Resumen
Asphalt pavement compaction quality control and quality assurance (QC/QA) are traditionally based on destructive drilled cores and/or nuclear gauge results, which both are spot measurements representing significantly less than 1 percent of the in-service pavement. Ground penetrating radar (GPR) is emerging as a tool that can be used for nondestructive continuous assessment of asphalt pavement compaction quality through measuring the pavement dielectric constant. Previous studies have established that asphalt pavement dielectric constant measurements are inversely proportional to the air void content for a given asphalt mixture. However, field cores are currently required to calibrate the measured dielectric constant to the pavement density. In this paper, a method is proposed to eliminate the need for field calibration cores by measuring the dielectric constant of asphalt specimens compacted to various air void contents. This can be accomplished with a superpave gyratory compactor (SGC), which is routinely used in the pavement industry to fabricate 6 in. (15.2 cm.) diameter specimens. However, this poses difficulties with the GPR antenna height, direct coupling, and the Fresnel zone in relation to the asphalt specimen dimension limitation. These challenges are overcome by employing a plastic spacer with a known dielectric constant between the SGC specimen and the antenna. The purpose of the spacer is to reduce GPR wave speed so that the signal reflected from the specimen is separated from the direct coupling effects at an antenna height where the Fresnel zone of the GPR is not affected by the specimen dimension. The specimen dielectric constant can then be measured using the reflection coefficient-based surface reflection method (SR) or the pulse velocity-based time-of-flight method (TOF). Also, The Hoegh?Dai model (HD model) is demonstrated to reasonably predict pavement density based on the results of field measurements and corresponding core validation, especially as compared to the conventional exponential model. Results are presented from multiple days of paving on one project, as well as a single paving day on a project with significantly different mix properties. The agreement between the HD model, coreless prediction, and field cores shows the promise for implementation of dielectric-based asphalt compaction evaluation without the need for destructive field core calibration.