Introducing Adhesion–Cohesion Index to Evaluate Moisture Susceptibility of Asphalt Mixtures Using a Registration Image-Processing Method

Abstract

Moisture damage is a major concern for evaluating the performance of asphalt mixtures. There are different types of experimental methods to determine the effect of moisture on mechanical and durability characteristics of asphalt mixtures. In this study, three different experimental approaches were implemented, including the boiling water test, the indirect tensile test, and the resilient modulus test, as well as the fracture energy analysis to evaluate moisture susceptibility of asphalt mixtures fabricated with different types of fillers including portland cement, limestone powder, and recycled concrete aggregates. Replacing the control filler material with these fillers resulted in improved fracture energy, which shows the stripping rate becomes slower by using them as fine aggregate. The fracture energy ratio of the asphalt mixture containing portland cement has the lowest rate of decrease for freeze-thaw cycles. Also, by applying a two-dimensional registration image-processing method as a low-cost soft-computing technique, an adhesion–cohesion index is introduced for determining the effect of moisture on adhesion between asphalt binder and aggregates as well as cohesion within asphalt mastic. Results have shown that there is a meaningful correlation between adhesion–cohesion index and number of freeze-thaw cycles, tensile strength ratio, resilient modulus ratio, and fracture energy ratio. To predict the adhesion–cohesion index of asphalt mixtures as an output of the image-processing method based on experimental results, a regression model was developed and verified in terms of the aforementioned parameters with an average prediction error of 4.46%.