Four ways to eliminate joints in wide-slab concrete floors

by Katie Daniel | January 6, 2016 10:44 am

All photos courtesy George Garber

By George Garber
What is it about ground-supported concrete floors that make owners and tenants complain? The answer depends on where the floors are located. In residential basements and garages, cracks generate the most complaints. In office buildings and retail stores, moisture-related problems top the list. In industrial buildings—warehouses and factories—the most common issues involve joints.

Joints are the seams dividing a concrete floor into multiple slabs. Modern industrial floors contain both construction joints and sawn joints. (The word ‘joint’ can also describe the location where a floor slab meets another building element such as a wall or column. The construction methods described here do not eliminate that kind of joint.) A construction joint, also known as a formed joint or a day joint, marks the line between two concrete pours made on separate days. A sawn joint is located within a pour, and is made by sawing partway through the concrete after it has hardened. The sawcut creates a weakened vertical plane where the concrete can easily crack. The goal is to control cracks by keeping them under the sawcuts, but this is not always realized. In most industrial floors, sawn joints far outnumber construction joints, and their sole purpose is crack control.

Joints can do a good job of controlling cracks, but there is a price. They are costly to make and maintain, and can have numerous issues including:

Given all these problems, it would make sense to try and eliminate these joints. Some designers do try, but most do not. The typical industrial floor in Canada still relies on closely spaced joints as its primary method of crack control. This is the same method used 50 years ago. Previously, it was common to space joints about 6 m (20 ft) apart, but these floors sometimes cracked between joints, so designers began recommending closer spacings. For a while, the standard rule of thumb for joint spacing was not to exceed 36 times the slab thickness. Meaning a 150 mm (6 in.) thick slab would have joint spacings up to 5.4 m (18 ft).

Today, the American Concrete Institute, (ACI) 360R-10, Guide to Design of Slabs-on-Ground, recommends joint spacings that would have previously been considered absurdly close. For a slab 150 mm (6 in.) thick, made of concrete with unknown drying shrinkage—a common situation—ACI 360 now calls for joints no more than 3.6 m (12 ft) apart. Even very thick slabs are subject to a maximum joint spacing of 4.5 m (15 ft) unless the concrete is tested for shrinkage and proves to have an ultimate drying-shrinkage rate below 0.078 per cent.


Cracks are still being controlled, but floors are still riddled with joints. Is this progress?

To remedy this situation, there are at least four ways to build industrial floors with far fewer joints. They include:

All four methods allow wide-slab (or ‘extended-joint’) construction, with joint spacings greater than 15 m (50 ft). Some call them jointless or joint-free floors, but these terms are inaccurate, because only 70 to 90 per cent of joints are actually eliminated.

Continuous heavy rebar
The continuous heavy rebar method was borrowed from highway engineers, who refer to it as continuously reinforced concrete pavement (CRCP). This method reinforces the floor with deformed steel bars running continuously the full length of the slab. In CRCP highways, the reinforcement ranges from 0.5 to 0.8 per cent of the slab’s cross-sectional area. Designers who adapted this method for interior floors around 1990 chose the low end of the range at 0.5 per cent, and it has been successful. ACI 360 endorses this, stating:

to eliminate sawcut contraction joints, a continuous amount of reinforcement with a minimum steel ratio of 0.5 per cent of the slab’s cross-sectional area in the direction where the contraction joints are eliminated is recommended.

The reinforcing steel is typically placed about 50 mm (2 in.) below the floor surface. Continuity is essential, so bar ends must be carefully lapped. A publication from the U.S. Federal Highway Administration (FHWA), Technical Advisory 5080.14, contains lap details and other useful information. (For more, see the United States Department of Transportation, Federal Highway Administration, Technical Advisory 5080.14 paper, Continuously Reinforced Concrete Pavement.)

Compared to other methods of wide-slab construction, continuous heavy rebar has several powerful advantages, including:

However, there is one large and often fatal drawback that needs to be considered. Continuous heavy rebar does not prevent cracks. On the contrary, it creates them. The goal of this method is to relieve shrinkage stresses at a large number of very narrow cracks. The distance between cracks varies by project, but can be as little as 300 mm (12 in). Crack width also varies, but usually stays below 0.3 mm and never exceeds 0.5 mm unless there is, by mistake, a break in the continuity of the rebar (Figure 3).

Though CRCP highways are common in Canada, industrial floors designed along the same lines remain scarce, probably because floor users have a strong prejudice against cracks. It could be argued a tight crack with rebar running through it is not a defect and will cause less trouble than a joint. While this is a reasonable argument, many users will not accept it.

Sawn joints in a new floor. These types of joints often look good when they are new, but they do not stay that way under heavy usage.

Shrinkage-compensating concrete
Ordinary concrete starts shrinking as soon as it is placed, and continues shrinking for a long time. The shrinkage puts the concrete into tension, causing cracks. Shrinkage-compensating concrete behaves differently. It starts out by expanding, putting the concrete into compression. After a few days the expansion ends and the concrete starts shrinking. However, the shrinkage does not put the concrete into tension—at least not right away; rather it just reduces the compression. If expansion and shrinkage are evenly matched, the concrete never goes into tension and does not crack.

The expansion comes either from expansive cement that replaces the portland cement found on ordinary concrete, or from an expansive component used with normal portland cement. ACI 223R-10, Guide for the Use of Shrinkage-Compensating Concrete, recognizes three kinds of expansive cement and four kinds of components, but not all are in regular use in Canada.

Expansive cement has become rare. Most shrinkage-compensating concrete in the country is made with an expansive component—Types K, S, or G. This work is mostly performed by specialist contractors, each of whom has their own favourite material. Fortunately, all the types produce similar effects.

To control cracks, the initial expansion must be restrained so the concrete goes into compression. Some restraint comes from subslab friction and other building elements, but the main source in most designs is conventional steel reinforcement. ACI 223 recommends rebar or wire mesh in two directions, equal to 0.15 per cent of the slab’s cross-sectional area. Some designers use less, however.

Floors made of shrinkage-compensating concrete require some special details if they are to remain crack-free. It is widely—though not universally—accepted the slabs should be roughly square, with an aspect ratio not greater than 3:1. Slabs must be isolated from walls and columns. The placing sequence needs to be carefully planned so each slab is free to expand.

Shrinkage-compensating concrete routinely allows joint spacings of 36 m (118 ft). It is unclear where the upper limit lies. ACI 223 reports slabs as big as 1850 m2 (19,913 sf), which, if square, would result in a joint spacing of 43 m (141 ft). However, in normal practice, spacings over 36 m are rare.

More than any other method of wide-slab construction, use of shrinkage-compensating concrete requires special knowledge and attention to detail.

Shrinkage-compensating concrete puts slabs into compression using chemical action. Post-tensioning does the same thing, but with steel cables instead of chemicals.

The cables run the full length and width of the slab, encased in plastic or metal tubes. After the concrete has set, but before it has had much time to shrink, the cables are stretched and locked off at each end, putting the slab in compression. The steel used is much stronger than normal rebar, with a tensile strength of 1830 MPa (270,000 psi). The size and spacing of the cables depends on slab thickness and length, subslab friction, and how much compression is desired. Where the sole purpose of the post-tensioning is to prevent cracks, the design normally provides for mid-slab compressive stress of at least 350 kPa (50 psi). Some designers specify higher values, either to raise the safety factor or to increase the slab’s structural strength. Design information, including the formula used to determine cable size and spacing, is available in the Post-Tensioning Institute’s (PTI’s) book, Design of Post-Tensioned Slabs-on-Ground.

Some joints look bad even when they are new. This one was sawn too early.

Though short slabs, up to 15 m (50 ft) long, are tensioned in one application, wide-slab construction usually involves a two-step process to head off early cracking. The first step involves applying one-third of the full design stress as soon as the concrete’s compressive strength reaches 8 MPa (1200 psi). The second step is to apply full stress after the concrete’s compressive strength exceeds 21 MPa (3000 psi).

As with shrinkage-compensating concrete, post-tensioning requires some special details. Slabs must be free to move horizontally, and be isolated from walls, columns, and any other sources of restraint.  A plastic slipsheet is essential to reduce friction between the slab and sub-base. Some designers even use two sheets of plastic to further reduce friction. Slab ends must be accessible for stretching the cables—this sometimes requires narrow gaps between slabs to be filled after stressing has been completed.

Post-tensioning has been used in slabs up to 183 m (600 ft) long, but a more practical limit is 120 m (400 ft), which still exceeds the limits imposed by shrinkage-compensating concrete or steel fibres. However, post-tensioning differs from the other methods because the cost increases with joint spacing—a longer slab needs additional or bigger cables. Sometimes the joint spacing is limited not by engineering rules, but rather the project budget.

This method offers a potential benefit not available from any other method of wide-slab construction. Post-tensioned slabs can be made thinner while supporting the same loads, thanks to the prestress provided by the cables. The compressive stress induced by post-tensioning cables has the effect of increasing the concrete’s flexural strength. Structural engineers can take advantage of this when calculating slab thickness. For slabs supporting heavy loads, post-tensioning often allows a reduction in slab thickness of 25 to 50 mm (1 to 2 in.). There is less benefit when loads are light.

Post-tensioning has two important drawbacks, both of which centre on the need to keep cables in place. The first involves access during the concrete pour. Cables need to be chaired up in their final position as concrete is placed around them. This makes it hard to discharge concrete directly from ready-mix trucks and it interferes with modern screeding machines. Floors made with continuous heavy rebar and shrinkage-compensating concrete also suffer from this drawback, since both require mats of reinforcing steel in the pour. However, the issue looms larger with post-tensioning because it is so important to keep the cables positioned correctly. In Australia, where post-tensioning is more common, contractors use special ramps to support a screeding machine above the cables.

The second drawback is unique to post-tensioning. Cables should not be disturbed after they have been tensioned. Users must be careful drilling holes in the finished floor. Certain kinds of floor repairs become harder, and floor demolition is complicated—though not impossible.

Steel-fibre reinforcement
Steel fibres are short, thin pieces of steel added to the concrete mix. They range in length from 12 to 60 mm (½ to 2 3/8 in.). When added in sufficient numbers, they prevent visible cracks by stopping invisible microcracks before they spread. In wide-slab construction, the fibre dose is typically about 40 kg/m3 (67 lb/cy). The method comes from Europe, where it has been popular for years.

As with post-tensioning, wide-slab floors made with steel fibres should be laid over a plastic slipsheet and be isolated from other building elements. Some designers believe it is important to specify a limit to the concrete’s drying shrinkage. A specified maximum of 0.035 per cent at 28 days, with testing to ASTM C157, Standard Test Method for Length Change of Hardened Hydraulic–Cement, Mortar, and Concrete, is typical.

Steel fibres allow joint spacings up to 38 m (125 ft), which is about the same as shrinkage-compensating concrete. However, the fibres allow bigger placements because some of the joints can be sawn. It is possible, for example, to place a rectangle 38 by 76 m (125 to 249 ft) and saw it at mid-length, creating two square slabs. With shrinkage-compensating concrete, every joint must be a construction joint. This means joint spacing dictates pour size.

The steel fibre method offers one large advantage over the other three methods because it does not get in the way of concrete trucks and screeding machines. Since the fibres arrive with the concrete, the area of the pour is left wide open. In contrast, the other methods require steel reinforcement or post-tensioning cables throughout the slab, restricting access.

The only serious drawback reported with steel fibres is the appearance of fibres at the floor surface. To hide the fibres, some designers call for a dry-shake finish. The use of shorter fibres, about 25 mm (1 in.) long, along with some care in floating and trowelling the concrete, keeps exposed fibres down to a low number. Most users find this acceptable.

Heavy continuous rebar[4]
The four methods of wide-slab construction are compared according to various factors affecting the slabs.

One- and two-way wide-slab construction
Most wide-slab floors have extended joint spacings in two directions. However, continuous heavy rebar and post-tensioning offer another option—a one-way floor with a wide-slab design in one direction, and a conventional jointed design in the other. This is common in very narrow aisle (VNA) warehouses that have superflat floors laid in long narrow strips (Figure 4). Each strip supports a single warehouse aisle, and the construction joints between the strips fall under the storage racks. The reinforcing bars or post-tensioning cables run only in the down-aisle direction.

Common ground
Though each method of wide-slab construction has its own benefits and drawbacks, all share certain features.

Any wide-slab floor will be more costly to build than an unreinforced floor full of joints. This inescapable fact is occasionally enough to make designers and owners reject all the methods described here. However, a full accounting must consider more than construction costs. After allowing for joint filling, and for joint maintenance over the building’s lifespan, a wide-slab floor may be cheaper than the conventional alternative. Additional benefits are harder to quantify, but are no less real. They include:

In general, the harder a floor is used, the easier it is to justify wide-slab construction.

Another consideration shared by all wide-slab methods is the likelihood of wide joints. This is a trade-off. When the number of joints is greatly reduced, the few joints left have more to do. Given the same amount of concrete shrinkage or thermal contraction, joints 30 m (98 ft) apart are going to open wider than those 3 m (10 ft) apart. (Perhaps not 10 times as wide, but wider.) For this reason, many wide-slab designs include armoured joints (Figure 5).

When looking at an industrial concrete floor with closely spaced joints, it is clear the floor would look better, do its job better, and last longer, if it had 1/8 or 1/10 as many joints.

Garber-headshot[5]George Garber is the author of Design and Construction of Concrete Floors, Concrete Flatwork and Paving with Pervious Concrete. Based in Lexington, Kentucky, he consults on the design, construction, and repair of concrete floors. Garber can be reached by e-mail at[6].

  1. [Image]:
  2. [Image]:
  3. [Image]:
  4. [Image]:
  5. [Image]:

Source URL: