Connecticut Department of Transportation (ConnDOT) specifications allow the routine use of Recycled Asphalt Pavement (RAP) in hot-mix asphalt (HMA) pavement at less than 15 percent by mass of the mix. Higher amounts can also be utilized with approval of the Materials Testing Division. A few of Connecticut's HMA contractors have used RAP in conventional Marshall mix designs. With the pending implementation of the Superpave system of mix design, ConnDOT officials felt that RAP must be allowed in these mixes. Since the original research did not address the use of RAP in Superpave mixes, it was desired to proceed with the use of RAP in Superpave mixes on a trial basis. A research project was developed to monitor and evaluate a Superpave mix which included 20 percent RAP on the westbound travel lanes of a 40 km long four-lane pavement overlay project.
Connecticut's first large-scale Superpave project was constructed on a 10 km section of State Route 2 extending from exit #21 to exit #23 and traversing sections of the townships of Colchester, Lebanon and Bozrah in southeastern Connecticut between May and October 1997. The construction project involved removal of the existing top 50 mm of an HMA overlay placed in 1986, the placement of 25 mm of a standard ConnDOT Class 2 leveling course meeting Marshall criteria, and the placement of a 63 mm Superpave surface layer. Traditional Class I mixes were used for control purposes. Six mixes, four Superpave and two conventional, were utilized for the surface layer. C was used for the design. Average annual melted precipitation in Colchester is 1220 mm, with approximately 750 mm of snow.
State Route 2 is a four-lane median divided highway functionally classified as a principal arterial. It carries from 15,000-18,000 vehicles per day with 10 percent trucks. It was originally constructed in 1970 as a full-depth HMA pavement and subsequently overlaid in 1986. For the Superpave system, a fifteen year design life of between 1 and 3 million 80kN ESALs was calculated. A maximum 7-day air temperature of 39°.
The RAP used in the project was material that was milled off the existing roadway. The 1986 pavement was composed of basaltic coarse aggregate with a maximum size of 25 mm, natural fine aggregate, and 5.2 percent AC-20 asphalt cement. ConnDOT's conventional approach for testing extracted asphalt from RAP for viscosity and penetration was used. Additional tests performed on the RAP by the design consultant included specific gravity of aggregate, gradation, coarse aggregate angularity and fine aggregate angularity.
Two Superpave mixtures with RAP and one conventional pavement with RAP were designed for the westbound direction of the project. In addition, three sections using virgin materials (two Superpave and one conventional design) were placed in the eastbound direction. Both Superpave RAP mixes had the same aggregate gradation, but different Superpave binder grades. The mix design called for the final Superpave binder grade, after addition of RAP and new asphalt, to conform to PG 64-28 and PG 64-22 requirements which correspond to 98% and 50% reliability, respectively. In order to meet the final Superpave binder requirements, a PG 58-34 with a modifier and an anti-strip agent was used to obtain a PG 64-28, while an unmodified PG 58-28 with anti-strip agent was used to achieve a PG 64-22. The anti-strip agent was required for both Superpave mixes after the result of AASHTO T-283 showed the potential for moisture susceptibility. This was an unexpected finding since the same aggregates have been used for many years in Connecticut with minimal stripping problems. The anti-strip agent was mixed at a rate of 0.375 percent of binder.
The method used by the designer for determining the PG grade of binder that would be added to the RAP to obtain the required PG 64-28 and PG 64-22 was empirical. In the past, an asphalt cement equivalent to an AC-10 was typically used with RAP mixes in Connecticut. An AC-10 is approximately equal to a PG 58-28. After blending, the extracted asphalt cement from the RAP with the virgin PG 58-28 asphalt, a PG 76-22 resulted which was deemed acceptable for the project. To meet the resultant PG 64-28 for the other section, it was decided to drop the low and high end one binder grade and use a PG 58-34 based upon guidance from the FHWA Superpave Mixtures Expert Task Group.
As many as thirteen trial blends were made in order to meet the criteria for voids, Nini, and field compaction. The final aggregate gradation of the two mixes passed below the restricted zone on the 0.45 power gradation chart. The final mix design was 20 percent RAP; 5.0 percent total asphalt (4 percent virgin); 3.9 percent voids; 14.3 percent VMA; 72.8 percent VFA; dust/asphalt ratio of 0.7; Gmm at Nini = 87.2 percent; Gmm at Nmax = 97.4 percent.
Milling of the existing pavement began on April 29, 1997. Paving of the Class 2 leveling course began on May 14, 1997. The first surface layer placed was a conventional Class I mix without RAP in the eastbound direction. Placement of the first Superpave RAP section occurred on August 11, 1997 after all of the virgin mixes were completed. All paving was completed by September 10, 1997. A total of 13 290 megagrams of Superpave with RAP were placed. Total tonnage of all mixes placed was 38 823 megagrams.
The contractor used a 3.6 megagram Cedar Rapids batch plant located in Montville, CT. On some days, the mixes were stored in silos before being transported to the project site. The batch plant was modified to allow the RAP to be incorporated into the pugmill. The RAP was loaded via front-end loader from the stockpile to aggregate bins at prevailing moisture. It was sieved through a 50 mm scalper screen and then transferred to the weigh hopper via aggregate conveyor belts. The virgin aggregate entering the batch plant mixing chamber at between 215°C - 230°C.
The asphalt fed to the plant already contained the anti-strip agent, and for one mix, the modifier. All blending of the asphalt cement with the anti-strip and asphalt modifier took place at the asphalt supplier in Rhode Island. At the job site, which was 15-25 km from the plant, conventional paving methods were used for placement of all the mixes. After application of a tack coat at 0.09-0.18 liter per meter squared, a Blaw-Knox PF 180-H paver was used for paving. An 11 megagram Hyster C766A double-drum vibratory roller was generally used for breakdown rolling. During some periods, a 12 megagram Caterpiller CB 614 vibratory roller was used for intermediate rolling. A 14 megagram Hyster C350C roller in the static mode was used for final compaction. The contractor was responsible for all quality control, which included the laboratory tests on molds from the Superpave Gyratory Compactor and monitoring of field density. ConnDOT also monitored density and performed laboratory tests for quality assurance.
This project is a participant in FHWA's LTPP SPS 9A study, "Verification of SHRP Asphalt Specification and Mix Design." As such, an extensive amount of monitoring is scheduled for at least four years. Pavement cores 150 mm in diameter are scheduled to be taken at intervals of 6, 12, 18, 24 and 48 months. These cores will be tested for maximum specific gravity, bulk specific gravity, asphalt content, aggregate gradation, volumetric properties, as well as tests on the recovered asphalt cement for penetration, viscosity, dynamic shear, creep stiffness and direct tension. Performance surveys will be performed annually for skid resistance, rideability rutting, deflections, and visual distress. Continuous traffic and weather conditions will also be monitored via a weigh-in-motion system and a Roadway Weather Information System installed at the project in Lebanon.
Overall, the pavements were placed without problems. However, achieving field density of greater than 92 percent maximum theoretical required more attention than the conventional mixes. Compaction appeared to be dependent on air and mix temperatures. The Superpave RAP mixes were more easily compacted when the ambient air temperature was below 24°C. The mix became tender when the mat temperature was between 93°C and 126°C. Note: On some Superpave mixes without RAP, a tender zone corresponding to mat temperatures between 93°C and 115°C (temperature range varies from mix to mix) has been found. This tender Superpave mix can be satisfactorily compacted above and below the tender temperature zone. The preferred compaction method is to obtain density before entering the tender temperature zone by adding additional rollers and increasing the compactive effort or changing the rolling technique. Another alternative would be to use a steel-wheel vibratory roller above the tender temperature zone, stop compaction efforts while the mat temperature is within the tender zone, and then finish the rolling process before the mat temperature reaches 80°C.
There is concern that blindly reducing the PG grade by one level on both the high and low end as recommended by the FHWA Superpave Mixtures Expert Task Group could lead to performance problems. The source, and particularly the age, of the RAP should ultimately determine the proper grade of virgin asphalt to be used. However, using blending charts and determining the PG grade of extracted asphalt cements proved difficult on this project.
In the past, stripping was deemed a problem only at isolated locations in Connecticut. There was some question on this project about the reliability of the AASHTO T-283 test for detecting moisture susceptibility.
For more information on this project, please contact Mr. Timothy Lewis , of FHWA at (202)366-4657, or Mr. Keith Lane, Director of Research and Materials at ConnDOT at (860)258-0371.
User Guidelines for Waste and By-Product Materials in Pavement Construction
Guidelines for Superpave Mixes Using RAP
Connecticut Department of Transportation