Hybrid force control for robotic friction stir welding

• Applicant: Professor Dr.-Ing. Alexander Verl
• Subject Area: Production Automation and Assembly Technology
• Term: since 2026
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 573040703

Due to their flexibility and cost efficiency, industrial robots (IR) are increasingly being used for manufacturing processes. One such process is friction stir welding (FSW), which places high demands on the manipulator due to the high process forces and force gradients involved. The high compliance of IR compared to traditional machine tools necessitates the implementation of force control for robotic FSW. To produce weld seams, state-of-the-art research for FSW with IR employs an external force control with an underlying position control. In this approach, the superimposed force controller computes additional position setpoints for the robot's inherent position control at the axis level. The main disadvantage of this approach lies in the low bandwidth of the force control, which results from its structural design. This, in combination with the rapidly increasing forces occurring in FSW, leads to insufficient control performance. Particularly during the plunging phase at the start of the process, this results in drawbacks such as reduced process speed and oversized equipment. The scope of this project is, therefore, to develop an adapted control structure with parallel rather than superimposed force control, which has the potential to significantly increase bandwidth compared to the state of the art, thereby optimizing both process speed and quality. The basis for this is a dynamic modeling and compensation of the typical gear coupling in the wrist axes of IR to prevent deviations from the path during force-controlled operation. The hybrid force control with integrated online adaptation will first be developed and designed in simulations using a process model. To transfer the control structure into industrial practice, the control architecture will be implemented and validated on a test bench.

• DFG Programme: Research Grants
• Co-Investigator: Dr.-Ing. Armin Lechler