Weaving 3D Shell-Shaped Fabrics Using Discontinuous and Curved Weft Rows

Paper Title

Weaving 3D Shell-Shaped Fabrics Using Discontinuous and Curved Weft Rows

Publication Info

• Topic area: Computational design and fabrication of 3D woven textiles.
• Keywords: 3D weaving, Variable Reed, discontinuous wefts, curved weft rows, computational design tools, textile engineering, soft materials, HCI, shell-shaped fabrics, parametric design.

Background and Problem

• Problem / challenge: Traditional weaving methods are limited to producing flat fabrics due to the perpendicular arrangement of warp and weft yarns. Existing techniques for creating 3D woven geometries often require specialized looms, complex post-processing, or are constrained in the shapes they can achieve.
• Significance: Achieving 3D geometries directly on the loom can eliminate the need for cutting, sewing, or molding, enabling more efficient and versatile textile production for applications in garments, upholstery, architecture, and composites.
• Motivation and related work: Prior work has explored multi-layer weaving, shrinkable yarns, and 3D looms, but these methods are limited in geometry, require specialized equipment, or involve post-processing. Discontinuous wefts and modified reeds have been used in artistic and functional contexts but not systematically for 3D shell-shaped fabrics. This paper builds on these efforts by introducing a novel weaving method and computational tools.

Solution

• Proposed approach: Development of a Variable Reed to enable weaving of discontinuous and curved weft rows, combined with computational design tools to translate 3D geometries into weaving drafts and stencil plans.
• Novelty:
  1. Introduction of the Variable Reed for dynamic dent positioning, enabling non-perpendicular warp-weft angles.
  2. Integration of discontinuous and curved weft rows to create 3D shell-shaped fabrics directly on the loom.
  3. Development of computational tools to generate weaving drafts and stencil plans from input 3D geometries.
  4. Demonstration of folding hinges (mountain and valley lines) for transforming flat fabrics into 3D forms.
• Procedure and key techniques:
  1. Replace the standard reed with the Variable Reed, which uses stencils to dynamically position dents.
  2. Weave discontinuous weft rows to define base shapes, followed by curved rows to achieve 3D geometries.
  3. Incorporate mountain and valley folding hinges using twill weaves.
  4. Use computational tools (Processing and Rhino/Grasshopper) to design weaving drafts and stencils based on input 3D surfaces.
  5. Manually fold the woven fabric into its 3D shape after removal from the loom.

Results

• Concrete findings:
  • Successfully created 3D shell-shaped fabrics, including trapezoids, arches, a half-dome, and a wing.
  • Maximum sustainable warp-weft angle determined to be 60° under the tested parameters.
  • 3D scan of the wing-shaped fabric showed a maximum deviation of 12mm from the input geometry.
• Advantage over baselines:
  • Eliminates the need for specialized 3D looms or post-processing steps like cutting and sewing.
  • Enables direct encoding of 3D geometry into the weaving process.
  • Supports iterative design workflows through computational tools.
• Experiments / evaluation:
  • Woven samples included trapezoids, arches, and freeform surfaces.
  • Evaluated folding hinges, angle sustainability, and alignment with input geometries.
  • Computational tools allowed iterative refinement of designs and facilitated the weaving process.
• Limitations and future work:
  • Limited working space of the Variable Reed restricts maximum shape height (~13cm).
  • Manual stencil replacement is labor-intensive; future work aims to develop automated actuation.
  • Challenges in maintaining weft tension and preventing slack in curved rows.
  • Plans to explore inverse computational methods for pre-distorting stencils and integrating functional yarns for sensing and actuation.
  • Future user studies to evaluate the creative potential of the computational tools.

Summary

This paper introduces a novel method for weaving 3D shell-shaped fabrics using the Variable Reed and computational design tools. By enabling discontinuous and curved weft rows, the approach eliminates the need for specialized looms or post-processing, directly encoding 3D geometry into the weaving process. Demonstrations include trapezoids, arches, a half-dome, and a wing, with computational tools facilitating iterative design workflows. While the method is currently limited by manual stencil replacement and working space, future work aims to address these challenges and explore applications in composites, architecture, and e-textiles. This research contributes to HCI by bridging traditional weaving techniques with computational fabrication and material programming.