Enhancing the laser weld seam quality with dynamic beam shaping and Bayesian optimization

This module contains a presentation on the efficient identification of process parameters for laser welding using dynamic beam shaping, which was presented by Kim Glumann at the LiM Conference in 2025. The presentation addresses the challenge of reducing welding defects such as spatter, undercuts and hot cracks, which are caused by instabilities in the keyhole and melt pool during the welding process. It shows how dynamic beam shaping based on Coherent Beam Combining can be combined with Bayesian optimisation to identify suitable parameter sets with a significantly reduced number of experiments. The presented study was conducted on AW5754 and demonstrates that the approach can be used to achieve weld seams with defined penetration depths and minimal defects, even at high welding speeds.