  |
   |
 |
 |
 |
 |
 |
 |
 |
 |
 |
November/December 2002
Digging into LTPP Pavement Data
by Antonio Nieves Torres and John J. Sullivan IVWhat can you do with the most extensive pavement study ever conducted? That's the question Roger Smith, associate professor at Texas A&M University, asks the students in his pavement engineering courses as they begin working with the Long-Term Pavement Performance (LTPP) database.
Smith and other faculty members in the United States and internationally recognize the unique value of the LTPP database to the pavement industry. The world's largest collection of pavement performance data ever compiled, the database contains nearly two decades of data on climate, materials tests, maintenance, rehabilitation, traffic, and pavement monitoring.
"We use the LTPP database as an exploratory data set," Smith says. "It gives us the data to develop the concepts and processes when we don't have the real data from a particular agency."
Each participant in the 2002 International Contest on LTPP Data Analysis received a copy of DataPave 3.0, a CD-ROM containing the most recent LTPP data.
To encourage greater use of the data to benefit the pavement industry, the Federal Highway Administration's (FHWA) LTPP product team, led by Monte Symons, and an American Society of Civil Engineers (ASCE) committee, chaired by Professor Fouad Bayomy, teamed up to sponsor a contest based on LTPP data. The International Contest on LTPP Data Analysis challenges students and professors from the United States and around the globe to come up with new and innovative analyses of the LTPP data. These analyses can enhance industry knowledge, improve pavement performance, or lead to new pavement technologies.
For the academic year 2001-2002, FHWA and ASCE recognized a total of five papers. One of Smith's students, for example, submitted a paper that describes a process for applying LTPP data to help researchers in California estimate road user costs due to pavement distress. Another student, from the University of Kentucky, used LTPP data to develop a climate map that could help highway engineers match pavement design, construction, and maintenance strategies to regions with similar climate patterns. Collectively, the winners demonstrate a variety of ways to use the LTPP data to reveal new insights about pavement performance.
Relating Moisture and Pavement Subgrades
Hassan M. Salem, a Ph.D. student in civil engineering at the University of Idaho, tied for first place in the graduate student category. Salem studied the effect of seasonal moisture variation on the resilient modulus of pavement subgrade soils. The changing seasons can lead to cracks and ruts in pavements. By developing a seasonal adjustment factor for the subgrade soil layer beneath asphalt pavements, Salem's research aimed to help designers determine the subgrade's resilience during any season and, ultimately, to improve the design of new and rehabilitated pavements.
Salem extracted elastic moduli, temperature, and moisture data representing different soil types in non-freeze zones at seven LTPP seasonal monitoring sites around the United States. After analyzing the data, Salem found that moisture and resilient modulus follow an almost sinusoidal function with different months of the year.
Salem, who learned how to use the LTPP database in a class offered by his advisor, relies on the data as an integral part of his graduate work. "The enormous quantity of data in the LTPP database can provide fodder for thousands of research papers—probably enough to keep a researcher busy for the rest of his life," he says.
Moisture content information from the LTPP database helped first-place winner Hassan Salem determine a seasonal adjustment factor for pavement subgrade. Here, Salem measures volumetric moisture content at a site in Idaho.
Developing Climate Maps
Yuhong Wang, a Ph.D. student in civil engineering at the University of Kentucky, tied with Salem for first place in the graduate category. As part of his thesis, Wang used environmental data from the database to develop a climate map of the United States and Canada, plotting the LTPP test sections according to their climate patterns.
"A potential use of the climate map," Wang says, "is to help highway engineers get climate pattern information for their geographical areas so they can apply the same design criteria, construction requirements, and maintenance strategies to those regions with similar climate patterns."
Wang selected tables containing climate information from the database, analyzed the data using clustering methods, and presented his results on a geographic information system (GIS) map.For Wang, the database already was contributing to his thesis, and the contest provided an opportunity to share some of his findings. "I wanted to get a chance to let other people look at my work and maybe get some feedback."
First-place winner Yuhong Wang used climate information to organize test sections from the database into 61 climate groups, some of which are shown on this map. Test section with similar numbers tend to have similar climate patterns.
Converting Distress Data
Shameem A. Dewan, a graduate research assistant at Texas A&M University, tied for second place in the graduate category. Dewan developed a method to convert LTPP distress data—such as cracking and rutting—to a format compatible with a pavement management system (PMS) developed by the Metropolitan Transportation Commission (MTC) of Oakland, CA.
"The MTC intended to develop a model for incorporating road user costs into decisions regarding maintenance and rehabilitation strategies for city and county streets in the San Francisco Bay area," Dewan says. "My objective was to estimate road user costs directly from the MTC-PMS pavement distress information."
Dewan extracted distress data for sites in California from the LTPP database and established a correlation between pavement distress and the international roughness index (IRI)—a measure of the roughness of a road. Since user costs depend on road roughness, Dewan developed a model for IRI as a function of pavement conditions, which enabled him to convert the LTPP distress data for use in the MTC-PMS software.
In his research, Dewan found that the LTPP data offered a suitable proxy for specific data unavailable from State and local agencies in California. "The cities and counties in the San Francisco Bay area did not have IRI data for local streets, so I used distress and IRI data collected from LTPP sites on California's highways and freeways instead," he says. "Although it would be more appropriate for the model to use data from city streets, the LTPP data offers a good starting point. In the future, we can refine the model using data from city streets."
Dewan first learned about the database while completing his master's degree. He attended a demonstration on how to extract and use data from the LTPP database offered by faculty members and FHWA staff members. He described the database as "very user-friendly once someone helps you start to use it."
Proofing Family Performance Curves
Mohammed Zulyaminayn, an exchange student from Bangladesh who studied at Texas A&M University, tied with Dewan for second place in the graduate category. Zulyaminayn developed a computer program to adjust performance curves for different groups, or families, of pavements to better predict the performance of individual pavement sections.
In prior research, Eun Sug Park and Clifford Spiegelman at the Texas Department of Transportation (TxDOT) developed sets of empirical performance curves to predict the development of individual types of damage for different families of pavements. Recognizing that site-specific environmental conditions and loading patterns cause the performance of individual pavements to veer from conditions projected by family performance curves, the Texas Transportation Institute developed an approach to adjust the projected performance to account for all of the observed performance data in TxDOT project 4186.
LTPP climate information can help researchers better understand how pavements react to environmental conditions such as soil type, humidity, rainfall, and nearby water sources.
Examining Influences on As-Built Roughness
Christopher M. Raymond, a Ph.D. candidate in civil engineering at Canada's University of Waterloo, placed third in the graduate category. Raymond used LTPP data to determine effects on the as-built roughness of asphalt overlay pavements. Because pavement roughness is a key measure of quality, designers and engineers need to understand the factors that influence the as-built roughness of a pavement so they can maximize their designs and develop smoothness specifications.
Raymond examined four factors: (1) extent of surface preparation prior to resurfacing, (2) overlay thickness, (3) type of overlay material, and (4) pavement roughness prior to resurfacing. Using data extracted from the LTPP database, he performed various statistical procedures, including paired data analyses, regression analyses, and a repeated measures analysis, to investigate the interactive effects. He determined that factors 1, 2, and 4 have statistically significant effects on the as-built roughness of a pavement, either directly or interactively with another variable. The overlay mix type was the only factor he found not to have an influence on as-built pavement roughness.
Raymond was introduced to the database through a workshop at the 1999 annual conference of the Transportation Association of Canada in St. John, New Brunswick. "In the area of pavement data, it's hard to obtain good data," he says. "It would be extremely difficult to coordinate the data required for my research without the LTPP database."LTPP in the Classroom
Associate Professor Neeraj Buch and Assistant Professor Karim Chatti, with the Department of Civil and Environmental Engineering at Michigan State University, earned first place in the curriculum category for inclusion of LTPP data in their undergraduate and graduate courses.
The professors found that using the LTPP database as a source for real pavement data has "enhanced considerably" the quality of their pavement rehabilitation, design, and analysis courses, according to Dr. Buch. Previously, in an undergraduate course, Buch and Chatti used pavement distress data from local and county roads. They found a number of drawbacks to using local data, including limited distress types, little or no deflection information, and lack of time series data. With these shortcomings in mind, the instructors decided to explore and use the LTPP database as a source of pavement distress and deflection data, as well as information on traffic growth, pavement inventory, and climate.
In December 2001, Dr. Fouad Bayomy taught a training workship at the University of Nevada at Reno on LTPP data analysis.
"In the past, we tried to create academic problems, but they didn't relate well to real life because we always found ourselves short of information," Buch says. "But with the LTPP database, we have real data—you can't argue with it."
Because of large class sizes in the undergraduate course, rather than provide direct access to the entire database, the instructors extract the raw data from the LTPP database and distribute it to small student groups to analyze. Armed with manageable amounts of raw data—including pavement geometry, distress, deflection, temperature at the time of testing, roughness, etc.—the students process the data, determine the pavement condition, and recommend rehabilitation strategies.
"By the end, the students will be able to recommend rehabilitation strategies based on the condition of the pavement," Buch says.
While the undergraduate course focuses on rehabilitation exercises, the graduate course is more analytical. Given the smaller size of the graduate course, the instructors provide each student with a copy of DataPave 3.0—a set of two CD-ROMs that include the most recent LTPP data. After a 2-hour tutorial, the students begin extracting pavement response and performance data, which they can use to evaluate existing performance prediction models.
"The LTPP database is such a wealth of meaningful data that graduate students in pavement engineering cannot afford not to know about it," Chatti says. "The possibilities of using the data to investigate different behaviors of pavement systems are almost endless."
About the Contest
Professor Fouad Bayomy chairs the ASCE-LTPP Task Committee.
“FHWA’s goal is to expose students—the future engineers and pavement practitioners—to the variety of data available in the LTPP database,” says Dr. Fouad Bayomy, chair of the ASCE-LTPP Task Committee that helped launch the LTPP contest in 1998. “With ASCE student chapters in virtually every university in the United States, ASCE is in a great position to deliver the data into the student’s hands.”
A win-win opportunity for all parties, the contest benefits FHWA by inspiring new transportation products, which will benefit managers of roadway systems throughout the world. ASCE stands to gain
new members, and the contest generates a multitude of new papers for possible publication. Most importantly, the LTPP contest puts the data in the hands of the students who will be the Nation’s future pavement engineers, managers, and administrators.“The partnership between FHWA and ASCE is a great match,” says Aramis Lopez, FHWA’s LTPP team leader. “The LTPP contest is a good tool to promote the use of the LTPP data to reveal new information about pavement performance.”
Students can enter the contest on their own or with other students in either the Undergraduate or Graduate categories, or partner with a State highway agency or industry group in the Partnership category. Students set the research objectives, conduct the research, analyze the data, and submit their findings in the form of a paper. The Curriculum category encourages educators to submit course curricula developed using LTPP data. All participants receive letters of recognition, and winners receive
cash prizes, plaques, and certificates. ASCE recognizes the winners at its annual meeting, and winning papers may be submitted for publication as well.A Golden Research Opportunity
Given four categories—Undergraduate, Graduate, Curriculum, and Partnership—virtually anyone involved with pavement research, development, or application can participate in the International Contest on LTPP Data Analysis. For example, FHWA and ASCE encourage State departments of transportation and others in the pavement industry to reach out to up-and-coming pavement engineers and academia through the Partnership category.
"The goal of the Partnership category," says Dr. Fouad Bayomy, professor of civil engineering at the University of Idaho and chair of the ASCE contest committee, "is to encourage State highway agencies and pavement experts to build or renew ties with local universities. While gaining valuable experience, student researchers can help agencies develop innovative solutions to real-world pavement problems."
To encourage further involvement of academia in using the LTPP database, Bayomy and ASCE's LTPP committee organized a training workshop in December 2001 at the University of Nevada in Reno to acquaint college professors with the LTPP program and provide hands-on training with the DataPave software. The workshop, which was sponsored by FHWA, attracted more than 20 faculty members from across the country. To learn more about the workshop or to download the instruction materials, visit http://www.uidaho.edu/engr/cedept/bayomy/bayomy.htm.
Antonio Nieves Torres is a graduate engineer from the Polytechnic University of Puerto Rico where he also worked as a professor. Nieves is a licensed professional engineer in Puerto Rico, and he has worked as a hydraulics and geometric design engineer in Puerto Rico and Florida. Over the past 18 years, Nieves has held positions in both the private and public sectors. A graduate of FHWA's Highway Engineer Training Program, Nieves has served as a research engineer for the LTPP program and as the executive director of the Pan American Institute of Highways.
John J. Sullivan IV is a contract writer and assistant editor for Public Roads magazine.
To learn more about the LTPP program, visit www.tfhrc.gov/pavement/ltpp/ltpp.htm. For future contest dates, visit www.tfhrc.gov/pavement/ltpp/contest.htm or contact Antonio Nieves at 202-493-3074 or antonio.nieves@fhwa.dot.gov. To obtain a free copy of the DataPave 3.0 software or to request technical support, contact LTPP Customer Support Services at 865-481-2967 or e-mail ltppinfo@fhwa.dot.gov.
Other Articles in this issue:
Measuring the Road to Improvement
Digging into LTPP Pavement Data
Does Your Interchange Design Have You Going Around in Circles?
From Small Beginnings Come Great Things
