Numerical validation of carbon fibre composite crush absorbers via finite element analysis for dynamic impact energy absorption

This research aimed to investigate the crashworthiness of carbon fibre composite structures for progressive crush energy absorption to be implemented as crush energy absorbers in the automotive sector. The study employed a combination of finite element simulation, experimental testing, and optimization of material cards and modelling approaches to validate the dynamic crush response behaviour of carbon fibre structures under various loading conditions. Five parametric studies were conducted to explore different aspects of carbon fibre crushing behaviour, material card optimization, sensitivity to changes in strengths and strains-to-failure, and failure mechanisms. The research highlighted the importance of material card selection, specifically the MAT54 card, and its parameters, boundary conditions, mesh quality, and element type on the behaviour of composite materials under crushing loads. The study determined that the MAT54 material card can accurately depict progressive crush behaviour but requires modification according to the failure mode of the composite structure to ensure a reliable representation of the results.

Parametric study 1 validated the dynamic crush response behaviour of carbon fibre structures under progressive crushing. The results showed a significant correlation between the experimental and simulation models, establishing that MAT54 material card as an accurate portrayal of the progressive dynamic crush behaviour for unidirectional carbon fibre tubes.

Parametric study 2 analysed the crush performance of a bi-directional truncated cylindrical carbon fibre tube with an angled crush trigger mechanism. The results showed that delamination and fibre fracture through axial compression were the primary modes of failure with an average stabilised crush force of 24 kN and that the FE model was effective in correlating the crush behaviour. Key parameters XC and DFAILC significantly influence element failure and average crush force, especially in 0° ply deletion. Lower DFAILC values reduce peak stress before ultimate strength, whereas higher values lead to plastic deformation until ply deletion. Similarly, smaller XC values result in plasticity under compression.

Parametric study 3 evaluated the crush performance of carbon fibre tubes under dynamic impact with crush trigger from the base and exhibited an average stabilised crush force of 35 kN. The results indicated that modifications to the material card definition will be necessary to correlate tensile shear fibre fracture failure mechanism as in the FE model the failure mode was observed to be changing from brittle to ductile behaviour. Notably, XT represents the stabilized crush force, and the reduced stabilized crush force observed suggests an inaccurate representation of failure mechanisms. Therefore, for composite structures that fails due to laminate tearing or fibre fracture, lower XT values are not recommended. Additionally for this failure mode DFAILC values must be kept low to avoid higher initial peak loads.

Parametric study 4 analysed the crush performance of a group of thin-walled carbon fibre tubes with an angled crush trigger which exhibited an average stabilised crush force of 18 kN. The findings demonstrated a strong correlation between the finite element (FE) and experimental models when triangular elements were excluded from the crush trigger zone, highlighting a significant insight. Consequently, for crush structures with crush triggers, optimal correlation capturing the initial peak load is achieved by employing quad meshing in the crush trigger region.

Parametric study 5 highlighted the challenges of manufacturing homogeneous carbon fibre samples with varying surface thickness. The results showed that carbon fibre parts with varying surface thicknesses can be effective energy absorbers in crush applications as the initial peak load is considerably lower than any other crush structure investigated. However, with an average stabilised crush force of 17.5 kN the results exhibited discrepancies and were unsuitable for accurate correlation with the FE model due to the inhomogeneity of the experimental test samples.

The research findings contribute constructive insights to simulation parameters for carbon fibre composite structures for progressive crush energy absorption. The insights gained from this study have the potential to be applied in the design and development of safer, more efficient vehicles with enhanced energy-absorbing capabilities.