No abstract

Ultrasonic inspection of large CFRP components requires an accurate positioning of the part with respect to the equipment’s coordinate system to ensure normal incidence during the scan. Fulfilling this condition often implies impractical and time-consuming manual adjustments of the part. A new method is proposed for ultrasonically determining the position of a complex part installed in an inspection system, and automatically correcting the offline-programmed path instead of iteratively and manually adjusting the part’s position. Two application strategies of our method are presented and experimentally tested on a complex CFRP landing gear component. The results suggest that even the simplest strategy using a planar learning scan of a randomly positioned part can produce C-scans of equal quality as those obtained with the manual alignment, in a fraction of the time that this process usually takes.