The current AASHTO specification for the determination of bulk specific gravity (G_mb) of compacted asphalt mixtures requires multiple equipment and weight measurements for different conditions which, along with the operator experience, influence the accuracy and variability of G_mb measurements. The use of a laboratory dielectric measurement system (LDMS) was explored in this study as an alternative procedure to the current AASHTO specifications for measuring G_mb. The LDMS measures the dielectric constant of a compacted asphalt concrete specimen using a ground penetrating radar (GPR) system. In Stage 1 of the project, a procedure to measure the laboratory dielectric value of a compacted asphalt concrete specimen was established. Asphalt mixtures with a range of aggregate and mixture properties (geology, nominal maximum aggregate size, binder type, RAP content, etc.) were collected. Compacted asphalt specimens of various specimen surface conditions and sizes were tested using the LDMS. The collected dielectric results were statistically analyzed to study dielectric measurement repeatability and reliability by grouping measurements at various orientations together and comparing them to all measurements of the corresponding asphalt specimen. Sensitivity analysis was also performed to study measurement accuracy and resolution for replicate dielectric measurements. It was determined that taking dielectric measurements at any 3 orientations 120 degrees apart (3 measurements total) yield repeatable dielectric measurements with acceptable accuracy. In Stage 2 of the project, a procedure to determine G_mb based on the laboratory measured dielectric value was developed. A select number of asphalt mixtures from Stage 1 work, along with their component materials (aggregates and RAP) were collected. The Al-Qadi Lahour Leng (ALL) model and the Hashin-Shtrikman (HS) model were investigated and analyzed to establish the relationship between G_mb and dielectric measurements. Based on the gathered data and statistical analyses performed in Stage 2, a procedure using the HS model and laboratory measured dielectric values of component materials was recommended to determine specimen G_mb.  The recommended procedure produces predicted values of G_mb from measured dielectric values that are within the acceptable range of two G_mb results in AASHTO T331 for most specimens evaluated.

The final report is available.