The Impact of Weigh-in-Motion Measurement Error on Mechanistic-Empirical Pavement Design Guide Reliability

Axle load spectra have a significant impact on predicted pavement performance. At the design stage, it is typically assumed that axle load spectra as measured by weigh-in-motion (WIM) systems are accurate. In fact, the quality of WIM-based data has inherent uncertainties due to inaccuracy and systematic bias in measurements. This paper investigates the impact of WIM measurement errors on axle load spectra (ALS) and quantifies the effects of these errors on design reliability. In the new Mechanistic-Empirical Pavement Design Guide (M-E PDG), the reliability procedure was developed based on the assumption that variability in performance prediction is approximately the same as the observed performance of the pavement sections used to calibrate the performance models. In this analysis, each distress type was approximated by a normal distribution and therefore, two parameters (mean and standard deviation) were determined to represent the expected value and associated variability by employing Monte Carlo simulation. The results show that the M-E PDG reliability analysis can compensate for negative axle load measurement bias for most of the distresses. However, a lower tolerance for negative bias needs to be enforced in order to ensure that both flexible and rigid pavements have the design reliability against cracking, especially for thinner pavements. While most of the findings further reinforce existing concepts, the study provides a systematic overview of WIM data accuracy and calibration needs and the impact of associated uncertainties on the pavement design process.