Long-life pavements have been constructed for years, but the use of the term "perpetual pavement" is fairly new. The concept basically assumes that once some minimum pavement thickness is reached, the pavement structure will never have to be replaced but will simply need to be maintained on some regular basis by milling and overlaying the wearing surface.

As the pavement structure becomes thicker, the strain under load at the bottom of the hot-mix asphalt (HMA) pavement layer decreases, thus reducing the damage due to loading and increasing the fatigue life of the pavement. It is believed that once the strain is lowered below some threshold, known as the endurance limit, then fatigue cracking as a result of tensile strain at the bottom of the HMA will never occur. A pavement in which the horizontal strain at the bottom of the HMA does not exceed the endurance limit and one that has sufficiently low stress on the subgrade to prevent deep structural rutting is considered to be a perpetual pavement.

There are many examples of pavements that were built years ago where the only maintenance that has been required to maintain the pavement has been milling to remove the surface deficiencies and application of a thin overlay to keep the pavement surface in good condition. These long-lasting pavements, while not designed to be perpetual pavements, are now considered as perpetual pavements based on their performance history. In fact, once some minimum thickness of HMA is exceeded in a pavement, the cracking and other performance issues will begin at the surface, not at the bottom of the HMA layer, and propagate down some depth into the pavement. This makes it easy to correct pavement surface problems by milling the top 2 to 3 in. of HMA and replacing the removed material with 2 to 3 in. of new HMA. For these thicker pavements, stresses from traffic and climatic conditions are more severe at the top of the pavement, resulting in the performance problems being initiated at or near the surface.

The best of the mix
When the NCAT Test Track was built in 2000, all of the pavement sections were designed with sufficient thickness to meet the requirements for a perpetual pavement. Actually, the thicknesses were greater than those needed for a perpetual pavement. The pavement thickness was selected to ensure that fatigue problems and structural rutting did not occur. The primary purpose of the experimental design during Phase I at the track was to investigate various material and mixture types so that state DOTs could use what they learned at the track to optimize the performance of HMA mixes by careful selection and control of various materials and mixtures. After 10 million equivalent single-axle loads (ESALs), there was no performance problem that was caused by an inadequate pavement structure; the only minor performance problems were caused by surface issues. Hence, these areas with minor surface problems could be maintained by simply removing the top 2 to 3 in. with a milling machine and replacing with 2 to 3 in. of HMA.

For flexible pavements, the primary requirements to meet the perpetual pavement concepts are for the strain level at the bottom of the HMA to be less than the endurance limit and the stress on the top of the subgrade to be sufficiently small to prevent structural rutting. However, another important factor is to ensure that the HMA surface provides a long life so that additional maintenance procedures are not required for a reasonably long period of time. If the surface has to be replaced too often, then, as far as the traveling public is concerned, this defeats the purpose of having a perpetual pavement structure.

Phase I at the track was designed to provide guidance on the best materials and mixtures to be used in HMA. This study showed that a wide range of materials could be used to produce a quality product as long as a good mix design is produced and as long as the quality of the mixture is adequately controlled during construction. Phase I of testing at the track was completed in December 2002.

Failure success
When the track was rebuilt in 2003, eight structural sections were included as part of the overall project, as shown in Figure 1. These eight sections were built using two binder grades and three different thicknesses so that failure of these sections would be expected at different times during the traffic cycle. Two sections were designed to have 5 in. of HMA, four sections were designed to have 7 in. of HMA and two sections were designed to have 9 in. of HMA. It was anticipated that these sections would fail at approximately 2 million ESALs, 5 million ESALs and 10 million ESALs, respectively. These designs were based on the engineering properties of the underlying materials using the 1993 AASHTO design procedures.

The properties of HMA should have an effect on the fatigue life, hence the variables in the mixes used in each of the structural sections were minimized. The primary mix variable for six of the structural sections was the asphalt binder performance grade (PG) level. It was decided that three of the eight sections would be constructed using a PG 64-22 grade of asphalt, three of the sections would be constructed using a PG 76-22 grade of asphalt and the other two sections would be constructed to evaluate other concepts related to perpetual pavements such as the use of stone-matrix asphalt (SMA) mixes on the surface and the use of a rich bottom layer.

Materials used in the structural sections were sampled to be included in a laboratory experiment to measure the endurance limit of the materials. Even though these sections were not designed to be perpetual pavements, the information collected will be very helpful in determining how much thicker these sections would need to be in order for us to consider them perpetual.

Strain gauges were installed at the bottom of the HMA layers in the field sections to determine the maximum horizontal tensile strain in the HMA. Data from these gauges, along with performance data, can be used to help establish the thickness needed for the pavement to be perpetual.

Pressure cells were installed on top of the subgrade and on top of the base course material. One use of this information was to ensure that the stress on top of the subgrade was sufficiently low so that structural rutting would not be expected.

Beginning in fall 2003, 10 million ESALs of traffic was applied to the Phase II track sections over a two-year period. The early traffic resulted in fatigue cracking and ultimately in failure in the 5-in. sections. The modified and unmodified sections for these thinner HMA sections failed at less than 2 million ESALs. Three of the four 7-in. sections performed all the way through the 10 million ESALs without any maintenance. However, these 7-in. sections had begun to crack at 10 million ESALs and it is expected that failure will occur with a relatively small amount of additional traffic. One of the 7-in. sections failed early and investigations are under way to determine the reason for this early failure. The two 9-in. sections had performed well after 10 million ESALs and it is expected that they will continue to perform well for many more ESALs. The plan for Phase III testing is to apply another 10 million ESALs of traffic to these 9-in. sections.

A gain on strain
Based on laboratory and test track work at NCAT, it appears that the endurance limit will be somewhere between 70 and 150 microstrain. Some mixes appear to have indefinite life at 150 microstrain while others fail reasonably quickly at 150 microstrain, indicating that the mix type has a significant effect on the fatigue life of HMA mixtures. At the conclusion of the Test Track Phase II research cycle, forensic investigations will be conducted to fully ascertain the factors affecting performance.

The structural sections were instrumented to measure the tensile strain at the bottom of the HMA. An example of the measured tensile strain under traffic is provided in Figure 2. The sections were subjected to traffic (Figure 3) on a near-continuous basis over a two-year period. Notice that, as expected, the measured strain is lower for the thicker sections and higher for the thinner sections. The strain for the 9-in. sections is approximately 150 microstrain, which indicates that these sections may be approaching the endurance limit for this pavement structure. The 9-in. sections will be subjected to additional traffic when traffic once again begins on the track near the end of 2006.

The work at the NCAT test track is helping to validate the perpetual pavement concept. As more sections are built and tested at the track and as additional data is obtained, design requirements needed to provide perpetual pavements can be more clearly identified.