Automatic weaving method for three-dimensional composite preforms using vision system. Issue 15 (August 2021)

• Record Type: Journal Article
• Authors: Li, Siyuan; Shan, Zhongde; Du, Dong; Zhan, Li; Li, Zhikun; Liu, Yunzhi
• Is Part Of: Textile research journal. Volume 91:Issue 15/16(2021)
• Page Start: 1876
• Page End: 1888
• Publication Date: 2021-08
• Subjects: Preform -- weaving -- machine vision -- digital manufacturing; Textile industry -- Periodicals; Textile fabrics -- Periodicals; Textile research -- Periodicals; Textiles et tissus -- Industrie et commerce -- Périodiques; Textiles et tissus -- Périodiques; Textiles et tissus -- Recherche -- Périodiques; Electronic journals; 677.005
• DOI: 10.1177/0040517520982385
• Languages: English
• ISSNs: 0040-5175

Abstract

Three-dimensional composite preform is the main structure of fiber-reinforced composites. During the weaving process of large-sized three-dimensional composite preform, relative rotation or translation between the fiber feeder and guided array occurs before feeding. Besides, the weaving needles can be at different heights after moving out from the guided array. These problems are mostly detected and adjusted manually. To make the weaving process more precise and efficient, we propose machine vision-based methods which could realize accurate estimation and adjustment of the relative position-pose between the fiber feeder and guided array, and make the needles pressing process automatic by recognizing the position of the weaving needles. The results show that the estimation error of relative position-pose is within 5%, and the rate of unrecognized weaving needles is 2%. Our proposed methods improve the automation level of weaving, and are conducive to the development of preform forming toward digital manufacturing.