Knowledge CentreTechnical Resources SearchConference PapersEvaluation of Darwin-ME Pavement Rutting Prediction Models Using Data from Alberta’s Pavement Management System

Evaluation of Darwin-ME Pavement Rutting Prediction Models Using Data from Alberta’s Pavement Management System


Research under the United States-Canada joint Strategic Highway Research Program and the Long Term Pavement Performance study has led to the development of the MechanisticEmpirical Pavement Design Guide (M-EPDG) and its software package known as DARWinMETM. An integral part of M-EPDG is the utilization of the predicted performance to evaluate the pavement structure being designed in its ability to carry the traffic loading in the design period. One of the key performance indicators being evaluated during structural design is the amount of permanent deformation, or rutting, in the pavement structure as a result of truck traffic. As with any research the study results of the M-EPDG may not be readily applicable to all jurisdictions. Earlier studies done by Alberta Transportation indicated differences in pavement material models adopted in the M-EPDG compared to those found in Alberta. It is important that the various models in M-EPDG and the associated DARWin-ME™ software package be validated against conditions in Alberta. Alberta Transportation’s Pavement Management System contains decades of pavement inventory and performance data. With the accumulation of automatically collected pavement roughness and rut data since the late 1990s, the PMS database could be mined to explore trends with pavement rutting. This paper provides the summary of the analysis results on these trends, and compares them with the predicted rutting results from the DARWin-ME™. The study found that for flexible pavements with non-stabilized granular base courses, when compared to the measured pavement rutting at the network level, the amount of total pavement deformation was over predicted by the default rutting model for flexible pavements in the new construction category. However, the predictions were much closer to the measured rutting with pavements after rehabilitation. The predicted rutting appeared to be generally lower than the measured values (under predicting) for those pavements treated by milling prior to overlay, while it compared reasonably well to the measured rutting for pavements treated with straight overlays. These study results could provide insights into the local calibration of the various performance prediction models in DARWin-ME™ as the new pavement design package is studied and evaluated for its potential adoption by Alberta Transportation. 

Conference Paper Details

Session title:
Wei He
Marta Juhasz
John Crockett
Venkat Lakkavalli