Comprehensive Approach for Development of Performance Models for Network-Level Pavement Management System Using LTPP Data (05-1414)**
Wael Bekheet, Alexandria University, Egypt
Khaled Helali, Stantec Consulting, Inc.
Halim Omar Abd El Halim, Carleton University, Canada
Jack H. Springer, Federal Highway Administration

Pavement Performance models are a key component of any Pavement Management system (PMS). Development of these models for a network-level PMS has been traditionally based on historical performance and inventory data. However, historical data may represent a challenge for the highways agencies, due to continuous upgrades in field data collection equipment and the possible incompleteness of inventory records. The Long Term Pavement Performance (LTPP) database can provide an alternate reliable source of data for the development of pavement performance models. LTPP data can be tailored to fit the scope of a PMS and used through a PMS approach for the development of base pavement performance models. These base models can then be adjusted to agency-specific experience and/or data to produce agency-specific models. In this paper, a mechanism for the development base pavement performance models in the context of a pavement management application is presented. As a case study to demonstrate this approach, IRI performance models for flexible pavements in different environmental zones are developed using data from LTPP SPS-5 test sites. The developed models successfully represented the expected general performance trends in term of the differential performance among rehabilitation activities and environmental zones. Models representing freezing environmental zones seemed to deteriorate faster than those developed for no-freeze environmental zones. Also performance models for thinner AC overlay rehabilitation activities seemed to deteriorate faster that those of the thicker AC overlay activities. State-specific data can then be used for final calibration and verification of these models.