Prediction and control of flaws in welds play an important role in design of complex structures against fatigue and fracture failure. Cold laps, outer lack of fusion, are small imperfections at the weld toe that appear under improper welding parameters and act as fatigue crack initiation sites which lead to a reduction in the life of a welded structure. The goal of this study is to investigate the influence of welding parameters and their variations on weld quality concerning formation of toe imperfections termed "Cold Laps" in tandem gas metal arc (GMA) welding. A statistic-based design of experiment for quality assessment in robotic tandem GMA welding was implemented. The experimental set up along with selection of welding parameters and their variations are based on the "Arc wise" method. To facilitate evaluation of weld quality a rapid destructive method was implemented. In this method weld toes could be opened and investigated longitudinally. Investigations of the fracture surface between the weld toe and base material was made, where two types of cold laps were observed, namely overlap and spatter. Measurements showed that these imperfections are in the form of semi elliptical micro cracks with depth between 0.1 and 1.5mm with a length between 0.1 and 4 mm. The statistical software Modde 5.0 was used to handle welding parameters, variations and quality assessment.
