Verification and local calibration/validation of the MEPDG performance models for use in Georgia.

TOOMRPRO130020The Georgia Department of Transportation (GDOT) is transitioning from empirical design procedures to the MEPDGprocedure for designing new and rehabilitated highway pavements. GDOT currently uses the 1972 AASHTO InterimGuide for Design of Pavement Structures as the standard pavement design procedure. As a part of the implementationprocess, GDOT has undertaken a project to verify the MEPDG global distress models and locally calibrate thesemodels for local field conditions of Georgia, if determined to be necessary by the verification process, using the LongTermPavement Performance (LTPP) and non-LTPP sections in Georgia.One objective of this project is to verify or confirm that the MEPDG transfer functions and global calibration factorsderived from NCHRP project 1-40D reasonably predict distresses and smoothness for the LTPP sites located inGeorgia using proper design inputs. This report includes a comparison of the predicted and measured distress andInternational Roughness Index (IRI) values measured over time and between different projects, pavement designfeatures, and/or site condition features. The confirmation process follows the procedure presented in the AASHTOMEPDG Local Calibration Guide (AASHTO, 2010). Specifically, this report documents use of the LTPP sites inGeorgia to determine the bias and accuracy of the MEPDG transfer functions in predicting the distress andperformance of those LTPP test sections.In summary, the number of Georgia LTPP sites and levels of distress exhibited on the test sections are consideredinadequate for the validation or confirmation process of the global calibration coefficients from a statistical perspective.More importantly, bias between the measured and predicted distress for some of the transfer functions of both flexibleand rigid pavements were found. Thus, it is recommended that GDOT proceed with the next phase of the study andselect projects to fill in the many key gaps so that the calibration process can be used to adjust the calibrationcoefficients for each distress