A Level-Set Based Representative Volume Element Generator and XFEM Simulations for Textile and 3D-Reinforced...

Area of Science:

• Materials Science
• Computational Mechanics
• Composite Materials

Background:

• Accurate prediction of mechanical properties in textile reinforced composites is crucial.
• Defining realistic representative volume elements (RVEs) is a key challenge in computational homogenization.
• Existing methods often lack detailed geometrical representation of textile reinforcements.

Purpose of the Study:

• To present a new framework for computational homogenization of textile reinforced composites.
• To develop a geometrically-based weave generator for realistic RVE creation.
• To integrate advanced geometrical processing for improved mechanical property derivation.

Main Methods:

• Development of a geometrically-based weave generator considering yarn contact and tension.
• Utilizing a level set-based post-processor to adapt yarn cross-section shapes.
• Integration with an extended finite element (XFE) scheme for property analysis.

Main Results:

• Generation of realistic 3D textile geometries accounting for complex yarn interactions.
• Implicit accommodation of yarn tension and explicit adaptation of yarn cross-sections.
• A seamless workflow from geometry generation to computational homogenization.

Conclusions:

• The proposed framework enables more accurate computational homogenization of textile composites.
• The geometrically-based weave generator significantly enhances RVE realism.
• This approach provides a robust tool for predicting the mechanical behavior of advanced composite materials.