Estimation of weld bead size in CO2 laser welding by using multiple regression and neural network

Authors
Citation
H. Park et S. Rhee, Estimation of weld bead size in CO2 laser welding by using multiple regression and neural network, J LASER APP, 11(3), 1999, pp. 143-150
Citations number
13
Categorie Soggetti
Optics & Acoustics
Journal title
JOURNAL OF LASER APPLICATIONS
ISSN journal
1042346X → ACNP
Volume
11
Issue
3
Year of publication
1999
Pages
143 - 150
Database
ISI
SICI code
1042-346X(199906)11:3<143:EOWBSI>2.0.ZU;2-M
Abstract
On the laser weld production line, a slight alteration of welding condition changes the bead size and the strength of the weldment. A measurement syst em is produced by using three photodiodes for detection of the plasma and s patter signal in CO2 laser welding. The relationship between the plasma or spatter and the bead shape, and the mechanism of the plasma and spatter sig nals were analyzed for the bead size estimation. Penetration depth and bead width were estimated using multiple regression analysis and the neural net work. (C) 1999 Laser Institute of America. [S1042-346X(99)00203-X].