Empirical-Markovian model for predicting the overlay design thickness for asphalt concrete pavement

Abstract

An Empirical-Markovian model has been developed to predict the overlay design thickness for asphalt concrete pavement from relevant design parameters. The Empirical-Markovian model mainly predicts the structural capacity of overlaid pavement as a function of the structural capacity associated with original pavement, annual traffic growth rate, rehabilitation scheduling time, and two calibration constants. The structural capacity is evaluated using either the structural number (SN) or gravel equivalent (GE) deployed by the AASHTO and Caltrans design methods for flexible pavement, respectively. The Empirical-Markovian model provides the practitioner with two options as related to the performance of overlaid pavement. The first option enforces the performance of overlaid pavement to be similar to that of the original pavement, an objective achieved by requiring the deterioration transition probabilities for overlaid pavement to be the same as the corresponding ones for original pavement. The second option imposes improved performance of overlaid pavement compared to that of the original pavement, an objective accomplished by requiring the deterioration transition probabilities of overlaid pavement to be lower than the corresponding values associated with original pavement. The two calibration constants can be estimated by either minimizing the sum of squared errors (SSE) applied to historical records of pavement distress (forward approach) or a backward solution of the developed Empirical-Markovian model mainly relying on historical records of pavement rehabilitation. Two case studies are presented to demonstrate the use of both forward and backward approaches with results seem to be in line with the common practice.