Mesoscopic Heterogeneous Modeling Method for Polyurethane-Solidified Ballast Bed Based on Virtual Ray Casting...

Area of Science:

• Civil Engineering
• Materials Science
• Computational Mechanics

Background:

• Mesoscale modeling of composite materials like polyurethane-solidified ballast beds (PSBBs) traditionally relies on X-ray computed tomography (XCT).
• Existing methods face limitations in specimen size, cost, and complex post-processing.

Purpose of the Study:

• To develop a novel mesoscale modeling methodology for PSBBs that bypasses the need for XCT.
• To establish a computationally efficient and accurate approach for analyzing the mechanical behavior of particle-binder composites.

Main Methods:

• Coupling the Discrete Element Method (DEM) with the Finite Element Method (FEM).
• Reconstructing high-fidelity discrete-element geometry from 3D laser scans.
• Utilizing a virtual-ray casting algorithm for heterogeneous material mesoscale reconstruction.
• Validating model accuracy and mesh convergence with laboratory uniaxial compression tests.

Main Results:

• Determined an optimal mesh size of 0.4 times the minimum particle size (0.4 Dmin) for model accuracy.
• Identified that a sleeper width of at least 0.73 times the ballast bed width (0.73 Wb) achieves optimal stress diffusion and displacement control.
• Demonstrated the method's effectiveness for analyzing ballast bed mechanical responses.

Conclusions:

• The developed DEM-FEM coupled method provides an efficient alternative to XCT for mesoscale modeling of PSBBs.
• The findings offer guidance on optimizing sleeper dimensions for enhanced ballast bed performance.
• The methodology is broadly applicable to other multiphase particle-binder composite systems.