Abstract
By means of ISO 9000 and other by law enforced rules the manufacturing industry is forced to document and control the quality of their products. With the aim to improve the quality of the manufactured products and the reliability of the corresponding process an improved understanding of the observability and controllability is desired. On-line process monitoring systems will be able to achieve reliably these requirements.
Process diagnostic devices commonly used in actual quality monitoring systems are spatially integrating plasma sensors. Analysis of the data is performed in different ways which will be shown on the example of commercially available sensors. Experimental results have shown that the reliability of the quality prediction can significantly be increased by using an image processing system with a two-dimensional spatial resolution instead. Based upon these results a different kind of online process control system was developed and can be adapted for different applications such as drilling, welding, cutting or surface processing. In the case of welding the system can monitor the penetration depth, a full penetration weld, deviations from the desired welding path, the seam width, stability of the capillary and defects of the seam caused by ejection of molten material. Recent results obtained with the system will be presented.