Simulation of unbound material resilient modulus effects on mechanistic-empirical pavement designs

Abstract

The variability of resilient modulus (M _R) of unbound materials and subgrade due to laboratory test conditions affect pavement performance and designs. The performance-based mechanistic-empirical pavement design guide (MEPDG) is gaining more popularity in recent years for pavement design use. However, limited research efforts have quantitatively studied M _R effects based on ME models. This research targets to evaluate the influences of M _R variability on pavement performance and designs based on the MEPDG performance models. With a normal-distribution of M _R seed values, pavement responses were computed with a layer-elastic analysis model, pavement performance was then predicted using MEPDG models, and design variability was studied via Monte Carlo simulation. Results indicate that the relationship between layer design thickness and M _R varies from almost linear to nonlinear, which is highly dependent on the pavement structure and material properties. For the evaluated specific pavement structure and range of M _R values, the least susceptible is the HMA design thickness as a function of M _R under fatigue with a design Coefficient of Variance (CV) of 7.51 %, while the most susceptible is the base design thickness as a function of M _R also under fatigue with a CV of 54.32 %. The combined effect of both rut depth and fatigue life considering the variability of both base and subgrade results in a design CV of 22.58 % for asphalt layer and 26.08 % for base layer. When using a weaker base layer or a thinner HMA layer, the modeled thickness design CV has changed −4.19 to 1.14 %.