Synthesizing dual-track road elevation data

Abstract

The ability to synthesize dual, parallel track road elevation profiles is essential for studying the multi-axial motion of vehicles. It is important that the profiles retain the random and variable character of real roads. This research introduces a novel method for synthesizing dual-track longitudinal pavement elevation profiles. Using measured data from asphalted roads, probability density functions were produced to describe the variations in roughness and correlation between the tracks. The correlation between the tracks was analysed using the coherence function and roughness ratio between the measured kerb-side and driver-side tracks. The results showed that variations in the roughness of the tracks were well described by the Weibull probability distribution function defined by the shape, scale and location parameters. The location and shape parameters were shown to be independent to road roughness, leaving the scale parameter as the main index to define the roughness distribution. This finding led to the creation of a new roughness classification scheme for asphalt roads which was used to select appropriate roughness levels for simulation. Variations in the coherence functions and roughness ratios were found to be independent of road roughness and were also able to be described using the Weibull distribution. Using the results from a detailed statistical survey, a novel method was developed to enable the synthesis of parallel track elevation profiles that emulate real roads.