Performance Evaluation of Asphalt Pavement Resurfacing Treatments Using Structural Equation Modeling

Abstract

The structural equation modeling (SEM) method was used to evaluate the performance of pavement resurfacing treatments. The pavement performance was regarded as a latent variable that cannot be directly observed, pavement condition data such as roughness, rutting, and cracking were regarded as endogenous observed variables, and influencing factors such as traffic and design features were regarded as exogenous observed variables. More than 200 asphalt pavement resurfacing projects applied in Tennessee were investigated for this study. Both the single latent pavement condition (LPC) factor model and the multiple-factor model were built. An exploratory factor analysis was conducted to determine the three latent condition factors LPC1, LPC2, and LPC3 representing ride quality, early age cracking, and excessive distresses, respectively. The results indicated that the multiple-factor model provided better model fitness than the single-factor model. Transverse cracking was the strongest indicator of LPC2, followed by longitudinal wheel path cracking and longitudinal nonwheel path cracking. The influence of rutting on LPC3 was the greatest, followed by fatigue cracking, block cracking, and longitudinal joint cracking. The analysis results also showed that highway class (interstates or state routes) was significant for the three LPCs. The factor loadings of age, mill, thickness, and traffic level on LPC3 were greater than the factor loadings on LPC1 and LPC2. The SEM method was proved to be able to quantify the effects of the five influencing variables on each pavement distress indicator.