Modeling flexible pavement overlay performance for use with quality-related specifications

Abstract

Recent advancements have been made relating the constructed properties of bituminous materials to the flexible pavement performance over time. The research in this study explores one of the newly-developed prediction tools to assess its potential for designing longer-lasting asphalt mixtures in the state of Rhode Island. A parametric study was first conducted to establish performance-related specification (PRS) relationships for hot mix asphalt (HMA) overlays, which make up the majority of pavement projects in Rhode Island, using the Quality-Related Specification Software (QRSS). A framework was developed for use in setting post-construction targets for rutting performance in asphalt pavement overlays, in order to achieve the pay factors selected by the Rhode Island Department of Transportation and to encourage better quality construction. Because the majority of pavement construction done in Rhode Island involves flexible overlays, assumptions involving mixture densification due to traffic, distress severity levels, and aging rates were made to enable the proposed modeling of an overlay as two newly constructed layers in order to overcome the limitations in the QRSS software. Regression equations showed a significant percentage of predictive model predicted life differences (PLD) can be captured by utilizing a relationship between PLD and the difference between as-built and job mix formula (JMF) effective dynamic modulus (|E*|) values. The analysis showed that structural thickness and binder viscosity are critical to predicting PLD. Results showed that the PLD in terms of overlay rutting performance does not vary significantly with realistic adjustments to the as-built targets. Judging by the level of importance associated with stiffness values throughout the program, a procedure seeking to relate stiffness directly to service life with a small number of variables appears to be an optimal choice for future endeavors. By relating stiffness, major input categories (structural design, traffic, etc.), and service life, software users can predict a service life parameter, without embarking on a full mixture design process or |E*| testing sweep. It is recommended that a model be developed for alligator cracking and top-down cracking in HMA overlays using structural thickness, binder viscosity, and the |E*| of the constructed asphalt mixture.