Quantification of the Spatial Gradient of Local Volume Fraction from Tomography Images

• Main Author: Wang, Yongping
• Other Authors: Steve Cai
• Format: Others
• Language: en
• Published: LSU 2002
• Subjects: Civil and Environmental Engineering
• Online Access: http://etd.lsu.edu/docs/available/etd-0417102-151129/

Asphalt concrete has three constituents: aggregate, asphalt binder and air voids. In the mixture, aggregates serve as a skeleton of the mixture; asphalt serves as a binder to hold the aggregates together; and air voids prevent asphalt concrete from bleeding to avoid plastic mixes. The properties of asphalt concrete are controlled by the interaction of these three components. In asphalt concrete, asphalt binder, voids and aggregates are not uniformly distributed. The local volume fractions of these constituents vary with spatial locations. As a result, there exist a spatial gradient of the local volume fractions between two locations. Due to the inhomogeneous distribution of the volumes of the constituents, the effective properties such as the modulus also vary with spatial locations, resulting in the stress concentration or strain localization. The purpose of this study is to make a preparation for modeling asphalt concrete in the continuum scheme by developing methods to quantify the local volume fractions of voids and the spatial gradients of the local volume fractions. X-ray tomography images were used to characterize the local volume fractions of voids and their gradients. By using Image-Pro Plus software as a platform, automated macros were developed to obtain the distribution of local volume fractions of voids as well as the gradients of local volume fractions. Statistical analysis of the experimental results shows that the field performance of the mixture is related to the mean local volume fractions of voids and their gradients. It may be concluded that local volume fraction and its gradient might be good field variables to characterize the internal structure of asphalt concrete. The experimental results are consistent with the predictions of several theories that use local volume fraction and the gradient of local volume fraction as field variables, and the field performance of the three mixes studied.