Local Calibration of Cracking Models in MEPDG for Ontario’s Flexible Pavement

The AASHTO Mechanistic-Empirical Pavement Design Guide (MEPDG) introduces a pavement design method which uses both the mechanistic analyses and empirical models to predict pavement distresses and performance. Due to the significant differences between the global and the local conditions, the empirical distress models need to be calibrated for local environments and local design practices based on local pavement performance data. This paper focuses on the local calibration of the various cracking models (fatigue bottom-up, fatigue top-down and thermal longitudinal) in MEPDG for flexible pavements on Ontario’s highways. First a local calibration database was developed based upon MTO’s Pavement Management System. After this, simulations were run in the software and the predicted data is compared to the observed data. Significant difference is found in the comparisons which need to be minimized by calibrating the distress models. A new regression model is used to optimize the calibration parameters by minimizing the standard deviations of the residuals between the predicted and observed distresses. The challenges encountered and solutions developed during the local calibration are discussed.