A stochastic approach on modeling the long-term unevenness of pavements considering spatially varying soil properties/Ein stochastischer Ansatz zur Modellierung der langfristigen Entwicklung von Fahrbahnunebenheiten unter Berücksichtigung räumlich variierender Bodeneigenschaften

Even within nominally homogenous soil layers, spatial variations of soil properties occur. Current design methods simplify varying soil properties by applying constant values. In order to achieve a reliable geotechnical design, the local variance of soil properties needs to be considered with respect to the long-term development of road unevenness. Hence, a variation of soil stiffness leads to spatially varying settlements and increasing dynamic forces on passing vehicles as time increases. This process continues steadily resulting in a noticeable reduction of driving comfort until a repair of the pavement is required. In this paper, a numerical model is applied to determine the long-term unevenness of pavements considering the spatial variation of soil properties. Thereby, the asphalt pavement, the subbase, and the subgrade are modeled by a set of coupled spring elements representing the soil behaviour of the respective layer. A stochastic distribution of soil properties based on a random walk is assumed to account for the spatial variation of the subsoil stiffness. The long-term unevenness development due to passing vehicles is extensively evaluated with respect to resulting dynamic forces and the impact of a deviating soil stiffness in individual soil layers. Finally, conclusions with respect to driving comfort are drawn.