FOR 5134:  Solidification Cracks during Laser Beam Welding: High Performance Computing for High Performance Processing

• Subject Area: Mechanical and Industrial Engineering ; Computer Science, Systems and Electrical Engineering ; Materials Science and Engineering ; Mathematics
• Term: since 2021
• Website: Homepage
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 434946896

Non-contact joining processes such as laser beam welding have become increasingly important in industrial production in recent decades due to increasing automation. The laser beam is the preferred tool for joining metallic materials due to the non-contact machining process, short cycle times and small heat-affected zones. A common problem with laser beam welding is solidification cracks, which significantly limit the range of materials that can be welded. Solidification cracks are caused by the complex interplay of thermal, metallurgical and mechanical factors and can only be detected inside the weld seam with considerable effort, although they represent a significant risk of the weld seam failing on the front side. So far, no models are known from the state of research that provide a quantitative statement on crack probabilities. This is the core objective of the research group: to provide a quantitative statement on crack probabilities as a function of the material and the process. To tackle this task, a highly interdisciplinary research group consisting of production engineers, materials scientists, mathematicians and computer scientists was brought together. In the first phase of the research group, the process was numerically modeled from the micro to the macro scale using a model material (stainless steel 1.4301) and compared and validated with experiments. This showed that, for example, the influence of the metal vapor on the absorption of the laser radiation and the melt pool geometry was underestimated, so that on the one hand there is still a need in the field of modeling and on the other hand further experiments are necessary in order to be able to model the solidification crack formation numerically. In order to prove the generalization of the methodology and to transfer it to materials that are highly susceptible to hot cracking, the material spectrum is to be extended to nickel-based alloys in the second phase, which poses completely new challenges in the field of material modelling, for example. Using numerical modeling, approaches will then be developed at different levels to derive crack probabilities based on material and process. In addition, this will also be done across the board using a data-driven, AI-based approach. Although the research group is working on a strongly application-driven issue from production engineering, significant progress in basic science can also be expected in the second phase due to the holistic and highly interdisciplinary approach.

• DFG Programme: Research Units

Projects

• Acceleration platform for the simulation of solidification cracks during laser beam welding (Applicants Koeppe, Arnd ;  Köstler, Harald ;  Selzer, Michael )
• Coordination Funds (Applicant Schmidt, Michael )
• Machine learning and high-performance computing for efficient simulation of thermo-elastoplastic solidification processes (Applicants Klawonn, Axel ;  Lanser, Martin )
• Macro-mechanical prediction of the prevention of solidification cracks during laser beam welding (Applicant Rethmeier, Michael )
• Massively parallel simulation of the melt pool and the solidification grain structure in laser beam welding (Applicants Köstler, Harald ;  Markl, Matthias )
• Microstructure simulation of solidification in the weld (Applicants Nestler, Britta ;  Schneider, Daniel )
• Multiscale thermoplastic analysis in the solidification zone (Applicants Scheunemann, Lisa ;  Schröder, Jörg )
• Thermo-fluid-dynamic investigation of the keyhole and the melt pool (Applicant Schmidt, Michael )

• Spokesperson: Professor Dr.-Ing. Michael Schmidt