Detecting and Correcting Localized Roughness Features

Data collected by inertial profilers on new asphalt pavements in Colorado in 2012 were used to analyze the effectiveness of the localized roughness specification in Colorado. For the analyzed projects, data were collected before any corrections were made as well as after diamond grinding had been performed to remove areas of localized roughness. The data indicated that localized roughness features having a half-car roughness index (HRI) lower than 175 in./mi were rarely addressed during correction. However, about half the localized roughness features that had an HRI of 175 to 200 in./mi were successfully addressed during correction. Localized roughness features having an HRI greater than 200 in./mi appeared to be successfully addressed during correction. The analysis indicated a significant difference in the localized roughness locations identified by AASHTO R 54 and the Colorado Department of Transportation (DOT) method of detecting localized roughness. The Colorado DOT procedure specifies a minimum length for a roughness feature that is to be corrected, but AASHTO R 54 does not. This paper shows that collecting accurate profile data and analyzing the data to determine localized roughness locations are not enough. The identified locations must be correctly marked on the pavement in the field so that the feature does not cause localized roughness. This paper presents a procedure not only for collecting accurate data but also for accurately marking the roughness features in the field. It is shown that it is possible to locate and correct localized roughness accurately to the current thresholds as set by AASHTO R 54.