An automatic methodology to CT-scans of 2D woven textile fabrics to structured finite element and voxel meshes

This paper presents a novel methodology to isolate the individual tow information contained within relatively low resolution mu CT scans of real textile specimens. The proposed method can process high fibre fraction (V-f > 0.60) datasets in a fast and automated manner with minimal user interaction. The methodology has been validated qualitatively and quantitatively. The geometric description of the isolated tows can be used to create structured finite element meshes and voxel meshes, that contain the local tow information such as fibre volume fraction and fibre direction using an in-house meshing algorithm. The structured finite element meshes can be used as initial geometry for textile compaction simulations. The voxel meshes can be used directly for a flow simulation to predict the textile permeability. Finally, a methodology to observe the geometrical evolution of the same tows from different mu CT datasets is also presented.