Abstract

Payment deterioration models are an important input for the efficient management of pavement systems, the allocation of cost responsibilities to various vehicle classes for their use of the highway system, and the design of pavement structures. This paper is concerned with the development of an empirical rutting progression model using an experimental data set from WesTrack. The data used in this paper consist of an unbalanced panel data set with 860 observations. The salient features of the model specification are: 1) three properties of the mix are sufficient to model the performance of the asphalt concrete pavement accurately, 2) the model captures the effects of high air temperatures at WesTrack, and 3) the model predicts rut depths by adding predicted values of the increment of rut depth for each time period, which is particularly advantageous in a pavement management context. The three mix properties are a gradation index, which is obtained from the aggregate gradation, the voids filled with asphalt obtained for the construction mix in the Superpave gyratory compactor, and the initial in-place air voids. The specified model is non-linear in the variables and the parameters, and is estimated using a random effects specification to account for unobserved heterogeneity. The estimation results and prediction tests show that the model replicates the observed pavement behavior at WesTrack well.