A binary logistic model for predicting the tertiary stage of permanent deformation of conventional asphalt concrete mixtures. (10th December 2019)

• Record Type: Journal Article
• Authors: Ali, Yasir; Irfan, Muhammad; Buller, Abdul Salam; Khan, Hammad A.; Gul, Hafiz M.F.
• Is Part Of: Construction & building materials. Volume 227(2019)
• Publication Date: 2019-12-10
• Subjects: Asphalt -- Permanent deformation -- Tertiary -- Flow number -- Logistic model; Building materials -- Periodicals; 624.18
• DOI: 10.1016/j.conbuildmat.2019.07.334
• Languages: English
• ISSNs: 0950-0618

Abstract

Highlights: Developed a binary logistic model to predict the tertiary stage of permanent deformation. Calculated probabilities for volume of air voids, |E*|, and gradation parameters. Sensitivity analysis was performed to determine the change in probability corresponding to the change in parameters. Compared performance of field and laboratory mixtures. Abstract: Since the advent of Mechanistic-Empirical Pavement Design Guide (M-EPDG), characterisation of asphalt concrete (AC) mixtures using flow number remains the topic of interest for most of the literature. Consequently, many flow number models have been proposed in the literature, which only caters AC mixtures that attain the tertiary stage of permanent deformation (PD). The literature is devoid of any study that ascertain, whether an AC mix would attain the tertiary stage of PD or otherwise. As such, this study develops a binary logistic model as a function of volumetric and performance parameters of field and laboratory prepared AC mixtures using the standard maximum likelihood procedure. The logistic model specification is best suited to predict the mix probability of attaining the tertiary stage of PD. More specifically, an increase in air voids and percent passing sieve #200 in the mix and relatively low value of dynamic modulus translated into an increased probability of attaining the tertiary stage of PD. The field and laboratory mixtures are compared using the binary logistic model and a sensitivity analysis is also performed on the model. The comparison reveals the difference in both the type of mixtures, and suggests optimising and customising the laboratory mixtures to the field conditions.