Approach for Predicting Cracking Deterioration in Sprayed Seals from Subjective Condition Ratings

Abstract

Cracking is the most influential distress on the performance of bituminous surfaces of granular pavements and ultimately that of underlying layers. The purpose of the study reported here is to describe the modeling approach adopted in developing cracking deterioration models of bituminous sprayed seals from historical time series subjective condition ratings. In this approach, a multilevel analysis has been applied to capture the variations between observations, segments, and highways. Further, it involved considering all possible contributing factors that affect cracking deterioration of in-service sprayed seals. Factors considered here include surface age, temperature, traffic volume, rainfall, shoulder seal width, and subgrade soil reactivity. The modeling approach has been applied to condition data from five rural highway networks separately then collectively. In the latter, only significant contributing factors from the individual networks’ models are considered. These networks have spray sealed granular pavements with different operating and environmental conditions. Predictions of the overall model have been compared with the currently used model. The latter has been developed for the same networks from two years of subjective condition data, using Markov chains (MC) and surface age as the only predictor. The overall model developed here using multilevel analysis and incorporating the significantly contributing factors predicts earlier deterioration than the MC model currently used. The latter predicts 70% of segments to be in good condition at the age of 5 years, whereas the first predicts only 49%.