Evaluating the Complex Relationship between Environmental Factors and Pavement Friction Based on Long-Term Pavement Performance

Abstract

Long-term pavement performance (LTPP) was used to investigate factors contributing to pavement skid resistance. The random effect model, with a Poisson distribution, was employed to analyze the relationship between various variables and pavement friction as a response, while accounting for the repetitive nature of panel-data observations. The results highlight a significant improvement in the model fit compared with the standard Poisson model. In this study, all pairwise interaction terms, instead of the additive impacts of various predictors, were considered. The results of this study highlight that the impacts of various predictors on pavement friction are not additive, but multiplicative. For instance, it was found that the impacts of pavement age, average annual temperature, number of lanes and annual Equivalent Single Axle Load (ESAL) on the pavement friction vary based on pavement type or on whether the pavement type is concrete or asphalt. The findings provide important information regarding the maintenance of pavement by paying the foremost attention to the pavement types for adjusting friction. This is one of the earliest studies that takes complex relations across various predictors and pavement frictions into consideration. A discussion regarding the implications of the findings is provided in the context of this study.