Abstract
This paper presents the development of a predicted performance model for a flexible pavement condition index for a network in hot asphalt concrete mix in the state of Bahia. Its starting point was a factorial design based on information from a database of highways, provided by the Department of Transport Infrastructure of Bahia. To verify the adequacy of the statistical model was used the straight fit to the observed data set, with determination of the correlation coefficient (r), residue analysis, and normal probability plot of the residuals. The model correlates the performance of a pavement condition index, used for pavement management, to pavement age, traffic and rainfall, factors considered having influence on the pavement deterioration process. The model developed has fit the observed data and showed r = 0.64, so that, there is evidence of the validity of its application for the road network of Bahia. There are no universal performance models and models that have been developed from data observed in studies in specific locations are not necessarily transferable, and have some inherent limitations.
Key words:
pavements management systems; performance models; distresses
