Fuzzy logic based quality monitoring in short-circuit gas metal arc welding

Conventional methods, e.g. destructive and non-destructive testing methods, are expensive and time-consuming; therefore, possibilities of online and automated quality control of a welding process during welding as such are investigated. The paper deals with the possibilities of application of fuzzy logic to the analysis of weld quality, particularly assessment of the weld surface condition by means of measurable electric signals emitted during welding. A simple fuzzy inference system was realised which could relatively efficiently assess the weld quality on the basis of time variations of the welding voltage and short-circuit time in a certain time window.