Introduction

The overall objective of the Mechanistic-Empirical Pavement Design Guide (MEPDG) is to provide the highway community with a state-of-the-practice tool for the design of new and rehabilitated pavement structures, based on mechanistic-empirical (M-E) principles. This means that the design procedure calculates pavement responses (stresses, strains, and deflections) and uses those responses to compute incremental damage over time. The procedure empirically relates the cumulative damage to observed pavement distresses. This M-E based procedure is shown in flowchart form in Figure 1-1. "MEPDG," as used in this guide, refers to the documentation and software package (NCHRP 2007).

Pavement distress prediction models, or transfer functions, are the key components of any M-E design and analysis procedure. The accuracy of performance prediction models depends on an effective process of calibration and subsequent validation with independent data sets. Pavement engineers gain confidence in the procedure by seeing an acceptable correlation between observed levels of distress in the field and those levels predicted with the performance model or transfer function. The validation of the performance prediction model is a mandatory step in their development to establish confidence in the design and analysis procedure and facilitate its acceptance and use. It is also necessary to establish the design reliability procedure. It is essential that distress prediction models be properly calibrated prior to adopting and using them for design purposes.

The term calibration refers to the mathematical process through which the total error (often termed residual) or difference between observed and predicted values of distress is minimized. The term validation refers to the process to confirm that the calibrated model can produce robust and accurate predictions for cases other than those used for model calibration. A successful validation process requires that the bias and precision statistics of the model for the validation data set be similar to those obtained during calibration. This calibration-validation process is critical for potential users to have confidence in the design procedure.

All performance models in the MEPDG were calibrated on a global level to observed field performance over a representative sample of pavement test sites throughout North America. The Long Term Pavement Performance ( LTPP) test sections were used extensively in the calibration process, because of the consistency in the monitored data over time and the diversity of test sections spread throughout North America. Other experimental test sections were also included such as MnRoad and Vandalia. However, policies on pavement preservation and maintenance, construction and material specifications, and materials vary across the United States and are not considered directly in the MEDPG. These factors can be considered indirectly through the local calibration parameters included in the MEPDG. The purpose of this guide is to provide guidance in calibrating the MEPDG to local conditions and materials that may not have been included in the global calibration process.