Abstract
Friction stir welding (FSW) is an innovative welding technology. It offers unique advantages due to the fact that the solidus temperatures of the workpieces are not exceeded. FSW is typically applied to lightweight aluminum structures where the joint quality is crucial. As the weld quality is mainly governed by the welding temperature, a temperature control system was developed to ensure homogeneous weld properties. In this study, two challenges were addressed. First, a novel temperature measuring system was introduced. For this purpose, an empirical correlation between the welding temperature and the process torque was suggested. Second, a L1 adaptive controller combined with an input linearization method was used to develop a torque-based temperature control system. This control structure was selected due to its robustness and fast adaption rates. Sufficient robustness was required, to account for the mechanical vibrations that occur when using industrial robots for welding. These vibrations can be detected as oscillations in the torque signal. The developed control system was successfully integrated into to an FSW machine and the achievable control quality as well as the absolute accuracy of the measuring method were investigated in experiments.
Original language | English |
---|---|
Pages (from-to) | 391-403 |
Number of pages | 13 |
Journal | Production Engineering |
Volume | 12 |
Issue number | 3-4 |
DOIs | |
State | Published - 1 Jun 2018 |
Keywords
- Friction stir welding
- L1 adaptive control
- Robotic FSW
- Temperature control
- Temperature measurement