Drop transfer mode prediction in pulse GMAW of aluminum using statistical model

Abstract

Processes like GMAW-P introduce additional complexity (with need of defining supplementary variables such as peak current, base current, peak time, base time, frequency and duty cycle) which needs intense process knowledge for correctly defining pulsing parameters to achieve good quality. For achieving good weld quality one droplet should detach in every pulse to produce best weld quality with minimal defects and spatter. But depending upon the input pulse parameters, drop transfer can result in either one drop or multiple drops per pulse which directly influence the final weld quality. The purpose of this study was to develop a statistical model able to estimate the number of drops transferred from the end of the electrode to the workpiece quantitatively using the pulsing parameters of the pulse waveform acquired during the real time operation. Model was proposed to estimate the number of drops transferred from the end of the electrode to the workpiece, performing the multiple regression analysis between model and quantified number of drops transferred from the end of the electrode to the workpiece.