Press forming of commingled yarn based composites: The preform contribution

• Published in: Composites science and technology Vol. 67; no. 3-4; pp. 515 - 524
• Main Authors: Wysocki, M., Toll, S., Larsson, R.
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01-03-2007; Elsevier
• Subjects: Applied sciences; B. Non-linear behaviour; Bundle response; C. Anelasticity; C. Computational simulation; Commingled yarn based composites; Compaction; Composite materials; Composites; Computer simulation; deformation; elasticity; Exact sciences and technology; fiber reinforced composite; Fluid pressure build-up; Forms of application and semi-finished materials; Mathematical models; Polymer industry, paints, wood; Porous materials; Porous media theory; preform; Preforming; Technology of polymers; Wetting; B. Non-linear behaviour; C. Computational simulation; C. Anelasticity; Porous media theory; Experimental test; Mechanical model; Stress strain relation; Propylene polymer; Fiber reinforced material; Theoretical study; Mechanical properties; Press forming; Mineral fiber; Modeling; Composite material; Consolidation; Porous medium flow; Non linear model; Anelasticity; Olefin polymer; Numerical simulation; Processing parameter; Glass fiber
• ISSN: 0266-3538; 1879-1050; 1879-1050
• DOI: 10.1016/j.compscitech.2006.08.016

The process of press forming commingled yarn composites is studied and modelled based on the theory of porous media. Two sub processes are considered: (1) the wetting and compaction of individual bundles and (2) the overall preform deformation. An experimental method is introduced, to study these two sub processes separately. The experiment suppresses the fluid pressure build-up, and thus the process of wetting and bundle compaction, by draining the resin into a porous wall. In consequence, the measured response consists only of the preform contribution. Finally, the total response is simulated and tested against independent consolidation experiments. It was demonstrated that the present approach gives better predictions compared to models based on the bundle response only.