Multiscale Viscoelastic Analysis of Asphalt Concrete

Area of Science:

• Civil Engineering
• Materials Science
• Computational Mechanics

Background:

• Numerical modeling of asphalt concrete is crucial for road engineering.
• Realistic material models and efficient computation are key challenges.
• Existing methods often lack integrated microstructure analysis and efficient solvers.

Purpose of the Study:

• To develop a novel numerical framework for asphalt concrete analysis.
• To integrate microstructure reconstruction, realistic material models, and efficient numerical methods.
• To validate the framework's accuracy and computational efficiency.

Main Methods:

• Reconstruction of asphalt concrete microstructure using image processing.
• Application of the Burgers material model for mastic and linear elastic model for aggregates.
• Implementation of a numerical homogenization method to accelerate finite element analysis.
• Integration of the Burgers model with numerical homogenization for small strain analysis.

Main Results:

• The developed framework reliably reconstructs asphalt concrete microstructure.
• The Burgers material model and homogenization method were successfully integrated.
• Homogenization error was below 7% for both elasticity and viscoelasticity problems.
• A reduction in degrees of freedom exceeding 510 times was achieved.

Conclusions:

• The developed methodology is applicable to analyzing viscoelastic materials under small strains.
• The framework offers a significant improvement in numerical efficiency for asphalt concrete modeling.
• This approach enhances the reliability and speed of road engineering simulations.