Raveling is a form of pavement distress involving the dislodgement of aggregates from the surface of an asphalt pavement. Monitoring of raveling occurrence, severity and extent in pavement management systems tends to follow a semi-automated approach, i.e. automated pavement data collection by imaging techniques and manual rating of pavement condition by trained raters. This leads to subjectivity in raveling condition rating within the existing pavement management system. Therefore, this paper explores the use of ordered probit models in relating raveling severity, extent and deduct value (subjective rater opinions) to texture measurements (objective laser measurements) collected within the network-level pavement management system. Data collection procedures for raveling and texture measurements and the rater evaluation of raveling distress are discussed. Using the data collected from highways in Indiana over a period of five years (2002 to 2006), ordered probit models for raveling severity, extent and deduct values are developed. These allow a convenient and objective determination of raveling distress severity, extent and deduct values within existing network-level pavement management system framework. The relative effects of factors affecting raveling are also studied. From the statistical analyses, it is found that root-mean-square (RMS) texture and change in RMS-texture over consecutive years are significant variables affecting raveling condition rating. Increases in these parameters increase the probability of raveling occurrence, higher raveling severity, raveling extent and deduct value. Raveling condition charts are also developed for pavement engineers to determine raveling deduct values of asphalt pavement at the network level conveniently.
Samenvatting