Maximum Asphalt-NCAT Test Track Findings

By E. Ray Brown, from the November 2006 issue of Better Roads Magazine

The test track was built to develop and evaluate better ways to design and construct hot-mix asphalt pavements.

Construction of the National Center for Asphalt Technology Test Track was originally completed in 2000 and t raffic was initiated shortly thereafter. The track is a 1.7-mile oval located near NCAT’s headquarters at Auburn University, Alabama. It is an accelerated loading facility where materials, mix designs, and structural designs are tested. In addition to specific studies funded by sponsors, the track supports research to solve national and local problems. Another benefit of the track is that it has helped to identify reliable tests that can predict rutting performance.

Testing for Phase I was completed in 2002. In those two years, the pavements were subjected to 10 million equivalent single axle loads, or ESALs. (One ESAL is equal to one 18,000-pound single axle load.) The track was rebuilt for Phase II in 2003 and testing was completed in 2005. Construction for Phase III of the track was under way at the time this article was written.

Funding for construction and operation of the test track was provided by the Alabama Department of Transportation, the Federal Highway Administration, and a number of state Departments of Transportation. Based on the amount of funding from each sponsor, each was provided with an appropriate number of test sections during the initial construction of the track. The sponsor was allowed to select the materials and mix types to be used in their sections. Some states shipped in their locally available aggregates and binders. For Phase I, the pavement design thickness was uniform for each section and could not be changed.

This drawing illustrates the layout for Phase II of the NCAT Test Track:
gold = structural sections; yellow = rutting studies; white = left in place.

Following the completion of the initial three-year cycle of testing, sponsors were again given the opportunity to plan research that best fit their needs for the second cycle. Many states elected to leave some or all of their original sections in place for additional traffic and extended performance evaluations. Twenty-three of the original 46 test sections were left in place for continued evaluations and 22 new sections were built in 2003. For Phase II, the sponsors had the option to modify the pavement thickness so that structural studies could be done.

Of the 22 new test sections, eight were used for a structural experiment by removing the existing pavement all the way down to uniform subgrade materials (approximately 30 inches) and rebuilding the pavement structure with varying thicknesses and materials. The other 14 sections were shallow mill and inlay (between 0.75- and 4-inches deep) rutting study sections.

Eight test sections were sponsored for construction
of a structural experiment on the 2003 track.

The primary objective of Phase II was to evaluate field performance of several experimental pavements. Rutting was expected to be minor in sections that were originally built in Phase I and subjected to a second round of traffic. The eight test sections (N1 through N8) in the structural experiment were monitored for structural distresses (primarily fatigue cracking). The goals of the structural experiment were to help validate mechanistic pavement design concepts and to learn more about characterization of pavement materials and evaluation of pavement responses. For the 14 sections that were mill and inlay sections, rutting was the anticipated distress. In addition to evaluation of field performance for individual sections, a goal of the project was to evaluate the potential to predict performance.

Experimental design

Many sponsors chose not to replace their test sections for the 2003 track so they could extend their rutting comparisons to 20 million ESALs and broaden performance comparisons to include durability. Weekly field-performance testing was conducted to characterize how rutting, roughness, texture, density, and friction changed as traffic accumulated beyond the 10 million ESALs originally applied.

Fourteen sections were milled to a depth of 0.75 to 4 inches as specified by the research sponsors. While some states wanted to conduct another full-depth (4 inch) rutting experiment, other states chose to compare the performance of shallow mill and inlay pavement preservation techniques. In cases where full-depth rutting mixes were placed, stone-matrix asphalt was a mix type that was often used. Several sponsors investigated the effect of relaxing aggregate specification requirements on the performance of SMA. For example, will mixes using softer aggregates exhibit more production and performance problems than mixes produced with harder aggregates? In cases where pavement preservation studies were planned, shallow mill and inlay methods were required. In these comparisons, sponsors evaluated various thin overlay options to determine which ones were most cost effective.

Eight test sections were sponsored for construction of a structural experiment on the 2003 track. The two primary experimental factors were HMA thickness and modified versus unmodified asphalt. The structural sections were instrumented to measure temperature at various depths in the HMA, longitudinal and transverse strain at the bottom of the HMA, and pressure applied to the top of the aggregate base and subgrade. The pavement was also instrumented to measure truck wheel wander as the trucks were driven over the eight structural sections.

Construction

The Phase II project was developed, let, and administered by ALDOT under the guidance of an oversight committee on which the sponsors were represented. East Alabama Paving Company was the low bidder on the track reconstruction project. The company chose to produce the hot mix at their plant located approximately 10 minutes away. They used the track’s prepared plant site as a staging area for out-of-state aggregate stockpiles. (Some of the sponsoring state DOTs sent their own locally available aggregates to the track for testing.)

Test sections were constructed to have three different thicknesses of
hot-mix asphalt so that the sections could be expected to fail at different times. 

When mix was produced for placement on the surface of the track, a number of samples were fabricated in the laboratory using a Superpave gyratory compactor. These samples were compacted to the sponsor-designated design gyration level for laboratory performance testing of various types. Additionally, a large amount of loose material was stored in metal buckets so that more samples could be tested at a later time.

This graph shows the different pavement layers
on eight test sections built during Phase II.

The structural sections were constructed to have three different thicknesses — 5, 7, and 9 inches of HMA. These thicknesses were selected so that the pavement would be expected to fail at approximately 1 to 2 million ESALs, 5 million ESALs, and 8 to 9 million ESALs, respectively. There were two sections at each thickness, one with a PG 76-22 modified asphalt binder and one with a PG 64-22 unmodified asphalt binder. There were two additional sections with 7 inches of HMA. One of these sections included an SMA surface and the other section included a rich bottom layer along with an SMA surface. The two thinner sections (N1 and N2) failed at approximately 2 million ESALs and there was no significant difference in the performance of the section with PG 64 and that with PG 76. The remaining sections had not failed at the end of 10 million ESALs, so they will be left in place for an additional 10 million ESALs in Phase III. It is expected that the 7-inch layers will have to be overlaid fairly quickly once traffic begins, but the 9-inch sections are expected to continue for a significant amount of time. Hence, the sections in the structural study have performed better than expected based on the 1992 AASHTO pavement design procedure, indicating that it may be conservative.

Findings and Implementation

For the test track to be beneficial, clear findings must be identified and implemented as a result of work at the track. Much of the supporting information can be found in various reports on the NCAT Web site at www.ncat.us. Additional information can be found at the test track’s Web site at www.pavetrack.com. Test track findings include:

Grinding to Improve Smoothness. Eleven of the transverse joints built during the original construction, and several additional joints built during the 2003 reconstruction, were ground to remove a bump at the transverse joints. The grinding process performed on these joints resulted in a very smooth surface across the joint. Of the 11 joints that were leveled with the grinding equipment during the first cycle of tests, none had performance issues during the initial two years of traffic. Some of these leveled areas have now been in place for up to six years with no performance problems. No sealing was provided to these treated surfaces.

Fine- and Coarse-Graded Mixtures. Based on observations at the test track, at least three state DOTs have begun to use more fine-graded mixes. The track showed that these mixes would provide good resistance to rutting and experience has shown that the fine-graded mixtures have less permeability than the coarse-graded mixtures.

Grade Bumping. Superpave guidelines recommend that the high-temperature PG grade be bumped for higher-traffic-volume roadways to minimize rutting. The results from the first cycle of testing indicated that on the average, there was more than a 50% reduction in permanent deformation when the high-temperature grade was bumped from PG 64 to PG 76. As a result of findings at the test track, some states have increased the number of projects requiring grade bumping.

Use of SMA. Stone-matrix asphalt mixtures have been used in the U.S. for almost 15 years with very good results. Since SMA was adopted in the U.S., one of the requirements has been to use only crushed stone. Evaluations were conducted at the track using crushed gravel in an SMA mixture, and it was determined that the SMA mixture performed well. As a result of this finding, two state DOTs have begun producing SMAs with their local gravel aggregates, thus making the use of SMA a viable option. Another DOT has begun to specify the use of SMA on projects. The test track work, as well as SMA performance in other states, has bolstered confidence in SMA mixtures for these states.

Open-Graded Friction Courses. As a result of the performance of open-graded friction courses at the track, two states that had not been using OGFCs have begun to use these mix types to minimize hydroplaning. The OGFCs at the track have performed very well, and it is clear during a rainstorm that drainage is improved and the amount of splash and spray kicked up by trucks is significantly decreased.

Predicting Permanent Deformation. There is interest in the pavement engineering field to identify a reliable test or tests to predict rutting performance. NCAT conducted several performance tests on the mixtures placed at the track including dynamic modulus, repeated load tests, and wheel tracking tests. As a result of the testing at the track, one state DOT has gained confidence in their newly implemented specification for a wheel tracking test. That confidence would have taken 10 to 15 years to develop without some type of accelerated testing. In general, the tests have shown that the two best methods for predicting rutting are the repeated-load permanent deformation test and the various types of wheel tracking tests.

Increasing Asphalt Content. At least two state DOTs have taken action to increase the amount of asphalt in their designed mixes. The track showed that additional asphalt can be added to mixes designed according to the Superpave method, especially when the asphalt is modified, without experiencing rutting problems. Generally the amount of asphalt is increased by lowering the number of gyrations specified for mix design. For Phase I at the track, the number of gyrations used was 100, but this was reduced to an average of approximately 75 (it varied from state to state) for Phase II. No problems were observed as a result of this reduced gyration level.

Aggregate Quality. Four state DOTs have evaluated local aggregates and made decisions about the utilization of those aggregates based on results from the test track. One state tested a relatively soft aggregate. It performed so well that the state now allows its use. In another case, an aggregate was shown to polish and it is not allowed in HMA.

Smaller Top Size Mixtures. Two states have evaluated performance of a 4.75-millimeter mix and have begun using these fine mixes based on performance at the track. The performance of these fine mixes on the track has been excellent. One advantage of these finer mixes is that they can be placed in thinner sections, resulting in a saving in cost to overlay a given roadway. Given recent increases in the price of asphalt, this approach of using thinner layers may become much more important.

Other Research. While the primary purpose of the track is to conduct the specific studies funded by the track sponsors, the track has supported additional research to solve national and local problems. Some of this work has involved effects of various mixtures on noise, friction, permeability, compactability, and performance testing. For example, the primary data for NCHRP 9-27, which developed recommendations for the minimum thickness of HMA layers, was developed during reconstruction of Phase II of the test track. Past and current work at the track is developing data that can be used in the design of Perpetual Pavements and warm asphalt as well. Other studies have also investigated performance of various types of tires and synthetic fuels.

E. Ray Brown, P.E., Ph.D., is director, National Center for Asphalt Technology.