Tube2FEM: a general-purpose highly automated pipeline for flow-related processes in (embedded) tubular objects

Area of Science:

• Computational fluid dynamics (CFD)
• Simulation sciences
• Scientific computing

Background:

• Simulating flow in tubular structures is crucial for biomedical and industrial applications.
• Generating simulation-ready meshes from raw 3D data is often labor-intensive and complex.
• Existing methods lack a streamlined, open-source solution for tubular geometry meshing.

Purpose of the Study:

• To present an open-source pipeline for automated mesh generation of tubular structures.
• To bridge the gap between raw 3D models (imaging, CAD, graph networks) and simulation-ready meshes.
• To facilitate finite element method (FEM) and CFD simulations in diverse tubular systems.

Main Methods:

• Developed a pipeline integrating surface mesh processing, center-line extraction, and anisotropic/volumetric meshing.
• Leveraged open-source tools: GIBBON, FEniCS, and Paraview.
• Applied the pipeline to diverse tubular structures including soil-root systems, lung airways, and vascular networks.

Main Results:

• Successfully generated simulation-ready meshes from various data sources for complex tubular geometries.
• Demonstrated pipeline versatility across five distinct application cases.
• Integrated CFD simulations and 3D-1D coupling strategies for enhanced analysis.

Conclusions:

• The open-source pipeline significantly streamlines the mesh generation process for tubular structures.
• It provides a flexible and broadly applicable framework for CFD and FEM simulations.
• Future work will focus on enhancing accuracy, reducing computation time, and advanced modeling strategies.