Investigation into Key Pavement Materials and Local Calibration on MEPDG

The release of Mechanistic-Empirical Pavement Design Guide (MEPDG) in 2004 has been leading a transition from empirically-based pavement design to a mechanical-empirical procedure. The pavement performance prediction models in the MEPDG combines design inputs such as material properties, traffic, and climate to the observed field performance. Since the prediction models were primarily calibrated through inputs and pavement performance data from Long Term Pavement Performance database, local calibrations were highly recommended due to the potential differences between national and local conditions.

Key properties of pavement materials were investigated for the local calibration of the MEPDG, including the coefficient of thermal expansion (CTE) of cement concrete and the resilient modulus of soils. CTE values and other properties of concrete from eight concrete plants were investigated. A micromechanical model was proposed to predict concrete CTE considering the time and energy exhausted experimental methods. The thermal stress analysis was conducted on a composite material composed of aggregate and cement paste. Aggregate gradation was incorporated into the concrete CTE prediction model. The proposed model was validated by experimental data. Sensitivity analysis was also performed to explore the major factors affecting concrete CTE.

The MEPDG utilizes the generalized model to describe the subgrade stiffness. Coefficients of the generalized model were regressed from the cyclic triaxial load test data for soils in Tennessee. Also the coefficients were correlated with soil physical properties and employed in evaluating the seasonal variation of subgrade resilient modulus. The influences of seasonal variation in subgrade resilient modulus on pavement performance were explored and found significant. Therefore, seasonal variation of soil resilient modulus should be considered in pavement design and analysis in MEPDG.

The highway pavement sections in the Tennessee pavement management system were analyzed using the MEPDG version 1.1. This analysis indicates that the national calibrated models predict pavement performance poorly in comparison with measured data. Local calibrations on rutting transfer functions were conducted on the two main types of pavements, i.e., asphalt overlay on cement concrete pavement and asphalt overlay on asphalt pavement. With the local coefficients provided, the MEPDG provides better agreement between predicted and measured rutting.