Shared control of robotic friction stir welding in the presence of imperfect joint fit-up

In this paper a shared control strategy is presented that allows a skilled operator to identify irregularities that occur during robotic friction stir welding (FSW) and assist the robotic system in producing an appropriate response. Human operators are adept at identifying disturbances; however, the complexity of the friction stir welding process makes it difficult for the operator to respond. While examining the capabilities of shared control in friction stir welding, this paper focuses on responding to defects that are caused by a lack of workpiece material during butt welding, such as gaps. A compensation strategy is presented that combines the human operator's perceptual strengths with an automated procedure for adjustment of the process parameters (i.e. travel angle and plunge depth). Experiments comparing four control strategies are performed while welding 5083-H116 aluminum. Through our experiments we demonstrate that if the FSW control task is appropriately shared between the human operator and the computer control system, the weld quality (strength) can be improved (from 9 ksi to 31 ksi for a gap size of 2.5 mm) as compared with the nominal case in which no corrections are made.