ABSTRACT

This paper investigates the parametric sensitivity in the Simplified Viscoelastic Continuum Damage (S-VECD) fatigue model of asphalt mixture by using model-based and experimental data-based sensitivity analysis. The model-based sensitivity analysis characterizes the change of model response with respect to individual parameters. The experimental data-based sensitivity analysis incorporates the parameter correlation and characterizes the change of model response with respect to correlated parameters. In this study, experimental data from a total of 31 specimens were used to generate parameters through the Markov Chain Monte Carlo (MCMC) technique to account for parameter correlation. Then, the simulated parameters were used to construct the experimental data-based sensitivity indices. The model-based analysis results in the parameter sensitivity ranking as C 11 > C 12 > D R > α > |E * |. The experimental data-based analysis gives the sensitivity ranking as D R > C 11 , C 12 > |E * | ≈ α.