Ground Penetrating Radar as a Quality Assurance Tool in Hot-Mix Asphalt Road Construction

Ground penetrating radar (GPR) is a nondestructive, rapid, and continuous measurement tool that can predict density in hot-mix asphalt. Despite the promise of GPR as a quality assurance (QA) tool, the technology has been slow to adoption. Significant hurdles include (1) whether the equipment can accurately predict density over several days of paving without constant recalibration, and (2) knowing which prediction models have the best precision and minimal bias. In this research, GPR was deployed on a paving project for three days. Air void content was predicted using both empirical and micromechanics models using different calibration methods. Overall, the empirical model had the lowest prediction error. Precision concerns with all models are mitigated by the sheer quantity of sampling possible with GPR. Bias problems may be resolved with daily testing of a reference material and, for the micro mechanics model, by using more calibration cores. These topics should be explored in future research. When comparing the QA results to the GPR predictions, there was a disconnect between the pay factors assigned to the contractor based on random sampling and theoretical pay factors from the air void distributions measured by the GPR. This disconnect highlights how risks in the acceptance decision -making process could be reduced with increased sampling. A laboratory mixture sensitivity study was also conducted. The GPR signal was sensitive to air voids and aggregate type, but not to asphalt content. In practice, on a given project, the dielectric is almost exclusively sensitive to air voids. GPR could revolutionize QA testing in road construction by reducing acceptance risk, improving safety, and decreasing project delivery time.