PREDICTIVE CATHODE MAINTENANCE OF AN INDUSTRIAL LASER USING STATISTICAL PROCESS-CONTROL CHARTING

The performance of a de-excited transverse-flow CO2 lasts welder can b e adversely affected by many factors such as electrode arcing, optical contamination, poor maintenance, etc. Many of these factors which occ ur in production environments are unpredictable, and therefore, diffic ult to simulate in the laboratory or the laser manufacturer's faciliti es. In this article the feasibility of using statistical control chart s for laser condition monitoring was demonstrated. in particular, its application for predictive maintenance for the cathode and for verific ation of the effectiveness of cathode cleaning were investigated. The same approach is readily applicable to other laser components. With mo re experimental data, optimal control limits can be defined to dictate the maintenance schedule. Future work will include the development of an on-line user-friendly monitoring system that can be used to indica te the condition of the laser and warn the user should any abnormaliti es be observed. The proposed monitoring system can also be used to det ermine when and what maintenance is required to keep an industrial las er functioning properly. (C) 1998 Laser Institute of America.