Correlation between process parameters and measured signals during laser welding of plates with artificial defects: Features of monitoring methods for laser welding and their application (3)

Summary

The authors are engaged in a research programme intended to clarify the behaviour of laser-induced plasma during laser welding and to establish its applicability in monitoring laser welding parameters.

The first paper describes direct observations of laser welding by CCD camera and clarification of shielding gas plasma by measurements of the spectrum emission, acoustic emission, and plasma potential signals. The second paper examines the appearances of the measured signals in response to changes in the basic welding conditions. The present paper seeks to establish the responsiveness of the signal behaviour to welding conditions, such as irregularities or defects occurring during laser welding of plates with artificial defects. The main results obtained are as follows:

1. In the shear cut identification test, the shear cut can be identified by the spectrum emission intensity and plasma potential. Through simultaneous measurement of the spectrum emission intensity and plasma potential signals, poor melting due to abnormal beam energy absorption by the Ar plasma can be identified.

2. In the vertical hole identification test, the vertical holes can be identified by the spectrum emission, acoustic emission, and plasma potential. The molten pool is perturbed by the hole before the laser beam reaches the hole, so that the keyhole is perturbed, and the output waveforms of the acoustic emission, spectrum emission, and plasma potential change before the hole is reached.

3. In Ni powder packed and Cu wire packed horizontal hole tests, data can be obtained on the fusion behaviour in the bead longitudinal section direction through observations being made of the fusion pattern in the Ni powder concentrated layer. Although a flow of molten metal can be clearly identified in this cross-section, the molten metal flows in a direction opposite to that of laser welding progression. If large defects or holes are present in front of the direction of laser welding progression, the flow is perturbed, and the keyhole behaviour is affected. Differences in the identification pattern of these cross-sectional signals can be determined by comparison of the spectrum emission and acoustic emission.

4. A comparison of the characteristics of the spectrum emission acoustic emission and plasma potential signal behaviour for identification of artificial defects reveals that the plasma potential signal responds highly sensitively to the presence of plasma and that the spectrum emission signal responds highly sensitively to migration of the plasma, including the keyhole.