Comparison of PMED Rehabilitation Levels for HMA Overlays

Abstract: Hot Mix Asphalt (HMA) over HMA overlay involves placing a new asphalt layer over an existing HMA pavement. This technique restores surface quality, improves rideability, and extends pavement life while maintaining structural integrity, making it cost-effective for pavement rehabilitation. HMA over HMA is a common overlay type used by several agencies. The Pavement Mechanistic-Empirical Design (PMED) can be used to design HMA over HMA overlays. However, the characterization of existing HMA layer is challenging.
This paper compares the global performance and local predictions (using calibrated models) for HMA over HMA overlay using PMED Level 2 and 3. These levels correspond to the overlay levels and not the material input levels. PMED Level 2 for HMA over HMA overlay requires the distress magnitude and severity, whereas PMED Level 3 requires the structural rating and environmental rating, each of which has five different levels (Excellent, Good, Fair, Poor, Very Poor). A total of 22 HMA over HMA sections were selected for which Level 2 and 3 data and performance data were available. These are existing sections obtained from the Michigan Department of Transportation (MDOT) database. Each section was run as an overlay in PMED for a design life of 20 years. Apart from the overlay level, all other input levels were kept the same. The following PMED models were locally calibrated: bottom-up cracking, top-down cracking, total rutting, thermal + reflective cracking, and IRI.
The global and local predictions were plotted over the line of equality (Level 2 vs. Level 3) for all models. The results show that the global and locally calibrated model predictions were comparable for all models except the thermal + reflective model. The differences between Level 2 and Level 3 for the thermal + reflective cracking model may not be entirely due to PMED rehabilitation levels but also due to the nature of the thermal + reflective cracking model. Any level (Level 2 or Level 3) data can be used depending on the agency's protocols.