Reliability-based design procedure for fatigue cracking in asphalt pavements

Abstract

The need to account for the effect of design input variabilities on predicted performance has led to many design procedures that address reliability for pavement applications. The Florida cracking model uses empirically derived reliability for fatigue cracking design of asphalt pavements. A reliability approach, which is based on probabilistic uncertainty quantification, is necessary to account properly and effectively for the contribution of the variability in each parameter to the overall variance. This paper presents a load and resistance factor design (LRFD) procedure for the Florida cracking model. By delivering designs of uniform reliability, LRFD provides the basis for developing quality control and quality assurance standards. A first-order reliability method that incorporates a surrogate model based on central composite design was used to compute the reliability and formulate the partial safety factors. The reliability calibration was based on field pavement sections that had a wide range of design inputs and target reliability. Illustrative designs based on the developed LRFD procedure show the effectiveness of the partial safety factors and further confirm the credibility of the reliability analysis methodology.