Summary of Cracking Calibration

The transfer function is in essence an empirical correlation between the fatigue or reflective cracking damage predicted by the ME models and the probability of surface cracking. To ensure the high accuracy of such empirical correlations, it is necessary to divide a highway network into sub-networks in which the pavement structures are similar and, in turn, have similar failure mechanisms.

The list of sub-networks for the field calibration is shown in the table below. As shown in the table, the division is mostly based on the structure type.

Before the performance data extracted from the Caltrans pavement management system (PMS) software program PaveM can be used for the field calibration, the pavements the data covers must be divided into short lane-by-lane segments with uniform construction histories, traffic, and climate. This results in approximately uniform explanatory variables for the associated performance time histories. Each of these segments with its associated performance time history will serve as a basic unit for field calibration and are hereafter referred to as “virtual projects.” Note that multiple virtual projects can occupy the same space but they must be from different time periods. A brief summary of the amount of data available for each of the sub-networks is listed in the following table.

A summary of the field calibration results is listed in the following table.

Note that these calibration factors are periodically updated when new data become available and new rounds of calibration are conducted. Some of the subnetworks (such as R-FDR-C) has very little available data so the calibration needs to be checked for reasonableness and adjusted accordingly.