedge is roughly equal to an FF 50. However, the standard’s Commentary states:

there is no direct equivalent between F-numbers and straightedge tolerances; the following table does give a rough correlation between the two systems.

Although there is a caution in the ACI 117-90 Commentary, most people use the table because it provides a simple method for comparing the two measurement methods.

The Commentary in ACI 117-10 contains further information on the correlation between the two measurement methods by stating a specified maximum gap of 3.2 mm (1/8 in.) under a 3.1-m (10-ft) straightedge could be equivalent to FF numbers ranging from 38 to 110. The F-numbers are sensitive to the number of 3.2-mm (1/8-in.) gaps, or waves, in the floor. As the number of waves in a 3.1-m (10-ft) length increases, the FF number decreases. This feature of the F-number measuring system enables specifiers to differentiate among floors with the same measured gap but with different numbers of waves.

The two different methods measure significantly different surface properties. Thus, even if concrete contractors satisfy Division 03 F-number requirements, and the floor does not change with time, flooring installers are unlikely to find their gap under the 3.1-m (10-ft) straightedge satisfies the Division 09 requirements. Additionally, there are still specifications with major floor flatness discrepancies—for example, specifying an FF of 20 in Division 03, but then specifying a Division 09 requirement of a 3.2-mm (1/8-in.) maximum gap under a 3.1-m (10-ft) straightedge.

Differences in floor flatness measurement locations
Although the concrete industry lauds F-numbers as a more precise approach to specifying floor flatness, the F-number measuring method does not meet the floorcovering industry’s needs. For instance, according to ASTM E1155 and ACI 117 the measurement should not be taken:

  • across construction joints;
  • within 0.6 m (2 ft) of a penetration; and
  • after 72 hours.

However, to provide the owner with a satisfactory floor finish, the floorcovering must be placed over construction joints and near penetrations on a floor that is certainly older than three days. Figure 5 shows a straightedge being used to measure the flatness directly across a construction joint and at the intersection of a column blockout and the floor slab. F-number measurements do not reflect the flatness variations indicated by the straightedge at these locations.

Although ASTM E1155 includes a procedure for measuring across construction joints, it is rarely used. If the floorcovering industry were to adopt F-numbers, the measuring method and acceptance criteria would have to change so measurements could be made at any location on the floor.

Dealing with multiple floorcovering flatness requirements
Owners and architects often specify multiple floorcovering products for use in facilities such as retail stores. The floor flatness requirement for each of these products can differ greatly. For instance, the Carpet & Rug Institute’s (CRI’s) 2011 Carpet Installation Standard does not have a floor flatness requirement. In contrast, the American National Standards Institute/Tile Council of North America (ANSI/TCNA) A108-2013, Specifications for the Installation of Ceramic Tile, states:

Tiles with all edges shorter than [380 mm] 15 in., shall have a maximum permissible variation of [6.4 mm in 3.1 m] ¼ in. in 10 ft from the required plane, and no more than [1.6 mm] 1/16 in. variation in [300 mm] 12 in. when measured from high points in the surface. For tiles with at least one edge 15 in. or longer, the substrate shall have a maximum permissible variation of [3.2 mm] 1/8 in. in 10 ft from the required plane, and no more than 1/16 in. variation in [610 mm] 24 in. when measured from the high points in the surface.

Floor flatness requirements for the Division 09 finishes vary for each specific floorcovering. Thus, it is possible to be comparing a Division 03 floor flatness specification with multiple Division 09 floor flatness specifications. To get the best price for owners, and meet their schedule, the concrete contractor must place 1400 to 3700 m2 (15,000 to 40,000 sf) of concrete daily. It is not feasible to have the concrete contractor meet separate floor tolerances and finish requirements for every area where a different floorcovering product will be used. Often, the owner has not even made the flooring product choices for different locations before the concrete slab is placed. Thus, Division 09 is unavailable.

Engineers often choose the floor flatness specification in Division 03 with or without input from the architect. The architect needs to give input to balance the needs of the floor flatness requirements for the specified floorcoverings. It might not be economical to just choose the highest floor flatness requirement for Division 09 and put that in Division 03 because, as previously shown, floor flatness decreases with time. Thus, the extra cost passed from the concrete contractor to the owner for achieving a flatter floor may not be of benefit to the flooring installer 12 months later. It may also not be economical to specify the lowest concrete floor flatness needed because that may increase the cost of grinding and patching later.

Flooring installers need to measure fl atness with a straightedge that crosses construction joints, column blockouts, and near penetrations. F-numbers measured in accordance with ASTM E1155 will not yield this information. The top photo shows a carpenter’s level placed across a construction joint and the bottom photo shows a straightedge being used to check fl atness at a column blockout.

Flooring installers need to measure flatness with a straightedge that crosses construction joints, column blockouts, and near penetrations. F-numbers measured in accordance with ASTM E1155 will not yield this information. The top photo shows a carpenter’s level placed across a construction joint and the bottom photo shows a straightedge being used to check flatness at a column blockout.
















Options for closing the gap
There are numerous options for closing the gap between Divisions 03 and 09 floor flatness specifications, but this article focuses on three:

  • design a long-term flat floor;
  • specify higher initial floor flatness; and
  • grind and patch as needed.

The goal is to balance the owner’s cost for producing the desired floorcovering quality by choosing one option or a combination of options.

Design a long-term flat floor
As shown in Figure 6, the goal is to design the floor to stay flat over time. ASCC Position Statement 30, Responsibility for Controlling Slab Curling, indicates both ACI and the Canadian Standards Association (CSA) recognize curling control is the designer’s responsibility. In 2003, when ASCC Position Statement 6, Division 3 versus Division 9 Floor Flatness Tolerances, was first published, there was not enough technical information or design experience for most engineers to design a floor to remain flat until the flooring installer arrived on site.

In 2014, however, some engineers are designing floors that remain flat by using one or more of the following options:

  • limit concrete drying shrinkage;
  • use shrinkage reducing admixtures;
  • lower concrete compressive strength;
  • use more non-prestressed reinforcement (from 0.5 to one percent);
  • use 3 to 4-kg/m3 (5 to 7-lb/cy) macrofibers in the concrete;
  • use shrinkage-compensating concrete; and
  • use post-tensioning.

All these options could be used for concrete slabs-on-ground to control curling, but some will be of limited value when controlling deflection for elevated slabs. Many engineers are not yet comfortable with the risk of designing a flat floor that stays flat, and will avoid this option.

Specify higher initial floor flatness
As shown in Figure 7, the specifier could ask the concrete contractor to produce a higher initial floor flatness with the intent that when the flatness decreases with time, it will still be usable without further remediation for the flooring installer. Most design teams are reluctant to employ this option as they are unsure of how much the floor flatness will decrease with time and when the flooring installer might arrive onsite.

When this strategy is pursued, there is a cost increase to the concrete contractor to provide the higher floor flatness. However, there remains a risk the floor flatness will decrease more than estimated, which means some grind and patch might still be necessary.

Grind and patch as needed
As shown in Figure 8, the concrete floor is designed as economically as possible (while balancing other design and owner concerns), before grinding and patching as needed to achieve the necessary flatness when the flooring installer arrives.

The cost of grinding and patching can add up to more than $100,000 on multi-story buildings. Some owners believe this is an unnecessary expense, but there are options for cost allocation. In other words, some design teams prefer to use an allowance the owner budgets at the start of the project. If the allowance is not needed, the owner keeps the money.

The money could also be spent in designing and constructing to keep the floor flat with time. However, this is a risk when the money is used in designing and building a flat floor that does not stay flat. In that case, the owner will be spending money twice—once for the flat floor option and then more for grind and patch as necessary to achieve a flat floor before floorcovering installation.













Specifying an allowance to bridge the gap
Since 2003, when ASCC Position Statement 6 was published, the ‘grind-and-patch-as-needed’ option has been used most often. The design team budgets it as an allowance so the owner need not spend the money if the concrete slab-on-ground or the elevated concrete slabs remains as flat as required by the flooring installer. The owner can then decide before flooring installation whether to use the allowance to ensure the desired quality of finished flooring.

The other benefit of this option is its adaptability to the requirements for different floorcoverings. For a concrete slab to receive carpeting, perhaps no preparation would be needed. However, for a concrete slab to receive 460 mm (18-in.) square ceramic thin-set tile, money used for prep work may be well spent.

1 One of this article’s co-authors—Bruce Suprenant—wrote a two-part article for ACI’s Concrete International in the spring of 2002. See “Why Slabs Curl−Part I: A Look at the Curling Mechanism and the Effect of Moisture and Shrinkage Gradients on the Amount of Curling” and “Why Slabs Curl−Part II: Factors Affecting the Amount of Curling.” (back to top)
2 See Suprenant’s July 2003 Concrete International article, “The Floor Tolerance/Floorcovering Conundrum.” (back to top)
3 See the authors’ Tolerances for Cast-in-Place Concrete Buildings (American Society of Concrete Contractors, 2009). (back to top)

Ward R. Malisch, PhD, PE, is concrete construction specialist for the American Society of Concrete Contractors (ASCC), an Honorary Member of the American Concrete Institute (ACI), and a member of ASTM International. He has been active in the concrete construction industry for more than 50 years, and has received the ASCC Lifetime Achievement Award, the National Ready Mixed Concrete Association’s (NRMCA’s) Richard D. Gaynor Award, and the Silver Hard Hat Award from the Construction Writers Association. Malisch can be reached at wmalisch@ascconline.org.

Bruce A. Suprenant, PhD, PE, is the technical director for the American Society of Concrete Contractors (ASCC) and a Fellow of the American Concrete Institute (ACI). He has taught concrete materials, construction, and structures for 15 years in universities and has been a consultant in that field for 20 years. Suprenant received ACI’s Roger Corbetta Construction Award and has authored or coauthored more than 100 articles and papers, including one that received ACI’s Construction Award in 2011. He can be reached at bsuprenant@ascconline.org.

Mix Design Fundamentals: Considerations for concrete for slabs-on-ground

Photo © BigStockPhoto/Theerapol Pongkangsananan

Photo © BigStockPhoto/Theerapol Pongkangsananan

by Paul Potts

“Concrete cracks and nothing can be done about it” is a common refrain when the material cracks unexpectedly. However, it is too often an excuse when poor design or improper placement has resulted in excessive random cracking. The real problem is too much mix design water, a lack of welded wire reinforcement, insufficient aggregate, or inadequate curing methods.

Grade-level concrete for hospitals, schools, and commercial buildings is something that can be walked on, supports vehicle traffic and other moderate loads, and provides a hard surface for floorcoverings. Interior concrete should not randomly crack or curl excessively to the point grinding is required before the floor finish can be installed. Similarly, exterior concrete should not randomly crack or deteriorate prematurely from freeze-thaw cycles. Using less water and more aggregate, making the right choices about reinforcement, and ensuring proper contraction jointing will improve the outcome.

This polished concrete floor was designed with synthetic fiber reinforcement but without welded wire reinforcement (WWR). Photos courtesy Paul Pott

This polished concrete floor was designed with synthetic fiber reinforcement but without welded wire reinforcement (WWR). Photos courtesy Paul Pott

It is a well-established belief among some architects, structural engineers, and contractors that strength is the defining characteristic predicting the quality of concrete. Another common erroneous assumption is synthetic fiber reinforcement can replace welded wire reinforcement (WWR) to make concrete less disposed to long-term shrink cracking. Others believe more, not less, cement is better, or the main impact of adding more aggregate to the mix involves making it more expensive. None of this is true.

Clearing incorrect info
It is not uncommon to base the specification for concrete on strength and slump. Unfortunately, strength does not get tested until several days after hardening; and slump, which has a strong relationship to placement, is only loosely related to the most desirable qualities of hardened concrete.

In this author’s first years as construction administrator, he was taught slump was an indication of water content, but there is not direct relationship between the two. As the Portland Cement Association’s (PCA’s) Design and Control of Concrete Mixtures points out, slump tests are simply measures of consistency—in other words, the ability of fresh concrete to flow.

The desirable qualities of slab-on-grade concrete are resistance to curling and shrinkage cracking, along with finishability, flatness, strength, and durability. In northern latitudes, ‘durability’ particularly means resistance to freeze-thaw cycles. To better ensure these qualities, the designer must consider total water content, size and quantity of well-graded aggregate, type of reinforcement, and timeliness of applying curing and cutting construction joints as controlling factors. The factors in designing concrete with these qualities are:

  • low water-to-cement (w/c) ratio;
  • minimum total cement content needed (with the understanding more cement requires more water);
  • size and weight of aggregate in a yard of concrete;
  • entrained air in exterior concrete; and
  • presence of WWR.

According to the PCA handbook, almost every quality of concrete will be improved by reducing the total water content in the batch. The w/c ratio is the weight of water divided by the weight of cement. To minimize total water content, designers must start by specifying a low water-to-cement ratio (e.g. 0.45 w/c ratios for interior concrete and either a 0.45 or 0.40 for exterior concrete). To further minimize total water, they should opt for the lowest practicable cement content (e.g. 5-1/2 sack). At the same w/c ratio, the less cement in a batch of concrete, the less total water needed.

According to Chapter 9 (“Designing and Proportioning Normal Concrete Mixture”) of PCA’s Design and Control of Concrete Mixtures:

For any particular set of materials and conditions of curing, the quality of hardened concrete is determined by the amount of water used in relation to the amount of cement. The less water used, the better the quality of the concrete—provided it can be consolidated properly. 

Concrete with a water to cement (w/c) ratio of 0.45 and below is stiff when it comes off the truck—otherwise, it may have been tampered with.

Concrete with a water to cement (w/c) ratio of 0.45 and below is stiff when it comes off the truck—otherwise, it may have been tampered with.


Using more—rather than less—large aggregate (e.g. 900 to 1000 kg [2000 to 2200 lb]) reduces the total cement and water needed, improves resistance to cracking and curling, and increases the strength at the same time. This is also economical, as aggregate is cheaper than cement.

It is important to keep in mind there is a limit to how low one can go with cement and water because there must be enough water (or substitute) to allow concrete to flow into place. Further, there must be enough cement to promote the finisher’s work. The effect low water has on finishing work can be counteracted with water-reducing admixtures (WRAs).

Water content, wire reinforcement, and cracking
Excess water (i.e. more than required for hydration) improves concrete flow, making the material more economical to place. However, water not hydrated by the cement eventually becomes ‘bleed water’ and evaporates. The evaporation reduces the total volume of the slab, causing the slab to shrink in overall dimensions. More excess water available for evaporation means more shrinkage and more potential for long-term shrink cracking.

When the concrete’s shrinking volume is restrained by a pipe, masonry corner, steel column, or the drag of an uneven subgrade, there may be stresses beyond the material’s tensile strength to resist—as a result, shrink cracks develop. (Tensile strength is only 10 per cent of compressive strength.)

Welded wire reinforcement buttresses the tensile strength of concrete by distributing the stresses over a wider area where it can be better controlled. Some architects and engineers believe synthetic fiber mesh can be substituted for welded wire in concrete for the same crack-resisting properties. However, synthetic fiber mesh is only effective at preventing plastic shrink cracking—hairline cracks that develop before final finishing.

To clear up confusion about the purpose of synthetic fibers, PCA published the following statement on its website:

Plastic fibers should not be expected to replace wire mesh in a slab-on-ground. However, although not affecting joint spacing, plastic fibers are used to reduce plastic shrinkage cracking.1

While finishing is not any more difficult with low w/c concrete, placement, and screeding require an ambitious crew.

While finishing is not any more difficult with low w/c concrete, placement, and screeding require an ambitious crew.

Mix design
‘Strength’ is the reciprocal of the w/c ratio—the lower the ratio, the higher the concrete strength. While it may be prudent for legal reasons to specify a minimum strength, it is just as well as to develop the mix design for slab-on-ground applications by specifying the maximum sack content, water-cement ratio, and aggregate quantity/size. Strength follows the water-to-cement ratio as surely as day follows night.

A w/c ratio of 0.045 and 5-1/2 sack content with 900-kg (2000-lb) of large aggregate with welded wire reinforcement is a good standby mix design for an interior slab-on-grade. The strength will fall around 30 N/mm2 (4500 psi). A w/c ratio of 0.040 and 5-1/2 sack specification with 900 to 1000 kg (2000 to 2200 lb) of aggregate and six per cent air with welded wire reinforcement is a durable exterior mix design, but this concrete is stiff and may need a water-reducer to aid placement. The strength will be around 35 N/mm2 (5000 psi).

Pop-outs in exterior concrete are the result of waterlogged soft stone or chert in concrete mixtures that, owing to their lighter density and porousness, absorb water and float up near the surface during finishing operations, then explode during the first freeze cycle. Most concrete mix designs limit soft stone and chert to less than one percent (if permitting it at all). However, even this one percent can leave an awful-looking mess on the surface of a new driveway or sidewalk. To avoid pop-outs in exterior concrete, one should use 21AA crushed limestone aggregate where it is available. When limestone is unavailable, the soft stone and chert in natural stone aggregate should be limited to less than one percent.

Contraction joints versus random cracking
Contraction joints in slab-on-ground designs are for aesthetic purposes—without contraction joints, cracks would occur randomly. Contraction jointing limits the cracking to where it is least objectionable and, in the case of thin-set terrazzo, epoxy, and urethane floorcoverings, at locations that can be coordinated with joints in the surface materials. Still, random cracking can be preferable in some scenarios.

For example, the appearance of carpet, vinyl, rubber, and linoleum can be improved by allowing the concrete to randomly crack instead of adding contraction joints with their inherent curling problems. To control the width of random cracking, one should start with a low water-to-cement ratio mix and require WWR. Random cracks in concrete are free of curling, and a low w/c ratio combined with welded reinforcement minimizes the width of cracks.

In a random crack system, attention must be paid to corridor intersections that may need contraction joints to prevent excessively wide cracks from developing where the long corridor runs and restraints at corners can produce quite large random cracks, regardless of the mix design and WWR. Concrete slabs for wood floors on sleepers or rubber cushions can be designed the same way.

It is not an option to allow random cracking under hard tile, epoxy floorcoverings, and thin-set terrazzo bonded to the concrete, because any cracks in the substrate will telegraph through to the surface of the bonded floorcovering (Figure 1). Concrete under bonded floorcoverings must be reinforced with WWR or a mat of steel reinforcement to reduce shrinkage cracking. Such components lower the risk of shrinkage cracking by restraining contraction during set time and reducing the width of cracks once the concrete sets. While some structural engineers may not see the need for WWR in non-load-bearing slabs-on-grade, its importance for reducing curling and cracking must not be overlooked.

telegraphed crack

This shrink crack telegraphed through the thin-set terrazzo and became the subject of litigation and costly court-ordered repairs. (A core sample can be seen at right/below.)

This shrink crack telegraphed through the thin-set terrazzo and became the subject of litigation and costly court-ordered repairs. (A core sample can be seen above.)

Vapor barrier and curing considerations
A vapor barrier should be placed under all slabs-on-ground that will receive floorcovering. Properly consolidated concrete is waterproof, but not vapor-proof. If a vapor barrier is not used, any moisture under the slab migrates upward by capillary action through the concrete to eventually degrade the floorcovering adhesive. Most floorcovering manufacturers consider the lack of vapor barrier a defective concrete installation—many void their warranties on that basis.

While some specifiers recommend a blotter layer of granular material between the vapor barrier and the slab, in practice this has met a lot of resistance, principally because of the difficulty of keeping the blotter layer intact while completing the mechanical and electrical underground and preparing the concrete pour.

According to Chapter 9 in PCA’s Floors on Ground handbook:

Other specifiers believe that no blotter layer is needed and that concrete should be placed directly on the vapor retarder. The idea is that concrete slabs should be cured from both the top and the bottom. A granular layer between the vapor retarder and the concrete creates a potential water reservoir that could cause moisture problems at a later date. 

A properly installed vapor barrier directly under the slab has other benefits beyond reducing the transmission of moisture vapor. First, the polyethylene film acts as a slip-sheet between the underside of the concrete and the sub-base; this reduces drag on the slab and decreases shrinkage cracking. Further, when combined with timely surface curing, a polyethylene vapor barrier under the slab is an asset, retaining moisture in the concrete to improve hydration during setting. Whenever a vapor barrier under the slab is used, the water-to-cement ratio should be 0.45 or less.

It is especially important to start curing as soon as practical after finishing operations are complete. Where concrete is placed directly on a vapor barrier, a double application of cure-only compound should be applied at right angles to each other. In cases where the floorcovering requires an adhesive, cure-only compounds should be used rather than cure-and-seal products—sealing the concrete interferes with the bond between the concrete and the floorcovering or topping.

Low water-to-cement concrete is harder to move around and requires considerably more handwork to get it in place. Placement and spreading of low water-to-cement ratio concrete can be improved by adding water-reducing admixtures (WRAs)—however, excessive use can cause more drying shrinkage. These are rarely required by the specification, but are included in the specification as an alternate product that may be used.

Low water-to-cement concrete is harder to move around and requires considerably more handwork to get it in place. Placement and spreading of low water-to-cement ratio concrete can be improved by adding water-reducing admixtures (WRAs)—however, excessive use can cause more drying shrinkage. These are rarely required by the specification, but are included in the specification as an alternate product that may be used.

Larger-size and greater quantities of coarse aggregates work in several ways to reduce cracking and curling, and improve the economy of concrete. Coarse aggregates, less than 50 mm (2 in.) in size, cost about half as much as the cement in concrete. The larger aggregates leave less room that must be filled with cement paste so they reduce the total water required in the mix. To avoid pop-outs in exterior concrete, the designer should specify crushed limestone aggregate where it is available.

More and larger aggregates reduce shrinkage cracking, and reduce curling. The weight and dimension of larger aggregates put a drag on the movement of the materials within concrete during the shrinkage stage, and reduce overall contraction thereby reducing shrinkage cracking. Greater quantities of coarse aggregate magnify these benefits—900 kg (1984 lb) of 32 mm (1 ¼ in.) coarse aggregate is a good starting point for specifying the aggregate in slab-on-ground concrete.

Concrete designed with low water-to-cement ratios and the maximum size and content of large aggregate, reinforced against shrinkage cracking with properly spaced contraction joints will have fewer random cracks and less curling. To that end, the purpose of this article is two-fold:

  • encourage the concrete mix designer to think in terms of the low w/c ratios and aggregate content instead of strength and slump as the starting point for the concrete mix design; and
  • encourage the use of WWR where crack mitigation is important, cautioning against misapplication of synthetic fiber as countermeasure with regard to long-term shrink cracking.

1 Visit www.cement.org/tech/faq_fibers.asp. (back to top)

Paul Potts has worked as construction administrator for Kingscott Associates, a school design/construction firm. A licensed MasterSpec writer and a consultant to the construction industry in Michigan, he is currently working on the reorganization of the Michigan State University Cabinet Shop and finishing construction administration work on the New Community Building and Pavilion project for the City of Potterville. Potts can be reached at paulpotts1@comcast.net.

ASTM celebrates concrete centennial

Concrete Finishers

ASTM International’s concrete-focused committee has worked to improve the material’s use and durability for a century. Photo © BigStockPhoto/FrenchToast

During last month’s round of standards development meetings in Toronto, ASTM International celebrated the 100th anniversary of the group that became Committee C09 on Concrete and Concrete Aggregates.

Since 1914, when a small group gathered to work on methods for making and testing field specimens, C09 has grown to more than 1400 members from 62 countries, maintaining a portfolio of more than 175 standards. Its 50 subcommittees focus on aspects ranging from self-consolidating concrete and chemical admixtures to supplementary cementitious materials (SCMs) and pervious assemblies.

C09 has developed global standards in construction, industrial, transportation, defense, utility, and residential sectors, but the group says its first standard remains one of its most important. ASTM C94/C94M, Specification for Ready-mixed Concrete, was first approved in 1933, but has kept pace with technology changes to present day.

“Looking at C09’s book of standards doesn’t tell the complete story of the committee’s success and accomplishments over the past century,” said committee member Richard Szeczy (president of Texas Aggregates and Concrete Association).

lobo_colin_2014 (2)

Colin Lobo, PhD, F.ASTM, received the ASTM International Award of Merit for Service to Concrete Committee. Photo courtesy ASTM International

“To produce the defining concrete industry documents stakeholders around the world rely on every day has taken countless hours of dedicated effort and cooperation from thousands of international experts,” he continued. “Over the years, C09 has embodied everything that is great about the ASTM process. That itself is truly worth celebrating.”

In related ASTM news, Colin Lobo, PhD, has received the organization’s Award of Merit for Service to Concrete Committee (along with Fellowship). Chair of the ASTM Cement and Concrete Laboratory (CCRL) executive group and senior vice president of engineering for National Ready Mixed Concrete Association (NRMCA), Lobo was lauded for his contributions to specifications for concrete materials, test methods for fresh and hardened concrete, data evaluation, and laboratory assessment.