Presentation of Predictive Models for Two-objective Optimization of Moisture and Fatigue Damages Caused by Deicers in Asphalt Mixtures

Abstract

The best way to deal with the freezing of the road surfaces is to use deicers, especially in cold areas. The presence of moisture causes various stresses in the pavement and reduces the strength of mixtures. Using anti-stripping agents can decrease the moisture sensitivity of asphalt mixtures. Researchers have evaluated the impact of different deicers on the moisture sensitivity of asphalt mixtures. However, fewer studies have been conducted on the effect of these materials on fatigue failure and thermodynamic parameters of asphalt mixtures. Moreover, fewer studies have been performed to find the exact optimum amount of additives for maximizing the two objectives of tensile strength ratio (TSR) and fatigue life ratio (NFR) concurrently in moisture and fatigue damages. So in this research, the moisture sensitivity and fatigue failure of asphalt mixtures under the influence of different deicers, including calcium magnesium acetate (CMA), potassium acetate (PA), and sodium chloride (NaCl), were investigated using nanohydrated lime (NHL) as an anti-stripping agent. The surface free energy (SFE) of materials and the permeability of asphalt mixtures were examined, and a boiling water test was applied. Finally, the prediction models of multivariate regression (MVR), group method of data handling (GMDH), and genetic programming (GP) were provided to obtain optimum additive percentage with two objectives of TSR and NFR. The results showed that GP had a higher R-value than the 2 other methods such that the R-value of GP for TSR and NFR was 98.8 % and 99.8 %, respectively. The optimization results showed that 1.17 %, 1.34 %, 0.87 %, 1.21 %, and 1.06 % NHL, respectively, were the best optimum values to maximize the TSR and NFR simultaneously in all samples and samples saturated in water, CMA, NaCl, and PA solutions.