Maine's Department of Transportation is actively researching how to improve road pavement quality and extend service life. Studies show that optimum pavement density is a crucial factor in increasing pavement life, and pilot tests on about a dozen projects show promising results.
Using GSSI's PaveScan RDM, a ground penetrating radar device that can accurately measure whether there is density uniformity throughout the pavement layer, researchers believe the equipment will be a powerful tool to help contractors make adjustments to ensure pavement quality.
Making the Most of Road Paving Budgets
Every state faces road paving funding restrictions. The State of Maine Department of Transportation devotes a significant portion of its budget to roadway and pavement improvements and needs to make the most of its limited resources. The state has a three-year $2.3 billion work plan; 2018 work included 48 miles of highway construction and rehabilitation at an estimated cost of $67.6 million; 356 miles of preservation paving estimated at $93.8 million, and 600 miles of light capital paving at a cost of about $22 million.
Dale Peabody, director of the Transportation Research Division at Maine DOT, explains that there is never enough funding to keep up with all the road pavement work needed. That’s why his team is constantly looking for ways to improve both the material mixes used and how they are laid in the field.
On the mix side, Maine DOT uses a variety of tools to test compacted asphalt mixtures, including the Hamburg Wheel Tracking Device (HWTD), and the Asphalt Mixture Performance Tester (AMPT), a computer-controlled hydraulic testing machine that subjects samples to cyclic loading over a range of temperatures and frequencies. They also recently began using MiST (Moisture Induced Stress Tester), a new method for testing moisture damage susceptibility of asphalt mixtures.
As for the materials themselves, Maine has conducted research on hot in-place recycling (HIPR); plant mixed recycled asphalt pavement (PMRAP) construction using asphalt emulsion and cement; and ultra-thin bounded wearing course (UTBWC) surface treatment, which nationwide research suggests reduces deterioration.
They are also using thin hot-mix asphalt (HMA) overlays to extend the life of pavement still in serviceable shape. Laboratory performance test equipment is used to ensure the mixtures will last longer and be more durable. They are also working on new lab tests that will be better at measuring performance with regard to moisture.
On the material lay down side, Maine DOT does the most it can to make sure the material is applied properly in the field. It specifies the equipment contractors must have, for example, materials transfer vehicles (MTVs), which ensure a more uniform and longer lasting road surface. They also include ride smoothness specifications to cut down on the undulations that would eventually lead to road performance issues.
Also included are specifications for joints; thermal profiling with an infrared (IR) scanner; and use of intelligent compaction (IC) rollers, which facilitate real-time compaction monitoring and timely adjustments to the compaction process by integrating measurement, documentation, and control systems. 
“We began implementing quality control/quality assurance (QC/QA) practices years ago, and we have trained and certified technicians sampling material on the roadway,” said Peabody. “It’s a continuous process, in which we have gone away from fixing the worst first to aggressive preservation practices. It’s just like how drivers don’t wait until a car breaks down before changing the oil.”
Uniform density throughout the pavement layer is critical
According to Maine DOT’s Peabody, optimum pavement density is the crucial factor in increasing pavement life. “Optimum density reduces oxidation, reduces moisture damage, and decreases rutting potential. It also offers improved fatigue life and increased load bearing capacity.” He cites past studies relating density to pavement life – the rule of thumb is that even a 1 percent decrease below minimum density results in a 10 percent loss of pavement life. 
Peabody has found that segregation (separation of the coarse aggregate particles in the mix from the rest of the mass) is the enemy of density – and is a major cause of premature pavement failures. There are two types of segregation: mechanical (physical or gradation) and thermal. Unfortunately, segregation is often only identified visually and is a subjective value that is difficult to quantify. It may not even be apparent at the time of construction, making it very difficult to enforce contractually.
In the past Maine has used nuclear density gauges to measure density but has moved away from this type of technology. Contractors now use a non-nuclear asphalt density gauge for QC purposes and cores are collected sent to a lab to determine voids/density acceptance. “The down side of cores is limited sampling and not having real time values.”
National Highway Research Solutions Include Use of GPR
The issue of pavement density has been extensively researched as part of the Federal Highway Administration’s Strategic Highway Research Program (SHRP2).  SHRP2 eventually came up with two non-destructive techniques for evaluating asphalt pavements during construction: infrared thermal scanning and the use of ground-penetrating radar (GPR), which uses electromagnetic wave reflection to “see” through materials.
According to SHRP2 studies, GPR can be used to measure uniformity and potential defect areas in asphalt pavements during construction. Significantly, it offers real-time testing of potentially 100 percent of the pavement area. Compare that to current density tests, in which typical random sampling measures only about 0.003 percent of pavement area.
Widely used for many applications, GPR is a common tool for utility location, measuring pavement thickness and bridge deck deterioration. For decades, researchers have been investigating whether it could be used for measuring pavement density, but they could never achieve the level of accuracy that would warrant recommending its widespread use. In addition, the use of GPR technique previously required specialized equipment, a great deal of data interpretation, and a number of manual steps.
In 2013, SHRP2 funded work aimed at developing a streamlined and operator-friendly GPR device that would provide real time profiling of asphalt mixture uniformity. The research was done by the Texas Transportation Institute (TTI), which worked with GSSI to develop the technology into the PaveScan RDM asphalt density assessment system.
The non-contact PaveScan RDM technology uses a sensor that typically outputs a measurement each half-foot along the lane traveled, so a mile’s worth of data includes roughly 10,000 measurements for each sensor used. This system is ideal for uncovering inconsistencies that occur during the paving process, including poor uniformity and significant variations in density.
To arrive at pavement density, PaveScan RDM measures the dielectric properties of the asphalt surface. The dielectric constant is the ability of a substance to store electrical energy in an electric field. For example, air dielectric is 1.00059; asphalt aggregate is about 3 to 6; while the dielectric of water is 80. With new pavement, the mixture is uniform; dielectric variation occurs primarily due to the percentage of air voids – which directly correlates to density. The measurement is based on the ratio of reflection from the asphalt surface to the reflection from a metal plate.
On the road in Maine
Maine is one of several states that have conducted pilot studies using PaveScan RDM equipment. The PaveScan RDM uses one or three 2 gigahertz (GHz) sensors mounted on a portable push cart that can scan a of up to a 6-foot width. Each antenna collects a continuous line of dielectric/density. An onboard computer captures dielectric values, which can be correlated to core densities.
Operators scan a pavement section and the device identifies high, low, and median density locations. They take a static reading directly over each location, obtain cores at each location, and then test the cores, entering the results in the software. Correlation accuracy depends on obtaining core densities over the entire range of measured dielectric values.
To date, Maine DOT has collected data on about a dozen recent paving projects and found that the data aligns well with what they have seen in the field. Says Peabody, “We were looking for a better way to ensure we’re achieving desired density ranges, since studies show that good density leads to improved service life. PaveScan RDM gives a much larger sample of the HMA mat and can be used as a QC/QA tool. The results to date have been really positive.”
It helped that Maine DOT had a crew that was already familiar with GPR, making it a relatively easy transition to get up to speed. The crew found the PaveScan RDM equipment to be easy to set up and use. The only limitation they found is that the battery does not last long enough for a full day of data collection. Other recent adjustments include a laser pointer that can be attached to the equipment to help operators align the scanning equipment, especially for use in night work. Operators found the extra lighting to be very helpful.
However, Peabody notes that the technology does have some boundaries. It is affected by surface moisture, does not work as well when temperatures dip below 40°F, and can be affected by mix constituents, which may happen with a change in aggregate source. Measurement accuracy for layers of less than 1-inch may be affected by the underlying layer, while layers that are 2.5 to 3-inches may be affected by density gradients within the layer.
Maine DOT is also looking for further enhancements. For example, some users are adapting the technology for use with vehicle mounts. This issue is one that GSSI is actively pursuing for Spring 2019. Other items on the wish list are better ways of using the technology for longitudinal joints; incorporation into the Veta intelligent construction software – a map-based tool for viewing and analyzing geospatial data; and better data analysis of intelligent compaction, thermal profile, and GPR density data.
Peabody’s colleague, Rick Bradbury, has presented the positive results of the pilot studies at several SHRP2 workshops and has shared the information with New England DOT colleagues. He also recently presented the information at the 2018 North East Asphalt User/Producer Group (NEAUPG), which is focused on improving the quality and performance of asphalt pavement applications in the Northeastern United States by promoting communication, knowledge, technology, and uniform solutions.
Special focus on data analysis
While not unique to this technology, one of the key factors affecting whether Maine DOT would want to implement the technology statewide is the issue of data analysis. “Who is going to collect all this data?” asks Peabody. “If there are 20 projects, you can’t have just one crew collecting data. We have thought about asking the contractor to purchase the equipment and use it but we are not ready for that, because for it to work you have to have people who really understand the technology.”
Ultimately, he believes it would be best if this type of data could be collected in a moving vehicle; the vehicle could then go project to project and collect data. However, Peabody acknowledges that this approach would have a downside – the lack of real time data contractors could use to make adjustments as they are going along. “If they collect data but don’t look at it until the next day, that does not really give the contractor a chance to make adjustments if there are density issues. I’d like a way to share data with the project team in real time, other than only being able to show the display on the Toughpad.”
According to Peabody, the larger question that must be answered is how to effectively manage the data to make near-real time changes during production, and to get timely reports to project personnel. “Currently, there is no good way to get data in the right people’s hands so they can make adjustments on the fly. There is a lot of benefit to having that ability. They could collect data and then go back to specific points on the road. They could select locations of high and low dielectric/density and then cut a core there and send it to the lab to determine the actual density.”
GPR Surveys Show Good Correlation Between Dielectric and Air Voids
The PaveScan RDM surveys show a good correlation between the dielectric value and the air void contents. Maine DOT can use the surveys to quickly identify and investigate low density areas and to check the compaction consistency. While data management is a huge effort, they believe the surveys will be very useful, in conjunction with other new technologies, including intelligent compaction and pave IR.
 What Is Intelligent Compaction?, http://www.intelligentcompaction.com/learn/intelligent-compaction-fundamentals/what-is-intelligent-compaction/, retrieved 11/20/18.
 HMA Compaction Assessment Using GPR Rolling Density Meter, Rick Bradbury, MaineDOT, Shongtao Dai, MnDOT, Atlantic City, NJ, October 17, 2018.
 Strategic Highway Research Program, Transportation Research Board of the National Academies, Stephen Sebesta and Tom Scullion, Texas A&M Transportation Institute, The Texas A&M University System, College Station, Texas, © 2014 National Academy of Sciences. All rights reserved.