Fuzzy modeling based estimation of short circuit severity in pulse gas metal arc welding

Avoiding short circuit is an essential condition for achieving good quality welds in Pulse Gas Metal Arc Welding (GMAW-P). Estimating short circuit in any welding process is dependent on proper selection and optimization of welding process parameters. Such optimization is critical in the GMAW-P wherein wire melting is closely dictated by numerous pulsing parameters in comparison to the conventional GMAW process. Fuzzy Logic based models are an excellent alternative in such situations where a complex relationship between the large number of predictor variables (independents, inputs) and predicted variables (dependents, outputs) exist and are not easy to articulate in the usual terms of correlations or differences between groups. In this paper, we have proposed an input output fuzzy model for estimating the short circuit severity in terms of number of shorts per pulse for GMAW-P process. Eighteen factors representing the characteristics of the pulse waveforms are employed as predictor variables and the short circuit severity (or number of shorts per pulse) is predicted on the basis of a modified exponential membership function fitted to the fuzzy sets derived from predictor variables. The exponential membership function is modified by two structural parameters that are estimated by optimizing the criterion function associated with the fuzzy modeling. The experimental data consists of GMAW-P welding of 6XXX group of aluminum alloys. The results demonstrate that proposed fuzzy model could estimate the short circuit severity with high accuracy.