Abstract

Online monitoring of keyhole during laser welding is of great importance to evaluate the stability of the live process and the final quality of the weld. Indirect methods based on visual imaging and thermal radiation are lack of strong correlations with keyhole behavior, resulting in a low accuracy. Optical coherence tomography (OCT) shoots beam coaxially with the processing beam into the bottom of the keyhole, enabling direct monitoring of keyhole behavior. However, current OCT keyhole monitoring systems suffer from the harsh noise and are limited to only evaluating the keyhole depth through statistical analysis. In this paper, an enhanced OCT system was built with a capacity of capturing keyhole open/collapse behavior, opening a way of quantitatively evaluating the keyhole behavior. First, the signal was denoised and filled given the nature of fluid dynamics in time domain. Then an algorithm was developed to automatically detect the keyhole open/collapse behavior, based on which the weld status (heat conduction mode and keyhole mode) is determined, and the metrics of interests are characterized quantitatively. The proposed method provides a solid method to quantitatively evaluate the keyhole stability, which is well suited for developing new laser welding process and monitoring laser involved high volume manufacturing.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call