Section 7: Joints

Concrete undergoes volume changes due to temperature and moisture changes. If these volume changes are not controlled properly, excessive stresses will develop, resulting in uncontrolled cracks. These cracks can be sources of distress and need to be controlled. Joints can be considered as intentional cracks and are provided where the cracking is most likely. These joints relieve stresses, thus preventing uncontrolled cracks. Provisions are made at the joints to provide wheel load transfer.

In CRCP, however, the concrete volume changes are controlled by random cracks, which are tightly held by longitudinal reinforcing steel. Figure 9-72 shows an example of random cracks in CRCP. These random cracks do not cause distress.

Random cracks in CRCP. (click in image to see full-size image) Anchor: #i1003396grtop

Figure 9-72. Random cracks in CRCP.

There are four types of joints:

  1. transverse contraction joint
  2. warping joint
  3. construction joint
  4. expansion joint.

Contraction joint is to relieve tensile stresses resulting from temperature drops and moisture variations in concrete. An example is shown in Figure 9-73. This type of joint is used only in plain jointed pavement (JCP or also called, in Texas, concrete pavement contraction design [CPCD]) at the transverse joint. In TxDOT, transverse contraction joints are provided every 15 ft. TxDOT’s design standards for JCP can be found in cserve/standard/roadway/cpcd94.pdf.

Transverse Contraction Joint. (click in image to see full-size image) Anchor: #i1003398grtop

Figure 9-73. Transverse Contraction Joint.

Problems can occur when contraction joints are sawed too late or too shallow. Attempts to saw-cut too early might cause raveling of aggregates, while too late saw-cutting might result in uncontrolled transverse cracking. There are a number of factors that determine the optimum time of saw-cutting. They include, but are not limited to, setting characteristics of concrete, ambient and concrete temperatures, and thermal coefficient of concrete, to name a few. Even though it is possible to theoretically determine the optimum time for saw-cutting, it is almost impossible to do at the job site due to so many variables involved. The best strategy is to saw-cut as early as possible without severe raveling. Item 360 allows some minor raveling due to early saw-cut.

Dowel bars are provided at the contraction joints to provide load transfer, which reduces deflections and stresses due to wheel load applications. To allow the slabs to move freely in the longitudinal direction due to temperature and moisture variations while providing efficient load transfer, dowel bars are lubricated at least one-half of their length.

The details requirements of dowels, such as dowel bar size, length, and spacing, are shown in the Standards CPCD-94. Figure 9-74 shows a picture of a dowel bar assembly.

Dowel Bars. (click in image to see full-size image) Anchor: #i1003400grtop

Figure 9-74. Dowel Bars.

Dowel bars are typically placed in heavy welded wire baskets that hold the bars in proper spacing, depth, and alignment. Full lane width baskets are typically used, although baskets that are one-half of the lane width have also been used. These wire baskets are shop fabricated. Typically, one end of the dowel bars is tack welded to the wire basket. The wire basket is nailed to the subbase to prevent the basket from being pushed along the subbase once submerged under concrete and under the paving machine.

Alternatively, dowel bars may be mechanically placed by the paving machine. When this method is used, it is important to ensure that the bars are properly aligned and positioned. When this equipment is first used, a few bars should be carefully exposed using a small trowel to verify alignment and positioning. This is done from a work bridge behind the paving machine.

Whether wire baskets or mechanical inserters are used, it is extremely important the dowel bars be parallel for the contraction joint to function properly.

Concrete saws are self-propelled and rotate their blades in a cut-down direction, unlike a handheld circular saw which rotates its blade in a cut-up direction (see Figure 9-75). Most concrete saws use diamond-tipped saw blades which require water to cool and lubricate the blade.

Riding Saw. (click in image to see full-size image) Anchor: #BGBFJIAFgrtop

Figure 9-75. Riding Saw.

Some saws are riding saws, like a riding lawnmower, and others are self-propelled walk behinds ( Figure 9-76). Some newer saw models do not require the use of water and can actually saw the concrete earlier than the concrete saws shown here.

Walk-behind Saw. (click in image to see full-size image) Anchor: #BGBFIAICgrtop

Figure 9-76. Walk-behind Saw.

Figure 9-77 shows that the initial saw-cut was being made from the top of the figure to the bottom. For some reason, the initial saw-cutting was delayed. When the saw was halfway across the pavement, the concrete cracked ahead of the saw.

Even the crack is sealed and allowed to stay, the concrete between the crack and the saw-cut will eventually spall out. To prevent fruther uncontrolled cracking, every second or third transverse joint should be sawed first then the intermediate joints. This is called skip-sawing.

Uncontrolled crack developed while initial
sawing. The left side of the slab needs to be removed and replaced. (click in image to see full-size image) Anchor: #GNFHHHHEgrtop

Figure 9-77. Uncontrolled crack developed while initial sawing. The left side of the slab needs to be removed and replaced.

Depth of the joints also plays a role in ensuring the proper working of the joints. Shallower cuts than specified will increase the chances of cracks somewhere else. Currently, TxDOT requires a minimum of 1/3 of slab thickness when siliceous coarse aggregate is used, and 1/4 of slab thickness when limestone coarse aggregate is used. The reason for this difference is that concrete with siliceous river gravel coarse aggregate has higher coefficient of thermal expansion (CTE) and modulus of elasticity than concrete with limestone coarse aggregates. For given temperature variations, concrete with higher thermal coefficient and modulus of elasticity result in higher stresses compared with concrete with lower values of CTE and modulus. Deeper saw-cut is needed to relieve these higher stresses. Item 360 states that “Saw joints to the depth shown on the plans as soon as sawing can be accomplished without damage to the pavement regardless of time of day or weather conditions.” Once the saw-cut is completed, it is cleaned and sealant materials are applied. TxDOT standards for sealing joints are found in

Figure 9-78 shows a pavement that has 13-in. slab thickness with siliceous river gravel coarse aggegate, which will require a saw-cut depth of 4.3 in. The actual saw-cut depth is about 2.5 in. As a result, a crack was not induced.

Crack was not induced due to shallower
saw-cut depth than required. (click in image to see full-size image) Anchor: #i1003408grtop

Figure 9-78. Crack was not induced due to shallower saw-cut depth than required.

In normal concrete pavement construction, the width of the concrete placement varies. It can be as wide as 50 ft. If the width of the slab is more than 15 ft., the chances of cracking, in longitudinal direction, increase. TxDOT standards require that when the concrete placement width is more than 15 ft., warping joint be provided in the longitudinal direction. Otherwise, longitudinal cracks occur and there are two primary causes for these uncontrolled cracks: late saw-cutting or shallow saw cut depth (as in contraction joints).

Transverse construction joints are needed when the paving operation is stopped in the middle of a lane. A stoppage can occur because of the end of the workday, or because of a mechanical breakdown.

When the placing of concrete is stopped, a bulkhead (header) of sufficient cross sectional area to prevent deflection, accurately notched to receive the load transmission devices and shaped accurately to the cross section of the pavement needs to be provided ( Figure 9-79). At transverse construction joints, additonal longitudinal reinforcing steel is required ( Figure 9-80).

Bulkhead (Header). (click in image to see full-size image) Anchor: #BGBHBFFFgrtop

Figure 9-79. Bulkhead (Header).

Additional steel is required at header. (click in image to see full-size image) Anchor: #i1003412grtop

Figure 9-80. Additional steel is required at header.

When two adjoining lanes are placed with separate concrete placements, the joint between the two placements is called a longitudinal construction joint. Across longitudinal construction joints are short pieces of reinforcing steel called tie bars. The tie bars keep the two adjoining slabs from pulling away from each other and keep the surface across the joints flat. In Item 360, either single piece tie bars or multiple piece tie bars are allowed at the longitudinal construction joints. The detailed requirements for tie bars, such as length, size, and spacing are shown in the Standards CPCD-94.

One piece tie bars may be acceptable when they don’t interfere with traffic or construction conditions.

Multiple piece tie bars are threaded in the middle to allow the two halves of the tie bar to be separated and then reconnected only prior to the casting of the second pavement. The tie bars shown in standard plan sheets CPCR(1)-03 and CPCR(2)-03 for continuously reinforced concrete pavement specify multiple piece tie bars. Unless shown on the plans, the spacing for multiple piece tie bars shall be equal to or less than that of the transverse bars.

Multiple piece tie bars shall develop a tensile strength over their entire length equal to 1-1/4 times the yield strength of the tie bars shown on the plans. Each end of multiple piece tie bars shall consist of deformed reinforcement of at least the size shown on the plans.

In the past, bent tie bars were commonly used to keep the tie bars from interfering with traffic or construction. This practice is no longer allowed, since there were problems such as lane separations.

Notice in Figure 9-81 that the spacing of the transverse reinforcing steel and the spacing of the tie bars match.

If the tie bars are inserted exactly at the transverse bar location, there is a higher chance of hitting the transverse bars. It would be a better practice to insert tie bars a few inches away from the transverse bar locations.

In Figure 9-82 half the tie bar length was embedded into the concrete. The other half sticks out, ready to be cast in the adjacent concrete placement. These tie bars were manually inserted into the plastic concrete during the paving operation.

Before paving. Tie bars (laying on base
in lower part of picture) will be placed at the paintmarks during
paving. (click in image to see full-size image) Anchor: #i1003554grtop

Figure 9-81. Before paving. Tie bars (laying on base in lower part of picture) will be placed at the paintmarks during paving.

After paving. (click in image to see full-size image) Anchor: #i1003556grtop

Figure 9-82. After paving.

Tie bars are also used in plain jointed concrete pavement. When these tie bars are used to connect two lanes, or a lane and shoulder, the tie bar may be mechanically inserted by the paving machine into the fresh concrete ahead of the float or it may be held in position by support devices. Unlike CRCP, there is not an available mat of steel reinforcement that can be used to position and secure the tie bars. Each tie bar will need to be separately supported and held in position by stakes or pins driven into the subbase. These supports need to be sturdy enough to hold the tie bar in position when concrete is placed and consolidated over it.

Expansion joints are installed to provide space for concrete to expand without causing excessive compression in concrete, which could result in blowups or structural damages in bridges. These joints are explained in “Terminal Anchors” described below.

Anchor: #i1008298

Terminal Anchors

Concrete experiences volume changes due to temperature and moisture variations. When the temperature goes down, concrete contracts and transverse contraction joints in CPCD and transverse cracks in CRCP open up and could allow incompressible foreign material to get into joints or cracks. This process can continue and more material could accumulate. When the temperature goes up in the summer, concrete expands and the incompressible material in the joints or cracks does not provide the space that once existed to allow for concrete volume changes. The result is the expansion of the total length of the concrete pavement, called “pavement growth.” Normally, concrete pavement is abutted to the bridge structures via approach slabs and if this “pavement growth” is not contained, the approach slabs will be pushed and serious damage could occur to the bridge structures. To prevent this problem, some form of system is installed at the bridge and pavement interface. Currently, there are three systems in use at TxDOT:

  • expansion joint system
  • wide-flange system
  • anchor lug system.

The basic premise of the expansion joint system is that the expansion joint width will be able to absorb any concrete pavement expansion without transmitting the compression forces to the bridge structure. A typical design standard for expansion joint system can be found at

Wide flange system is similar to the expansion system, except that the expansion joints exist under the wide flange and are not seen from the pavement surface. One advantage is that, since the joint is not exposed to the pavement surface, joint maintenance is minimized. Typical design standard for wide-flange system can be found at

Anchor lug system tries to contain the concrete movement at the interface between bridge and pavement by providing five anchor lugs. The design standard for this system can be found at In all three systems, expansion joints are provided.

At this point, it is not known which of the three systems is the most efficient in minimizing damage to the bridge structures. There will be a research project that will compare the effectiveness of these three systems. It is expected that, based on the findings from the research study, TxDOT will develop one system for statewide use.

Previous page  Next page   Title page