Date of this Version
Asphaltic materials are classical examples of multi-phase composites in different length scales. The understanding of the mechanical behavior of asphaltic materials has been a challenge to the pavement mechanics community due to multiple complexities involved: heterogeneity, anisotropy, nonlinear inelasticity, and damage in multiple forms. The micromechanics-based models based on numerical methods have been receiving attention from the pavement mechanics community because the modeling method can account for those complexities of asphaltic materials by considering the effects of material properties and geometric characteristics of individual components on overall performance behavior of mixture or structure. As a step-wise effort, this study intends to identify some of key relevant mechanical characteristics such as linear viscoelastic, non-linear viscoelastic, and fracture properties of asphaltic materials in two different length scales, e.g., mixture scale and component scale. More specifically, this study developed testing-analysis methods to rigorously define the stress-dependent nonlinear viscoelastic material characteristics at various stress levels and the viscoelastic mixed-mode fracture properties at different loading rates and testing temperatures.
The results from three-dimensional finite element simulations of the pavement structure presented significant differences between the linear viscoelastic approach and the nonlinear viscoelastic modeling in the prediction of pavement performance with respect to rutting. This implies that differences between the two approaches are considered significant and should be addressed in the process of performance-based pavement design.
The Semi-circular Bend (SCB) fracture test presented reasonable and repeatable results. The test and analysis results in this study suggest that the rate-, temperature-, mode- dependent fracture properties are necessary in the structural design of asphaltic pavements with which a wide range of strain rates and service temperatures is usually associated.
Advisor: Yong-Rak Kim