Buch, Neeraj; Chatti, Karim; Haider, Syed W; Baladi, Gilbert; Brink, Wouter; Harsini, Iman.
Implementation; Mechanistic-empirical pavement design; Overlays (Pavements); Pavement maintenance; Pavement management systems; Pavement performance; Rehabilitation (Maintenance)
The main objectives of Task 2 of the project were to determine the impact of various input variables on the predicted pavement performance for the selected rehabilitation design alternatives in the MEPDG/DARWin-ME, and to verify the pavement performance models for MDOT rehabilitation design practice. In general, for HMA over HMA, the overlay thickness and HMA volumetrics are the most significant inputs for the overlay layer while the existing thickness and pavement condition rating have a significant effect on pavement performance among the inputs related to the existing pavement. For composite pavements, overlay thickness and HMA air voids are significant inputs for the overlay layer. In addition, among the inputs related to the existing intact PCC pavement, the existing thickness and PCC layer modulus have a significant effect on pavement performance. For rubblized pavements, the HMA air voids and effective binder content are the most significant inputs for the overlay layer. Furthermore, for longitudinal cracking and IRI, existing PCC thickness is more important as compared to the existing PCC layer modulus. However, existing PCC layer modulus is more significant for alligator cracking and rutting. For unbonded overlays, all overlay related inputs significantly impact the cracking performance while the PCC elastic modulus is the most important among inputs related to existing layers. The interaction between overlay air voids and existing pavement thickness significantly impacts all performance measures among HMA rehabilitation options. The interaction between overlay thickness and existing PCC layer modulus is the most significant effect on unbonded overlay performance. It should be noted that all analyses were conducted using the inputs ranges reflecting Michigan practices. The verification of the performance prediction models based on the selected projects for different rehabilitation options show the need for local calibration.