In-line Quality Monitoring System by Using Multiple Regression Equation in Remote Laser Welding

Abstract

Remote laser welding between coppers is used in our HV inverter. Moreover, many studies of in-process monitoring system using such as reflected light, plasma light, and heat radiation have been reported. However, it is difficult to apply their monitoring systems in product line due to large high frequency noise and not satisfactorily measuring weld area. In this paper, we showed that plasma light not only can detect lack of penetration, but can calculate weld area in quantitative terms using multiple regression equation. These results provide improvement of laser process monitoring accuracy as well as new quality assurance system.