Skip to Main content Skip to Navigation

Analyse/Synthèse tridimensionnelle de textures fibreuses

Abstract : The aim of this work is to study fibrous woven materials and to develop morphological analysis of their texture enabling simulation of fibrous layout in yarns. The ―analysis‖ stage consists in data extraction to characterize the geometry of fibers and yarns constituting the woven material, based on tridimensional images generated by high resolution X-ray tomography. An image processing workflow to separate and identify fibers and yarn is proposed. Then, diameters, fiber orientations and local fiber density in yarns are computed. The ―simulation‖ stage targets to generate virtual fibrous materials. An algorithm based on the resolution of a dynamic model allows placing objects representing fibers into a yarn. It uses as input the characteristic parameters previously extracted in the analysis stage or arbitrarily chosen ones. Consequently, all fibers are synthesized according to these constraints. These two stages are successfully applied to characterize and synthesize woven ceramic matrix composites. Geometrical characteristics, extracted from the high-resolution tomographic images, highlight yarn phenomena like shearing planes. The synthesis simulates fibers placement, merging geometrical parameters extracted on high and low resolution. Indeed, using fiber parameters extracted from high resolution images and yarn envelopes from low resolution ones enable generating a result with a more representative woven length.
Document type :
Complete list of metadata
Contributor : Abes Star :  Contact
Submitted on : Friday, March 13, 2015 - 10:23:05 AM
Last modification on : Saturday, October 5, 2019 - 3:29:28 AM
Long-term archiving on: : Monday, April 17, 2017 - 2:29:18 PM


Version validated by the jury (STAR)


  • HAL Id : tel-01131194, version 1


Cedric Chapoullie. Analyse/Synthèse tridimensionnelle de textures fibreuses. Matériaux. Université de Bordeaux, 2014. Français. ⟨NNT : 2014BORD0100⟩. ⟨tel-01131194⟩



Record views


Files downloads