Kinematic weld-pool control for additive manufacturing with high-performance materials by gas tungsten arc welding

• Applicants: Professor Dr.-Ing. Burkhard Corves; Professor Dr.-Ing. Uwe Reisgen
• Subject Area: Production Automation and Assembly Technology ; Joining and Separation Technology ; Primary Shaping and Reshaping Technology, Additive Manufacturing
• Term: since 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 442454814

Additive manufacturing using arc welding processes belongs to the DED-Arc (Direct Energy Deposition Arc) process category in accordance with ISO/ASTM 52900. From a sustainability perspective, additive manufacturing using plasma welding is particularly interesting for repairing (e.g. wear compensation) or extending existing components (e.g. pipe flanges) made from high-performance materials. In order to additively manufacture on any surface, a combination of trajectory planning based on real surfaces and stable material application is required. Both aspects are largely unexplored in relation to TIG or plasma welding processes. The ROS2-based system to be developed offers an excellent basis for this, as kinematics, sensor technology and actuators are merged in an open system. Many existing functions can be further developed to achieve this goal (in the course of the continuation). The central enabler is access to the three levels of offline, in-process and real-time planning/control and the cooperative implementation between welding technology and robotics. In contrast to the previous project, the process chain is to be broken up, supplemented by the use of dual quaternions, and the planning cycles are to be run through for each path segment. This opens up the planning for analyses (e.g. based on a laser line scanner) and a subsequently readjusted in-process path and trajectory planning. Real-time access to the trajectory also allows the trajectory to be adjusted during the process (e.g. as a distance control between the component and the welding torch, comparable to AVC but in 6D). As a prerequisite for an error-free starting point for offline planning, the components (robot, welding head and sensor) must first be calibrated to each other in an automated process.

• DFG Programme: Research Grants
• Co-Investigators: Dr.-Ing. Markus Schmitz; Dr.-Ing. Rahul Sharma