U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
|Accelerating Infrastructure Innovations|
Publication Number: FHWA-SA-98-022
Date: May 1998
Oil and water don't mix-and nowhere is that more evident than in asphalt pavement. The asphalt binder, a petroleum-based product, "glues" the aggregates in a hot-mix asphalt pavement together. But water that seeps into the pavement can cause the asphalt to strip away from the aggregate. This type of moisture-induced damage, known as stripping, leaves the pavement extremely vulnerable to cracking, rutting, and other serious damage.
This is not news to materials engineers and pavement designers, who routinely test mixes in the laboratory to determine if they are susceptible to moisture-induced damage and thus would benefit from the addition of anti-stripping agents. What is news to some, however, is that the test method routinely used in the laboratory is not calibrated for use with mix samples produced using the Superpave gyratory compactor.
The test procedure, known as American Association of State Highway and Transportation Officials (AASHTO) T283 ("Resistance of Compacted Bituminous Mixtures to Moisture Induced Damage"), was designed to be used on 100-mm-diameter (4-in) mix samples produced using the Marshall method of compaction. With the Superpave system, however, 150-mm-diameter (6-in) mix samples are produced using the Superpave gyratory compactor, which uses a kneading action rather than a series of blows for compaction.
Although the T283 test procedure allows the use of larger samples, it's not known how these larger samples and the different method of compaction might affect the test results, says AASHTO's Haleem Tahir. In addition, he says, some States have found that T283 does not always identify moisture-susceptible asphalt mixes designed using the Marshall method.
The possible incompatibility between T283 and specimens compacted with the Superpave gyratory compactor may explain why laboratory tests failed to indicate that the Superpave test sections on U.S. Route 93 near Kingman, Arizona, would be susceptible to moisture damage. The test sections, which were built in August 1993 as part of the long-term pavement performance (LTPP) program's Specific Pavement Studies 9 experiment (see sidebar), developed severe fatigue cracking after serving less than 5 years and carrying less than one-third of their design traffic load.
In January 1998, shortly after the fatigue cracking was found on the test sections, representatives from the Arizona Department of Transportation (DOT), the Federal Highway Administration (FHWA), and industry traveled to the site to investigate the problem. They found that the Superpave pavements had severe moisture-induced damage, while adjacent pavements built at the same time but using Arizona DOT's conventional mixes had no moisture-induced damage.
Arizona DOT had run the T283 test on the conventional mixes for the Route 93 project and found the mixes to be susceptible to stripping. Portland cement was thus added as an anti-stripping agent, which helps explain why the conventional pavements show no signs of moisture-induced damage. The Superpave mixes for the Route 93 project were also tested using T283 and passed, indicating that they did not require anti-stripping agents, according to Arizona DOT's Larry Scofield.
To prevent experiences like Arizona's, AASHTO is funding a National Cooperative Highway Research Program (NCHRP) project to develop recommendations to make T283 compatible with the Superpave gyratory compactor. The project, "Evaluation of Moisture Sensitivity Tests" (NCHRP 9-13), now in progress at the University of Nevada-Reno, is scheduled to be completed this fall.
Several factors can make a pavement susceptible to stripping, but the most common are the characteristics of the aggregate and the way the aggregate and asphalt interact. That's why FHWA's John D'Angelo, who investigated the Route 93 project, says highway agencies should use caution when working with aggregates that are known to cause moisture susceptibility problems. "If a DOT knows from prior experience that the aggregate being used can create problems, even if tests show the mix to be acceptable, they should take corrective action," such as adding anti-stripping agents to the mix.
However, highway agencies do not always have enough experience with particular aggregates to know if they are likely to cause moisture susceptibility problems. When highway agencies use these aggregates in Superpave mixes, Scofield advises preparing 100-mm (4-in) samples for T283 testing using conventional compaction procedures.
For more information, contact John D'Angelo at FHWA (phone: 202-366-0121; fax: 202-366-7909; email: firstname.lastname@example.org).
This SPS-9 project in northwestern Arizona suffers from severe moisture damage.
During the Strategic Highway Research Program (SHRP), researchers recognized the need for a new way to test hot-mix asphalt mixes for their resistance to moisture damage. Their solution was the environmental conditioning system (ECS).
Although the ECS looked promising, it was costly and in the end less accurate at predicting a mix's susceptibility to stripping than the relatively inexpensive and widely accepted T283 test procedure, says Gerry Huber of Heritage Research Group and one of the original SHRP researchers.
SHRP therefore opted to stick with the T283 test procedure. In this pass-fail test, compacted samples of a mix are saturated with water, and then subjected to an optional freeze cycle if appropriate for the project location; the samples are then warmed to 60°C (140°F) for 24 hours, and the samples' stiffness is measured. If a mix becomes too soft, it fails the test, and the mix designer knows anti-stripping agents should be added to make the mix resistant to moisture damage.
Feeling confused about the long-term pavement performance (LTPP) program's Specific Pavement Studies 9 (SPS-9) experiment? The SPS-9 experiment was originally designed to provide field validation of the Superpave asphalt binder specification and mix design procedures. Today, it's a little more complicated-there are three types of SPS-9 sites.
In 1995, the original SPS-9 experiment was split into two related experiments, called SPS-9A and SPS-9B. The main purpose of the SPS-9A experiment is to validate the Superpave binder specification. All SPS-9A sites include a control section built using a State's conventional asphalt mix, a section built using a Superpave mix with a binder that meets the Superpave specifications for the conditions at the site, and a section built using a Superpave mix with a binder of a higher or lower performance grade than required for the site. The SPS-9A sites also provide an opportunity to compare the performance of Superpave mixes with States' conventional mixes. So far, 20 SPS-9A sites have been built in the United States and Canada.
The SPS-9B experiment was to focus on pavement structural factors and the Superpave performance prediction models. Because the performance prediction models are still under development, the SPS-9B experiment is on hold, says Monte Symons of FHWA's pavement performance division, which manages the LTPP experiments. Symons says the SPS-9B experiment will be designed with input from the Superpave models project (see Focus, April 1998) to ensure that the experiment will provide the needed data.
Nine SPS-9 sites, including the U.S. Route 93 project, were built when the Superpave specifications were still under development. These sites are therefore labeled SPS-9P, for "pilot."
For more information, contact Monte Symons at FHWA (phone: 703-285-2730; fax: 703-285-2767; email: email@example.com).