Controllable parameters as the essential components in the analysis, manufacturing and design of 3D woven composites. (10th November 2022)

• Record Type: Journal Article
• Authors: Sitnikova, Elena; Xu, Mingming; Kong, Weiyi; Li, Shuguang
• Is Part Of: Composites science and technology. Volume 230(2022)Part 1
• Publication Date: 2022-11-10
• Subjects: Textile composites -- Elastic behaviour -- Finite element analysis (FEA) -- Multiscale modelling -- Woven composites design; Composite materials -- Periodicals; Composite materials; Fibrous composites; Periodicals; 620.118
• DOI: 10.1016/j.compscitech.2022.109730
• Languages: English
• ISSNs: 0266-3538

Abstract

Abstract: The comprehensive design feasibility for woven composites has been established by identifying the missing component in the design tool for these materials, the so-called controllable parameters. These are the parameters that are involved, directly or indirectly, in manufacturing of woven preforms, and they include the tow densities, the number and the size of filaments in a tow and intra-tow fibre volume fractions. Controllable parameters have been related through a simple procedure to the conventional geometric parameters of woven composites, which eliminates the need for their costly, inefficient, and often unreliable measurements. The controllable parameters provide sufficient representation of woven geometry in terms of both modelling and manufacture, thus offering common terminology to the woven composite designers and manufacturers. This also allows to naturally involve the practical considerations and manufacturing restrictions in material design exercises. With controllable parameters being incorporated in woven composite design framework, a direct analogy can be drawn to classical lamination theory-based design for conventional laminates. The wide applicability of the design tool has been demonstrated via a series of systematic material characterisation exercises carried out with woven composites of sufficiently different internal architectures and constituents, which also showed good predictive capability of the models involved. Graphical abstract: Image 1