Asphalt compaction quality control using Artificial Neural Network

Abstract

Adequate compaction of asphalt pavements during their construction is essential to the long term performance of the pavement. Under/over-compaction during the construction process leads to the early deterioration and failure of the pavement. Current quality control techniques in the field involve the extraction of roadway cores and the measurement of density using point wise measurement techniques. Such tests determine the quality at discrete locations typically after compaction is complete and are not indicative of the overall quality of the pavement. Conversely, analyzing the vibrations of a vibratory compactor during asphalt pavement construction is a proven indicator of the complete compaction quality. The compaction of asphalt pavement is a complicated process resulting in the absence of a closed-form solution for estimating stiffness or density of the mat being compacted. In this paper, we present Intelligent Asphalt Compaction Analyzer (IACA) as a decision making device for operators to treat the asphalt pavement in an appropriate way to control the quality of compaction. IACA is a classification tool that uses Artificial Neural Network (ANN) to give an estimate density value of the road underneath the roller drum during compaction process. The Fast Fourier Transform of the roller vibrations is used by the ANN to estimate density values. Currently available Intelligent Compaction (IC) techniques provide a measure of quality that is hard to relate to any physical measurement. In contrast, the proposed method gives a density measurement that can be verified either by the extraction of roadway cores or through the use of conventional density gauges. As a result, the performance of the proposed method can be validated during the construction process.