Abstract
The AASHTO Mechanistic-Empirical Pavement Design Guide requires local calibration to account for local conditions, materials, and engineering practices. Previous local calibration studies in Ontario focused mainly on permanent deformation models for pavement rutting. The objectives of this study are twofold. First, to provide an enhanced calibration for the rutting models by using more vigilantly cross-verified input data and updated observed rutting data. Second, to perform a trial calibration for the international roughness index (IRI) model by considering three different calibration methods. Cracking models calibration, being performed by another colleague, has not yet been finalized; therefore, the IRI model calibration cannot be finalized in this study. Based upon 63 Superpave sections, the local calibration coefficients were found to be βAC = 1.7692, βT = 1.0, βN = 0.6262, βGB = 0.0968 and βSG = 0.2787 , which reduced the standard deviation of residuals to a value of 1 mm. The IRI calibration study found that the initial IRI value plays an important role in the calibration. Keywords: International Roughness Index (IRI) model; local calibration; Mechanistic-Empirical Pavement Design Guide (MEPDG); rutting model; Superpave.
Highlights
It was the first road test to quantify an empirical connection between pavement performances and influencing factors such as structural design and varying axle loads
Considering following enhancements which differs from previous studies, at the beginning stage of this research, preliminary evaluation on the selected Superpave sections was performed to assess the prediction performance of global rut depth model: This study used the most recent Ministry of Transportation of Ontario (MTO) default parameters to execute the latest version of Mechanistic-Empirical Pavement Design Guide (MEPDG) software (V.2.3) on the expanded database included Stone Mastic Asphalt (SMA) sections
Performing a cross-sectional local calibration on the rutting model resulted in the following calibration parameters: βAC = 1.7692, βGB = 0.0968, βSG = 0.2787 with layer contributions of 19%, 9% and 72%, respectively
Summary
At the end a non-bias and precise performance prediction model was introduced using least square linear regression technique The focus of this part of the study was on performing a preliminary local calibration of the current global IRI model for Ontario’s Superpave pavements. One important factor that needs to be discussed is that the cracking model calibration has not yet been finalized For this reason, the IRI study focused on an exploration and presentation of the best calibration method. For each selected section, observed historical IRI measurement data was obtained from Ontario’s ARAN database. It is important to understand that for each proposed calibration methods, data selection criteria varied as following: Method 1 database selection was based on the availability of measured IRI0 values, only 40 sections were found to be appropriate for its analysis. Method 3 did not have any constraints all of the 52 sections were included in its analysis
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
