• Yi, Qiuji
• Hughes, Robert R.
• Wilcox, Pd

Abstract

The mechanical performance of carbon fibre reinforced polymer (CFRP) composites is determined by the stacking sequence of specifically orientated ply lamina layers. In these multi-layer structures, global ply misalignment and local fibre waviness can occur during the manufacturing process, leading to reduced performance and structural failure during operation. High-frequency eddy-current testing (ECT) has previously demonstrated its capability for detecting orientation-related features, including fibre orientation and waviness. However, accurate sensor optimisation and inversion of orientation & material structure cannot be achieved without proper, validated modelling techniques to simulate CFRP features. The lack of suitable models is in part due to CFRP exhibiting highly-complex electromagnetic interactions between fibres, lamina and the ECT sensors, which are challenging to integrate. This work proposes a novel finite element modelling approach to simulate the ECT response to planar multi-layered CFRP components. The fibre tow structure of each unidirectional ply is modelled using orientation dependant 2D conductivity tensor waveforms, and virtual 2D ECT scans are simulated by shifting the waveforms within the model mesh. The results demonstrate that idealised electromagnetic characteristics of the CFRP structure can be successfully modelled compared with experimental data and that 2D ECT data of complex CFRP layers structures can be simulated with improved computational speeds, up to 5x faster compared to standard approaches. Automated data-analysis tools, including Radon transform (RT) and 2D FFT, are employed to validate the simulated 2D scan data through the characterisation of fibre orientations and simulated 2D scans used to evaluate the orientation inversion techniques. The results demonstrate that RT analysis detects fibre orientations with better accuracy, precision and consistency than equivalent 2D FFT analysis techniques. The simulation also demonstrates the reduced resistivity losses compared with isotropic materials caused by the different heterogeneous and multi-layer structures. It predicts high current densities at interfaces of plies with orthogonal orientations, resulting in an effective interface skin-depth.

Topics

• impedance spectroscopy
• polymer
• Carbon
• resistivity
• simulation
• layered
• composite
• isotropic