Modelling and evaluating moisture susceptibility of laboratory prepared asphalt concrete mixtures

Abstract

Moisture susceptibility, one of the key reasons of asphalt concrete (AC) pavement deterioration, lowers the durability of AC pavements. Analyzing and evaluating the combined effect of aggregate gradation and asphalt on moisture susceptibility is critical for avoiding and/or minimizing premature failures of AC mixtures; however, evidence of such combined effect is scant. As such, this paper examines the vulnerability to moisture in laboratory-prepared AC mixtures to establish the impact of differential aggregate gradations (both wearing and base) and bitumen grade. Superpave gyratory compacted mixtures are subjected to indirect tensile strength (ITS) test and are evaluated using tensile strength ratio (TSR). Four wearing course gradations, two base course gradations and two binder grades are considered in this study. The findings of the laboratory tests suggests that Asphalt Institute Manual Series (MS-2) wearing course mix and Superpave mix (SP-2) base course mix prepared with 40/50 binder grade result in higher TSR values, implying relatively better resistance to moisture damage of these AC mixtures compared to other tested mixtures. In addition, the study found that wearing mixtures prepared with 40/50 binder grade reveal 40% more tensile strength compared to the same mixtures prepared with 60/70 binder grade. To study the relationship of moisture susceptibility, a first order multiple linear regression model is developed as a function of percent sieve passing, nominal maximum aggregate size and binder grade. The model being proposed captures data variability with high degree of accuracy. The model is also validated and sensitivity analysis is carried out in order to determine changes in TSR with regard to the explanatory variables. The findings of this research will help in selecting the AC mixtures while considering their sensitivity to moisture damage and the developed model will contribute in evaluating moisture susceptibility of AC mixtures without performing hectic experimental tests.