Automatic process control of an automated fibre placement machine

Deposition defects arising during automated fibre placement (AFP) reduces production rate and creates wastage from rejected parts. This issue can be alleviated by identifying and reacting to those defects in real-time during deposition. To demonstrate the concept, this study augmented an AFP process with profilometry sensors and software control. Description of the machine, its sensors, and the control system is provided. In the experiment, representative bonding defects at the tow level representative of a typical AFP layup were introduced. These defects were picked up by the profilometry sensors, and the machine automatically adjusted the process parameters when the defects reached the nip point. Microscopy revealed that such a procedure can influence the material microstructure and reduce in-plane waviness in the final cured laminate. This demonstrates the feasibility of the observe-think-react concept in AFP and provides another step toward real-time defect monitoring and correction in composites manufacturing.