Cracking and raveling are the two main distress conditions, which are commonly encountered at the longitudinal joints of multi-lane hot-mix asphalt (HMA) pavements. These distresses are usually caused by relatively low density at or near the vicinity of the longitudinal joint. A density gradient exists across a typical longitudinal joint. Such a density gradient is caused by the low density at the unconfined edge when the first lane is paved, and a relatively high density at the confined edge when the adjacent lane is paved. There is a need to identify suitable joint construction techniques, which will result in more uniform and high density across a longitudinal joint and, therefore, minimize or eliminate cracking and raveling problems at the joint.

The National Center for Asphalt Technology (NCAT) initiated a national research project in 1992 to evaluate different longitudinal joint construction techniques in four states in the U.S. Seven techniques were used on I-60 in Michigan (1992); eight techniques were used on State Route 190 in Wisconsin (1992); seven techniques were used on I-25 in Colorado (1994); and eight techniques were used on State Route 441 in Pennsylvania (1995). A total of 12 different techniques including three different rolling techniques were used in these four state projects. In all 30 experimental test sections, pavement cores were obtained right on the joint immediately after construction. The performance of these test sections has been evaluated annually by a team of engineers.

Conclusions drawn

The following conclusions can be drawn at the present time from this research project:

• The performance ranking of the different joint construction techniques appears to have been influenced by the overall density at the joint obtained by the technique. The joints with high densities generally show better performance than those with relatively low densities.
• On the Michigan project, the Michigan joint (12.5-mm vertical offset and 12:1 taper) has given the best performance after three years in service.
• On the Wisconsin project, the edge restraining device and the Michigan joint have given better performing joints after four years in service.
• On the Colorado project, the tapered joint (25-mm vertical offset and 3:1 taper), the cutting wheel, and the rubberized tack coat (joint adhesive) have given better performing joints after two years in service.
• On the Pennsylvania project, the cutting wheel, the rubberized tack coat (joint adhesive), and rolling from hot side have given the best performance after only one year in service.
• Among the three different joint rolling techniques used in all four projects, rolling the joint from hot side generally gave the best performance followed by rolling from hot side 152 mm (6 in.) away from the joint.

Recommendations

These projects will be visually evaluated for at least five years after construction. The following general recommendations are made at this time:

• The Michigan joint has the best potential of obtaining a satisfactory longitudinal joint. The vertical offset is considered very essential to its performance.
• Both the cutting wheel and the edge restraining device have a good potential of obtaining a satisfactory joint. However, these techniques are operator dependent and, therefore, may not give consistent performance results.
• The hot side should always overlap the cold side by 25-38 mm (1-1.5 in.) at the joint. A butt joint is not desirable.
• Rolling of the longitudinal joint should be done from the hot side with a vibratory roller as soon as possible. The objective should be to obtain the highest possible density at the joint to ensure best performance.
• Paver manufacturers should consider installing some additional tamping or vibrating mechanism near the edge of the paver screed to obtain a relatively high density at the unconfined edge. The preceding modifications to the paver will make the longitudinal joint construction less dependent upon the paver and roller operators.
• Highway agencies should specify minimum compaction levels to be achieved at the longitudinal joint. This will further ensure the best possible performance of the longitudinal joint. It is recommended that the density at the joint be not more than 2% lower than the density specified in the lanes away from the joint.

Kandhal is the associate director at the National Center for Asphalt Technology at Auburn University. You may write him care of the editor. A copy of the interim research report can be obtained from NCAT by calling 334/844-6228.