Application of Taguchi methods to laser process engineering

A principal goal for laser job shops in recent years has been the consistent achievement of acceptable cut quality, irrespective of the work piece or cutting conditions. At present, in order to achieve a good quality cut edge, control parameters such as cutting speed, shield gas pressure and gas type have to be adjusted individually in factorial experiments, i.e. one variable at a time. Their effects also have to be tested on the material to be processed so that the cut quality can then be optimised. This is both costly and time consuming. In addition, optimal parameter settings are still frequently determined from the operator's own experience of the process. The application of Taguchi methods, based on Genichi Taguchi's approach to Parameter Design, not only reduces the number of experiments required but also determines the optimal parameter settings in a more rigorous manner since it takes into account noise factors which full-factorial experiments do not. This paper details a generic methodology for Taguchi methods applied to the CO₂ laser cutting process. To demonstrate this concept a particular application relating to the processing of decorative ceramic tile is used. This outlines the form of the necessary experiments, analysis of results and determination of the optimal control parameter settings, as well as the development of a predictive model.