Development of Traffic Inputs for PavementME Using Weigh-In-Motion Data in Alberta

Traffic characterization to develop appropriate traffic inputs for pavement design using Mechanistic Empirical Pavement Design procedure (PavementME) is a crucial and challenging task.  The purpose of this study was to develop traffic inputs such as vehicle class distributions for different axle configurations, growth factor, monthly adjustment factors, number of axles per truck, monthly and hourly distribution factors, and axle groups per vehicle, which can be used as inputs for pavement design, using PavementME, in Alberta.  Weigh-In-Motion (WIM) data from eight different stations in Alberta, collected from 2007 to 2019, was analyzed to characterize traffic loads as inputs for PavementME.  Traffic inputs were developed at two levels: Level 1 (site-specific) and Level 2 (provincial averages).  Simple statistical analyses like minimums, maximums, averages, and standard deviations both for individual stations and for all stations were completed to develop confidence in the Level 2 inputs.  The effects of the developed Level 1, Level 2, and PavementME default Level 3 traffic inputs were studied on the predicted performance at many highway segments of asphalt pavements.

Among all traffic parameters, vehicle class distribution (VCD) and monthly adjustment factor (MAF) varied the most between the hierarchical input levels.  In addition, Axle Load Distribution (ALD) showed moderate differences between the three input levels. Alberta‘s diverse truck traffic, changing during different times of the year, lead to the differences in MAF and VCD.  A sensitivity analysis was conducted of the impact of the developed Level 1, Level 2, and PavementME default Level 3 traffic inputs on the predicted performance of flexible pavements.  In this context, the results indicate that flexible pavement performance is most sensitive to vehicle classification.  The results also indicated more variance in total permanent deformation, AC top-down and total fatigue cracking for different input levels.  This study recommends the Level 1 (site-specific) inputs be used for pavement analysis and design.  For projects where site-specific data are not available, a Level 2 provincial average traffic inputs should be used, which will provide more representative inputs than the default traffic inputs.