High-performance carbon fibre reinforced polymers (CFRPs) are increasingly used across many industrial sectors, leading to the adoption of more automated manufacturing processes to meet demand. However, quality control and characterisation of the in-process manufacturing are rudimentary. In this paper, a bi-directional differential eddy current testing (ECT) probe was developed for in-line defect detection and implemented to detect subsurface features in an automated CFRP Filament Winding (FW) process. The ECT probe was installed within an industrial FW manufacturing system and performed helical scans after each deposited layer of an example pressure tank layup, and was used to map successive layers of CFRP tape, successfully demonstrating the capability to identify and map defects and tape structures to a 3D geometry, even when buried under multiple layers. These results demonstrate the effectiveness of the ECT sensing probe for defect detection, and its surprising capacity to detect simulated waviness defects under six hoop-layers, with the defect signal-to-noise ratios increasing as more layers are added. This ECT sensing approach enables real-time quality monitoring and improved manufacturing rates. The probe offers a non-contact, non-destructive approach to identify features of interest and subsurface defects under varying conditions. Deployment of in-line ECT sensors in automated composite manufacturing processes could enable immediate rework of the part during manufacturing and reduce the need for costly post-inspection.
