Degree Name

Doctor of Philosophy


School of Electrical, Computer and Engineering - Faculty of Informatics


This research sets out to explore the mass and thermal transfer mechanisms involved in gas metal arc welding. This analysis was undertaken to enable improved process control strategies to be developed and evaluated. As a result of the investigations a new current control approach is proposed for dip transfer GMAW for use with conventional primary inverter power sources. This new control approach offers improved spatter reduction and fusion control compared to conventional constant voltage control systems. The control approach is based on the premise that if optimum droplet transfer conditions exist, then the short circuit current can be minimized through clamping, the droplet transfer taking place primarily under the influence of surface tension. Unlike previous controlled dip transfer strategies this approach places no reliance on pre-emptive detection of the short circuit rupture or on rapid peak current reduction. Novel adaptive control and monitoring techniques have been developed to assist in assessing the degree to which current clamping is introduced to the process, as well as identifying the type and degree of spatter generated. A study of optimized welding conditions has identified a relationship of use for the selection of welding parameters. The study led to the derivation of a hypothesis that defines the optimum short-circuiting event for droplet transfer. The investigation of the welding process has also identified inherent process irregularities when using CO2 shielding gas, which has led to the development of control strategies whose actions minimize the affect and enhance the welding process stability. To validate the proposed control approach a DSP based power source controller was designed and built. A comprehensive experimental program gave quantitative indication of the benefits likely to be achieved in practical welding situations.