In this paper we present the development of an empirical pavement overlay crack initiation model using condition survey data for the highway system in the state of Washington. We use a semiparametric Cox model and show that its hazard function represents the sample hazard function, which follows a bathtub shape. This is a shape that commonly used parametric models cannot predict. Empirical data complications, such as the endogeneity of overlay design variables and data censoring, are corrected for using appropriate econometric techniques. Our research shows that a specification that captures the main factors responsible for the overlay crack initiation process, combined with careful analysis of the data, can produce models of sufficient realism for maintenance and rehabilitation action selection purposes.
|Original language||English (US)|
|Number of pages||8|
|Journal||Journal of Infrastructure Systems|
|State||Published - Sep 2008|
- Stochastic models
ASJC Scopus subject areas
- Civil and Structural Engineering