At Boston-Logan International Airport in East Boston, Mass., it’s all about pavements for a recruited team of engineers. More specifically, it’s about maintaining the pavements for Logan’s complex arrangement of runways, taxiways and aprons.
Owned and operated by the Massachusetts Port Authority (Massport), Logan Airport is the 19th busiest airport in the nation, handling over 400,000 aircraft operations and 27 million passengers annually. The majority of the pavements are constructed of hot-mix asphalt (HMA), supplemented with portland cement concrete aprons for aircraft parking at the terminals.
For the safety of aircraft and passengers, it is most important to keep these pavements in prime condition. Pavement rehabilitation and reconstruction projects can cause delays for the airlines and increase the workload of air-traffic controllers.
To minimize the effects of these expensive disruptions of service, the pavements must be designed and built to withstand the effects of traffic as well as harsh environmental conditions. Recent problems with moisture-induced damage in their HMA pavements prompted Massport to initiate a collaborative research effort that included reaching out to the highway community.
Supporting role
The pavements at Logan must support loads up to 875,000 lb for a Boeing 747 at maximum takeoff weight. Tire pressures can be in excess of 200 psi and the traffic is highly channelized. Combine these factors with heavy aircraft turning sharply, taxiing slowly, braking or standing in queue during hot summer weather and you have a recipe for pavement distress.
Pavements are designed according to Federal Aviation Administration (FAA) standards for a 20-year service life, but in reality the surfaces of Logan’s most heavily used HMA pavements last only 10 to 12 years. Prior to the proliferation of wide-body aircraft, pavements were rehabilitated primarily for cracking and oxidation due to age. By the early 1990s, with increasing numbers of heavier aircraft, Logan’s pavements constructed with AC-20 binder, the same binder used for highway work, were experiencing rutting and shoving, particularly in the summer months.
Since 1995, various modifiers such as Trinidad Lake Asphalt (TLA) and polymers (e.g., SBS) have been added to the liquid asphalt binder to stiffen the hot mix with some success. However, modifying the asphalt binder seemingly makes the HMA sensitive to moisture damage in the form of stripping, where there is a loss of cohesion between the binder and aggregate.
Aggregate Industries of Saugus, Mass., which supplies most of the HMA for Logan Airport and the surrounding streets and highways, confirmed that road mixes using an unmodified PG 64-28 binder (equivalent to the old AC-20 binder) continue to pass the specified moisture sensitivity requirements without the aid of an antistripping agent and are performing well. Not so for the modified HMAs used at Logan between 1995 and 2003.
While these mixes are performing adequately in less-traveled areas, the typical failure on a high-volume taxiway is the development of a pothole. Either the modified mix at the surface fails or the underlying layer fails, indicating that moisture has penetrated the modified surface course.
This moisture appears to be trapped in layers that were constructed after the mid-1970s when rubber-tired rollers, which help to seal the surface, were replaced with steel-wheeled vibratory rollers. Investigation of some of the early failures indicated jet fuel and hydraulic fluid spills were not the culprit.
Unfortunately there is no lack of moisture at Logan Airport, which is surrounded by water on three sides and was built on fill in Boston Harbor. Boston also receives the equivalent of 3 to 4 in. of rain every month of the year. With summertime pavement temperatures in excess of 130ºF, abundant moisture and heavy slow-moving traffic, stripping is just about inevitable in a moisture-sensitive pavement. The heavy traffic areas at Logan are being permanently repaired by milling and inlaying the wheel tracks with a modified recycled asphalt pavement (RAP) mix.
Peeling the problem
The stripping began in earnest in 2000, resulting in the partial reconstruction of the taxiway that services the hot weather departure runway. In 2001, Massport switched to a polymer-modified PG 76-28 binder mix for all paving projects. To combat the stripping problem, Massport, with the FAA’s concurrence, modified the specifications to include the addition of a liquid antistripping agent and a higher retained tensile strength for the moisture susceptibility test (AASHTO T283)—the freeze-thaw test. Despite using an antistripping agent and changing the minimum acceptable tensile strength ratio (TSR) from 75 to 92%, it was evident by the end of 2003 that even this HMA was vulnerable to moisture damage, and its use was discontinued.
Faced with the dilemma of a prematurely failing mix that satisfied enhanced FAA specifications, Massport initiated an ongoing comprehensive research program. The partners and participants are Project Manager Bob Pelland of Massport’s Capital Programs Department, consultants Ned Dawes of Edwards & Kelcey and Chris Bowker of Bowker Consulting, Prof. Rajib Mallick of the Worcester Polytechnic Institute (WPI) Civil Engineering Materials Testing Laboratory and Civil Engineer Victor Lung of the FAA. Their goal is to develop further revisions to the FAA specifications for screening out poor-performing HMAs and to develop more durable HMAs that are readily achievable with local materials.
Help from the highway side came with the recommendation to use lime as an antistripping agent. Conversations with Ray Robertson at the Western Research Institute led to discussions with Tom Aschenbrenner of the Colorado DOT, Ron Collins formerly of the Georgia DOT, Dale Rand of the Texas DOT and Julie Nodes of the Arizona DOT. Their state roads have benefited by the introduction of lime to combat stripping. Besides being an effective antistripping agent, Robertson explained, lime also will stiffen a mix to make it more rut-resistant and slow down the aging effects of oxidation. Dry hydrated lime can be added at the plant during batching, or the aggregate can be marinated with a lime solution.
Research began in the WPI laboratory by fabricating samples of an SBS-modified PG 76-28 mix with aggregate from several local sources, including the traprock used predominantly at Logan Airport. No antistripping agent was used. The FAA requires only one freeze-thaw cycle to determine the TSR value, but the WPI test subjected the samples to 10 cycles. Even after 10 cycles, with no antistripping agent, the Logan sample still met the FAA minimum TSR value of 75.
Similar samples were then subjected to an accelerated loading test under hot, wet conditions. The Model Mobile Load Simulator (MMLS3), developed by Prof. Fredrick Hugo of South Africa, is a one-third-scale device that uses four rubber tires (100-psi pressure, 600-lb load) and a 140ºF water bath. After 80,000 cycles, the test on the Logan sample was terminated when loose particles were clogging the hot-water circulation system. Visual inspection confirmed severe stripping. All the samples also were evaluated for rut depth and indirect tensile strength after 80,000 cycles. The first phase of testing clearly showed that the freeze-thaw test alone, which measures distress caused by volumetric contraction and expansion, is not a good indicator of performance of the local mixes.
For the second phase of testing, the same tests were re-run with 1.1% of lime as an antistripping agent. The addition of lime unquestionably improved the performance of all the samples, regardless of the aggregate source, with reduced stripping and rutting under hot, wet conditions and improved freeze-thaw TSR values.
The beneficial effects of lime are well documented, but for Massport it was reassuring to know that local mixes not known for stripping until modified could be improved with the addition of lime. All the HMA used at Logan Airport since 2003 has included lime. The test results also show that the hot, wet accelerated loading test is a useful tool for determining distress caused by traffic loading and moisture at high temperatures.
Of the dozens of paving projects done at Logan in the last 30 years, all but four were supplied by the asphalt plant closest to the airport. With this in mind, additional research at WPI concentrated on studying HMAs using aggregate from this local supplier in the quest for the “everyday” HMA. Various polymer-modified binders with PG ratings of 70-28, 76-28, 82-28 and a fuel-resistant binder by Citgo were tested as well as aggregate gradations of 3?4-in. and 1-in. maximum. The fuel-resistant mix performed extremely well in the lab and has been used at Logan successfully on the first 400 ft of the hot-weather departure runway and in a busy terminal area. Because of its high cost, this mix is used for special applications only.
Also tested was a RAP mix, known locally as the “Ronnie” mix for its developer, Ron Tardiff of Aggregate Industries. It is composed of an unmodified PG 64-28 binder, 18% RAP, latex at 4% of the liquid binder content and 1% lime.
With the requirement for a tough mix made from readily available materials, the original RAP mix with a liquid antistripping agent was first used in 2001 for a repair on one of Logan’s busiest taxiways and it is still performing well. Evolving from a mix used just for repairs, this mix, now with lime, has become Logan’s everyday mix since 2004. In the laboratory it compares well with the PG 76-28 mixes for rut and moisture damage resistance.
Testing has shown little difference between the 3?4-in. and 1-in. max mixes, so Massport has opted for the finer gradation with its tighter finish for surface course applications. Massport continues to weigh the merits of a RAP mix verses an all-virgin material mix with regard to cost/benefit and quality control, where variations in RAP sources could affect performance.
Newcomers welcome
Although there is some assurance that a good performing mix can be produced by local suppliers, there is always the possibility that new aggregate and liquid asphalt sources will be used. To vet new and old materials, Massport would like to include the MMLS3 hot, wet accelerated loading test in the specifications as a supplement to the freeze-thaw test.
The question becomes choosing the performance criteria and the pass/fail values needed to distinguish between a good and a bad HMA. A database must be accumulated in order to correlate damage observed in the loading device to actual field performance. State DOTs that use the Hamburg Wheel Tracking Device, an accelerated loading machine not available in New England, have had to do likewise.
For the MMLS3 performance criteria, Massport is focusing on a maximum allowable rut depth of 0.12 in., a minimum retained indirect tensile strength of 145 psi and a maximum allowable stripping of “moderate” (defined as retained coating on 60 to 90% of the sample’s surface). These pass/fail values were developed from tests at WPI on laboratory-fabricated samples and field cores taken from known good- and bad-performing airfield pavements. Before the new specification is put in place, these values will be further refined as the database grows. The new specification will require cores to be taken from the contractor’s test strip for testing in the MMLS3. The samples will have to meet the established criteria before the contractor will be allowed to swing into full production. This process will add approximately one week to the approval process prior to the start of construction.
While much has been accomplished by Massport’s task group, they recognize that the testing and research program started three years ago will continue to be a work in progress. Some of the lessons learned are that a well-established pavement specification may not always account for changing local conditions and that help from outside the airport community can be just a phone call, or perhaps a road trip, away.