The AASHTO pavement design method was developed based on road tests conducted in the 1950’s with a limited variation in test conditions including the traffic. Extrapolation was required for conditions outside the experiment boundary. This may involve uncertainty in design and may lead to an over or under designed structure. In fact, over the last several decades, highway agencies experienced a reduction in pavement life for large increase truck traffic. Highway agencies were eager to improve the way the pavement is currently being designed. The new Mechanistic Empirical Pavement Design Guide (MEPDG) has been developed based on fundamental properties of materials and the physical observations of performance. It can be used for all truck volume and axle load scenarios. However, for a more reliable design, local material properties, climate data, truck volume and distributions, and axle load spectra (ALS) are critical. This paper presents the experience of Manitoba Infrastructure and Transportation (MIT) with the MEPDG in using the local truck traffic data with an example of a flexible pavement design. The sensitivity of the program for changes in truck volume, ALS and truck distributions are presented. Analysis/experience showed that MEPDG produces designs with similar or thinner pavement structures for low truck volume but it overestimates the pavement structures for moderate to high truck volumes compared to the AASHTO 1993 and surface deflection methods. A significant variation in required structure was also noted for a within province variation in the truck class distribution. This emphasizes the importance of calibrating the performance models to local conditions. The issues and challenges in calibrating the MEPDG performance models are also discussed.
