Accurate real-time density monitoring is crucial in quality control and quality assurance during the asphalt concrete (AC) pavement construction process. Ground penetrating radar (GPR) technology has shown great potential in the continuous real-time density prediction of AC pavement. However, it is not accepted as a routine method by transportation agencies in the United States due to the lack of validation under field testing conditions. In this study, three field tests were performed using GPR to estimate AC pavement density. The Al-Qadi-Lahouar-Leng model was used to predict the density from GPR signals. The reference scan method was used to remove the effect of surface moisture during construction. The gradient descent-based non-linear optimization method was used to reconstruct the overlapped GPR signals result from the use of thin AC overlay, which has been widely implemented as an AC pavement rehabilitation technique. Digital filtering and other signal processing methods were used to de-noise the signal. GPR results using the proposed methods were compared with field core data and nuclear gauge results. The results show that the proposed methods were effective in estimating in-situ AC pavement density using GPR. Continuous density estimation by installing GPR on the roller is suggested to provide real-time compaction monitoring during the AC pavement construction process.
ASJC Scopus subject areas
- Civil and Structural Engineering
- Mechanical Engineering