The Mechanistic-Empirical Design Guide (MEPDG): An Analysis of Different Level Inputs for Alabama Pavement Surface Mixtures

Abstract

Since the beginning of pavement design, there have been several methods that can aid in determining and estimating a pavement’s construction, rehabilitation, and end of life. Current design methods include but aren’t restricted to empirical, limiting shear-failure, limiting deflection, and regression methodologies. A newer method was introduced that combines observed destructive and nondestructive pavement performances and reconciles those results to identify physical distresses within the pavement structure. This approach can be used more accurately during pavement design due to the addition of pavement performance simulations and calibrations. These design considerations can give a better understanding of how pavements deteriorate over time and could eventually lead to advancements in new design technologies, which will aid in the mitigation of future distresses when a new road is being constructed or if an existing road needs rehabilitation. The results of various pavement performance tests can be modified and used as inputs into design guides, creating better design criteria in comparison to previous methods. The collection of information from locally gathered materials, climactic data, and performance calibrations can be utilized in creating more accurate pavement designs, given the moniker of the Mechanistic-Empirical Design Guide (MEPDG). This thesis aims to highlight the importance of MEPDG by collecting respective data that would be used as Level -1 and -2 inputs for pavement design while also comparing the sensitivity of different mix test results throughout the state of Alabama. Asphalt mixes were tested in accordance with ALDOT and AASHTO methodologies, where results from different tests correlated with their susceptibilities to mixture rutting, mixture stiffness, and creep compliance and strength. Each mix’s data was compared with one another to determine errors, similarities, or differences between mix designs while also utilizing transfer functions within the modeling program of the Pavement Mechanistic-Empirical Design (PMED) to determine and judge the accuracy of MEPDG against previous accumulated state designs. Between 2021 and 2022, nine asphalt mix designs from five regions in Alabama were distributed to the National Center for Asphalt Technology (NCAT) for performance testing. The extracted/recovered binders were tested with the Dynamic Shear Rheometer (DSR), Bending Beam Rheometer (BBR), and Rotational Viscometer (RV) to determine the Superpave Performance grade (PG), Delta Tc (ΔTc), and viscosity. The binders were also tested with the MSCR test per AASHTO T 350 for evaluating elastic response and rutting resistance; the LAS test per AASHTO T 391 for evaluating fatigue resistance; and the Frequency Sweep test followed by master-curve analysis for evaluating ductility and block cracking resistance. Mixture results were taken and analyzed from flow number (FN) or repeated load permanent deformation (RLPD), dynamic modulus (E*), and indirect tension (IDT) testing, while mix verification was performed to accurately record key mechanical properties that were later used in MEPDG (performance grade, complex modulus, gradation, specific gravities, aggregate and binder content verifications). These results showed that there was overall less discrepancy between Level-3, and combined Level -1 and -2 designs, when the distresses that were focused on were more driven to accommodate problems that were associated with high temperature plus increased loading time design potentials. As a result, there are more discrepancies in the designs when cold temperature-controlled distresses are associated.