Simplified low order composite laminate damping predictions via multi-layer homogenisation. (1st April 2022)

• Record Type: Journal Article
• Authors: Mace, T.; Taylor, J.; Schwingshackl, C.W.
• Is Part Of: Composites. Number 234(2022)
• Publication Date: 2022-04-01
• Subjects: A. Carbon Fibre -- A. Laminates -- B. Internal friction/Damping -- B. Vibration; Composite materials -- Periodicals; Materials science -- Periodicals; Composite materials; Periodicals; Electronic journals; 620.118
• DOI: 10.1016/j.compositesb.2022.109641
• Languages: English
• ISSNs: 1359-8368

Abstract

Abstract: The increased adoption of composite laminates in modern engineering requires advancement in the prediction of their dynamic behaviour. Current damping prediction techniques can be prohibitively time consuming and computationally expensive for application during early design stages, and to abstract three-dimensional geometries. A novel, lower order methodology for damping prediction is proposed, which uses a higher-level of homogenisation than established composite damping prediction techniques to provide a reasonable damping prediction without requiring a detailed model of a laminate's internal structure. Principal loss factor components are harvested from a set of base layup specimens and used to predict the modal loss factors and frequency response of a set of geometrically abstract single layup validation specimens. A numerical study shows the low-order approach to produce approximately equivalent strain energy distributions to a well-established 'layered' approach at reduced computational cost and for a third of the CPU time. Furthermore, the damping and amplitude predictions produced by novel methodology are shown to closely match experimental measurements, providing scope to expand the application of this approach to more complex, multi-layup laminate components. Highlights: Low-order 'smearing' of properties predicts equivalent energies to existing methods. Validation tests show low-order amplitudes/loss factor predictions to correlate well. Approach simplifies modelling, improving computational cost with no loss of accuracy.