An experimental and numerical study of industrially representative wrinkles in carbon fibre composite laminates. (October 2022)

• Record Type: Journal Article
• Authors: Pilato, Nicholas R.; Butler, Richard; Trevarthen, James; Lunt, Alexander J.G.
• Is Part Of: Materials & design. Volume 222(2022)
• Publication Date: 2022-10
• Subjects: Wrinkle -- Finite Element Analysis (FEA) -- Carbon Fibre Composites -- Compression Testing of Composites -- Fracture; Materials -- Periodicals; Engineering design -- Periodicals; Matériaux -- Périodiques; Conception technique -- Périodiques; Electronic journals; 620.11
• DOI: 10.1016/j.matdes.2022.111055
• Languages: English
• ISSNs: 0264-1275

Abstract

Graphical abstract: Highlights: A novel methodology is presented for rapid creation of finite element models for wrinkled composites from measurable non-destructive testing data. Compression test data presented for coupons extracted from a laminate section of an industrial product with non-intentionally induced wrinkling. Wrinkle severity (wrinkle angle in the loading direction) and wrinkle extent were identified as key parameters through presented sensitivity studies. Good agreement is found between the presented modelling method and compression results from commercial coupons. Abstract: Wrinkles (out-of-plane waviness) are a common carbon fibre composite manufacturing defect that causes significant reductions in the components' mechanical properties. The substantial cost implications of empirical testing has led to industrial demands for reliable and representative wrinkle models. This paper presents a novel modelling method for laminates with wrinkle geometries measured from non-destructive testing and/or sample cross-sections. The resulting finite element modelling pre-processor offers a self-contained system of mesh generation, local material orientation calculation, material assignment, boundary condition selection and solver specification. The approach was validated by comparison with pre-existing analytical models and experimental data. Through the resulting rapid model generation, the influence of various physical wrinkle parameters were assessed at an unprecedented rate and sensitivity. For example, this highlighted the dominant dependence of wrinkle angle on the laminate's compressive strength. Additionally, increases in wrinkle extent and reductions in amplitude decay rate were shown to be detrimental to mechanical properties. The model was then validated against an experimental study on industrially relevant carbon-epoxy laminate samples to understand the effect out-of-plane waviness on compressive properties. The model results showed good agreement to the experimental values ( < 15% error) and at least a 32% reduction in compressive strength.