Classification of laser weld penetration based on measurements of selected process emissions was investigated. Optical, acoustic and charged particle emissions were measured during CO2 laser welding of thin sheet steel butt joints over a range of laser power and travel speed. The values of the operating parameters were chosen so as to produce both full and partial penetration welds. The various sensor signals were compared to each other and were found to be significantly (but not totally) correlated. The frequency spectra of the signals from the full and partial penetration welds were processed using linear discriminant analysis. This statistical technique is designed to separate data corresponding to different conditions, facilitating recognition of the different conditions from measured signals. It was found that linear discriminant analysis of combinations of the plasma charge plus acoustic signals as well as all three sensor signals in combination produced a good separation between signals from full a...
[1]
Eckhard Beyer,et al.
On-Line Plasma Diagnostics For Process-Control In Welding With CO2 Lasers
,
1989,
Other Conferences.
[2]
W. M. Steen,et al.
Multi-frequency fibre optic sensors for in-process laser welding quality monitoring
,
1993
.
[3]
Hongping Gu,et al.
Resonant acoustic emission during laser welding of metals
,
1996
.
[4]
Dave F. Farson,et al.
Frequency–time characteristics of air‐borne signals from laser welds
,
1996
.
[5]
D. J. Brookfield,et al.
Plasma charge sensor for in-process, non-contact monitoring of the laser welding process
,
1996
.
[6]
Hongping Gu,et al.
A statistical approach to acoustic monitoring of laser welding
,
1996
.