Calibration of the AASHTO MEPDG for Flexible Pavements to Fit Nevada's Conditions
Civil and Environmental Engineering
AltmetricsView Usage Statistics
The Mechanistic-Empirical Pavement Design Guide (MEPDG) consists of transferring pavement mechanical responses such as stresses and strains into predicted distresses. The nationally calibrated models for rutting, bottom-Up fatigue cracking, top-down fatigue cracking, International roughness Index (IRI), thermal cracking and reflective cracking need to be recalibrated to properly fit Nevada’s local conditions for materials, traffic, and climate. This study focuses on the local calibration of the fatigue bottom-up cracking and the rutting models. For this purpose, data was collected from the Nevada Department of Transportation (NDOT) Pavement Management Systems (PMS) database and converted to match the MEPDG models requirements. Additionally, field-produced mixtures were sampled from 45 paving projects to develop a materials database. These mixtures were collected from all three districts and tested for dynamic modulus, binder properties, rutting, and fatigue. This was completed to characterize the polymer-modified asphalt binder mixtures technologies in Nevada which was one of the main factors that mandated a local Pavement-ME calibration as the nationally calibrated models used unmodified binders.The calibration was performed by optimizing the local calibration factors to reduce the sum of error squared between predicted and measured distresses data. The calibration for rutting was conducted for new and rehabilitated sections from the three districts. On the other hand, the fatigue calibration seperated new and rehabilitated sections but combined between district II and III as most mixes from these districts use the PG64-28NV binder as opposed to District I where PG 76-22NV binder is predominantly used. The final calibration sets for rutting and fatigue cracking were 6 and 4 respectively.This thesis recommends additional performance monitoring of the polymer-modified paved sections as the calibration was validated using only 10 years of performance data. Future recalibration could be undertaken to increase the accuracy of the models.