Concrete Mesostructure Modeling via Random Radius Field and Rigid Body Dynamics Packing

Area of Science:

• Materials Science
• Computational Mechanics
• Civil Engineering

Background:

• Accurate mesostructure generation is crucial for simulating heterogeneous materials like concrete.
• Existing methods often struggle to capture realistic aggregate morphologies and packing densities.

Purpose of the Study:

• To develop an efficient framework for generating realistic 3D mesostructures.
• To integrate stochastic geometry with physically based packing for improved simulations.
• To validate the framework using recycled aggregate concrete (RAC) fracture simulations.

Main Methods:

• Developed a Random Radius Field (RRF) method using multi-scale noise and topology optimization for aggregate generation.
• Implemented a Rigid Body Dynamics (RBD) based packing strategy simulating physical casting processes (gravity, vibration).
• Generated multi-phase mesostructures and performed fracture simulations on RAC.

Main Results:

• Generated 3D aggregates with realistic and controllable morphologies.
• Achieved high-density aggregate skeletons through physically based packing.
• Successfully reproduced crack propagation patterns and damage evolution in RAC simulations.

Conclusions:

• The proposed framework is an effective tool for mesoscopic modeling of heterogeneous concrete.
• The integration of stochastic geometry and RBD packing enhances simulation accuracy.
• Validated framework shows potential for analyzing material behavior under various conditions.