Accurate knowledge of pavement layer thicknesses is important in many aspects of pavement management. Often this information is unknown, and records are inaccurate, out of date, or difficult to access. To date, the only method for obtaining pavement layer data has been core sampling. That is time-consuming, labour intensive, intrusive to traffic, and limited in coverage. This study investigated the capability of ground-penetrating radar to provide accurate subsurface pavement profile information. Eleven sites were selected to represent the population of pavement types present in Kansas, with a particular emphasis on variations in base type and road history. The radar results show substantial variations in pavement thickness within each 1,000-ft test section, and in general, higher values of pavement thickness than were reported in available records. These predictions, when correlated with data from 73 ground truth cores, show an accuracy of plus or minus 5 percent to 10 percent, depending on the treatment of the data. The asphalt thicknesses in this study ranged from 2.5 to 20 in. The radar data were analysed automatically using software that operated directly on the raw radar waveforms. This software is based on an electromagnetic model of the pavement layer structure. The resulting predictions were correlated with core samples obtained in co-operation with the Kansas Department of Transportation. The results show that a radar system that combines air-launch horn antenna equipment with appropriate software can provide an effective alternative to coring for pavement thickness measurements. In addition, this system provides more information to the agency, is cost competitive, and is safer to use because it does not require lane closures. (Author/publisher)
Abstract