Calibration of CalME models using WesTrack Performance Data

Since 2000 the UCPRC has been developing a mechanistic-empirical (M-E) pavement design method for Caltrans. The software associated with the flexible pavement design methodology is termed CalME. This report presents the results of a validation and calibration study for the CalME software using the performance data from the FHWA’s WesTrack accelerated pavement test (APT) program. This APT was conducted in the period between March 1996 and February 1999 at a site near Carson City, Nevada; all 26 initial test sections were used in the study. These sections, constructed with Superpave-type hot-mix asphalt (HMA) consisting of a single source for the aggregate and asphalt binder (PG64-28), were subjected up to approximately 5 million ESALs. HMA variables included: aggregate gradation, asphalt content, and degree of compaction. The calibration procedure used both falling weight deflectometer (FWD) field test data and laboratory-measured properties of all the mixes. Laboratory-measured characteristics included: mix stiffness, flexural fatigue characteristics, and rutting response as determined in the SHRP-developed shear test. Results of this CalME validation and calibration exercise using the WesTrack data indicated that the models selected are able to predict the performance observed on this full-scale APT experiment using actual trucks and environmental conditions at this site in western Nevada. Differences between actual and predicted performance were identified and recommendations for reducing these differences were made; e.g., modifications to laboratory testing and to shift factors relating laboratory data to field performance.